RES!DENT ADJUSTMENT PATTERNS N THE RURAL-URBAN FRINGE Thesis for fine Degree of DH. D. MICHIGAN STATE UNIVERSETY Richard Alien Kurtz 1959 llllllllmlll Ill11mTIIWWIIU’HIWIW L THE—5.. 3 1293 ooesa 3372 ———-—.__ ______ This is to certify that the thesis entitled /) _4 . ,_ ’1', , / 7520 (/14! ”(C/1 [St L7fl0(+%( 71/ C/ [I/ 5?, 9/32, ’L 7L-’2/'/ ._ " ”7 , / c— ’ ' a” L/Ze / {éiccf/ - (fie/ed- Mr ch K70 presented by 66/9144?" 5214 ~ / {6/62/17 a ‘7 has been accepted towards fulfillment of the requirements for /2 A5" degree in 5&4: 57 [1255/5 -a LIBRARY Michigan State University PLACE II RETURN BOX to remove this checkout from your mood. TO AVOID FINES return on or baton dd. duo. DATE DUE DATE DUE DATE DUE LL 1;”: a ’MAR 23 [_ MSU In An Afimdlve Action/Equal Opportunity lnetltmlon cmms-m RESIDENT AMUSEMENT PATTERNS IN THE RURAL-URBAN FRINGE by Richard Allen Kurtz A THESIS Submitted to the School of Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MOTOR OF PHILOSOPHY Department of Sociolog and Anthropology Year 1959 Approved AN ABSTRACT Recent information has suggested that the traditional descrip- tion of urban social life, which emphasizes the lack of cohesive and in- formally organized interaction patterns, must be questioned. Since in- vestigators in the rural-urban fringe area have reached conclusions which parallel the urban stereotype the possibility exists that this view is also distorted. Therefore, the present study was addressed to an empirical analysis of social life in the fringe area. The study site chosen for analysis was the fringe area surround- ing Lansing, a typical industrial city of the North Central States. A random sample of residents was drawn, representing famers, part-time farmers, and nonfarmers. A battery of questions was submitted to sample members by utilizing a field schedule. Questions were directed to two dimensions: 1) participation patterns in the area, and 2) identifications with the area. The first variable was termed "objective integration" and the latter, "subjective identification." By utilizing scalogram analysis an integration scale, composed of four questions, was derived. Scalogram techniques were not successful for the identification dimension, nor was the method of sum- mated ratings. The failure of scaling methods created a need for a to- tally new approach to the analysis of identification. Therefore, typ- ologies based on the direction, content, and depth of identification were constructed. Frequency distributions of both the integration scale and the identification typologies suggest that the conclusions of investigators who described the fringe area as an "institutional desert" cannot be accepted. Over forty percent of the respondents exhibit high partici- pation patterns and more than ninety percent may be classified as posi- tively identified with the area. It is suggested that the discrepancy is a result of the fact that previous investigators conducted their re- search in areas which are more clearly suburban than fringe and, in ad- dition, that sociologists working in areas surrounding central cities have usually concentrated on newcomers, while ignoring long-time resi- dents . Having established the existence of both integration and iden- tification, the study was focused on attempting to account for differen- tiale. It was ascertained that long-time residence, relative stability, rural and non-Lansing background, and commitment to the area are all as- sociated with a high degree of within-area participation. Direction of identification could not be accounted for but it was found that both content and depth are associated with the same general variables as is integration. In addition to accounting for adjustment the study was addressed to determining the influence of integration and identification on com- munity orientations. It was found that definitions of the area, as either rural or urban, are affected by adjustment. Respondents who ex- hibit high degrees of integration, whose identifications with the area are based on friendship patterns and familiarity with it, and who had no prior choice for living in the area, define their place of residence as rural, rather than urban. . The data suggest that patterns of integration and identification do exist in the fringe, and that these are influenced by the residential experiences of residents and by commitment to the area. In addition, adjustment patterns affect some community orientations. ACKNOWLEDGEMENTS The writer's interest in areas surrounding central cities devel- oped when he was a graduate student at the University of Connecticut, where Dean Nathan L. Whetten was pioneering sociological interest in suburbanization. He wishes to express his gratitude to Dean Whetten for initial inspiration. Dr. Joel Smith was instrumental in guiding the writer through every phase of the research. He is indebted to Dr. Smith for his con- stant willingness to contribute ideas and criticisms. To Drs. William H. Form and J. Allan Beegle grateful acknowledge- ment is made for their assistance throughout the research. Gregory P. Stone contributed many ideas when the project was first being formulated. Finally, to his wife, Patricia, the writer wishes to express deep gatitude for her understanding whenever he found it necessary to devote himself to the study, instead of his family. In addition, the writer's wife has contributed her knowledge of research to all phases of the study. Cheuptusxr III. III. III. Tr. Vifilo TABLE OF CONTENTS Page {EHEDRETICAL FRAMEWORK, STATEMENT OF THE PROBLEM, AND momLmY O D O D O O O O O O O O O O O O O O O O O O O 1 Social Life in the Fringe: General Comments Social Life in the Urban.Area IDefinition of the Fringe Area Social Life in the Fringe: Specific Propositions Statement of the Problem Methodology and Techniques A DESCRIPTION 0? TE AREA. 0 O O O O O O 0 O O O O O O O O 29 Characteristics of the Central City Characteristics of the Lansing Fringe Area AN ANALYSIS OF OBJECTIVE INTEGRATION AND SUBJECTIVE IDmTIFICATION O O O O O O O O O O O O O O O O O O D O -. 57 Sample Distributions The Relationship between Objective Integration and Subjective Identification RESIDmTIAL HISNRYe O O O O O O O O O O O I O O O O O O 0 81 Accounting for Integration and Identification Summary: Accounting for Adjustment Patterns CHARACTERISTICS ormpomanon. . . . . . . . . . . . . 113 Accounting for Integration and Identification Summary: Accounting for Adjustment Patterns COWITYORIENTATIONS..................140 Index Construction .Accounting for Community Orientations Summary: Accounting for Community Orientations summrmconcmsmns..................163 Background of the Study Results Generalizations APPENDIXES A. INTERVIEW SCHEDULE. . . . . . 3’ TABLES. O O O O O O O O O O I C. THE MEASUREMENT OF ADJUS'DIENT BIBI‘IWRAPHYO O O O O O O O O O O O O 176 194 227 250 .Q LIST OF TABLES Table Page 1 Sample-Universe Goodness-of Fit Test 25 2 Percentage Distribution of Male Workers by Industry Groups for Lansing, 1950 Census 194 5 Percentage Distribution of Male Workers by Major Occupational Groups for Lansing, 1950 Census 194 1+ Percentage Distribution of Household Heads by Age for Lansing Sample 195 5 Percentage Distribution of Male Workers by Educational Attainment for Lansing Sample 195 6 Percentage Distribution of 1949 Family Income for Lansing, 1950 Census 195 7 Percentage Distribution of Birthplace of Workers for Lansing Sample 196 8 Year Moved to the Area 196 Difference Between Year Moved to the Area and Year Moved to Present Address _ 197 10 Number of Residences Occupied Since 1940 197 11 Place of Previous Residence 197 12 Rural and Urban Nature of Previous Residential Area 198 15 Area in Which Head of Household was Raised 198 11+ Rural and Urban Nature of Area in Which Head of Household was Raised 198 15 Occupations of Sample Members 199 16 Lansing and Non-Lansing Place of Work 199 17 Tier of Residence 199 13 Tenure Status 200 19 Age of Head of Household 200 20 Schooling of Head of Household 200 21 Number of Children Under Seventeen 201 {Dable 22 23 21+ 25 26» 27' 28 29 31 32 33 35 56 0'9 Family Gross Weekly Income Why the Residential Area was Chosen Opinions of Whether the Area has a Name Name Given to the Area Opinions of Whether the Area is Part of a Larger Area Name of the Larger Area Opinions of Whether Rural, Urban, and "In-between” Areas Exist, and the Type Chosen to Describe the Fringe Opinions of Whether there are Differences Between People in the Residential Area and City People Stated Differences Between Residential Area People and City People Opinions of Whether there are Differences Between People in the Residential Area and Rural People Stated Differences Between Residential Area People and Rural People Opinions of Whether Farm.People Think Differently from City Dwellers Opinions of Whether People in the Area come Closer to the Farm or City Way of Thinking Opinions of the Size of the Local Area How "Area Around Here" Boundaries were Decided Upon Opinions of Whether Respondents Would Like to Continue Living in the Area Reasons for Wanting to Remain in the Area Reasons for Wanting to Leave the Area Opinions of Whether the Area is Getting Better or Worse Opinions of Why the Area is Getting Better Facilities Desired in Area Opinions of Whether Living in Lansing has Advantages Over Living in the Area Stated Advantages of Lansing and Present Residential Area Opinions of Whether Living in the Outlying Farm Area has .Advantages Over Present Residence Area Stated Advantages of Outlying Farm Area and of Present Residential Area Opinions of How Well People in the Area.Know Each Other Page 201 202 202 202 203 203 204 204 210 211 211 212 212 ikfldle 37 39 Jul 142 #3 89 ‘+5 OU‘Q #7 49 SO 51 0'9 52 UN 53 5# How Many People in the Area Respondents Know by Name Number of Families Respondents Come into Contact with Frequently Number of Families with which Respondents Spend Afternoons or Evenings WithinpArea Lending Patterns Areas in which Friends are Visited Residential Area of Three Best Friends Number of Organizations Belonged to IMeeting Places of Organizations Belonged to Type of Organizations Belonged to Participation in Local Informal Facilities Participation in Voluntary Community Groups Type of Voluntary Groups in Which Participated Participation in Local Issues Type of Issue in Which Participated Reasons for Participating in Local Issues Reasons for not Participating in Local Issues Areas Utilized for Shopping Sources Utilized for News About the Area, the Township, the County, and Lansing Leisure Time Activities Fbod Raising by Nonfarmers Proportion of Family Food Cost the Production Covers Attitudes Toward the Formulation of the Lansing.Metro- politan Planning Commission Reasons for Endorsing the Formulation of the Commission Reasons for Opposing the Formulation of the Commission ‘Attitudes Toward a Commission Proposal for Annexation Reasons for Endorsing an Annexation Proposal Reasons for Opposing an Annexation Proposal Distribution of Sample Members Among'Objective Inte- gration Scale Types JDistribution of Sample Members Among Direction of Sub- jective Identification Types 213 214 214 214 215 215 215 216 216 216 216 217 217 217 218 221 221 221 222 222 222 58 59 Table 55 Distribution.of Sample Members Among Content of Subjec- tive Identification Types 0 56 Distribution of Sample Members Among Depth of Subjec- tive Identification Types 57 Illustration of the Prediction Improvement Measure 58 Relationship between Objective Integration and Direction of Subjective Identification 59 Relationship between Objective Integration and Content of Subjective Identification 60 Relationship between Objective Integration and Depth of Subjective Identification 61 Summary of the Relationship between Degree of Integra- tion and Two Aspects of Identification 62 Summary of the Degree of Association between Degree of Integration and Two Aspects of Identification 63 Relationship between Depth and Content of Identification 6# Summary of the Relationship between Objective Integration and Residential History 65 Summary of the Relationship between Objective Integration and Indexes of Residential History 66 Summary of the Relationship between Direction of Iden- tification and Residential History 67 Summary of the Relationship between Direction of Iden- tification and Indexes of Residential History 68 Summary of the Relationship between Content of Identi- fication and Residential History 69 Summary of the Relationship between Content of Identi- fication and Indexes of Residential History 70 Summary of the Relationship between Depth of Identifi- cation and Residential History 71 Summary of the Relationship between Depth of Identifi- cation and Indexes of Residential History 72 Summary of the Relationship between Objective Integra- tion and Characteristics of the Population Page 60 6O 63 69 70 74 76 79 83 91 92 96 101 104 108 116 Table 73 7A 75 76 77 78 79 80 81 82 85 8h 85 86 ‘87 88 89 90 Summary of the Relationship between Direction of Identi- fication and Characteristics of the Population Summary of the Relationship between Content of Identifi- cation and Characteristics of the Population Summary of the Relationship between Depth of Identifica- tion and Characteristics of the Population Distribution of Sample Members Among Localite Index Types Distribution of Sample Members Among Lansing Orientation Index Types Distribution of Sample Members Among'Urban Definition of the Area Index Types Distribution of Sample Members Among Rural Definition of the Area Index Types Distribution of Sample Members Among Satisfaction Index Types Summary of the Relationship between Objective Integration and Indexes of Community Orientation Summary of the Relationship between Direction of Identifi- cation and Indexes of Community Orientation Summary of the Relationship between Content of Identifica- tion and Indexes of Community Orientation Summary of the Relationship between Depth of Identifica- tion and Indexes of Community Orientation Distribution of Sample Members Among StabilityéMobility Index Types Distribution of Sample Members Among Rural-Urban Experi- ence Index Types Distribution of Sample Members Among Lansing Experience Index Types Distribution of Sample Members Among Urban and Lansing Experience Index Types Distribution of Sample Members Among Commitment Index Types Objective Integration Guttmanptype Scale, Indicating Ob- served and Chance Response Pattern Marginals Page 125 224 224 146 151 153 157 225 225 225 Table Page 91 Derivation of the Discriminating Power of Items in a Likert-type Scale 245 LIST OF CHARTS 1 The Interrelationship between Identification and Integration, Both Dichotomized into Positive and Negative Categories 19 2 Perfect Scale Types for a Four-Item Guttman-type Scale 229 CHAPTER I THEORETICAL FRAMEWORK, STATEMENT OF THE PROBLEM, AND METHODOLOGY Social Life in the Fringe: General Comments Investigations of social life in the rural-urban fringe area have tended to emphasize the paucity of informal associations and the lack of feelings of solidarity among residents. Speculations and con- clusions indicate that the area lacks the basic elements of cohesive community organization, and this is generally attributed to the pecu- liar nature of fringe area residents.1 Investigators have concentrated most frequently on residents who migrated to the fringe while maintain- ing their economic and social ties with the area from which they moved, usually the city. Residents are characterized as subjectively and ob— jectively oriented to the nearby urban area, utilizing the fringe com» munity as a barracks or sleeping quarters. Fringe dwellers are viewed as beinggig but not 2; the fringe community. Thus, sociological descriptions of the fringe indicate that so- cial relationships in the area are at a minimum, that social life in the area is impersonal, and that there exists a lack of voluntary asso- ciation among residents. Fringe dwellers are characterized as express- 1Specific propositions within this framework are being reserved for a later section of this chapter. Only a general statement of the problem is presented here, to introduce the orientation. - 1 - - 2 - ing*little interest in their communities of residence since they have maintained ties with areas outside the fringe. The fringe is not char- acterized as an area of social disorganization by these investigators, rather the descriptions indicate that it lacks social integration. The basic requisites of an integrated community life are viewed as absent. Social Life in the Urban Area The traditional view of social life in the urban area parallels the fringe description in emphasizing the lack of cohesive, informalby organized social relationships. The classical expressions of the nature of urban social organization are presented by Simmel2 and Wirth,3 who generalized from the rural-urban contrasts presented by Durkheimk and TBnnies.5 The point made by these theorists is that social life in the urban area is of a different order, or quality, than rural patterns. They were observing that urban living necessitated a social life quite different from rural patterns.. Both Durkheim and T6nnies emphasized differences in quality but they were also quick to point out that in the empirical case, as distin- guished from model constructs, elements of the cohesive, integrated life still exist. In the "Preface" to the second edition of The Division 9; Labgguin Society, for example, Durkheim stresses the cohesion and.com- mon solidarity which the occupational group contributes to the social 2Georg Simmel, "The Metropolis and.Mental Life," Cities gggfl§gr elegy, ed. Paul K. Hatt and Albert J. Reiss, Jr. (Glencoe, Illinois: The Free Press, 1957), pp. 635-6u6. 5Louls Wirth, "Urbanism as a Way of Life," ibig., pp. 46-63. “Emile Durkheim, The Division 9; abor _i_n Society (Glencoe, Illinois: The Free Press, 1 7 . 5Ferdinand Tdnnies, Fundamental Concepts 2; Sociolo (Gemein- schaft gnd|Geeellschaft) (New York: American Book Co., 1955;. - 3 - ILife of the society characterized by organic solidarity.6 In a similar manner, T6nnies observes that As free and arbitrary products of thinking, these normal concepts are mutually exclusive; rational will and natural will are strictly separate entities....[However,] observation and inference will easi- ly show that no natural will can ever occur empirically without ra- tional will by which it finds expression, and no rational will withp out natural will on which it is based. [In addition, the development of Gemeinschaft]...tends toward an approach to Gesellschaft in which, on the other hand, the force of Gemeinschaft persists, although with diminishing strength, even in thg period of Gesellschaft, and remains the reality of social lifee In a recent article SJoberg points out, in agreement with Durk- heim and T6nnies, that urban centers foster special types of social or- ganization.9 Sjoberg is concerned with the pro-industrial city but he also indicates that industrial cities require a special kind of social structure. However, he emphasizes that the forms of social organization in industrial and pre-industrial cities are substantially different. "This . . . suggests that many phenomena which were thought to be gener- 10 to cities are generic only to urban industrial places."10 Thus any recent statements directed to urban life-styles should more precisely be directed to an "industrial Bryan? way-of-life. In addition, Duncan points out that the nature of social organ, ization within cities varies with city-size.11 He presents data which 6Durkheim,‘gp..git., pp. l-Bl. 7mm», 22. gi_t., p. 162. 81mm, p. 272. 9Gideon Sjoberg, "The Preindustrial City," Hatt and Reiss,‘gp. 9—22,, pp. 179-1880 10This is a comment made by the editors in reference to Sjoberg's article; see ibid., p. 177. 11Otis Dudley Duncan, "Optimum Size of Cities," ibid., pp. 759-772- ..on»9‘ p.073" I n I... I 0‘ . . I...‘.'. as: .q I...I'. :0..n .4 I'LIQ; P! (I 3 u l“ / 1‘?" ." )3” ti - h - indicate that marked differences exist between size categories and the facilities, organizations, municipal efficiency, and physical plans of cities. Duncan's extensive analysis may be viewed as an empirical sub- stantiation of the Durkheim—T6nnies position that social life in the ur- ban area is of a different order than rural patterns. In addition, Duncan's findings are an extension of this position since social life is demonstrably different according to a continuum of city-size.12 Prom.the DurkheimpTonnies frameworks Simmel, and.more recently Uirth, constructed models that deal specifically with large-size indus- trial urban social intercourse. These recent theorists ignored the in- timate and cohesive elements of urban social organization referred to by Durkheim and TBnnies in favor of presenting'a one-sided accentuation of social organization in construct terms; they deal with the special type of social structure necessitated by industrial urban living. Simmel's model construct of urban social organization indicates that the social relationships of the city resident are segmental and.im- personal. Participation on an effectual basis is ruled out because of the multiple, rapidly changing, impersonal contexts of urban social sit- uations. 'Uithin these contexts a spontaneous, affective way of life is not possible, rather, the individual must respond on a conscious, "inp tellectual" level. Punctuality, calculability, and exactness are terms used by Simmel to denote the life style; the clock and time schedule are symbolic of these traits. Expediency, sophistication, and efficiency / replace the emotional feelings of social relationships. ‘34:? I Wirth, in the same tradition, concentrated on urbanism as a way- 12The same conclusion may be reached by examining another of Duncan's articles; see his ”Community Size and the Rural-Urban Cons timmll,” ibide, pp. 35-“50 - 5 - of—life. He indicates that social organization in the urban area can be described as consisting of: ... the substitution of secondary for primary contacts, the weaken- ing of bonds of kinship, and the declining social significance of the family, the disappearance of the neighborhood,13nd the undermine ing of the traditional basis of social solidarity. In addition, he describes urban social contacts as impersonal, superfi- cial, transitory, anonymous, and segmental and he states that the urban- ite loses ". . . the spontaneous self-expression, the morale, and the sense of participation that comes with living in an integrated socie- ty."1# With this understanding of urban life, sociologists undertook an examination of social organization in the city. However, an emphap sis on the segmental, impersonal, secondary nature of urban social rela- tionships directed sociologists to the disorganizing, to the neglect of organizing, aspects of city life. Therefore, "early studies of cities focused on the personal and social problems which arise in, and contri- bute to, the 'disorganized atmosphere' of cities."15 Within this frame of reference, early investigations were oriented to a study of social disorganization in urban areas. This approach may be illustrated in an article by J.J. Rhyne.16 Rhyne's view of the city is expressed as follows: The highly artificial life of the great city, the stimulus which comes from close association of vast numbers, the excitement and strain under which many in the large cities live, and the highly developed competitive struggle to maintain status have combined to 1“:bid., p. 5h. 1&1.er 910 '23-‘50, Fe 600 15This is a comment made by the editors in ibid., p. a. 16J. J. Rhyne, "Social Man and His Community,” Cities £22.52f normal, ed. Elmer T. Peterson (Normal: university of Oklahoma Press, 19R): PP- 156-157. - 6 - give the greatlcity a disproportionate share of the major ills of human society. It is evident that the city is viewed as a social area in which the so- cial Problems of society are concentrated. Rhyne points to the preva- lence of unemployment, delinquency, crime, and mental disease in the ur- ban area. Essentially, the city is viewed as the center of personal and social disorganization. Recently, these descriptions of the city, i.e., descriptions which indicate that the city is a center of disorganization and which emphasize formal integration exclusively, have come under re-examina- tion.18 Investigations have revealed that the traditional view of the city presents a distorted picture of urban life. The traditional view- point has tended to orient urban sociologists to the dysfunctional ele- ments in urban society while the integrative aspects of city life have been ignored. This distorted view of urban life was largely responsible for a lack of research oriented to the intimate and personalized charac- ter of many urban relationships. Attention was directed to delinquency, crime, divorce, mental illness, and other social problems, while studies 17Ibid., p. 136. 18See, for example, Robert C. Angell, "The Moral Integration of American Cities," American Journal of Sociolo , LVII, Part 2 (July, 1951), pp. l-l#O; Morris Janowitz, The Communit Press inhan Urban.§gtr g5 (Glencoe, Illinois: The Free Press, 1952;; Gregory P. Stone, "City Shoppers and Urban Identification: Observations on the Social Psychol- ogy of Urban Life," American Journal of Sociolo ,LX (July, l95#), pp. 36-45; William H. Form, 23 gl., "The Compatibility of Alternative Approaches to the Delimitation of Urban Sub-areas," American Sociologi- gal Review, 19 (August, 1954), pp. #311440; Joel Smith, William a. Form, and Gregory P. Stone, "Local Intimacy in a.Middle-sized City," American Journal 9; Sociolo , 1.1 (November, 1951+), pp. 276-284; Joel Smith and William H. Form, "Urban Identification and Dis-identification" (unpub- lished paper); and Joel Smith and William H. Form, "Urban Identifica- tion: Orientations and Mechanisms" (unpublished paper). - 7 - of intimacy and solidarity in the urban area were not undertaken. The source of the distorted view is not clearly evident but it seems to have been based, to some degree, on a definite pro-rural bias. Tannies and Durkheim had based their theories on a rural-urban dichotomy which emphasized the calculative, rational, and instrumental orientation of city dwellers. In the rural community, on the other hand, social in- teraction was considered intimate, cohesive, and familistic.19 Such a dichotomy, even if stated objectively, will invariably be judged eval- uatively. - In an interesting article, Tomars attributes the negative eval- uation of the urban area, when compared to the rural area, to the sur- vival of rural values in American culture.20 He makes the point in his discussion by stating that ". . . emphasis upon rural values carries with it corresponding scorn for many urban values."21 Tomars' view is more clearly stated in the following passage: ". . . stress upon rural characteristics and values always implies its obverse--the deprecation of urban traits and values."22 The point here is that the descriptions of the city by Simmel, Hirth, and Rhyne are rooted in the theoretical rural-urban contrasts presented by Durkheim and Tbnnies. These descriptions, then, emphasize the non-integrative aspects of urban existence, an orientation seemingly based on a pro-rural bias. This bias may be attributed to the survival —__ 19See T6nnies, 22. git., pp. 119-154. Durkheim,‘gp..gi§., does not deal specifically with rural-urban contrasts but his theory lends itself to interpretation in these terms; see pp. 70-146 and, especially, pp. 17#—190. 20Adolph s. Tomars, "Rural Survivals in American Urban Life," Sociological Anal sis, ed. Logan Wilson and William L. Kolb (New York: Harcourt, Brace and Company, l9h9), pp. 371-378. 21Ibid. p. 376. 22Ibid., p. 377. IAA‘O' f.. “ .mt a. ..H u. I... V!II‘ is...“ ...‘n I t""-. - a - of rural values in American culture. Definition of the Fringe Area The preceding discussion of social organization and relation- ships in the urban area was presented in an attempt to provide an orien- tation to a more precise analysis of social life in the fringe area. Hewever, before fringe area social life is discussed in greater detail it will be necessary to define the fringe and indicate the context with- in which the social relationships take place. The usual definition of the fringe as an area which lies on a continuum somewhere between rural and urban.polar constructs23 is re- jected for five reasons: 1) The questionnable assumption that only one type of continuum exists; it is possible to fermulate a number of different constructed typologies for social structures. 2) As a corollary to this first point, there is also an implied assumption of a unidimensional continuum, i.e., a continuum.in which all parts are so interconnected that if change takes place at one point, all others follow suit. L 23s”, for example, Rural Sociolo , 18 (June, 1953), "The So- ciological Significance of the Rural—Urban Fringe," pp. 101-120. Charles E. Lively, ”Introduction," p. 101; Stuart A. Queen.and David B. Carpenter, ”From.the Urban Point of View," pp. 102-108: and Walter C. HoKain, Jr. and Robert G. Burnight, "From the Rural Point of View," pp. lOS-ll#. Lively presents the following hypothesis on p. 101: "The rural- urban fringe represents a position on a continuum within a single rural- urban distribution.” Queen and Carpenter state on p. 103: "The rural-urban fringe would, were it consistent with the gradient pattern, display character- istics intermediate between those of the center and the periphery; . ." McKain and Burnight state on p. 110: "The limited fringe and the extended fringe thus lie along the rural-urban continuum, the lim- ited fringe being closer to the urban end of the continuum." - 9 - 3) Rural and urban constructs, when defined, differ according to the individual or the needs of the study. #) This type of definition indicates what the fringe is‘ggp, rather than what the fringe i3. 5) Exhibited empirical limitations of the continuum hypotheses of continuous gradation and consistent variation.2u This rejection of the continuum approach means that another type of definition is necessary. In an effort to come to an understanding of the nature of the fringe a listing of its characteristics will be pre- sented: this listing will serve as a structural definition of the area.25 The most obvious characteristic of the fringe area is its loca- tion, usually beyond the limits of the legal city. In most cases the fringe is beyond the limits of what the United States Bureau of the Cen- sus refers to as the "urbanized area."26 The fringe is located in the area generally referred to as the "agricultural hinterland" or, in Cen- sus terms, in the rural area. A unique characteristic of the fringe is the presence of both rural and urban land use: much of the area is in 2“Otis Dudley Duncan, "Community Size and the Rural-Urban Con- tinuum," Hatt and Reiss, 9p. git" pp. 55-45. 25For a more extensive treatment of the characteristics of the fringe area see: Richard A. Kurtz and Joanne B. Richer, "Fringe and Suburb: A Confusion of Concepts," Social Forces, 57 (October, 1958), PM 32-37- 26The United States Bureau of the Census defines the urbanized area in the following manner: "Each urbanized area contains at least one city with 50,000 inhabitants or more in l9#0 or according to a spe- cial census between l9h0 and 1950. Each urbanized area also includes the surrounding closely settled incorporated places and unincorporated areas. . . . The boundaries of these . . . areas were established to confbrm.as nearly as possible to the actual boundaries of thickly set- tled territory, usually characterized by a closely spaced street pat- tern." United States Census 2; Housi , 1250, California, Bulletin H‘AS, pe VI. I.- e y I I. 0' ‘ l I I. ,4.— _ I m e- M. “‘0 Os: he! 0“ 0 Int. .. u u... I'lle. .. ' V ‘M u 0' Id.‘ : I . .Ie'. ‘ D One-.4, : m .; i“ l‘!‘ - 1o - farmland but a large number of nonfarm dwellings are interspersed among the farms. Since the area is unincorporated, governmental responsibil- ity is in the hands of township political officials (in the Midwest) and is usually under the jurisdiction of ordinances which were enacted to meet the problems of a predominantly rural area. This situation has fostered relatively lax zoning regulations which have permitted the ran- domly spaced construction of such business establishments as gas sta- tions, drive-in restaurants, and.outdoor movies along major traffic ar- teries. Municipal services such as street lights, sewage systems, and paved roads are usually absent. In addition, the fringe is usually an area of rapid population growth due to migration: it exhibits the po- tentiality for future growth and increasing density ratios since there is still much open land available for expansion. _ Thus the physical features of the fringe mark it as an area 13; cated in the agricultural hinterland of the city in which there is a lack of continuity of land use and no regular pattern of farm.and none farm residence can be detected. The distinguishing physical features in- dicate that the area is "in transition" since the pattern of land use seems to be undergoing a change from strictly rural to a mixture of rural and urban. The social features of the fringe area point to a mixture of rural and urban ways of life, stemming primarily from the fact that people of rural and urban occupations and backgrounds reside in close proximity. Thus the area is one in which rural and urban orientations are intermixed. The key variable in both the physical and the social character- istics of the area is a mixture of rural and urban. Perhaps the most - 11 - important elements can be summarized under the terms "land use" on the physical level and "orientations" on the social level since, basically, the fringe area is one in which the use of land is mixed and the orien- tations of individuals are mixed. This is a consequence of the parti- cular type of residential propinquity which is unique to the area. Social Life in the Fringe: Specific Propositions It is now possible to return to the discussion of social life in the fringe area. The previous discussion, which presented a brief over- view of social life in the fringe, was designed as an orientation to the problem. A more precise review of research on the fringe, emphasizing the specific propositions of previous research, will now be presented. As a guide to this review the concept of the community will be utilized since this concept will focus attention on the nature of social life in the area. Several statements dealing with the nature of the community were examined, in an attempt to clarify the important variables. It was found that the emphasis of authors varied. Hawley, Reiss, and Sanders, for example, tend to emphasize interaction,27iMacIver and Page refer to 28 "community sentiment," and Spykman speaks of "community loyalty."29 27Amos Hawley, Human Ecology: ‘A Theory‘gg Community Structure (New York: The Ronald Press 00., 1950); Albert J. Reiss, Jr., "The Come munity and the Corporate Area," University 2; Pennsylvania‘ggy Review, 105 (February, 1957), pp. #45-k65; and Irwin T. Sanders, The Community: Agilntroduction 32.5,Socia1 System (New York: The Ronald Press 00., 1958 . 28Robert M. Fischer and Charles R. Page, Societ : in Introduc- tog Analysis (New York: Rinehart and Co., Inc., 19139). 29Nicholas J. Spykman, "A Social Philosophy of the City," 1h; bean Community, ed. Ernest W} Burgess (Chicago: The University of Chicago Press, 1926), pp. 55-61.» - 12 - However, the literature has been conveniently summarized by Hillery, who made a significant contribution by analyzing ninety-fbur definitions of the community in an attempt to find convergence.)o Hillery concluded that sociologists agreed on only three elements: physical space, social interaction, and emotional bonds. An.examination of various studies of social life in the fringe indicates that the latter two, i.e., the strictly social elements, are not present in the area. More specifically, researchers have indicated that social interaction and emotional bonds are nonexistent in the fringe area. This viewpoint is expressed in general terms by Firey, who observed the social life of three subdivisions, and concluded that the fringe is "a kind of institutional desert."31 This conclusion, somewhat evaluative, is reiterated by Kimball: ”The traditional rural community in both its economic and social characteristics showed evidences of bee ing shattered"52 while newcomers are not being accepted in their new area of residence.53 Similarly, Beagle and Schroeder conclude that ". . . those who live on the fringes of the community are also on the fringe socially."5u These general comments on social organization are derived from 50George A. Hillery, Jr., "Definitions of Community: Areas of Agreement,” Rural Sociolog, 20 (June, 1955), pp. 111-123. ”Walter Firey, Social Aspects _t_o_ Land Egg PM _i_._x_l_ 31;; Coun- t -Ci Fringe: ‘ghg Case 2; Flint, Michi , A.E.S Special Bulletin 539 Michigan State College, June, 19 , p. 58. 32$olon r. Kimball, $33 is}; Social Frontier: The Egg)? A.E.S. Special Bulletin 360 (Michigan State College, June, 19%), p. . ”Ibid. , p. 6. 3“J. Allan Beegle and Vidick Schroeder, Social Or ization.ip {323,North Lansigg Frggge, A.E.S. Technical Bulletin 251 (Michigan State University, September, 1955), p. 26. v fi 1.. 01...... '..o- -s I... ...-CE: .0. (.1 In be. .r‘:'l . 'Oee' : I ~ ‘3' Vin. ‘ :“u . Qi'.'.‘ '0‘. .. l l ell" . .\ .m - 15 - the specific findings of those investigators who dealt with social in- teraction and identification patterns in areas which they designated as fringe. However, a word of caution is necessary: if the definition of the fringe area presented in this chapter is accepted, the studies by Firey and by Beagle and Schroeder were conducted in areas which cannot be considered fringe. Nevertheless, these studies are accepted by many sociologists as studies in the fringe area, even if they do not meet the criteria of definition presented above. Therefore, it has been from these studies that an understanding of fringe area social life has been derived}5 Specifically, Firey describes the fringe as an area which pos- sesses a very limited degree of social life; he presents the proposition that the residents ". . . lack stable groupings of their own."36 This specific proposition is in agreement with Kimball's finding that "in the fringe there is no proof that established groups are uniformly extend- ing the bond of welcome or integrating the newcomers into community par- ticipation. . . ."37 Beagle and Schroeder are in agreement with this general proposition: "the data . . . suggest that this fringe area is exceedingly loosely organized. . . ."38 This conclusion is drawn by Beegle and Schroeder after examining data which indicate that only a small proportion of the residents are active participants in voluntary community associations, that few of the residents reported shopping in 35 Fbr an extensive discussion of the confusion of the fringe area with other residential patterns see Kurtz and Eicher, pp. git. 36Pirey, 22s 93-20, Pa 58. 37Mb311, .92. £1.30, pe 6e 38Beegle and Schroeder,'gp.‘gi§., p. 28. -14- the area, and that many residents indicated that they have no friends residing in the neighborhood. In addition, Noel P. Gist studied social participation in a fringe area and indicated that the tendency is for ". . . city families to move to the open country while retaining their occupational and other connections with the urban community."39 In summary, there is evidence of consensus on the part of re- search men who have concentrated on social relationships in the fringe area. The major proposition which is reiterated by each investigator in his unique way is that‘phg fringe area lacks thg patterns‘gf social 29;: gygpgd community. Based upon the specific propositions discussed above, the conclusion.must be reached that the fringe area fails to meet the social interaction criterion mentioned by Hillery. Emotional bonds, which will be interpreted as "identifications" in the present context, is the second social criterion of a community mentioned by Hillery. As was the case with social participation, fringe investigators present specific statements on the nature of the fringe area identification patterns. Of course, Firey's characterization of the fringe as "a kind of institutional desert" has implications on this level also but, more specifically, he indicates that neighborhood bonds are weak in the area.40 Kimball is more specific in the following pro- position: ". . . for many who remove to the new locality the change is one of space and not of social identification."#1 This statement is certainly to the point.- Beegle and Schroeder concentrated on social 39Noel P. Gist, "Ecological Decentralization and Rural—Urban Relationships," Rural Sociology, 17 (December, 1952), p. 528. “Garey, 22o 93$.” pm 38. #J'Kimball, 22s Ego, pe 29o - 15 - participation in the fringe area but they also mention the lack of pat- terns of identification or emotional bonds: ". . . this fringe area. . . possesses little unity and solidarity.”2 Gist's conclusion is similar; he finds that fringe dwellers are closely tied to areas outside of the 45 area of residence. In summary, the conclusions are once again consistent. The pro- positions presented by researchers indicate that th_e, £13952 _a_l;§_a_ 132k; existence‘g§,§ community. Based upon the specific statements discussed above, the conclusion must be drawn that the fringe area fails to meet another of the essential criteria of an actual community. Thus, according to the propositions derived from empirical in- vestigations, the fringe possesses neither the social interaction nor the identification patterns which characterize a community. In the true sense of the term, fringe dwellers are characterized as ig,but not 9; the residential area. The determination of where the fringe dweller does participate and identify is secondary to the finding that it is not in the area of residence. However, the question is still an important one, and it is one that is easily answered: according to the findings of the investigators, the fringe dweller tends to participate in, and identify with, the central city to which the fringe area is tied by economic and social dominance. #ZBeegle and Schroeder, pp.‘gip., p. 28. #BGiSt, 22s Cite, pe Bali‘s ##This is a very confusing finding in view of the tendency to think of this person as not identified with his community when he lived in the city. It is possible, of course, that identification may be by hindsight; this is a question which calls for further research. ._L_ l" 9,. n. b“ -16.. It is evident that descriptions of social life in the fringe area are quite similar to the stereotyped picture of urban life des- cribed above. In both cases a lack of integration and identification has been emphasized.“5 In view of this similarity of conclusion, the possibility exists that the description of fringe social life may also be the product of the same pro-rural bias that seems to have influenced the urban stereotype. Thus, the assumption that rural life is charac- terized by a high degree of community integration and identification may have influenced biased investigators to accept the implication that this is an exclusive quality of the rural community. Statement of the Problem In view of the points made in this discussion it is felt that an empirical study, specifically addressed to a determination of the na- ture of social life in the fringe, is needed. Since previous investiga- tors have concentrated on the social relationships and identifications of fringe area residents, these are the dimensions that need special at- tention. 'With this rationale, the present study has been undertaken. Objectives The object of this study is to investigate the adjustment pat- terns of residents in the rural-urban fringe. The problem will be in- vestigated through an assessment of: l) the specific subjective rela- tionship of the fringe dweller to the area of his residence, and 2) the specific objective role behavior of the fringe dweller in his area of #SThere is, however, a very important difference: according to the urban theorists, the nature of urban social life is of a different quality than rural life; the fringe investigators, on the other hand, indicate a complete lack of social life. - 17 - residence. The subjective and objective dimensions will be analyzed in terms of their relationship with each other, and with an emphasis on their influence on community interactions. In addition to the determin- ation of the subjective and objective dimensions of adjustment and an analysis of their relationships, certain correlates of these groupings will be investigated. Thus the study will utilize two concepts to refer to the basic components of the individual adjustment pattern. The first dimension is "subjective" and the second is "objective"; these will be referred to as "subjective identification" and "objective integration." Subjective identification "Subjective identification" is a summary concept which will be used to identify the basic orientation of the individual to the area of his residence. Perhaps the use of the term "subjective" along with "identification" is superfluous since the basic element within the cone cept is the individual's attitudinal evaluation. Thus the "subjectively identified" fringe resident is one who will make positive statements about his area of residence and who will reject the notion of residen- tial superiority of other types of areas, viz: the city and the outly- ing farm area. The person who is subjectively identified will indicate his preference for the area in which he is living when presented with certain alternatives. These alternatives are presented in the form of questions which are designed to test the individual's degree of positive choice for his area of residence, for the rural area, and for the city. Subjective identification will be considered as an overall characteris- tic which is the product of very specific components which can be de- termined by an examination of answers to questions which are designed -18.. specifically to evaluate this general dimension. Objective integyation "Objective integration" is on quite a different level than sub- jective identification. Whereas the latter is concerned with opinions and attitudes, and is determined by an analysis of the individual's re- sponses to a battery of questions designed to test his degree of posi- tive choice for the fringe area relative to other possible residential areas, the former focuses attention on the specific activities of the individual within his area of residence. Thus the differentiation be- tween the two dimensions is well expressed in the accompanying adjec- tives—-the latter is "subjective" whereas the former is "objective." Objective integration, like subjective identification, will be considp ered as an overall characteristic which is the product of very specific components. These components are on a behavioral level; they were chos- en in an attempt to determine the degree to which the individual parti- cipates in the area of his residence. \ Ehg relationship between subjective identification gag objective integgation In addition to a determination of the degree to which identifi- cation and integration exist, an analysis of their patterns will be un- dertaken. The next step will be an analysis of the relationship between the subjective and objective categories and an attempt to explain the resulting patterns. Finally, the dimensions will be handled independent- 1Yin an attempt to determine the factors that influence each. The relationship between the subjective and objective dimensions (”no be clarified by a presentation of their cross-classifications in chart form. Chart 1 is designed to indicate the nature of the relation- - 19 - CHART 1 THE INTERRELATIONSHIP BETWEEN IDENTIFICATION AND INTEGRATION, BOTH DICHOTOMIZED INTO POSITIVE AND NEGATIVE CATEGORIES Subjective Identification Identified Not identi- with the fied with fringe the fringe +) (-) Integrated A B in the + + + - Objective fringe (+) Integrati°n Not inte- c D grated in - + - - the fringe (-) Explanation of cells: Cell A: The individual who is both integrated and identified Cell B: The individual who is integrated but is not identified Cell C: The individual who is not integrated but is identified Cell D: The individual who is neither integrated nor identified Cell A illustrates the case of the individual whose consistency is such that he is fully a member of his residential community in both pat- terns of behavior and evaluations. Cells B and 0 illustrate discrepant adjustment patterns since they exp hibit inconsistencies between integration and identification. Cell D illustrates the case of the individual whose consistency is such that his behavior is carried on in an area other than his residential community and whose evaluations are such that they cannot be consid— ered positive toward the fringe area. .. 20 - ship by dichotomizing each dimension into positive (+) and negative (-) categories. In the chart, Cell A illustrates a strong positive area ad- justment pattern. To the degree that residents exhibit such an adjust- ment pattern the fringe area can be considered to possess communities in all respects. Cell D illustrates the adjustment pattern indicated by Pirey, Kimball, Beegle and Schroeder, and Gist. Cells B and C illustrate discrepant adjustment patterns since they exhibit an inconsistency be- tween positive identification or integration with the fringe area and positive patterns in either the rural or the urban area. The possibility exists, of course, that in Cells B, C, and D, a positive identification and/or integration may be entirely lacking. If such an adjustment pattern actually exists it would be a completely unique phenomenon which raises very interesting questions that sociolo- gists have previously attempted to answer in the context of the social "isolate." An analysis of this particular type of adjustment pattern is beyond the scope of the present study. 1'93 relationship between adjustment patterns sad. £93.13}. use. The point of concentration that is crucial in this study is the determination of specific adjustment patterns. Once this is achieved, the analysis will shift to an attempt to explain the adjustment cate- 80ries. In addition, a series of generalizations will be presented in an attanpt to indicate the influences that such patterns have on social life in the fringe area. If the fringe area is in fact an "institution- a1 desert,” this phenomenon will actually be a reflection of the subjec- tllVe and objective categories. The point here is that the fringe area Will be analyzed in terms of the identification and integration patterns - 21 .. of the residents. An analysis of these patterns will indicate the na- ture of fringe area social life since the residents' patterns of adjust- ment influences their behavior in the area. The patterns peg; §_e_ may also suggest that adequate social life exists in the fringe area. Adjustment patterns and social variables: some perspectives In addition to analyzing the nature of social life in the fringe area the attempt to account for the specific adjustment patterns will be of first importance. Therefore, it will be necessary to examine the so- cial variables that are related to the types of patterns found. Among the perspectives which will be used in addressing this problem are resi- dential experience and demographic factors (e.g., occupation, tenure status, age, education, and income). Methodolog and Techniques Research designs are frequently influenced by practical necessi- ties which are not related specifically to the problem under investiga- tion.“6 This point is relevant to the present study since the data were collected in conjunction with a second study, conducted in the same spa- tial area, but dealing with a different problem!+7 This latter study dealt with fringe area farmers only. Methodological decisions were made in such a manner that neither study was affected adversely by the other. The Michigan Agricultural Experiment Station provided the finan- %For an extended discussion of this point, see Marie Jahoda, Morton Deutsch, and Stuart A. Cook, Research Methods .i_n_ Social Relations, Part One: Basic Processes (New York: The Dryden Press, 19515, Appen- (113: A, pp. 342-5610 A7The results of this study are reported in Joel Smith and Richard A Kurtz, Pripge Area Farmers' Use _o_i_‘_ Market News Information (tentative title; unpublished manuscrip-a. mo. and, 01.. to... u... . 0'. I - 22 .. cial support necessary for the collection of information from the resi- dents of the rural-urban fringe. The field schedule administered to farmers was constructed to handle both research problems as separate units, although the data for both studies were collected at the same time. The nonfarmer interview schedule, administered at a later date, consisted only of that section of the farmer schedule which dealt with adjustment patterns. Sampling Farmers Since the related study concerned one aspect of the behavior of farmers, it was necessary to adopt sampling techniques that would assure a sample representative of farmers in the area, without introducing bias into the nonfarmer sample. Therefore, although some of the initial sam- pling decisions were designed to solve problems specific to the sample of farmers, the procedures selected did not endanger the randomness of any component of the sample. The first step was to define the universe of study in precise terms. On a general level, the universe was defined as "all the farm residences in the eight-township area surrounding the City of Lansing. " However, it was necessary to demarcate the universe more precisely since only farm residences located in the rural-urban fringe were within the Boom of the study. In the redefined universe, areas which contained an OVerwhelming proportion of nonfarm residences were eliminated from the ullilverse of study since these areas are actually homogeneous "pockets" °t villages and suburbs, rather than fringe.“8 \ “A justification for this procedure is presented in Kurtz and Eicher, 22c fie - 23 - General Highway Maps, published for each Michigan County by the State Highway Department in 1951, were utilized to obtaininfomation about the residential composition of the eight townships in the study area. These maps listed all roads and residences in each county; a sys- tem of symbols was used to identify farm and. nonfarm dwelling units. An examination of the maps depicting the eight-township area indicated the necessity of eliminating some sections from the rural-urban fringe study area since they did not meet the criteria of definition; this reduced the number of universe sections” in the eight-township area from 288 to 255. In several other cases, small parts of the sections were eliminated from the universe of study for the same reason. When the universe of study was defined precisely, the next step was to determine how closely the defined universe corresponded to the fringe area definition. It had already been established, before the sampling process was initiated, that the area met the growth and densi- ty, governmental, and locational criteria. The land use and occupation- 91 criteria remained to be applied. After the universe was redefined, tabulations indicated that the number of farms in the universe was 1569 (11.8.1 percent) compared to 1692 nonfarms (51.9 percent); this calculation fiuggelted that the universe of study fits the land use and occupational criteria. A random stratified technique was used to draw the sample of farm residences. First, each of the 255 sections was classified accord- ing to the number of farms it contained. On the basis of an examination 01‘ this classification, a decision was made to stratify the sections into \ 49The word ”section" is used here to refer to the Midwestern one- I"innate-mile area. Each township is made up of thirty-six sections. .. 24 .. three groups; those with less than five farms, those with five to seven farms, and those with more than seven farms. The universe proportion of all sections in the area falling in each of these three categories was obtained from the general Highway Map: and was used as the criterion for the proportionate size of each of the stratified areas in the sample. Within each of the stratified categories, the specific sample sections were drawn randomly. Instead of planning to interview at every fam residence within the sample sections, it was decided to draw a sample of sections twice as large as was needed and. to interview at every second farm residence. Thus, it was possible to draw a sample of sections with a wide areal scope which maximized the chances that all six tiers of sections in the area would be represented. Practical and. statistical considerations indicated that a. sample of approximately 200 fans residences, 12.7 percent of those in the uni- verse, would be adequate for the purposes of the study. The universe was prepared for sampling by filling out a card for each of the 255 sections. These cards were sorted into three categories according to the number of farms in the section, and were numbered consecutively. Sixty-five cards were selected at random by choosing those cards which had numbers corresponding to the last three digits of a random page of 8 book of random numbers. This procedure produced a sample size of 20h.5 farm residences, located in 65 sections. Sample representativeness was evaluated statistically by means 01‘ a chi-square goodness-of-fit test. The results of this test are re- DOrted in Table l and indicate that the sample may be accepted as rep- Ii'eaentative of the universe. -25- TABLE 1 SAMPLE-UNIVERSE GOODNESS-OFBFIT TEST Stratified Sample Universe category frequency frequency More than 7 farms 18.5 21.5 5 - 7 farms 97.5 96.1 Less than 5 farms 88.5 86.9 Total 209.5 209.5 X2 3 el‘l'68 ~ e90 > p > e80 Degrees of freedom: 2 Before selecting every second farm residence in the sample sec- tions, it was necessary to make revisions in the General Highway Map classification of farm and nonfarm dwellings because: 1) mic-classifi- cations were possible in an area containing a mixture of farm and non- farm dwelling units; and 2) the study site, like most fringe areas, was undergoing a rapid change in residence composition. It was possible that dwelling units which had been correctly classified as farms in 1951 were no longer farms in 1954:, and that many new nonfarm dwellings had been constructed in the study area during the three-year period follow- ing publication of the maps. The maps were checked by observation in the field. Two observ- ers drove through each of the sixty-five sections in the sample and made Changes suggested by their observations. when a problem of classifica- tion arose, the resident of the dwelling unit was questioned. These a»<1.1ustments reduced the final sample total from an expected 20’+.5 farm units to an expected 156. After this map revision was completed, every -26- second farm dwelling in each section was chosen as a sample farm. Nonfarmers With two exceptions, the nonfarm sample was derived in the same manner as the farm sample: 1) no attempt was made to stratify the non- farm group, and 2) the nonfarm sampling ratio was 1:1: instead of 1:2. Thus, while the same sections were used in both cases, farm dwellings were chosen randomly be selecting every second unit while nonfarm dwell- ings were chosen randomly by selecting every fourth unit. Special problems While original plans had called for sampling farm and nonfarm dwelling units at the same rates, a modification in the nonfarm sampling ratio became a practical necessity because of the rapid increase in the number of nonfarm dwellings in this fringe area. Whereas the 1951 Gen- eral Highway Map indicated that farm and nonfalm dwelling units occurred in approximately equal proportions, the corrected map indicated that the proportion of nonfarm dwellings was now sixty-four percent. Under these conditions, if the 1:2 ratio had been used for the nonfarm sample, the time and money resources would have been exhausted before all sample members could have been interviewed. The changed residence composition of the fringe area reduced the expected size of the farm sample and increased the expected size of the nonfarm sample. After the map was completely revised, the sample sec- tions contained a total of 276 farms and 492 nonfarms. Sampling at a 182 rate produced a final sample of 138 farm units, while sampling at a his rate produced a final sample of 123 nonfarm dwellings. Since farmers and nonfamers were sampled at different rates, - 27 - the collected data had to be adjusted before statistical analyses could be undertaken. A simple comparison of farmer and nonfarmer responses to any given question was possible only if the comparison involved percent- age distributions exclusively. However, any statistical test of associ- ation or significance would be affected by the difference in sampling rates. The data could have been adjusted by either of two methods: 1) inflating the nonfarmer data to equal the farmer sampling rate, or 2) deflating the farmer data to equal the nonfarmer sampling rate. Since the farmer sampling rate was double the nonfarmer ratio, the first alter- native would necessitate a doubling of all nonfarmer data whereas the second alternative would necessitate a reduction of all farmer data by one-half. The second alternative is the more cautious procedure since an inflation of the sample would imply smaller sampling errors than those known to be the case while a deflation of the sample would have the re- verse effect. Therefore the second alternative, i.e., the deflation of the farmer data, was applied to the data on which all of the tables pre- sented in the following chapters are based. The Schedule Schedule construction followed the usual procedure of developing a set of questions which could provide the information suggested as ne- cessary by the theoretical framework. Questions were organized around the following variables: subjective identification, objective integra- tion, and population characteristics. Access to schedules which had been constructed by other investigators helped immensely in developing - 28 - 5O appropriate and workable questions. A pre-test of the field schedule, carried out in sections of the fringe area which had not been selected as sample sections, proved to be an invaluable indicator of the workability of the schedule. Modifica- 51 tions and rearrangements of some questions were adopted as a result. Interviewing Intensive interviewing was conducted throughout 195# and 1955; "clean-up" interviewing took place in 1956.52 Since the farmer schedule was constructed to provide data for two separate research problems, farm- er interviews took about twice as long to complete as nonfarmer inter- views. The average farmer interview lasted more than two hours whereas the average time of a nonfarmer interview was slightly over one hour. Males were chosen as respondents almost exclusively. The rate of interview completion was extremely high. Only one farmer refused to participate in the adjustment study, while two nonfarms era could not be located. Thus, the interview completion rate was over ninetybnine percent in the case of farmers, and over ninety-eight per- cent in the case of nonfarmers. The final sample size figures are 136 in the case of farmers and 121 in the case of nonfarmers. 50These questions had been developed by William H. Form, Joel Smith, and Gregory P. Stone for their analyses of the urban community. 51The final field schedule used in the interviewing is repro- duced in Appendix A, infra. 52The interviewers were the author (approximatel half of the farm interviews and almost all of the nonfarm.interviews ; Siegfried Mallenkott, a hired interviewer (approximately half of the farm inter- views); and Frank Sim, William D'Antonio, and Frank Nell, graduate stu- dents (about ten nonfarm interviews during the "cleanpup" phase). CHAPTER II A DESCRIPTION OF THE AREA This chapter will be devoted to a description of the Lansing fringe area and its social life. Such factors as prior residential exr perience of fringe sample members, characteristics of the population, and respondents' definition of the area, evaluations of the area, and social participation will receive attention. An understanding of objec- tive and subjective aspects of the social life in the Lansing fringe can be gained by examining each of these variables within this context. However, before presenting data on the Lansing fringe, a descrip- tion of the central city will be presented.1 Since the rural-urban fringe area surrounds the City of Lansing, and since it is dependent up- on the central city economically (and possibly socially), an understand- ing of the nature of the urban area seems necessary. In its most direct form, characteristics of the Lansing fringe area may be a function of its location near Lansing. Characteristics of the Central City Since Lansing is typical of the medium-sized industrial cities 1The description of Lansing has been derived from the following sources: U. S. Bureau of the Census, U. S. Census‘pf Po ation: 1250, Vbl. II, Characteristics pf ppg Popplation, Part 22, Michigan Chapter B, P-BZZ (Washington: U. S. Government Printing Office, 1952); J. Allan Beegle and Widick Schroeder, Social Orgppization ipippg,florth Lansipg Pripge, A.E.S. Technical Bulletin 251 Michigan State university, Septem- ber, 1955); and Joel Smith, William H. Form, and Gregory P. Stone, "Local Intimacy in a Middle-sized City," American Journal 9; Sociolo , Lx (November, 1954), pp. 276-28'+. - 29 - - 30 - in the North Central States,2 and since it is located conveniently to Michigan State University, it was decided to investigate the problem in the rural-urban fringe area surrounding this central city. In 1950 the population of Lansing was over 90,000 and its urbanized area housed more than #0,000 people; one-half of the urbanized area population reside in East Lansing. The city is located in south-central Michigan, on the north-western corner of Ingham County, a standard metropolitan area of more than 170,000 inhabitants. These population figures represent a 1940-1950 increase of seventeen percent for the city and thirty-two per- cent for the county. More than forty percent of the employed.males were working in manufacturing industries in 19505(M+.2%; see Table 2); of these, three— quarters were engaged in the manufacture of motor vehicles and motor ve- hicle equipment (75.5%). The Oldsmobile Division of General Motors, Fisher Body Company, and Rec Motors are all located within the city. Of all employed males, over one-half were classified by the Census Bureau as being foremen and in skilled, semi-skilled and unskilled occupations (55.3%; see Table 3). Beegle and Schroeder found that the age distribution of heads of households, reported by a sample of more than 500 Lansing workers, was a fairly young one. Over twenty percent of the sample were found to be under 50 years of age and approximately the same proportion reported 2Beegle and Schroeder, pp.‘p;p., p. 9, make the following obser- vation: "With the exception of Michigan State University to the east, Lansing possesses characteristics similar to numerous industrial centers of comparable size in the North Central States. SAll tables on which this chapter is based will be found in Appendix B. - 31 - that their ages were between 50 and 39 (21.2% and 25.Q% respectively; see Table #). The median age is slightly more than thirty years. The same sample, including all male workers, indicated that ap- proximately one-third have not attended high school (32.2%; see Table 5) and over forty percent reported that they attained a high school diploma or went beyond (#5.fi%). The median education attainment for the Lansing sample lies within the "some high school" category. Census data indicate that the l9#9 median income of Lansing fam- ilies was 8#O97. One out of ten families earned less than 82000 in that year (10.7%, see Table 6), and approximately the same proportion earned 87000 or more (12.3%). An examination of the income distribution indi- cates that the modal family income is in the 83000-83999 category. How- ever, the distribution of family incomes by 81000 categories is a fairly even one, with no category represented by more than twenty-five percent of the population (the modal income category accounts for only 24.B% of the total distribution). Information about the birthplace of workers for the Lansing sam- ple is available in the Beegle and Schroeder study. Slightly less than two-thirds of the workers reported that they had been born in Michigan (63.65; see Table 7); of this group one-third were born in Lansing. Workers in the Lansing sample were born in various sections of the coun- try but three-quarters were born in Michigan or in an adjacent state (76.2%) In addition to the demographic information provided by Beegle and Schroeder and the Census Bureau, information pertaining to social life in Lansing has been collected by Smith, Form, and Stone. In the context of the present fringe study, perhaps the most significant finding - 52 - of these investigators was that intimate social relationships which con- tribute to social integration ". . . are prevalent in the city and are found both throughout the city and within local areas of residence."1+ Thus the Lansing rural-urban fringe area surrounds a city which does not fit the stereotyped picture of urban social life.5 Characteristics of the Lansing Fringe Area 2319; residential egperience As is generally true of most fringe areas, the large influx into the Lansing fringe is a recent phenomenon. More than one-half of the respondents (57.2%: see Table 8) moved to the area after the second world war, while the depression years and war years were each periods during which less than ten percent of the present residents arrived (9.3% and 8.5% respectively). Approximately one-quarter of the respondents (2#.9%) have been living in the area since before 1930. The vast major— ity of present residents are non-natives, only twenty percent of the re- spondents having been born in the area. Three-quarters of the fringe residents have had only one resi- dence since living in the area. (See Table 9.) Among those who have moved within the area, most waited more than five years before making such a move. Residential stability within the area is by no means sur- prising, since most of the residents are recent arrivals. More than half of the Lansing fringe population are recent in-migrants who own their own homes and have lived in them for only a short time. The hoary representation of fringe dwellers who have experienced #Smith, Form, and Stone, pp. p;1., p. 283. 5For a discussion of this stereotype see suppa, pp. 2ff. - 33 - more than a five year difference between the tune they moved to the area and the time they moved to their present address is also influenced by length of residence since one-quarter of the residents were in the area before 1930. These "oldtimers" have had more than a twenty year period during which residence could be changed. An indication of the degree of residential stability of Lansing fringe residents becomes apparent by the fact that three-quarters of the respondents had lived in no more than three different homes since 19h0, with these being approximately equally distributed among the three alter- natives. (See Table 10.) Respondents have not moved very often since 19b0, nor have they moved often after locating in the Lansing fringe area. The movement to the Lansing fringe is primarily a local migra- tion (i.e., centrifugal). More than half of the residents moved to the area from Lansing or East Lansing (57.6%; see Table 11) and over twenty percent moved to their present location from one of the three study coun- ties (22%). In total, eighty-five percent moved to the fringe from ar- eas not more than one county distant. The movement to the fringe area also represents a process of urban decentralization. More than eighty percent of the respondents in the sample reported that they moved to the Lansing fringe from cities (82.2%; see Table 12). Of these, seventy percent migrated from Lansing and East Lansing. No evidence of movement from fringe area to fringe area was found. The data do not lend them- selves tc an analysis of whether a suburb-to-fringe migration is taking place since residents who moved from suburbs undoubtedly classified their previous residence as "urban." The fringe area is populated primarily by residents who were raised in close proximity to the area. (See Table 13.) One out of five -31.- residents of the Lansing fringe were raised in the immediate area, anoth- er one-quarter were raised in Lansing or East Lansing, and approximately fifteen percent were raised in the three study counties (l#.5%). Slight- ly under an additional ten percent were raised not more than one county removed from the study counties (7.2%). In total, two of every three residents of the Lansing fringe were raised right in, or in close prox- imity to the area. Thus, the movement to the Lansing fringe area seems to be a relocation of people who were raised in the general area. In addition to the name of the place in which they were raised, respondents were also asked whether the area in which they were raised was rural or urban in character. Approximately equal groups indicated that they had been raised in each type of area (h6.q£ rural and 53.I$ urban; see Table l#). When viewed in the context of the fact that urban locations were so frequently reported as the place of previous residence it may be concluded that the movement to the Lansing fringe area is often a movement from rural to urban to fringe area. §Egg§£xz residential histogz characteristics The goal of this section was to describe the residential exper- ience of the Lansing fringe residents. Residential history characteris- tics may be summarized in seven points: 1. The movement to the Lansing rural-urban fringe area is large- ly a post-second world war phenomenon. More than half of the residents moved to the area after 1945. 2. Movement to the fringe area does not usually bring about subsequent moving experience. More than three-quarters of the residents have maintained only one home in the area. 3. The movement to the Lansing fringe is essentially a local - 55 - migration. Over half of the residents moved to the area from Lansing or East Lansing and eighty-five percent moved from not more than one county distant. 4. The movement to the Lansing fringe is essentially a movement from the city outward. More than eighty percent of the residents moved to the area from cities. 5. Residents of the Lansing fringe area have not experienced wide-scale moving since 1940. Three-quarters have lived in less than four homes since this date. 6. The Lansing fringe area is composed of residents who are na- tive to the general area. Two-thirds of the residents were raised with- in the study counties (including Lansing and East Lansing) and counties contiguous to these. 7. Lansing fringe residents are neither predominantly rural nor predominantly urban in background. Approximately one-half were raised in each type of area. Demoggaphic characteristics Occupational heterogeneity, which is characteristic of fringe areas in general,6 is evident in the Lansing fringe. (See Table 15.) Almost one-quarter of the Lansing fringe residents are full-time farmers (23.2%), more than a third are employed in factory work (36.0%), twelve percent combine farm work with another occupation, and the remaining one-quarter of the population are engaged in a diversity of typically urban occupations. Almost two-thirds of the Lansing fringe residents work in the central city (62.2%: see Table 16). This is an expected 6See su rs, pp. 8ff. - 55 - pattern which is a corollary of the occupational structure of the area. The daily commuting of a majority of residents to the city creates a situation which undoubtedly has consequences for the social life of the fringe area since a large part of each week day is necessarily spent in Lansing. The tier of residence of respondents indicates the distance fringe residents live from the central city. Only five percent of fringe dwellers occupy homes in the first tier of sections surrounding the out- er limit of Lansing (see Table 17) while almost one-third live in the tier of sections furthest from the city (30.9%). It is, however, neces- sary to take account of the fact that the closest tier is composed of twenty-eight sections while the furthest tier contains sixty--eight.7 Therefore, the mean number of respondents in each tier was computed; these figures are presented in the last column of Table 17. No pattern of section density can be detected; sections three tiers from Lansing have the highest density figure and sections in the tier closest to Lansing have the lowest. Almost ninety percent of the Lansing fringe residents own their home (88.11%; see Table 18). This is an extremely high proportion of home owners, but it is to be expected in an area of this nature. The low number of renters in the Lansing fringe area is a result of the fact that many of the "oldtimers" are members of families that have been in the area for several generations and most of the newcomers are buying homes that have been constructed recently. Growth in the fringe area is \ 7The term "section" refers to the Midwestern one-square-mile a-I‘ea. Each township is composed of thirty-six sections. - 37 - taking place primarily through the building of many single family dwell- ing units that are intended for sale. Approximately two-thirds of the respondents are between thirty and forty-nine years of age (65.1%; see Table 19). Looking at the ex- tremes of age, less than ten percent are under thirty years old (8.5%) and over fifteen percent are sixty years of age and over (15.8%). The newcomers contribute heavily to the young age groups while the oldtimers contribute to the older age categories. The young age structure of the area is striking, with a median age of forty-four years. A fairly young age structure is typical of an in-migration area since migrants are usu- ally in the young age categories.8 With a young age structure it is to be expected that respondents would exhibit a fairly high education distribution. The data confirm the expectation: less than ten percent did not graduate from grade school (7.7%), more than forty percent either graduated from high school or went beyond (411.8%), and more than one out of ten attended college (15.0%). (See Table 20.) The median education figure is over three Years of high school. The age structure also leads to the expectation that the number 01‘ children would be high. That this is not the case is evident upon eInsemination of the data. (See Table 21.) Almost forty percent of the 8«maple have no children (58.9%) and less than one-quarter have more than two (22.79%). With two-thirds of the sample between thirty and forty- 11in» years of age (see Table 19) it is a logical expectation that the I‘esidents would exhibit fairly high birth rates; instead, more than \- 8fies Warren S. Thompson, Population Problems (New York: McGraw- Hill Book Company, Inc., 1955), p. 3 . 0" la. la - 58 - three-quarters have less than three children. It was reasoned that per- haps the marital status of the population has led to this situation; however, a check of this factor revealed that only seven percent of the respondents are unmarried. Thus, the low birth rate of the Lansing fringe dwellers remains unexplained. The median gross family income is more than 8107 and twenty per- cent reported that the family gross weekly income is over 3200. (See Table 22.) The data are not available, but it is possible that one of the major factors influencing the income distribution is the large num- ber of families that have no children (see Table 21); this situation al- lows both husbands and wives to work. gapgpggzz characteristics p£,the popplation This section was devoted to a description of the demographic characteristics of the Lansing fringe population. These characteristics can be summarized in eight points: 1. The Lansing fringe exhibits the expected heterogeneous occu- jPational structure which is characteristic of such areas. Almost one- quarter of the respondents are full-time farmers, one out of eight is a Part-time farmer, more than.one-third are employed in a factory, and one- quautsr are employed in a diversity of urban occupations. 2. The occupational structure of the area affects place of work Patterns: almost two-thirds of the Lansing fringe residents work in the central city. 5. Mere fringe residents live in the tier furthest from Lansing mMan in any other. However, this is due to the fact that the last tier 13 larger than any other. In terms of density, the furthest tier is the Second highest while the closest tier to the city has both the lowest - 39 - number and lowest density of residents. #. Mest Lansing fringe dwellers own their home: only one re- spondent in eight rents in the area. 5. Fringe dwellers are in the fairly young age groups. Approx- imately two-thirds are between thirty and forty-nine years of age; the median age is forty-four. 6. The educational attainment of fringe respondents is fairly high. Less than sixty percent did not graduate from high school; the median educational attainment is three years of high school. 7. The number of children in the area is not as high as was ex- pected. Almost forty percent have no children under seventeen and less than one-quarter have more than two. 8. The income distribution in the Lansing fringe is high. Twen- ty percent reported that the family gross weekly income is over 8200, and the median family weekly income of respondents in the area is over 8107. Respondents' definitions o_f t_h_e_ _s_._r_eg Respondents were asked why they chose their particular area of residence rather than some other part of the Lansing area. The responses tO this question indicated that over half did not actually have positive ctuaice for the specific area (51.8%); rather, the majority answered that 13heir choice was based on economic considerations such as "best buy." (See Table 25.) Less than ten percent indicated a positive choice for the area (8.2%). Responses that seem to be based specifically on a de- ‘Ilsion to move to the fringe because of the unique character of the area Were not very frequent. These responses are either that the respondent wanted to live in an unsettled area (15.9%) or that the respondent found - 40 - the fringe area convenient as a compromise that would permit him to live in the country while being near town (9.Q%) In an attempt to determine whether fringe residents consider their residential area as an independent unit respondents were asked whether the area has an independent name and whether they consider the area part of a larger unit. Thirty percent felt that the area does have an independent name. (See Table 2k.) When asked to supply the name more than one-third supplied a local name (59.8%), over one-quarter sup- plied the name of the township (25.9%), more than one-quarter replied with the name of a local town (26.8%), and one out of eight supplied the name of the school district. (See Table 2#a.) One-third did not perceive their residential area as a complete- ly independent unit; these respondents indicated that the area is part of a larger area. (See Table 25.) When asked to indicate the name of the larger area these residents provided a variety of answers, the most prevalent being the township name, a bordering town name, and Lansing (51.1%, 51.1%, 20.6% respectively; see Table 25a). Perhaps the most in- teresting group is composed of those thirteen respondents who felt that their residential area is part of Lansing even though they live outside the Lansing city limits. In order to determine whether respondents thought they were liv- ilug in a rural, urban, or an "in-between" area, sample members were pre- 8eluted with these alternatives. The procedure was as follows: inter- viewers handed each respondent a card with the following inscriptions (“I it: 1) rural-~urban 2) rural-~urbane-in-between The interviewer stated: "Here are two ways of classifying the types of - 41 - areas in which people live. Which do you think makes most sense in terms of the way in which things actually are?" Over sixty percent chose the first alternative (62.9%; see Table 26). After choosing between the two alternatives, respondents were asked to indicate the type of area in which they live. As can be seen in Table 26, two-thirds of the sample felt that they live in a rural area, almost thirty percent felt that they live in an "in-between" area (28.8%), and only five percent felt that their residential area was urban. Those respondents who indicated that they resided in an ”in-between" area were asked to provide a name for this amorphous designation. However, the question had so biased re- spondents that most could think of no name other than the "in-between" phrase. When asked whether they thought there were differences between area people and city dwellers, two-thirds of the respondents who did not think they were living in an urban area replied that there were no diff- erences. (See Table 27.) Those respondents who felt that area people and city dwellers were different gave a variety of differences, with al- mmst sixty percent stating that the people in the area were more friend- 13 or neighborly (57.9%; see Table 27a). Of those respondents who did not feel that they were living in a Immal area, more than two-thirds felt that there was no difference be- tWeen residential area people and rural dwellers (68.5%; see Table 28). ReSpondents who indicated that a difference exists between area residents 3“"1 rural dwellers were asked to state the nature of these differences. (See Table 289..) However, the base is so small that the opinions of this group must be examined very cautiously. As a filter to determine whether respondents felt that area .. 1+2 - people think like farm or city dwellers, the sample was asked whether they thought that farm and city dwellers think differently. The ninety- two respondents who felt that a difference in thinking exists (see Table 29) were then asked whether residents in the area think like farm or city dwellers. Over three-quarters of these respondents indicated that area dwellers thought like rural inhabitants (76.176; see Table 29a). Thus, once again, respondents indicated that their area of residence was simi- lar to a rural area. Respondents were asked several questions in which the phrase "area around here" was utilized, this phrase being used consistently to refer to the immediate vicinity, or the neighborhood. After the inter- viewer had used the phrase several times, respondents were asked the following question: "We've been referring to the 'area around here' for some time now, and I'd like to get an idea of what you've been referring to. How far north, south, east, and west does it go?" Responses to each direction were added together and it was found that almost two-thirds of the sample felt that the "area around here" encompassed a total area of eight miles or less (65.9%; see Table 50). In addition, respondents were asked to indicate the criteria they utilized for delimiting these boundaries. Sixty percent replied that they were defining the area in terms of where their friends and people they knew lived, more than ten Percent utilized the school district boundaries (12.4%), and more than ten percent indicated that their choice was based on the fact that there are many houses in this area (12.7%; see Table 50a). m: respgndents' definitions gt; the. area In this section an attempt was made to investigate differential Perceptions of the Lansing fringe area. The findings can be summarized - 43 - as follows: 1. When sample members were asked to indicate their reasons for choosing the Lansing fringe area as a place of residence, there was a lack of responses indicating positive evaluation of fringe areas in gen- eral, and the Lansing fringe in particular. The most prevalent response indicated economic motivation. 2. Independence of the area was not perceived. More than two- thirds of the respondents could not assign a name to the area. Approxi- mately one-third felt that their residence area was part of some larger area. 5. In seeming contradiction, two-thirds of the sample members felt that the area was rural while a majority saw no difference between area residents and city people. Thus, it seems that residents defined the Lansing fringe as a rural area that was inhabited by city people. 4. Those respondents who did not classify the area as rural (65.5 residents) felt that the residents are rural: two-thirds saw no difference between area inhabitants and rural people. In agreement, of those who thought that rural and urban people think differently (92 residents), three-quarters felt that area inhabitants think like rural Peeple. 5. The neighborhood was defined as a fairly small area. Almost two-thirds of the respondents defined the "area around here" (i.e., the neighborhood) as an area encompassing a total of eight miles or less (litilizing the sum of four sides). 1:d\'._a.luations pf 35g area When asked whether they would like to continue living in the area, respondents overwhelmingly replied in the affirmative. (See - 41+ - Table 51.) Only eight percent of the sample failed to give a positive answer to the question. This one-sided evaluation indicates that fringe dwellers are indeed positively oriented to the fringe area. The question designed to determine whether respondents would like to maintain their residence in the area was followed by questions which asked those who replied in the affirmative what they like about the area, and those who answered in the negative what they dislike about the area. Those who would like to remain gave a wide variety of reasons, the most frequent of which implied a simultaneous positive choice for the area and a negative evaluation of the city. (See Table 51a.) Picking out only those responses which were endorsed by more than ten percent of the sample, the most frequent reasons for liking the area, in descending or- der, were the following: the area was not crowded (26.8%); it was quiet and peaceful, and allowed for freedom and fresh air (20.8%); the neigh- bors were friendly (16.5%); the area was advantageous for farming (15.5%); children benefited from residence in the area (11.9%); and having been tern or raised in the area is sufficient reason for liking it (10.7%). Since those who gave a negative reply to the initial question are so few in number, discussion of their reasons will not be presented. Since the Lansing fringe area is changing so rapidly as it goes through a transition from rural to a mixture of rural and urban, it was decided to give fringe dwellers an opportunity to evaluate this change. 1"1218 was accomplished by asking sample members whether the area is get- ting better or worse. Over half of the respondents felt that the area ‘flis improving (56.8%), more than one-quarter indicated that they could not decide (26.7%), one in eight replied that it is staying the same, and less than five percent thought that the area was getting worse. -45- (See Table 32.) Those respondents who thought that the area was getting better were asked what made them feel this way. The most frequent replies in- dicated a positive evaluation of the in-migration: almost one-half felt that the area is improving because nice houses are being built (h9.1%) and one out of five endorsed the in-migration strongly by replying that the area was getting better because of its rapid growth. (See Table 52a.) In an attempt to determine evaluations of area facilities sample members were asked whether they felt that the facilities in the area were adequate. This question is of particular interest since such a large proportion of the residents moved to the area from Lansing, where paved roads, sewage, and other municipal services are taken for granted. Con- trary to expectations, only fifty-six percent of the sample felt that new facilities were needed in the area. (See Table 55.) Most of the facilities desired by residents were of the type that are present in the urban area, with improved roads and a sewage system most frequently men- tioned (21.1% and 11.1% respectively). Respondents were asked whether they thought that living in Lan- sing would have advantages in comparison to their present residential area, and whether the Lansing fringe has advantages in comparison to residence in Lansing. Overwhelmingly, sample members indicated that their present residential area had advantages (96.5%) but thirty percent also felt that Lansing residence would have certain advantages. (See Table 5h.) Thus, a number of respondents saw advantages in both areas. Sample members were also asked to indicate the nature of the ad- vantages in each area. The Lansing advantages seem to be of an instru- mental character, with closeness to work, available facilities, and .. 46 - closeness to stores heading the list of replies (58.5%, 57.6%, and 17.1% respectively; see Table 54a). 0n the other hand, respondents who felt that the fringe area was advantageous seemed to imply a negative evalu- ation of fringe life. Leading the list of responses are: the area was not congested; it was a healthy place to live, especially since it was quiet, had lots of fresh air, and was clean; there was room for the children; there was more freedom and privacy than in a city; and it was possible to garden or farm in the area (47.0%, 40.9%, 16.8%, 15.1%, and 11+.8% respectively; see Table 54a). Respondents who thought that they are not living in a rural aree (see Table 26) were asked to indicate whether their present residential area had advantages over living in outlying farm areas. Eighty percent felt that fringe area residence was advantageous, principally because of its facilities, closeness to town, and closeness to work (92.1%, 55.5%, and 25.11% respectively; see Table 55). The outlying farm area advantages will not be discussed since few respondents are in this category. My: evaluations 9_1_‘ the. 9.339 This section has dealt with respondents' subjective evaluations 0f the area. A battery of questions was devised whereby sample members Were put into a position of making judgments about the Lansing fringe area p_e_l; _s_e_, and the Lansing fringe in comparison to alternative areas of residence. The findings can be summarized-in the following five POints: l. Overwhelmingly, respondents indicated a desire to remain in the Lansing fringe area. The most prevalent reasons for remaining Seemed to imply a positive choice for the area and a negative evaluation 01‘ city life. - 1.7 - 2. More than half of the sample members felt that the area was getting better. The most prevalent reasons for this attitude suggested a positive evaluation of the large-scale in-migration taking place. 5. In spite of the fact that many of the residents are former urban dwellers, only fifty-six percent of the sample members expressed a desire for facilities in the area. it. Overwhelmingly, respondents indicated that Lansing fringe area residence had advantages over residing in the city. Thirty percent felt that Lansing had advantages in comparison to fringe residence (many respondents indicated advantages for both areas) with facilities and closeness to work and stores mentioned frequently. 5. Of those respondents who did not think that they are living in a rural area, eighty percent felt that fringe residence was advanta- geous to living in a rural area. Facilities and closeness to the city were the most prevalent reasons given for this attitude. Social pgticipation _ip 3313 M frigge Sample members were given alternative choices on the question of how well people in the area knew each other. Thirteen percent of the respondents indicated that the people in the area knew each other "not so well," one-half felt that Lansing fringe dwellers knew each other "fairly well,” nineteen percent replied "quite well," and thirteen per- cent felt that the residents knew each other "very well." (See Table 35-) Thus, half of the respondents chose the middle alternative, indi- cating that the extremes of friendship patterns do not exist in the area. It is significant that only one respondent felt that the residents in the area did not know one another at all. Another checklist was pro- . '1de for indicating how many people in the area they knew by name. The - ha - extremes of "none" and "all" are represented by four percent and twelve percent respectively; the most frequent choice was "a few,” with over forty percent answering in this manner. (See Table 37.) Once again it seems that the extremes of friendship patterns were not prevalent. In an attempt to determine the extent of family contacts in the area, three questions were suhnitted to respondents. The first asked sample members to indicate the number of area families with which they come into contact for a few minutes each day (see Table 38), the second was designed to determine the number of area families with whom most of an afternoon or evening would be spent (see Table 39), and the third asked whether respondents participated in lending-borrowing relationships with other area families (see Table #0). Looking at lack-of-contact only, the data indicate that thirty percent do not come into contact with area families very frequently, forty percent do not spend after- noons or evenings with area families, and two-thirds neither lend nor borrow. Thus, a sizeable proportion of sample members have little or no contact with area families. Respondents were asked to list the areas to which they go to Visit friends and to name the area in which their three best friends live. The answers to both questions suggested that while friends in the fringe area are prevalent, residents' friends are more frequently lo- cated in Lansing. All respondents indicated that they do possess at 1east three best friends, though this might be the result of interviewer urging, rather than a measure of the actual situation. Table 1&1 points out that almost sixty percent of the sample mem- bers visit friends in Lansing (58.2%); it is assumed that this high fig— are is partially a result of maintaining friends in the previous resi- - 1,9 - dential area. In comparison to this proportion, one-half of the sample replied that they visit friends within the area. Thus, the visiting pattern of fringe residents is such that slightly more trips are made to the city for visiting that are made within the area. One-quarter of the sample members located all three best friends in Lansing, while eigiteen percent gave their neighborhood as the loca- tion of all three. Almost twenty percent replied that their three best friends are in both areas (18.5%; see Table I+2). In total, fifty-seven percent indicated that at least one of their three best friends is lo- cated in Lansing, while forty-seven percent listed at least one in the neighborhood. Thus, a slightly higher proportion of best friends are Lansing residents than are neighborhood residents. One-half of the sample members do not belong to any formal or- ganizations and approximately one-quarter belong to one (22.5%; see Table #3). Of those respondents who are members of organizations, most belong to groups that meet in a local town (53. 2%), more than forty per- cent are members of organizations that meet in the neighborhood (H.595), and over one-third are in organizations that meet in Lansing. (See Table Ida.) Respondents are almost equally distributed among fraternal, Church, school, and farm organisations (all around 50.0%; see Table lab). In a series of three questions designed to determine the extent of participation in informal community groups, respondents were asked if they get together with others in such local establishments as community hfills, taverns, or gyms; whether they participated in such volunteer community groups as community fund raising organizations, civil defense, 01‘ volunteer fire departments; and whether they were involved in such 100-31 issues as schools or zoning. Over ninety percent replied that - 50 .. they did not utilize local informal facilities (91.0%; see Table 44). The seventeen who mentioned participation indicated that a local commun- ity hall or a church basement is the most frequent place of participation. Approximately one-quarter of the respondents have Joined with other com- munity members as volunteers in local groups (23.5%; see Table 1+5). The type of volunteer group mentioned most frequently was civil defense, with a comunity fund-raising campaign, and the volunteer fire department also mentioned “7.2%, l110.9%, and 19.1% respectively). Finally, more than one-quarter of the sample members have participated in local issues (28.0%; see Table #6), all of them mentioning a school problem, thirteen percent mentioning a fire department issue, and ten percent indicating participation in an issue concerning zoning problems. (See Table 1+6a.) The most frequent reasons given for participating was that the issue con- cerned their children (44.3%), or affected the tax rate (23.6%), or im- proved the community (20.8%). The most frequent reason for not partici- pating was "no knowledge of the issue" (89.7%). Respondents were presented with a list of items which included groceries, gasoline, movies, clothing, drug items, barber shop, and nightclub, and were asked to list the specific areas to which they go for these services. The most frequent reply was "all in Lansing" (ll-0.2%; See Table #7). Twenty percent replied that more than half of their shop- Ding took place in Lansing and one-quarter that more than half took Place in the area ("area" includes surrounding towns). Once again the dominance of the central city is evident. As one means of determining the extent to which fringe residents u“ii-lined local resources, sample members were asked for their sources of news about the immediate area, the township, the county, and Lansing. - 51 .. One-quarter indicated that they didn't keep up with news of the immedi- ate area. The same proportion did not keep up with township news. In contrast, only ten and four percent, respectively, did not keep up on news of the county or Lansing. (See Table 1+8.) Friends and neighbors are utilized as news sources by over one-half of the respondents in the case of immediate area news (5}. 2%), by more than one-quarter for town- ship news (28.6%), by one out of eight for county news, and by less than ten percent for Lansing news (8.7%). Overwhelmingly, the newspaper, ra- dio, and television are utilized for information about the city (91.8%). Hatching television was the most frequently mentioned leisure time activity of sample members (#8.9%). One-third either participated in or observed sporting events, twenty percent watched movies, and the same proportion worked in their garden. (See Table ‘09.) Honfarm sample members were asked whether they raised any food and, if an affirmative reply was received, the proportion of the family food cost covered by the production was determined. A majority of the nonfarm residents in- dicated that they did raise food (57.9%; see Table 50). Two-thirds of those who raised food reported that their production accounted for less than one-quarter of their food cost. (See Table 50a.) Thus, most of the gardening is on a small scale. However, there are a large number of respondents who reported that their gardening accounted for one-quarter, and between one-quarter and one-half of their food cost (18.6% and 10.0% respectively) . m: social participation _i_n_ 1h; Lansi_n_g gm A battery of questions was submitted to sample members to deter- mine the extent and type of social participation in the Lansing fringe a”Ga. The results of this aspect of the examination of social life in - 52 - the area can be summarized in the following nine points: 1. In the estimation of respondents, extremes of friendship pat- terns are not prevalent in the Lansing fringe area. When provided with a checklist to determine respondents' opinions of how well people in the area know each other, one-half of the sample members chose the middle category, indicating that people in the area know each other ”fairly well." The extreme categories ("not at all" and "very well") were not frequently chosen. 2. When asked to indicate how mamr residents are known by name, the most frequent checklist choice was "a few." One out of eight sample members indicated that they know all people in the area by name. 3. Family interaction patterns were examined: thirty percent of the sample members did not have frequent contacts with area families, forty percent did not spend afternoons or evenings with area families, and two-thirds did not lend anything out. 1+. Friends were visited in Lansing to a larger extent than in the area itself. In addition, the largest proportion of respondents re- POrted that their three best friends are located in the city. Thus, friendship patterns of Lansing fringe residents seem to be dominated by the city. 5. One-half of the Lansing fringe residents belong to formal 01‘88n1zations; the most frequently mentioned meeting place was local tom, 6. Local informal activities are not frequent: over ninety per- cent have not participated in local informal facilities, approximately °ne~quarter have participated as volunteers in local groups, and one- quaafter have participated in local issues. - 53 - 7. Lansing is the most frequently used shopping area: forty percent of the sample members did all their shopping in the city and another twenty percent utilized Lansing for more than half of their shop- pins. 8. Friends and neighbors were the most frequently used sources for obtaining local news. One-quarter did not keep up with local news and the same proportion did not keep up with township news. 9. Watching television was the most frequently reported leisure time activity. Over half of the nonfarmers raised food, with two-thirds of these reporting that their production covered less than one-quarter of their total food cost. g unique case: reaction. .29 _a_ threat t_o_ independence Some time before the interviewing started, a Lansing radio exec- ‘utive attempted to institute an organization to be known as the Lansing Metropolitan Planning Commission. The function of the Commission was an- “01111096. to be the centralization of a unit which would plan for the cen— tlnil city and its tributary area, which included the area in which the fI‘ILnge study was conducted. The proposed Commission received its normal Stuare of publicity but the attempt failed due to a variety of factors. One of the important factors leading to the demise of the proposed Plan- ning Commission was the rather critical attitude taken by governmental “Writs outside the City of Lansing; the proposal was frequently inter- Pl‘eted as the beginning of an insidious attempt on the part of Lansing tO’dictate to the surrounding area and to lead this area into an unan- nminced course which would eventually culminate in annexation. It was decided to take advantage of this situation when the study .. 5i... - was projected by asking sample members whether they would endorse or re- ject the idea of a commission of the nature described. However, very few of the respondents recognized the name of the Commission in the in- terview situation. Therefore, the organization was described for respon- dents in the following manner: About two years ago an attempt was made to organize a group called the Lansing Metropolitan Planning Commission. The purpose of the group was to help coordinate the development of the City of Lansing and Ingham, Eaton, and Clinton Counties by collecting information and fermulating plans for the total Metropolitan Area. The group was nonrprofit and had no political power. This area would have been one of those covered by the Commission's work. It was felt that respondents who were concerned with maintaining the political independence of their residential area would reject the idea of such an organization since the only threat to independence exist- ing at the time was domination by the city. The question of endorsement or rejection was followed by a query about how respondents would react if the Commission recommended annexation. Again it was felt that sample members who were concerned with maintaining independence would reject such a recommendation. One-half of the sample members endorsed the Commission, one-third took a negative view toward its organization, and the remainder could not decide on the issue. (See Table 51.) Respondents who felt that the idea of the organization was a good one supported their opinions with such reasons as it being good to plan, that it would lead to orderly growth, and that better facilities would result (50.5%, 28.9%, and 22.5% respec- tively; see Table 51a). These most frequent answers were really only a positive evaluation of planning, without reference to the particular sit- uation. Host of the other answers indicated that the specific community would benefit from the planning process. - 55 .. Respondents who were opposed to the Commission gave as their rea- sons (see Table 51b) the lack of understanding of their area on the part of the city, the distance of their area from Lansing, and a desire to re- main rural (27.7%, 16.2%, and 16.2% respectively). That many of the re- spondents interpreted the proposed Commission as a threat to the indepen- dence of their area of residence is evident in answers such as the de- sire to remain rural, an opinion that Lansing would try to dictate to the area (10.0%), an expression of the ability of the area to handle its own problms (13.1%), and the opinion that such an organization would eventually lead to annexation (13.8%). Overwhelmingly, sample members rejected the idea of annexation (83.3% opposed, 11.9% in favor; see Table 52). The most frequent reason for rejecting an annexation proposal was not in terms of a threat to in- dependence, but in terms of the effect such an action would have on the tax rate (50.8%; see Table 52b); in other words, a personal rather than comunal reason was most prevalent. Other frequent responses were that the residents prefer to maintain the area as rural (19.7%), and that it was too far from Lansing (14. 3%). In addition to the desire to remain rural the only clear-cut answers that reflected a concern with preserv- ing independence are: "we can handle our own problems" (2.2%), "I moved to get out of the city" (5.11%), answers indicating concern that the city would attempt to dictate to the area (9.8%), and answers indicating that the area would become too crowded (3.8%). All in all, responses imply- ing a concern with a threat to the independence of the fringe area were nOt prevalent. m: reactions 33 a threat 3:3 independence In this section respondents' attitudes were examined within the - 55 - context of a threat to the independence of their community. Sample mem- bers were informed that an attempt had been made to organize a commission that had as its goal the formulation of plans for the total Lansing Met- xxipolitan Area, an area which included the study site, and were then asked to indicate how they felt about the formulation of the organiza- tion and for their reactions if the Comission recommended annexation. It was felt that respondents who were concerned with possible Lansing encroachment would interpret the situation as a threat. The reactions of respondents can be summarized in two sections: 1. Only one-third of the respondents rejected the idea of such El commission; one-half favored it, and the remainder could not decide. Interestingly, those who expressed opposition to the plan did interpret the situation as a threat, while those who endorsed the plan most fre- quiently indicated a positive evaluation of planning, without relating the situation to the specific areas involved. 2. The idea of annexation was soundly opposed. However, the “kbst prevalent opposition reason was not an indication of the recognition of a threat at all--rather, more than half of the opponents to annexa- t1cm indicated that one of their reasons was their fear that the tax lirte would climb. The opinion that annexation is a threat was mentioned, but it took second place to the tax issue. CHAPTER III AN ANALYSIS OF OBJECTIVE INTEGRATION AND SUBJECTIVE IDENTIFICATION Sample Distributions An examination of the distribution of sample members among the integration and identification categories will provide a description of the empirical patterns of these dimensions.1 In the case of integration the distribution can be interpreted in a positive and negative context since the scale continuum runs from "low" to "high" integration; in the case of identification only direction provides immediate information of this type since neither content nor depth lend themselves to such an in- terpretation. 0b ’ ective integation Almost sixty percent of the respondents fall into the extremes 0f the objective integration scale, with most representing the "low" end. (See Table 53.) Since Type it contains more than twice the number of TYPO 2 respondents, approximately the same number of sample members pop- “late the two "high" scale types as the two "low" scale types. Only one r‘=’3P<'.~lndent in eight occupies the center position. \ 8:: 1'See Appendix C for a discussion of the theory and method of ale analysis and a presentation of the derivation of the integration and identification scales. -57- -53.. TABLE 53 DISTRIBUTION OF SAMPLE MH'IBERS AMONG OBJECTIVE INTEGRATION SCALE TYPES Scale m Number Percent 1 (LOW) 64e5 %e1 2 l7 9 .O 3 25 12. 2 it 39.5 20.9 5 (High) #5 23.3 Total 189 100.0 Direction 9}: subjective identification The overwhelming preponderance of residents with positive iden- tifications is immediately apparent upon examination of Table 5‘}. More than ninety percent of the respondents are positively identified with the fringe area, a finding that is not consistent with the conclusions reached by other investigators of the fringe.2 Pirey, Kimball, Beegle and Schroeder, and Gist have all indicated that fringe area residents lack identification with their area of residence, with Kimball‘s state- ment most directly to the point: ". . . for mamr who remove to the new l~‘393ality the change is one of space and not of social identification."5 CeItainly, expectations based on the findings of these men are at radi- 0&1 divergence with the distribution indicated in the table. Because of the:ir inconsistency with the conclusions of previous investigators, the data. strengthen the suspicion that those conclusions may in large mea- 8“re reflect the theoretical positions which guided the research. Posi— t1Ve identifications of residents with the fringe area do exist, and to ‘ 2See an ra, pp. llff. S 3Solon T. Kimball, The Egg Social Frontier: The Fring' e, A.E.S. peeial Bulletin 360 (Michigan State College, June, 19:9), p. 30. -59.. an overwhelming degree. TABLES‘} DISTRIBUTION OF SAMPLE MEMBERS AMONG DIRECTION OF SUBJECTIVE IDENTIFICATION TYPES Direction Number Percent Non-positive l6 8. 5 Positive 17} 91 . 5 Total 189 100. O Content _o_i_‘ subjective identification The distribution of the respondents among the various identifica- tion categories is presented in Table 55. Approximately two-thirds of the residents can be classified in pure content types, with the personal category accounting for well over half of these. The personal category, combined with its mixed types, is present in forty-seven percent of the cases. The ecological category, together with its mixtures, is repre- Sented in about one-quarter of the cases while the social categories ac- c’C’unt for approximately twenty percent. On the whole, identification takes such a wide variety of forms fO-‘P the respondents that little can be found in the way of clusters. The notable exception to this observation is personally based identification. Among the reasons that such residents desire to maintain residence in the area are the following; they feel that it is good for raising chil- dren, they have freedom there, it is a good place to retire to, they are anti-city, it is quiet, it is peaceful, it has good facilities, they can Burden, or there is room for children. Social, economic, ecological, coereed, and aesthetic reasons are quite scattered. .. 60 .. TABLE 55 DISTRIBUTION OF SAMPLE MEMBERS AMONG CONTENT OF SUBJECTIVE IDENTIFICATION TYPE Content Number Percent Ecological l6 9 . 2 Economic 6. 5 3.8 Ecological-economic 6 3. 5 So cial 11 6 . ’+ Social-ecological 10. 5 6 . 1 Soc ial-economic it 2. 3 Soc ial-personal 8. 5 4. 9 Personal 66 38. 2 Ecological-personal 7 ’+. O Coerced 19 11. O Other combinations with ecological or economic 10. 5 6. l _ojher 8 he 6 Total 173 100. 1 ml; 53 subjective identification More than half of the Lansing rural-urban fringe residents wish t<3 remain in the fringe community) for reasons other than those for which they first moved to the area. (See Table 56.) Over one-quarter of the sample members expressed consistent reasons for moving and staying and InOre than one in five indicated that their residence in the area was not 8- result of prior choice. TABLE 56 DISMBUTION OF SAMPLE MEETS AMONG DEPTH OF SUBJECTIVE IDENTIFICATION TYPES m Number Percent 3'0 prior choice 57.5 21.7 Sther reasons 88.5 51.2 We reasons as staying 1+7 27.2 ...; Total 173 100.1 .. 51 - It is evident that among those residents who chose to move into the fringe area, about two-thirds indicated that their reasons for stay- ing are not consistent with their reasons for the initial move. Thus the identification pattern is such that present orientations are frequent- ly at variance with orientations held before the move was made. 131% An examination of the objective integration distribution indi- cates that the spread among the five scale types is such that all are fairly well represented, with each extreme type contributing most heavi- ly to the distribution. The median scale type is represented by only twelve percent of the respondents, with the other eighty-eight percent almost equally divided on both sides of the center score. More than ninety percent of the respondents indicated that if they had their choice they would continue living in the fringe community; these respondents are classified as exhibiting positive identification. Respondents who exhibited such positive identification were examined in an attempt to determine the focus of support for this feeling. The scatter was quite diverse, with heaviest representation among the person- a]. reasons. The ecological and social reasons seem to be next in impor- tance, but the two combined do not occur as frequently as personal ori- errtations. As to the depth of identification, more than half of the re- 8Pcn'idents who exhibit positive identification responded to questions abOut why they moved to the area and why they would now prefer to re- main with different types of answers. Among those who had exercised a Preference in moving to the fringe, approximately two out of three sup- plied. different reasons for first moving into the area and now staying. - 62 - The Relationship between Objective Integration and Subjective Identification Before an attempt is made to account for the various patterns of objective integration and subjective identification, an analysis will be undertaken to determine the nature of the relationship between integra- tion and the direction, content, and depth of identification. This analysis will attempt to indicate whether or not the two dimensions are independent and, if not, will suggest the pattern and degree of relation- ships that may exist. The chi-square test is particularly suitable for answering two of these questions because: 1) it indicates whether independence exists, and 2) it indicates the direction of deviation from chance expectation in each cell of a contingency table. A measure of the degree of association, as well as the chi- square indication of the existence of association, is also necessary. Two different measures of association were used, the first being con- cerned with the predictive function of variables and the second with the degree of association among variables. 1. Prediction improvement measure: given an attribute classi- fied.into exclusive sub-categories, the best estimate of any individual's POSition is that he is a member of that sub-category which is largest. FOI‘ example, given a distribution that classifies a population into Category in (75%) and'category x" (25%), errors would be at a minimum if every randomly chosen individual were predicted to be a member of Cate- “W X'. In this case the probability of error is l - 75%. or 25%, while the prediction of Category 1" would lead to 757‘ error. If the population is further classified into a second group of -63-— exclusive sub-categories which may be cross-classified with the first attribute referred to above, and if the two variables are associated, the error in prediction can be reduced. Table 57 is presented to illus- trate the point.l+ The added information of knowledge about Y' and Y" makes it possible to reduce the total error by predicting each X' to be a Y' and each X" to be a I". In this case, by taking the additional in- formation into account, the amount of error is reduced from 1 - 75% to l - 85%, i.e., from 25% error to 15% error. TABLE 57 ILLUSTRATION OF THE PREDICTION IMPROVEMENT MEASURE B Classification X' X" Total Y' 65 5 70 A Classification Y" 10 20 30 Total 75 25 100 In a recent article Goodman and Kruskal point out that the in- crease in efficiency of prediction provided by the introduction of the second attribute may be utilized for devising a measure of association between the two attributes.5 The measure may be defined as follows # #Table 57 is presented in 2 X 2 form for reasons of parsimony. ghe Prediction improvement measure is applicable to tables larger than X 2 if both variables possess exclusive sub-categories. 5Leo A. Goodman and William H. Kruskal, "Measures of Association for cross Classifications," Journal of the American Statistical Associ- $’ 1+9 (December, 195%), pp. 732-767;. This article expands on the Gala originally presented by Louis Guttman, "An Outline of the Statisti- pau'l'heory of Prediction," in The Prediction 3f Personal Ad ‘ustment, ed. pp. 2:01.86, gt _a_1_. (Social Science Research Council, Bulletin E8, 19%), 8-2 3. -64- (Goodman and Kruskal assign the Greek letter "lambda" to this formula; for convenience L will be substituted here): - Probabilit of error in case 1 - Probabilit of error in case 2 I‘1': Probability of error in case 1 In the example presented above, the following figures may be substituted: .2 -.1 Lb'JTzT'i"“°0 Thus .400 is the relative decrease in probability of error in predict- ing sub-categories X by taking sub-categories Y into account. Or, Lb indicates the proportion of errors that can be eliminated by utilizing information from the A classification. An examination of Table 57 also indicates that the prediction of X from Y is clearly a different problem from the prediction of Y from X. Utilizing the information presented in the table, it is evident that in the A classification only, error would be at a minimum if all randomly chosen individuals are predicted to be Y'. In this case the error would be 1 - 70% or 30%. The addition of the information in the B classifica- tion makes it possible to reduce the total error by predicting each Y' to be p and each r" to be x". By utilizing the additional information the amount of error is reduced from 1 - 70% to l - 85%, i.e., from 30% error to 15% error. Following the formula presented above, the follow- ing figures may be substituted: La _ .3O.30.l§ _ .500 Thus 0“300 is the relative decrease in probability of error in predicting I by 1ialcing X into account. Or, La indicates the proportion of error in predicting A that can be eliminated by utilizing the B classification. Whether one attempts to derive Lb or La in a given cross—classi- f ication depends on the problem. If the B classification is the inde- ‘ - 65 - pendent variable in a given study, it would seem logical to determine the gain in efficiency of prediction provided by utilizing the infome- tion of this grouping, i.e., the determination of La would be appropri- ate. If the reverse situation holds, Lb would seem most crucial. If both classifications are judged of equal importance, the improvement in both directions must be taken into account simultaneously. This can be accomplished by utilizing the information provided by both Lb and La' For example, if the Table 57 case is judged symmetrical, the following figure may be derived: ,gg - .1 + .30 - ~12 . L ' .25 + .30 ’455 2. The contingency coefficient: Goodman and Kurskal point out that "association" is an ambiguous concept which must be interpreted in terms of its contextual meaning.6 As indicated, their article points to the utility of the "lambda" statistic as an appropriate measure of the degree of prediction improvement when one group of sub-categories is utilized to clarify the relationship with a second classification. Another group of measures of association is based on the chi- squéire statistic; when used in addition to chi-square, a given relation- ship can be interpreted with greater clarity. Whereas chi-square indi- cates the existence of association, these measures indicate the W of relationship. As Siegel points out, "the problem of measuring $289.22 or aaaociation between two sets of scores is quite different in charac- ter 1from that of testing for the existence of an association in some pop- ulation. ..7 \ 61bide, p0 735e en. 7am, Siegel. W Was $2.15: the “Behavioral as;- Q (New York: McGraw-Hill Book Compamr, Inc., 1956): p. 195. -66.. One of the useful measures of the degree of association is known as the contingency coefficient, which is symbolized as C, and is defined as follows: 0...}... The contingency coefficient is based on the reasonable idea that the amount of correlation between variables classified in a contingency table depends upon the divergence of observed frequencies inacells from the frequencies expected if there were no relationship. Thus, the larger the discrepancy between expected and observed frequen- cies, the larger is the value of C, i.e., the larger is the degree of association. C varies between zero and one; therefore, if there is no associ- ation between values, C will equal zero. However, C understates the de- gree of association present since the upper limit, which is a function of the number of rows and columns, approaches, but does~not reach, the value of one. In the 2 X 2 case, for instance, the upper limit of C is -707; in the 3 X 3 case the upper limit of C is .816. To obviate this limiting factor to some degree, Peters and Van Voorhis have developed a 9 correction factor for C which is based upon row and column values; by m311:5.zing the correction factor C is derived. The upper limit of C does approach one and the correction thus yields a measure which makes contin- gency coefficients from dissimilar tables somewhat comparable. Since even the correction factor does not place the limits of C precisely \ c t 8Merle W. Tate and Richard C. Clelland, Nonparametric _ar_lc_1_ Short- -“~ Statiotics (Danville, Illinois: Interstate Printers and Publishers, In“ ’ 19575. po 17- ce 9Charles C. Peters and Walter R. Van Voorhis, Statistical £111- \ “1‘98 and their Mathematical Bases (New York: McGraw-Hill Book Com- Deny -- -——-———-— ...—_- . inc.’ 19 :05, Fe 598e —67- between zero and one, the investigator must still proceed with caution in interpreting two values of C from tables with varying rows and columns. By reasoning from an analogous measure, the product-moment corre- lation of coefficient, C- may be understood as an estimate of the propor- tion of variance in one variable accounted for by its relation to anoth- er variable. The square of the correlation coefficient‘10 estimates the proportion of the total variation in a variable that is explained by its association with another variable which has varied. Likewise, the square of the contingency coefficient may be utilized to estimate the propor- tion of variance that is explained. Thus if the U for a given group of data. equals .500, the variance accounted for is twenty-five percent (1. e., twenty-five percent of the variance in one variable is accounted for by variance in the other). Unfortunately, the ambiguity in the term "association" makes a Precise interpretation of the meaning of a 6' value impossible. This is evident if an attempt is made to interpret the meaning of a given 5, 908- o .500. Whether .500 is a "low," a "moderate," or a "high" degree 0f association between variables depends upon a subjective evaluation 0f these terms. All one can say with confidence about a given C is that it is: or is not, significantly different from zero. However, this con- fident statement is not a conclusion based on the value of 6; rather, it is 9' cOnclusion drawn from the value of chi-square. If the chi-square d°riVed from a cross-classification is significant, any value 'of C de- rived 1‘ rom the same data is significantly different from zero; if the chi'aquare is not significant, the derived C is not significantly diff- erent from zero. \ mi 10This is frequently referred to as the "coefficient of deter- na-tiOn. II -68.. The data presented in Table 58 indicate that the degree of objec- tive integration and direction of subjective identification are indepen- dent. The probability value of chi-square is of such a magnitude that whatever relationship is observed may have occurred by chance. The im- portance of this observation is crucial to the study since, on the basis of this finding, it is possible to take the position that integration and identification are completely separable. Or, in its null form, a hypothesis that there is no relationship between integration and direc- tion of identification cannot be dismissed. Such a finding is unexpected since it had been assumed that the behavior patterns of Lansing fringe residents would be consistent with their subjective evaluations. On the basis of the findings it is neces- sary to question this assumption. The social life of the Lansing fringe area seems to be of the following pattern: those residents who are in- tegrated, i.e., who interact in the area, are not positively identified with the area (of course the converse to this statement also holds). Lansing fringe residents thus tend to exhibit discrepant adjustment pat- terns if these patterns are viewed within the framework of integration and identification. W and content pf identificatigg The chi-square computation presented in Table 59 indicates that inte8'1‘ation and the content of identification are not independent, i.e. , the ditllensions are significantly associated. When coupled with the non- s:I'gn‘ifzimant relationship between direction and integration (see Table 58): this suggests that identification is a complex phenomenon, and h e193 to explain the inability to treat this variable as a unidimensional .An.:v .An.mv .H «cowpepsnaoo chance Iago on» you mzoaaom we dmawnaoo who: weak» endow :OfipmHMoan obwpoonno N ”sovmonu no momHMon on. A a A on. ... :34 .. Nx mod n: «.mm mm NH m.:m fleece ma m.~ m.n n a m coauomuae o>aawmom . Icoz 0/ AU nafl m.m: mm om ea m.mm monsoonae . o>apamom Haves m e m N H Anmfimv Asogv omhy madam nowumnmoadH o>aoomnno onadloHezmnH MbHaomhmDm mo ZOHBUMMHQ 924 20HBaeoonpsm -70- TABLE59 RELATIONSHIP BETWEEI OBJECTIVE INTERATION AND CONTENT OF SUBJECTIVE IDENTIFICATION Objective Integration Scale Type Subjective (1“) (High) Identification l 2 5 4 5 Total Ecological 6.5(+)* 1 (-) 2 (+) 2 (-) u.5(+) 16 Economic 2 (-) l.5(+) -(-) 1 (-) 2 (+) 6.5 Ecological- Economic .5(-) .5(-) 2 (+) .5(-) 2.5(+) 6 Social 3 H .5(-) .5(-) 2 <-) 5 (+) 11 Social- Ecological .5(-) 1 (+) 2.5(+) u (+) 2.5(-) 10.5 Social- Economic .5(-) - (-) l.5(+) l (+) l (+) 1+ Social- Personal 1 (-> .5(-) - (-) 3.5M 3.5M 0.5 Personal 51+ (+) 6.5(+) 6 (-) 12 (-) 7.5(-) 66 Personal- Ecoloaoax 1+ (+) 1 (+) -<-> 1 (-) 1 <-) 7 Coerced th5(-) 1 (-) 3.5(+) 505(-) 6-5(*) 19 Other Eco- logical or Economic combinations 1 (-) 1.5(+) 2 (+) 3 (+) 5 (+) 10.5 filler 1 H 1 (+) -(-) 2.5(+) 3.50) 8 Total 58 . 5 16 20 56 (+2.5 17} 2 .. X - 25.195 ~ .01 > p > .001 c - A26 L - .079 D""3'1‘388 of freedom: 9 Objective Integration Scale types were combined as follows for the chi- square computation: l, (2.3), 1+: 5° Subj ective Identification types were combined as follows for the chi- squaare computation: (Ecological, Economic, Ecological-Economic, Other gombinations with Ecological or Economic), (Social, Social-Economic (ocial—Ecological, Social-Personal), (Personal, Personal-Ecologicals, c>erced, Other). E)this and all subsequent cross-classifications in which they appear, 1:10 and (-) indicate the direction of deviation from chance expecta- no -71.. phenomenon.n The sharp differences exhibited by an analysis of identi- fication as it is related to integration point to the advisability of isolating more than one mode of identification level. Besides indicating whether the degree of integration and content of identification are significantly related, the chi-square analysis in- dicates the pattern of any such relationship. The pattern is indicated by the series of (+) and (-) signs in the table, which signify for each cell the direction of deviation from chance expectation. The patterns of deviation from expectation seem unclear at first glance but a closer examination indicates that near-patterns do exist. The clear-cut pat- terns are as follows: the social, social-economic, social-personal, and some of the other combinations with ecological and economic categories exhibit a consistent pattern of an under-representation in the "low" integration scale categories and an over-representation in the "high" integration scale types; the personal and ecological-personal content categories are consistent in the opposite direction. All other identi- fication categories, with the exception of the ecological and social- ecological, possess one discrepant case which, if removed, would indi- cate under-representation in the "low" integration scale type in all cases. The ecological content category is unique since the inconsisten- 03' of pattern is clear; the social-ecological identification category is “1221113 since it exhibits an under-representation in each of the polar Bria-en” of integration scale types, with an over-representation in the 11h:- 98 middle categories. Looking at extreme scale types only, the ecological, personal, and ecological-personal identification categories are over-represented \ _L “See the scalogram analysis in Appendix C. .. 72 .. in the "low" integration scale types while the social-ecological, per- sonal, and ecological-personal are under-represented in the "high" inte- gration types. All other content categories are of an opposite nature. Considering the patterns as a whole, only the ecological category con- tributes as many as two deviant cases. Generalizing from the data, this means that Lansing fringe resi- dents who exhibit a high degree of within-fringe interaction tend to al- so be identified for friendship, location-economic, friendship-economic, friendship—personal, and coerced reasons. On the other hand, respondents who conduct few activities within the area are inclined to possess ori- entations based on personal reasons such as the perceived freedom in the area; the fresh air, peacefulness, and quietness in the area; the bene- fits for children; and for various anti-city attitudes. This is a some- what confusing array of relationships from which patterns are difficult to derive. The complexity of the situation is evident. At this point the situation may be summarized best by stating that the differential contents of identification exhbitied by Lansing fringe residents are Significantly associated with their behavior patterns. The contingency of coefficient between integration and content or identification is .426. As was indicated previously, this coefficient can be utilized to estimate the proportion of variance in a dependent Variable accounted for by variations in a related independent variable. This is accomplished by squaring the '5 value. Thus, since (A26)2 - .181, almost one-fifth of the variance in one of these variables is ac- counted for by variations in the other. Since integration and identification were Judged to be of equal importance, i.e., since neither could be accepted as an independent -73.. variable, L was calculated rather than LB or Lb" As indicated in Table 59, lambda is .079. This means that knowledge of the degree of integra- tion and content of identification provides for a mutual prediction im— provement of only eight percent. Intemtion and depth of identification A chi-square test performed on the data in Table 60 indicates that degree of objective integration and depth of subjective identifica- tion are not independent, i.e., there is a significant association be- tween objective behavior patterns and respondents' stated reasons for staying in, and moving to, the fringe area. In this respect depth is similar to content, and different from direction, of identification. The patterns exhibited by the data are quite clear: respondents who had no prior choice about coming to the fringe area contribute heav- ily to the ”high" side of the integration scale; respondents who moved to the fringe area for reasons that differ from the reasons they are staying, and respondents whose staying’and moving reascne are consistent, contribute heavily to the "low" integration side of the scale. In all cases the patterns are clear. In other words, those respondents who tend to participate in the fringe did not have a choice for their resi- dence location; those who do not participate in the area have moved and are staying for the same reason, or for discrepant reasons. At this point the meaning of this relationship is unclear, since the variables that account for the depth patterns have not been analyzed. HOwever, it is possible to conjecture on the meaning of the situation. Since respondents who were born and raised in the Lansing fringe area were assigned to the no prior choice category, it may be Mpothesized that l) long-time residents interact within the area to a greater degree - 74 - em a: .AnONV “H «cowpcvsnaoc onesvmnano on» you msoaaom mm escapees one: mommy oadom cowvdnmoanH o>wpoonno w .aouoonm no wconmen mac. n a awe. n_m. Hoe. A a A Ho. 2 me~.H~ a «x men m.~¢ em on a” m.mm Hence 5 3 a 33 39m 3?. 392 was: ad Snowman" oasm coauccncaccceH m.mm A-v an Anvm.ea A-vm.~ A+v on A+Vm.an ccccccc ccnpccncsm scape m.em A+vm.ma A+v Ha a+v a Auvm.n A-vm.a coacac hogan on Hence n a n m H Afisv cad snag madam conceamcpem c>aaccnno onedoHMHBzmQH HbHaomhmDm mo mammn and 20H9 p > .01--moderately significant .01 > p > .001-~highly significant .001 > p—-extremely significant See Margaret J. Hagood and Daniel 0. Price, Statistics f9; Sociologists (New York: Henry Holt and Co., 1952), p. 525. -86- expected direction: fringe residents whose last residential area was rural are under-represented in Scale Type 1 (the "lowest" integration type) and are over-represented in the remaining four types while those respondents whose move was made from a city exhibit an opposite pattern. Neither degree of association measure is as impressive as those computed for the specific place of previous residence. The coefficient of contingency value computed for the rural-urban previous residence and integration relationship points out that less than twenty percent of the scale variation is explained (52 - .173) while the predictive gain for scale types is less than ten percent (Lb - .081). The findings up to this point have shown that previous residence is associated with integration differentials. These results suggest that the nature of the area in which respondents were raised would also be significantly associated with the integration dimension. The data confirm the hypothesis: integration and area in which fringe residents were raised are significantly associated. The chi-square value is ex- tremely significant since it is beyond the .001 level. (See Table 64.) An examination of the patterns of association yields the following in- formation: those respondents who were raised in the area and those who were brought up in the study counties contribute more than chance expec- tation to the "high" end of the scale while all other respondents are over-represented in the "low" end. One-quarter of the variance in scale types is accounted for by variations in place raised (52 - .252). Knowledge of the place raised provides a seventeen percent improvement in the prediction of integra- tion scale types. Again, the data suggest that rural and urban upbringing is -37- significantly associated with integration differentials. The data sup- port this hvpothesis (see Table 6‘!) and are consistent with the findings regarding place raised. The association is extremely significant and the patterns of association are clear-cut: those respondents who were raised in a rural area are concentrated in the three "highest" integra- tion scale types; respondents who were raised in an urban area exhibit the converse pattern. Thus, rural upbringing is associated with "high" objective integration while urban background is associated with "low" integration. Well over one-hundred percent of the variance in objective in- tepation is explained by the residential history variables utilized in the analysis. In addition, the lambda values indicate that, in total, prediction of integration by taking all the variables into account is improved almost to perfection. Obviously, some of the specific variables must be highly interrelated, suggesting the existence of some more gen- eral processual factors in previous residential experience that affect the degree of objective integration. Therefore, the seven specific var- iables were examined in an attempt to find clusters of relationships and to determine how to best construct from them indexes for which the clus- ters stand. This examination suggested that three general themes are present in the residential history of sample members. These are: stability- mhility, rural and urban experience, and Lansing experience. It was deczi-<3.ed to construct indexes from the variables which would reflect these clusters. A stability-mobility index was constructed by assigning weiehts on the basis of l) lifetime and non-lifetime residence in the I"Sinai—mg fringe and 2) the number of residences occupied since 1940. A -88.. rural-urban experience index was constructed on the basis of 1) whether respondents moved to the Lansing fringe from a rural or urban area and 2) whether respondents were raised in a rural or urban area. A Lansing experience index was constructed by assigning weights on the basis of 1) whether respondents moved to the fringe area from Lansing and 2) wheth- er respondents were raised in Lansing. Obviously, the rural-urban and Lansing experience indexes are interrelated; therefore, a fourth index, composed of both, was also constructed.3 All associations between the indexes and objective integration were found to be significant. (See Table 65.) High integration is as- sociated with stability, rural background, and lack of Lansing experience. Conversely, low integration is associated with mobility, urban background, and experiences with the central city which dominates the area. This means that stable, rural, non-Lansing people tend to participate in the Lansing fringe area to a significantly higher degree than residents who exhibit converse characteristics. These findings may be interpreted as meaning that those residents who are moving into the Lansing fringe are not becoming integrated in the area. Residents who are "oldtimers" by fringe area standards are the area participants. The non-integrated respondents probably partici- Pate in the Lansing urban area, an area with which they are already fam- 11ia.r. Thus, it is the recent migrants who are in, but not of, the area. Looking at the extremes only, the Lansing fringe seems to be populated by two groups of inhabitants: one group has had relatively stable and rural experiences; the other has had relatively mobile and urban experi- ences. The former group is relatively integrated in the community; the \ 5 The index distributions will be found in Appendix B. -89- latter goup exhibits a lack of integration. Thus, the Lansing fringe area seems to be populated by one group which participates in the area as an on-going community, while the other group is living in a "social hiatus" or an "institutional desert." TABLE65 SWARY OF THE RELATIONSHIP BETWEEN OBJECTIVE INTEGRATION AND INDEXES OF RESIDENTIAL HISTORY Residential Degrees history 2 of _ index x freedom p 0 Lb Rural-urban experience 28. 818 8 . 001 .#51 . 121 Lana ing-non Lansing experience 26.506 8 .001 .455 .056 Stability-mobility 1+2. 052 8 . 001 . 5 51 . 115 Rural-urban , Lens ing- non Lansing experience 57. 716 12 .001 .‘+75 . 101 Speculation on an interesting question is now possible: as in- habitants tend to retain their area of residence by remaining in the fringe (thus exhibiting relative stability), and as they drift further away from their urban backgrounds, will these residents tend to become more integrated? This question may become meaningless if the fringe area is actually in transition to suburban or urban. If, however, it is an area that will stabilize as we currently know it, then the more re- cent fringe area residents will probably become more highly integrated. W 9;; W identification In this section the problem to be analyzed is: in what way do r esidential history variables influence positive and negative evaluations or the area? It will be useful once again to think in terms of the .. 90 .. long-time, relatively stable residents who are rural in background. This unique group may be expected to evaluate the area positively because of their residential experience. In a manner, it would seem that these res- idents are in a semi-coerced residential position since their experiences are quite limited. whereas recent mig'ants to the area are in a posi- tion to make comparisons, long-time fringe residents do not have this opportunity. Specifically, it may be expected that respondents who have re- sided in the fringe for many years would tend to be positively identi- fied with the area. This may be tested directly by utilizing two of the residential history variables: length of fringe residence and lifetime and non-lifetime fringe residence. However, when the relationship be- tween each of these variables and direction of identification was exam- ined, no significant association was found. (See Table 66.) Thus, the hypothesis that long-time fringe residence and direction of identifica- tion are significantly associated cannot be accepted. An indication of the spatial mobility of fringe dwellers is a- vailable from information concerning the number of homes occupied by the respondent since 19140. (See Table 66.) In view of the evidence pre- sented elsewhere} it may be hypothesized that residential stability con- tributes to direction of identification. However, statistical analyses revealed that there was no significant association between mobility ex- periences and direction of subjective identification with the Lansing fringe. Still another posited relationship between residential experience and direction of identification is that previous residence is associated \ 1... Smith, Form, and Stone, pp. 931;. - 91 - TABLE 66 SUMMARY OF THE RELATIONSHIP BETWEEN DIRECTION OF IDENTIFICATION AND RESIDENTIAL HISTORY Residential Degrees history 2 of variable X freedom p Length of residence 1.629 2 .25 Lifetime and nonslifetime * residence .22 number of residences occupied * since l9#0 .21 Place of previous residence 1.099 2 .55 Rural and urban nature of previous residence .215 1 .55 Area in which head of household was raised 5.027 2 .15 Rural and urban nature of area in which head of household was raised 1.25# 1 .15 _— ‘I’ The p value was computed by utilizing an exact probability test. The Chi-square test could not be used because theoretical frequencies were too low. -92.. with direction of identification. It was expected that residents who came to the fringe from Lansing or from city areas in general would lack positive identifications with the fringe area and that all others would exhibit the converse pattern. However, the data do not support this con- tention: neither place of previous residence nor rural-urban nature of previous residence are significantly associated with direction of sub- jective identification. (See Table 66.) In attempting to account for direction of identification differ- entials, the last general factor to be considered was the place and the type of area in which the head of the household was raised. The hypothe- ses considered are basically similar to those utilized in an attempt to discern the previous residence-integration relationship. Specifically, it was expected that residents who were raised within the general area (excluding Lansing), and those who were raised in a rural area would ex- hibit positive identification patterns. However, in both cases, the analyses indicated that the variables are not significantly associated with direction of identification. (See Table 66.) Thus, utilization of the area-in-which-raised categories to account for direction of identi- fication differentials was not successful. It is possible that, although the specific residential history Variables are not significantly associated with direction of identifica- tion, the derived groupings obtained by the index classifications are re- lated. Therefore, the more general processual factors reflected in the indexes were examined in the direction context. Once again, however, the tests indicated that direction of identification is not significant- 1y a3300::iated with the prior residential experiences of sample members. (See Table 67.) This means that mobility experiences, rural-urban - 93 - experiences, and Lansing experiences of sample members do not signifi- cantly influence positive and negative identifications. TABLE 67 SUMMARY OF THE RELATIONSHIP BETWEEN DIRECTION OF IDENTIFICATION AND INDEXES OF RESIDENTIAL HISTORY Residential Degrees history 2 of index X freedom p Rural-urban experience .985 l .25 Lansingbnon Lansing experience .01# l .#8 Stability-mobility .#25 2 .45 Rural-urban, Lansingbnon .392 2 .#S Lansing experience The consistent lack of association between direction and other factors is cause for concern since this aspect of identification has 5 been interpreted as crucial to the analysis of adjustment patterns. Two general interpretations of the situation are possible: 1) perhaps the direction variable itself is non-differentiating, or 2) perhaps the Positive and negative evaluations of the Lansing fringe area are not in- fluenced by the prior residential experiences of residents but, rather, by some other experiences. In terms of the possibility that the dimension itself is non- differentiating, the most obvious reason may be that the question uti- lized for the direction classification may have been misunderstood. The queation wording may have "begged" the answer since it was presented in \ 5See an ra, pp. 17ff. -94.. the following manner: "If m M mu; 9119323, would you continue liv- ing around here?" In reality, most residents do not have a choice of this nature and, therefore, they may not be able to think in these terms. It is also possible that respondents interpreted the question in terms of a comparison between residence in the Lansing fringe and residence in the city itself, i.e., sample members may have indicated a positive choice for the fringe area by default, in terms of anti-city bias. Another possibility remains: perhaps respondents were influenced by a felt pressure to reject a negative position after a commitment of fringe area residence had already been made. It is evident that the above speculation is as; m; like most speculation of this nature, the reasoning is in terms of what seems logical. This is not a completely satisfactory method since the same type of logic may be applied to almost any factor that did not vary ac- cording to expectations. A much more satisfactory approach is to address the question em- pirically, by examining experiences other than residential history to find whether relationships exist. It has been demonstrated that deci- sions to move are not completely based on satisfaction with one's neigh- bfibrhood.6 Among the factors involved such variables as satisfaction with 0118's job, wages, or health may be influential. Information about these 1' actors was not obtained in the present study. However, the relationship bfitween direction of identification and certain demographic factors can \ 6See Peter H. Rossi, m Families Move (Glencoe, Illinois: The FrQe Press, 1955). Rossi demonstrates that in addition to neighborhood 8‘EL'hisfaction, such factors as dwelling unit space: housing utilities; sisal environment; distance from friends, relatives, work, and church; 1 ‘3- housing costs must be taken into account in order to explain mobil- ty inclinations. -95.. be examined directly since information of this nature was obtained in the Lansing fringe study. Analyses of the relationship between demo- graphic factors and adjustment patterns will be undertaken in the next chapter. Content 9_f_ subjective identification Content is the second component of subjective identification to be considered. Again, differentials in this aspect of resident adjust- ment patterns will be examined in reference to residential experiences, with the object being to account for variance in content of identifica- tion. The specific relationships between content of identification and residential experience are summarized in Table 68. Before the empirical determination of the relationship between content and residential history is undertaken it may be illuminating to discuss the influences that prior residential experience might have on adjustment. It would seem logical to expect that long-time, relatively stable residents who are rural in background will exhibit identifications based upon being born in the area and upon friendship patterns and a sub- Jective knowledge of the area. In other words, it would be expected that residents who have spent their lives in the area would indicate 1”fiat their identifications are based upon their knowledge of the area and some of its inhabitants. On the other hand, relatively short-time, mobile residents who are urban in background would be expected to have chosen the area for reSidence on the basis of its assets for children, its lack of conges- tion, and its location relative to the central city. Their knowledge or the area and its inhabitants would not be a crucial factor in their content of identification. -96- TABLE 68 SUMMARY OF THE RELATIONSHIP BETWEEN CONTENT OF IDENTIFICATION AND RESIDENTIAL HISTORY Residential Degrees history 2 of __ variable X freedom p C Lb Length of residence 55.197 8 .001 .509 .094 Lifetime and non-life- time residence 57.792 4 .001 .562 .095 Number of residences occupied since 1940 55.555 12 .001 .485 .048 Place of previous residence 14.915 6 .05 .560 .005 Rural and urban nature of previous residence 40.275 6 .001 .562 .052 Area in which head of household was raised 15.499 8 .10 Rural and urban nature of area in which head of household was raised 26.557 6 .001 .470 .O .k - 97 - The content category into which residents who indicated that they would like to continue living in the area because of an investment of money or because of economic advantages (i.e., the economic content category) is unique since it would be expected that both extremes of residential experiences would contribute to such an adjustment pattern. Thus the relatively longbtime, stable, rural background residents would view the area as one in which their economic assets are committed while the relatively short-time, mobile, urban oriented residents would per- ceive the area as one in which an inexpensive home could be bought and an area in which the tax rate is low. Since there are numerous content and length of residence cate- gories (the original table is composed of ninety-nine cells) broader length of residence categories were constructed. It may be reasoned that longbtime fringe residence is associated with friendship-based and area familiarity identification (i.e., social and coerced content) while short-time fringe residence is associated with the other orientations. Computations indicated that a significant association between length of residence and content of identification does exist (see Table 68) with ‘the expected patterns present. By utilizing the deviation patterns in ‘the contingency table the following patterns become clear: Before 1940: social and coerced content Befbre 1946: social-ecological and social-economic content After 1959: ecological content After 1945: economic, personal, and social-personal content The contingency coefficient value of .509 means that more than ‘3llte-quarter of the variance in content of identification is accounted for by length of residence. The gain in prediction improvement, as meensured by lambda, is slightly under ten percent. This means that by tiflaring length of residence into account, prediction of the content of -93.. identification is only slightly improved. Generalizing from this relationship, it may be expected that lifetime and non-lifetime residence is significantly associated with the content of identification, social and coerced content types being related to the lifetime category. The hypothesis of a significant association between the variables is verified (see Table 68) but content based on friendship did not deviate in the expected direction. Specifically, lifetime residents in the Lansing fringe exhibit identifications based on perceived inability to move (coerced), economic investment and gain (economic), friendship and locational ties (social-ecological), and a combination of locational and economic orientations (ecological-economic). While the lambda value is fairly low (Lb - .095), special atten- tion should be given to the coefficient of contingency (C - .562). It indicates that almost one-third of the variance in content of identifi- cation is explained by variance between lifetime and non-lifetime resi- dence. This is a considerable improvement over utilizing length of res— idence for this purpose since it accounts for only one-quarter of the Variance. Spatial mobility was taken into account by examining the rela- tionship between the content categories and mobility since 1940. It was found that the number of residences occupied since 1940 is significantly aisaociated with content of identification. (See Table 68.) As was ex- Pected, all of the content categories that are associated with lifetime 1‘ Seidence (i.e., the coerced, economic, social-ecological, and ecologi- cal-economic categories) are associated with relative stability since 19%, the latter category being composed of residents who made no moves, or only one move during this period. This pattern was expected since -99- lifetime residents of the fringe have been in the area for a long period of time and tend to maintain their home. In addition to these content types those based on friendship patterns and personal reasons also ex- hibit relative stability. The thirty-seven residents who were born in the area were elimi- nated in the analysis of the relationship between place of previous res- idence and content of identification. The association is moderately sig- nificant. (See Table 68.) The patterns of relationship are difficult to determine since the table has almost one hundred cells, many of which contain less than one unit in both expected and observed frequencies: in the list of patterns which follows, only those cells in which either the observed or the expected frequency is higher than one are considered: Lansing or East Lansing: locational, personal, locational-personal, and friendship—locational content. Other Michigan city of 25,000 or more: personal content. Study counties: social, economic, coerced, friendship—locational, and locational-economic content. Contiguous counties: friendship and friendship-economic content. Southern state: personal content. Generalizing from these data, it was reasoned that fringe dwell- Ora' last place of residence, dichotomized into rural and urban place, "Ould also be significantly associated with content. In addition, the o"erwhelming preponderance of movers from Lansing and East Lansing led to the hypothesis that content types associated with urban previous res- idence would also be associated with previous residence in these two urban areas. The data support the hypothesis on both counts: the re- la-"-‘-:Lonship is extremely significant (see Table 68) and precisely those content categories that are associated with previous residence in Lan- sing, i.e., types which are based on the locational qualities of the fringe, personal reasons, personal-friendship orientations, and a person- ~100- al-locational combination, are related to previous residence in a city. A knowledge of both specific area of previous residence and rur- al-urban previous residence improves the prediction of content only slightly (Lb - .005 and .052 respectively). The coefficient of contin- gency, however, points to the superiority of utilizing the information provided by the rural-urban distribution: specific place of previous residence accounts for only thirteen percent of the content variance while rural-urban previous residence explains almost one-third of the content variance. It may also be expected that place of upbringing, both specifi- cally and in terms of rural and urban experience, would be associated with content. However, no significant association was revealed between specific upbringing place and content. (See Table 68.) Therefore, the hypothesis that specific place of upbringing is significantly associated With content cannot be accepted, while the hypothesis that rural or ur- ban upbringing is significantly related to content cannot be dismissed. Residents who were raised in an urban area exhibit contents based on the location of the area, qualities of the fringe, and a combination °f qualities and friendship patterns. These categories are consistent with the types of content exhibited by residents who migrated to the LeIlsing fringe from urban areas; the lone exception is that the personal- l°<>a.tional category is present in the "moved" category and is not pre- Bent in the "upbringing" category. However, the striking consistency seems to indicate that urban experience affects content of identifica- tj-On patterns. The lambda coefficient indicates that knowledge of whether re- 8pcridents are products of a rural or urban area provides no improvement ~101- in the prediction of content. The contingency coefficient, on the other hand, points out that more than twenty percent of the variation in con- tent is accounted for by rural-urban upbringing (CZ - .221). All the relationships between the residential history indexes and content of identification were found to be significant. (See Table 69.) This means that the general residential processual factors, as measured by the indexes, influence the type of content exhibited by sample members. Therefore, content differentials have been accounted for to some degree. TABLE 69 SUMMARY OF THE RELATIONSHIP BETWEEN CONTENT OF IDENTIFICATION AND INDEXES OF RESIDENTIAL HISTORY Residential Degrees history 2 of _ index X freedom p C Lb Ru‘l‘al-urban experience 46.451 6 .001 . 587 . 061 IJansing-non Lansing GXperi ence 17. 524 8 . O25 . 575 . 0 Stability-mobility 65.785 8 .001 .644 .077 R“rial-urban, Lansing- 30:: Lansing experience 41. 965 10 .001 .537 .056 Prior residential experience influences the content categories in the following manner: Respondents who have experienced relative stability exhibit contents based on a perceived inability to move, the economic advantages of the area, a combination of economic and locational area qualities, and a combination of locational factors and patterns of friendship. Respondents who have experienced relative mobility exhibit contents based on qualities of the area, both alone and in conjunction with social and locational factors. Respondents who have had a large degree of rural and non-Lansing ex- perience tend to possess positive identification for friendship, economic, and coerced reasons, with social, ecological, ecological- -lO2 - economic, and social-economic contents also present. Respondents who have had a large degree of urban and Lansing exper- ience exhibit contents based on the location and qualities of the fringe area and on friendship-personal and locational-personal fac- tors. Special attention should be given to the stability-mobility in- dex since it explains approximately forty percent of the variance in content differentials (Ce - .415). Such a high value suggests that it has been possible to isolate a crucial factor influencing this one as- pect of identification. The pattern of relationship is not strictly as expected since the social content type (expected to be related to sta- bility) and the ecological content type (expected to be related to me- bility) exhibit inconsistent deviations from expectation, both contri- buting a surplus to the middle category of the stability-mobility index. In addition, the economic content type is associated with stability. The reasons for the failure of these content types to deviate in the ex- Pected directions remains unexplained after careful scrutiny of the data. However, what is important here is to point out that a higily significant relationship has been found, with stability and mobility 8"Sl‘ongly influencing the content of Lansing fringe dwellers' identifica- tions. Relatively stable respondents identify with the fringe area ha~81cally in terms of their knowledge of the area and their economic contaitments; relatively mobile respondents identify basically in terms or personal reasons, such as the stated assets for children, the quiet- ness, peacefulness, and lack of congestion in the area. km 9_f_ identification The final component of identification to be considered in this analysis is depth. An attempt will be made to determine the nature of -lO3- the relationship between the three depth categories (i.e., had no prior choice for fringe residence, staying in the fringe for the same reasons as moved, and staying in the area for reasons that differ from the ini- tial move reasons) and differential residential experiences. The exis— tence and depee of relationships are summarized in Table 70. The most obvious general relationship to be expected is that long-time, relatively stable residents who are rural in background will exhibit the no prior choice depth type. The reasons for this expecta- tion are apparent: residents who have been in the Lansing fringe for many years were in the area when it was still rural in nature. In addi- tion, these residents are not apt to move very frequently. Basically, the essential dichotomy seems to be between the "no prior choice" depth category on the one hand, and the "same reasons" and "other reasons" categories on the other. It is expected that these lat- ter two depth types will vary in the same direction, both in opposition to the "no prior choice" pattern. After much deliberation it was not Possible to arrive at adequate differentiating characteristics for the "Same reasons" and "other reasons" variables. Therefore, the empirical analysis will provide an understanding of the differentials involved. As in the case of the other two identification components, it may be expected that length of residence is significantly related to depth. In terms of the direction of the association, it may be reasoned that those who had no choice for fringe area residence have been in the area for many years. The data support the hypothesis of a significant asBociation (see Table 70) and verify the expected pattern: residents who had no prior choice were in the area before 1940, residents who in- dicated that their reasons for staying and moving do not differ came -104- TABLE 70 SUMMARY OF THE RELATIONSHIP BETWEEN DEPTH OF IDENTIFICATION AND RESIDme HISTORY Residential Degrees history 2 of ._ variable X freedom p C Lb Length of residence 45.122 8 .001 .552 .225 Lifetime and non-life- * time residence .912 1 .50 Number of residences occupied since 1940 70.258 6 .001 .691 .256 Place of previous r881dence 38 e 954 l} e 001 eu99 e 11 3 Rural and urban nature of previous residence 55.555 2 .001 .719 .202 Atrea in which head of * household was raised 6 . 909 4 . 20 Rural and urban nature of area in which head of household was raised 42.425 2 .001 .648 .107 * {The "no prior choice" depth of identification category was eliminated 111 'the chi-square computation because those respondents who were born and. brought up in the area were coded as lifetime residents who were raised in the area; this would artificially result in a significant chi- aquare value. .. 105 - after 1959, and residents who responded with different reasons for mov- ing and staying came to the area after 1945. The lack of a significant association between the lifetime and non-lifetime distribution and depth (see Table 70) is largely a function of the fact that the "no prior choice" category was excluded from the analysis since those residents who were born in the area (i.e., many of the lifetime residents) had been classified in this manner. Thus, the lack of a significant association means that neither group of residents who had a choice for residence in the area is influenced by whether or not they are lifetime residents. As expected, a significant association was found between real- dential stability, as measured by the number of homes occupied since 1940, and depth. (See Table 70.) Overwhelmingly, those respondents who had occupied only one home since 1940 are not in the area by choice; con- versely, those respondents who had a choice in coming to the fringe area have lived in more than one home since 1940. Almost half of the variance between the variables in the cross- classification is accounted for ('52 e .477); this is one of the highest contingency coefficient values obtained in the analysis of the relation- 8hip between residential history and integration-identification. How- eVar, a knowledge of the number of homes occupied since 1940 leads to only a twenty-four percent improvement in the prediction of depth. It may be posited that previous residence and depth are signifi- c’antly associated, with the local area and rural residence composed of regidents who had no prior choice for fringe residence. The data here °°nfirm the expected relationship. (See Table 70.) The pattern how- e"er, possesses one unexpected relationship: residents who moved to the ~106- Lansing fringe from contiguous counties provide more than chance expec- tation to the category composed of residents who came and are staying for different reasons and are under-represented in the other two depth types. Residents who did not have prior choice for fringe residence came to the area from other parts of the study counties. The coefficient of contingency value is the highest of all those obtained in the analysis of the residential history-identification re- lationship. A coefficient value of .719 means that more than one-half of the variation in depth is explained by whether residents moved to the Lansing fringe from a rural or an urban area (C2 - .517). The lambda value indicates only a twenty percent improvement in the predic- tion of depth types from a knowledge of migration from a rural or urban area. No significant relationship exists between place-raised and depth. (See Table 70.) As was the case with content of identification, the specific place of upbringing is not significantly associated with depth, although rural and urban background is related. The evidence gathered pointed overwhelmingly to the presence of 8- more than chance representation of residents who had been brought up in a rural area in the category composed of residents who had no prior choice for the area. The coefficient of contingency value is fairly h18h: forty-two percent of the variance in depth types is accounted for. A knowledge of the upbringing distribution leads to an eleven percent 1“11>:rovement in the prediction of depth. It was consistently found that the basic differentiation among the depth categories lies not among the three aspects of depth, but in terms of a division between the "no prior choice" category on the one -lO7- hand and the remaining two categories on the other. This relationship was affected by the fact that long-time residents were generally coded in the no prior choice category while they also tended to be classified lifetime, stable, rural background inhabitants. Therefore, it was de- cided to examine the relationships between the residential history in- dexes and all three categories and, in addition, to examine the depth- indexes relationships with those residents who had no prior choice for Lansing fringe residence eliminated. This approach helped immensely in clarifying the relationship between depth and residential history. With the category remaining, an extremely high degree of relationship was found (mobility-stability .62 - .657; rural-urban, Lansing-non Lansing experience CZ - .591); with this group eliminated, significant associations between the variables are lacking. (See Table 71,) Therefore, the disproportionate influence of residents who had no choice for fringe area residence has been demon- strated. Thus, it has been ascertained that prior residential experience has influenced respondents who had no choice in fringe residence but has not significantly affected respondents who moved to the Lansing fringe area by choice. Relatively stable, non-urban residents have been in the fringe area for many years while relatively mobile, urban residents have re<>ent1y made a choice for fringe residence. Whether respondents have ”Oved and are staying for the same or different reasons does not seem to be influenced by their prior residential experiences. The most crucial residential history variable affecting depth or identification seems to be stability in the area itself. It is quite possible that as inhabitants remain in the area for a longer period of ~108- TABLE71 SUMMARY OF THE RELATIONSHIP BETWEEN DEPTH OF IDENTIFICATION AND IND“ OF RESIDENTIAL HISTORY Residential Degrees history 2 of _ index X freedom p 0 Lb with "no prior choice" category included Rural—urban experience 58 . 509 4 . OOl . 682 . 214 Lens ing-non Lansing experience 24. 879 4 .001 .481 .0 Rural-urban , Lans ing- non Lansing experience 59 . 519 8 . 001 . 625 . 201 with "no prior choice” categogy eliminated Rural-urban experi ence 2 . 556 2 . 50 Lane ing-non Lansing experience 5. 782 2 . 20 Stability-mobility . 192 2 . 95 Rural-urban, Lansing- ncn Lansing experience 6.168 4 . 20 — 109 - time this aspect of their adjustment patterns wi1l be affected to a larg- er degree. Perhaps at this later period it will be possible to isolate the factors that have contributed to the types of depth patterns that they exhibit. Summary: Accounting for Adjustment Patterns In this chapter an attempt was made to utilize residential his- tory variables to account for the patterns of objective integration and the three subjective identification categories. This section will be devoted to a summary of what was determined in the attempt. a) Objective integration The analysis has indicated that a high degee of objective inte- gration is associated with the following variables: residence in the fringe area prior to 1942; life residence; migration from areas other than Lansing, East Lansing, and other Michigan cities of over 25,000 pop- ulation; migration from a rural area; less than three homes occupied since 1940; upbringing in the area itself or in the study counties (ex- cept in Lansing or East Lansing); and rural raising. Three main trends seem to be implicit in the above listing: long-time residence in the area, residential stability, and non-urban in be"~3k81‘ound and recent experience. Thus, in accounting for patterns of Ob.‘l<=4‘<:tive integration, it has been found that a high depee of within- area participation is associated with long-time residence, infrequent moving, and rural background. In view of these findings, four indexes of residential history were constructed. As expected, the indexes indicated that relatively stable respondents who have had rural, non Lansing backgrounds tend to -llO- participate in the fringe area to a higher degree than sample members who exhibit converse characteristics. b) Direction of identification Attempts to account for direction of identification patterns have met with failure. It has been found that direction of identifica- tion is not significantly associated with any aspect of residential his— tory. The analysis suggests that the direction category has little util- ity in any determination of fringe area adjustment patterns. c) Content of identification Content of identification is significantly associated with some of the residential history variables. The significant associations found between content and residential history are as follows: year moved, life or non-life residence, place moved from, rural or urban pre- vious residence, moving experience since 1940, and rural or urban up- bringing. The indexes of residential history are significantly associated With content of identification. Relative stability is associated with content types based on the economic assets of the area, initially co- erced residence, a combination of locational qualities and friendship Patterns, and a locational-economic evaluation; relative mobility is re- lated to contents based on personal preferences for the qualities of the ”98-. this same preference combined with a desire to maintain area frieruiships, and personal preference combined with locational qualities Of the area. Rural, non Lansing experience is related to an emphasis on area friendships, perceived economic qualities of the area, initially coerced residence, combinations of locational qualities with friendship - lll - and economic reasons, and a combination of personal reasons with loca- tional factors. d) Depth of identification Depth of identification is significantly associated with five of the residential history variables: year moved to the area, place moved from, rural or urban nature of previous residence, moving experience since 1940, and rural or urban upbringing. In attempting to account for the depth of identification sub-categories the following associational patterns have been found: 1. No prior choice pattern: these respondents generally moved to the area before 1940 from a rural area. The specific areas from which they came are the study counties, other Michigan areas of less than 25,000 population, and other Midwestern states. They have lived in the same home since 19% and were raised in a rural area. 2. Same reason as staying pattern: these respondents moved to the area after 1959 (except for l946-l9‘+8) from Lansing, East Lansing, and Midwestern states. These residents have lived in more than one home since 191:0 and were generally raised in an urban area. 3. Other reasons pattern: these residents moved to the area after 19l+2 and in the latter half of the depression from an urban area. They have lived in more than one home since 19% and were raised in an urban area. The indexes of residential history and depth were found to be Sig"1~"-f.’:.cantly related. However, a closer examination of the relation~ ship revealed that the "no prior choice" pattern was contributing dis- proportionately to the significance of the association. When this - 112 - pattern of depth was eliminated, it was found that there is a lack of significant association between the depth-index variables. Thus it has been possible to account for the "no prior choice" pattern and not for the "same reason" and "other reasons" patterns. Residents who are not in the area by choice are relatively stable and of rural, non Lansing backgrounds e CHAPTER V CHARACTERISTICS OF THE POPULATION As in Chapter IV, the object of this chapter is to analyze the relationship between a group of social variables and the integration and identification adjustment patterns. Attention will be focused on the following variables: occupation factors, tier of residence, tenure sta- tus, age, education, children in respondents' families and family gross weekly income. Akin to the analysis in the previous chapter, the data may be utilized to explain the differentials of integration and identification. This chapter and the previous one are thus complimentary; methodological 1}" this chapter will proceed in precisely the same manner as did Chap- ter IV. The basic problem is to determine the influence that demograph- ie factors have upon the nature of adjustment patterns. As in the pre- vious chapter, each of the four adjustment variables will be examined independently to determine the specific influences of population charac— teristics on these dimensions. Accounting for Integration and Identification Ob e<-"1=:i.ve intemtion Looking at the occupation factors first, it would seem that farm- er 3 Should exhibit high integration. This expectation is based on the 3 up pQBition that farming is an occupation that tends to root people to -113 - -114- their residential area, thus orienting them away from the most obvious alternative in the fringe area--the central city. Directly related to this expectation, one may also expect that respondents who do not work in Lansing would also exhibit a relatively high degree of integration. It may also be expected that owners have committed themselves to the area to a higher degree than renters. It would seem that this com- mitment would influence behavior patterns by providing residents with a feeling of "attachment" for the area. Therefore, owners would tend to manifest a relatively higher degree of integration than renters. Since the elderly are more apt to be compelled to remain in their present line of work, it would be expected that they would tend to per- ceive their present residential area as more-or-less permanent. This attitude of permanence may influence behavior patterns by leading to a relatively high degree of integration. In a way, respondents with school age children are also in a Bani-coerced position since their predisposition to mobility would tend to be minimal. The anticipated readjustments incumbent upon moving a Child from one school to another may influence parents to perceive their Present area of residence as semi-permanent. In addition, the various specialized organizations which have been created for the benefit of chi 1dren and their parents may tend to influence this grouping toward relatively high degrees of integration. Tier of residence, education, and income are factors that may influence integration but the direction of influence cannot be antici- Dated at the present time. Rather than taking a position at this point, it was decided to see what the data contribute to an understanding of Int eB'ration patterns . -115- Since farmers have had longer experience in the fringe area it may be reasoned that they would exhibit "high" integration relative to nonfarmers, i.e., farmer behavior patterns would tend to be performed in the fringe area in comparison to nonfarmers who, very possibly, are in— tegrated in the urban area where they are employed. Under these condi- tions, it is hypothesized that the analysis of the relationship between occupation and objective integration will indicate a significant associ- ation. The data support this hypothesis (see Table 72) and the expected pattern of relationship does exist: farmers are clearly under-represen- ted in the two "lowest" integration scale types and contribute more than chance expectation to the three remaining types while nonfamers are over-represented in Scale Type 1 (the "lowest") only. The twenty-three part-time farmers follow a pattern similar to the full-time farmer re- spondents. The degree of association measures are not high: the coefficient 01‘ contingency squared indicates that occupation accounts for approxi- miii-tely one-quarter of the objective integration variance (7:.2 . .272), while lambda shows only a fifteen percent improvement in the prediction 01‘ scale types. Since the basic distinguishing factor utilized in the analysis °f Occupation was found to be the farmer, part-time farmer, and nonfarm- er trichotomy, it was reasoned that farm experience m _s_e_ may be re- lated to objective integration. Therefore, the 121 nonfarmers were clasSified into two groups composed of those who have had farm experi- ence and those who have never worked on a farm. These groupings were found to possess almost equal distributions: slightly over one-half have worked on a farm and slightly less than one-half have not. However. - 116 - TABLE 72 SUMMARY OF THE RELATIONSHIP BETWEEN OBJECTIVE INTEGRATION AND CHARACTERISTICS OF THE POPULATION Degrees Population 2 of _ characteristics x freedom p C Lb Occupation factors a; occupation #7.s92 12 .001 .522 .149 b farm work exper- ience 7.566 # .20 a) place of work 28.580 A .001 .481 .10# Tier of residence 28.298 12 .Ol .hl9 .O6# Tenure status 7. 093 2 . 05 . 277 . 0 Jugs of respondents 19.359 12 .10 Ekiucation of respondents 9 . 267 9 . 50 (zliildren £1 number #.04# 9 .95 13 number under 5 2.920 3 .50 0 number 5-17 4.978 8 .80 Family goes weekly income 18 . 869 12 . 10 * -117- when the test of association was made, no significant association was evident. (See Table 72.) Another lead was provided by the analysis of occupation: since most nonfarmers work in Lansing, it may be expected that a significant association would be found between integration and whether respondents work in Lansing. The data support this contention: the association is significant. (See Table 72.) Looking at the pattern of association, the data point overwhelmingly to the fact that Lansing workers do not participate in the fringe area while non-Lansing workers exhibit the con- verse pattern. Ecologists, in their analyses of the space-time-cost trinity, have pointed to the significance of location in reference to the city (senter.1 Martin deals with this factor in a study by formulating the fol- lowing null hypothesis: "There is no significant relationship between distance from the city center according to one-mile concentric zones and degree of satisfaction with residence location in the rural-urban fringe."2 However, in his empirical study Martin found that he could not reject this null hypothesis.3 Nevertheless, it was decided to test a similar hypothesis: limitations of the data did not allow for utiliz- ing- concentric zones; rather, respondents were classified according to \ 1See, e. ., James A. Quinn, Human Ecology (New York: Prentice- Hall. Inc., 1950 , esp. pp. 285-288. zwalter '1'. Martin, The Rural—Urban Frye: A Study o_f Adjust- .1‘;& 39 Residence Location (Eugene, Oregon: University of Oregon Press, 53) , p. 21. It is important to point out that Martin's study was not nudeI‘taken in an area that meets the rural-urban fringe criteria of def- B tZion utilized in this dissertation; see Richard A. Kurtz and Joanne Eieher, "Fringe and Suburb: A Confusion of Concepts," Social Forces, 37 (October, 1958), pp. 52-37. 3Martin, gp. _c__i_1_:., pp. tBf. -118- tier distance from the central city. (See Appendix B, Table 17.) It was found that the association between integration and tier of residence is significant, but the pattern of relationship is somewhat confusing: tiers one and two (the closest-to-Lansing tiers) have a surplus of re- spondents in the "lowest" scale type, while tier four has a preponder- ance in the two "highest" scale types; these three tiers are the only ones which exhibit clear-cut patterns. Tiers three, five, and six ex- hibit almost a complete lack of consistency. It may be posited that tenure status is significantly associated with objective integration, with renters being the non-participators. Previous evidence seems to support this position: old-time residents who are farmers are integrated, and this group is in the "owner," rather than "renter," category. The data confirm the contention: the associ— 8'tion is significant (see Table 72) and there is a surplus of renters in the lowest integration type. Owners contribute less than chance expec- ta-1=:I.on to the scale type in which non-participators are located and high- er than chance frequencies to all other parts of the scale. Both measures of the degree of association are quite low: the coefi‘icient of contingency is .277 and the lambda coefficient is zero. This means that tenure status accounts for only eight percent of the val‘18.nce in integration (52 = .077), and that a knowledge of tenure sta- tug provides no improvement in the prediction of the scale distribution. No significant association was evident between the age distribu- tion of heads of households and objective integration. (See Table 72.) It was reasoned that since old-time fanners populate the older age groups, “Ge itself would be a significant differentiating variable which could be used to explain integration. However, no relationship can be discerned -119- between age categories and integration. After closer scrutiny of the data, the lack of a significant relationship remains unexplained. The relationship between the schooling of respondents and inte- gration was also found to be non-significant. This finding is not com- pletely unexpected since the lack of a significant association between age and integration had already been found. It may also be expected that the presence of children in a fami- ly would have an effect on integration patterns. This influence may op— erate in several ways, the most important factors being that respondents with children would be more limited in opportunities as far as time is concerned, but would have more opportunities to come into contact with other families and with organizations that are designed for children and their parents. However, the data do not support the contention that the presence of children is significantly associated with integration. (See Table 72.) In an attempt to clarify the relationship between the presence of children and integration, it was decided to refine the analysis by differentiating between families that have children under five years of age and families in which children between the ages of five and seven- teen are present. The justification for dichotomizing respondents into these groups is the different types of experience involved: children under five years of age are usually pre-school and frequently demand nor e parental supervision; children of five years of age and over are lls‘latlly in school and have fewer demands. In addition, many formal o:K‘ga-Itizations, such as the P.T.A., are designed for families with school- age Children. Therefore, it may be reasoned that the refined groupings would be significantly associated with integration. However, the data -120- do not support this contention: neither the presence of children under five nor the presence of children between five and seventeen years of age are significantly associated with objective integration. (See Table 72.) Thus, integration is not significantly associated with the presence of children in any way. The last factor to be considered in this section is family in- come: it may be expected that this variable is significantly associated with integration. However, no significant relationship was evident be- tween the variables. This finding may have been anticipated since edu- cation which, no doubt, is related to income, also revealed a non-sig- nificant association. The paucity of significant relationships between population char- acteristics and integration brought into question the advisability of treating the demographic factors as independent, unrelated units. It was reasoned that perhaps the individual population characteristics are interrelated as some more general processual factor which cannot be de- tected in an atomistic approach. In other words, the focus of analysis was shifted to a more general level, with the question of what each of the demographic factors reflect now receiving attention. After some deliberation it was decided to view several of the 8Pettific demographic variables in the general context of commitment to the area of residence. This commitment is a manifestation of the fact that people have investments inan area which tend to keep them from antici- pating a change of residential area. In other words, it was reasoned that relative degree of investment in the area would tend to influence the nature of adjustment patterns. In general it can be expected that Lansing fringe residents who manifest a relatively high degree of -121... investment in the area would exhibit a low predisposition to mobility, with these factors acting as a causative element in relatively high par- ticipation. It was decided to construct an index which would reflect degree of investment and to determine the nature of the association between this index and the adjustment dimensions. In this manner adjustment patterns may be accounted for and explained. The commitment index was constructed by taking account of the following variables: a) occupation, b) tenure status, 0) age, and d) the Presence of children between five and seventeen years of age. Each of these index components was dichotomized on the basis of whether or not it reflects an investment in residence location. The specific interpre- tations are as follows: a) Occupation: an examination of the occupa- tion distribution (see Appendix 13, Table 15) indicated that two of the Occupations may be viewed as rooting individuals to the area. These oc- cuPations, which are assigned to the high investment end of the index, 31' e ”full-time farming" and "own business." b) Tenure status: owning, as Opposed to renting, one's home is interpreted as a reflection of an 1nVestment in residence location. c) Age: this characteristic was in- cluCied in the commitment index since the older age groups are placed in a. "1‘ orced investment" because of the difficulty faced by these people in finCling new jobs which may necessitate new residences. Therefore, the older age groups tend to be rooted in their area of residence. It was arbitrarily decided to assign respondents fifty years of age and older to 13he high investment end of the index. d) Presence of children be- tween five and seventeen years of age: this last variable was included In the index because of the coercion incumbent upon the presence of -122- school-age children in a family. Such families may be affected by the fact that a move frequently necessitates readjustments on the part of children since it means a new school and new friends. Respondents with children of school age were assigned to the high comitment end of the index. Dy weighting each response in the manner indicated, the commit- ment index, ranging from O to it (low to high commitment) was obtained.“ It remained now to analyze the relationship between types of confluent and patterns of adjustment to determine the degree to which investment in the Lansing fringe affects adjustment differentials. The first aspect of adjustment patterns to receive attention was 0b.? ective integration. It may be expected that the association between mment and integration would be significant, with high investment acting as a causative factor leading to high objective integration. data support the hypothesis since the association is significant5 and However, the degree of associa- The the existing patterns are as expected. tion measures are rather low: less than twenty percent of the variance in integration is explained (C2 .. .172) and the gain in prediction is only ten percent (Lb - .096). nevertheless, it has been possible to account for part of the integration differentials. Interpreting the results, this finding sug- geats that respondents who are committed to residence in the Lansing fringe, because of various investments, tend to participate in the area. A eon‘rersely, low investment in the area leads to a low degree of within- \ “for the marginals see Appendix B, Table 57. 8~ l5'1'he findings are as follows: 1:2 - 24.057; degrees of freedom: ’ p < 0005; C - e'+15; Lb . ewGe .. 12 3 .. area participation. Thus, residents who conceivably find it most diffi- cult to make a break with their fringe location participate in the area to a relatively high degree. It may be reasoned that fringe area residents who have a minimum of contact with the City of Lansing would tend to manifest positive iden- tification with their residential area. This expectation is based on the proposition that the slight contact would provide little opportunity for positive identification with the city to develop, while the within- fringe contact would provide many such opportunities. This reasoning is most crucial in the case of farmers and non-Lansing place of work. Residents who own, rather than rent, their homes may be expected to be tied to the fringe since they have made a major commitment to the 81333.. For this reason, owners may be expected to manifest positive iden- tII-f'ication. In addition, elderly residents would tend to exhibit a low Predisposition to mobility since they are in a semi-forced position of atailbility. In general, this situation is fostered by an inability on their part to change jobs at their pleasure. This condition may influ- ence their identification patterns by almost forcing a positive evalua- t1°n of what they already possess, with fringe area residence one of the crucial "possessions." Residents with children may be expected to come into contact with 0151191. area residents more frequently than childless residents. In addi- tion. their predisposition to mobility may be low. Under these condi- t1°ns it may be expected that the presence of children would influence 1‘ eaidents to manifest positive identifications with the fringe area. The influences of income, education, and tier of residence are unclear. Rig: income may lead to either satisfaction with the area or -124... it may be a manifestation of upward mobility, with fringe residence per- ceived as a temporary condition. High education may be interpreted in the same manner. The tier of residence furthest from Lansing might in— fluence residents to feel that the city facilities are too distant, thus leading to a negative evaluation of the area, or such location may lead to positive evaluation of the fringe because of an anti-city bias. How- ever, the data do not support any of these hypotheses. (See Table 73.) Therefore, it was decided to examine the relationship of this dimension to the commitment index to determine whether the index may be utilized to explain direction differentials. Consistent with previous findings, a non-significant association was found.6 Although the association approaches significance, the pat- tern of relationship seems to be one of chance, with observed and expec- ted frequency deviations showing no clear pattern. Thus once again it has not been possible to isolate the factors that influence respondents' Positive and negative identifications with the Lansing fringe area. This finding means that the previously stated position must be matsinedz" either the direction variable itself is non-differentiat- ing, or positive and negative evaluations of the Lansing fringe are not influenced by demographic variables. Unfortunately, a choice between these alternatives presents one with a difficult problem, since both pc- 81tions can be defended. Until further evidence is available, it would “9m that the first alternative, i.e., the position that the direction variable is simply non-differentiating, is the most logical one to ac- cept. However, it must remain an open question until other variables \ 6The findings are as follows: x2 - 4.988; degrees of freedom: 2‘ D < .10. 7See an ra, pp. 96f. -125- TABLE?) SUMMARY OF THE RELATIONSHIP BETWEEN DIRECTION OF IDENTIFICATION AND CHARACTERISTICS OF THE POPULATION Degrees Population 2 of characteristics X freedom p Occupation factors a. occupation 10.7% 2 .10, b farm work experience .20 0 place of work 1.745 1 .20 Tier of residence .804 2 .70 i Tenure status .29 Ace of respondents 5.857 2 .20 Education of respondents 1.553 2 .50 Children a. number .548 2 .80« 5 number under 5 .th 0 number 5-17 .928 2 .70 .521} l .50 Fanily gross weekly income ‘- EXact probability measure used. -126- can be brought to bear on the question. At the present time the most fruitful research leads have been presented by Rossi, who has demonstra- ted the relationship between predispositions to mobility and indexes of dwelling unit space complaints, utilities complaints, physical environ- ment complaints, social environment complaints, distance complaints, and housing costs complaints.8 Unfortunately, such information was not ob- tained in the present research undertaking. Content of identification The specific influences of demographic variables on content of identification are difficult to anticipate, because of the extensive Variety of content types. However, looking at pure content types only, some influences may be expected in some specific cases. Farmers as a group are more-or-less rooted in the area and have been for many years. Under these conditions it may be expected that their mode of identification would be expressive of coerced and social These orientations refer specifically to forced ties to On the other orientations. the area and to the importance of friends in the area. hand, nonfarmers may be expected to view the apparent assets of the area on the basis of a comparison with the city from which they migrated; this orientation has been classified as personal in the content distri- butione Home owners and the elderly may also be expected to manifest a coerced and a social orientation to the area. In a manner, both of these gl‘oupings have a residence investment which would keep their mobility Predispositions to a minimum. \ 8Peter H. Rossi, fly; Families Move (Glencoe, Illinois: The Free Press, 1955). See especially Chapter V and Appendix B. -127- Families with children would be expected to have more frequent contacts with area residents. Under these conditions it may be antici- pated that they would tend to perceive the social climate of the area as basic. Therefbre, it may be expected that the social content type would take precedence. Occupation was the first general factor to be considered in the analysis of content. It was hypothesized that a significant relation- ship exists between occupation and content; the data support this hypoth- esis, with an extremely significant association. (See Table 74.) By trichotomizing the occupation category into nonfarmer, part-time farmer, and full-time farmer categories the patterns of relationship become clear (<3ells which are occupied by less than one unit in both observed and the- oretical frequencies have been ignored): Nonfarmers: content based on perceived qualities of the area (per- sonal) this same type combined with friendship patterns (personal- social), and friendship coupled with locational assets (personal- ecological). Part-time farmers: content reflecti locational qualities (ecologi- cal), economic advantages (economizg, forced ties (coerced), a combination of the first two mentioned types (ecological-economic), and a combination of the importance of friends with locational and personal factors (social-ecological and social-personal). Farmers: content based on the social climate (social), location (ecological), economic assets (economic), perceived inability to move (coerced), a combination of the second two factors (ecologi- cal-economic), and friendship ties coupled with both locational and economic qualities (social-ecological and social-personal). It is interesting to note that farmers and part-time farmers exhibit ainnilar content while nonfarmers seem quite dissimilar to the other two occupation categories. The coefficient of contingency deserves special notice since a Value of .569 means that the occupation grouping accounts for one-third 01‘ the variance in content (62 . .324). On the other hand, the lambda -128- TABLE 71+ SUMMARY OF THE RELATIONSHIP BETWEEN CONTENT OF IDENTIFICATION AND CHARACTERISTICS OF THE POPULATION Degrees Population 2 of _ characteristics X freedom p C Lb Occupation factors a; occupation 50.727 9 .001 .569 .131 b farm work exper- ience 20237 ll’ e70 0) place of- work 20.305 7 .01 .#16 .0 Tier of residence 15.798 8 .05 . 357 .0 Tenure status 6.618 2 .05 .280 .0 Age of respondents 15.120 12 . 30 Education of respondents 17.969 6 .01 .#02 .010 Children a number 6. 580 9 .80 13 number under 5 3.077 3 .50 0 number 5.1? 8e207 8 .50 Family gross weekly income 11.801! 12 .50 ¥ -129- measure is not particularly high: a knowledge of occupation provides only a thirteen percent improvement in the prediction of content. Once again it has been found that while occupation is signifi- cantly associated with adjustment, whether or not respondents have worked on a farm is not. This means that present-day fanning affects content, rather then whether such work was performed in the past. From the significant association found between occupation and content it may be reasoned that place of work is also significantly re- lated to this aspect of identification. Furthermore, it is expected that the resulting pattern would more-or-less follow the occupation anal- ysis, with Lansing workers exhibiting the nonfarmer pattern and non- Lansing worders consistent with farmers. The data confirm the hypothe- sis and, except for one deviant case, the pattern is precisely as ex- Pected. In the lone case of deviation, content based on friendship is related to working in Lansing, although it was over-represented among farmers in the analysis of occupation. Tier of residence and content are significantly associated (see Table 74) with a pattern of relationship that was completely unexpected: in eight out of ten cases, the two tiers closest to Lansing and the two tiers furthest from the city are completely consistent in their effect on content. This seems to indicate that location in the two closest and two furthest tiers of residence effects content in almost precisely the Sallie manner. Both degree of association measures are low: tier of res— 1Clence accounts for slightly more than ten percent of the content var- imioe (32 - .127), while a knowledge of residence location provides for no improvement in the prediction of content. An examination of the data indicates that a moderately signifi- -130... cant association exists between tenure status and content. However, the degree of association is slight: only eight percent of the content var- iance is accounted for and the prediction of this identification aspect is not improved by a knowledge of tenure status. Although age of respondents is not significantly associated with content, the education of respondents is simificantly related to this variable. This is an unexpected finding since it was assumed that these variables would more or less vary in the same manner; the young age structure was thought to be the crucial factor affecting the fairly high education distribution. The degree of association between education and content is fairly low: only sixteen percent of the variance in content is explained by education and a knowledge of the latter distribution Provides only a one percent improvement in the prediction of content. The analysis of the relationship between number of children, number of children under five, and number of children five to seventeen years of age indicated non-significant relationships in all cases. The relationship between family gross weekly income and content of identifi- cation was analyzed but the chi-square value also indicated a lack of significant association. (See Table 7t.) As expected, the relationship between content of identification and. the commitment index was found to be significant.9 This means that 1‘81ative degree of investment in the area influences the types of con- tent exhibited by respondents. The degree of influence can be approxi- mated by the degree of association measures: about one-quarter of the variance in the content distribution is accounted for by extent of \ B 9The findings are as follows: X2 -= 52.789; degrees of freedom: ; D < .001; c - 01"93; Lb . 00610 .. 131 .. investment in the area (5"2 - .245) and utilizing the commitment categor- ies to predict content makes the prediction only slightly more efficient (Lb 3 0061). Specifically, residents who are highly committed indicated that they tend to evaluate the Lansing fringe basically in terms of their friendship patterns and because of their economic ties in the area; in addition, certain mixed orientations are present. 0n the other hand, a low degree of investment in the Lansing fringe tends to produce an ori- entation based on private reasons for evaluation purposes; among these private reasons are that it's good for children, the area is not con- gested, and the area is both quiet and peaceful. Thus the relationship can be summarized by stating that residents who are committed to the area tend to feel that their friendship and economic ties are the impor- tant area assets while low commitment residents tend to evaluate the area for personal reasons. Lentil. of. identification Before reporting the specific empirical relationships found be- tween population characteristics and depth, a general discussion of d~emog'raphic influences on this adjustment dimension will be presented. This discussion will concentrate on the general processual factors that may influence only one aspect of the depth distribution, this aspect be- ing the ”no prior choice" category. This procedure is being followed lZn’MCzsmse previous findings have indicated that the basic meaningml dif- rerentiation is the dichotomous, rather than the trichotomous, one. Since many farmers were raised in the area itself it would be 10See an re, p. 111. -132 .. expected that their orientations would lead to a sus-coerced evaluation of the area. Home owners, the elderly, and residents with children have investments in the area which would tend to lead to a semi-coerced eval- uation of the area. Under these conditions it may be expected that each of these factors would tend to influence respondents toward a semi-forced evaluation. The function of the tier, education, and income variables is not understood at the present time. Therefore, it will be necessary to examine the empirical relationships of these variables and depth for a. meaningful understanding. As in the case of the other adjustment dimensions, it may be ex- pected that occupation and depth are significantly associated. The evi- dence gathered points to the presence of the association. (See Table 75.) Both the degree of association measures are the highest found in the entire analysis of the relationship between adjustment patterns and population characteristics: the coefficient of contingency value indi- cates that more than forty percent of the variance in depth is explained by occupation (52 - .#25); while lambda suggests that the prediction of depth can be made more than twenty percent more efficient by considering Occupation (Lb - .219). The pattern of relationship, utilizing the broad occupational categories of nonfarmer, part-time farmer, and farmer POints out that farmers had no prior choice in coming to the fringe, nonfarmers moved to the area by choice, and part-time farmers combine b0 th forms. Once again the data seem to provide a research lead: since the farm-nonfarm dichotomy so strongly affected the depth pattern, it was decided to determine whether farm work experience was itself the influ- erItial element involved in depth. However, the data do not support the -133- TABLE 75 SUMMARY OF THE RELATIONSHIP BETWEEN DEPTH OF IDENTIFICATION AND CHARACTERISTICS OF THE POPULATION Degrees Population 2 of _ characteristics X freedom p C Lb Occupation factors a; occupation 66.819 8 .001 .652 .219 b farm work exper- ience .151 l .80 0) place of work 50.485 2 .001 .565 .107 Tier of residence 8.405 8 .50 Tenure Status 0958 2 e70 Age of respondents 28.256 10 .01 .445 .101 Education of respondents 11.224 6 .10 Children a number 5.885 6 .50 1: number under 5 4.120 2 .50 0. number 5-17 9.644 6 .20 Family gross weekly income 16 . 942 8 . 05 . 570 . 102 ¥ -13/+- contention that farm experience affects depth patterns: whether or not nonfarmers had ever worked on a farm has no significant affect on depth. (See Table 75.) Since occupation is significantly associated with depth, it was hypothesized that employment in Lansing would also affect the adjustment category. Under these conditions, it was expected that the pattern of relationship would be such that Lansing workers would vary in the same manner as nonfarmers. The data support the hypothesis of a significant association (see Table 75) but the pattern deviated slightly from expec- tation since Lansing workers gave discrepant reasons for first moving, and now staying, in the area. The pattern may have been slightly upset by the part-time farmers, most of whom work in Lansing. The coefficient of contingency value is fairly high, indicating that one-third of the variance in depth is explained by whether or not respondents work in Lansing (52 - .519). Lambda is fairly low: predic- tion of depth is improved by only ten percent. In neither case does the degree of association value reach those obtained in the analysis of the Occupation-depth relationship. (See Table 75.) Tier of residence is not significantly associated with depth. Tlherefore, where respondents live, relative to Lansing, does not affect their modes of depth. In addition, no significant association was evi- dent when tenure status and depth were analyzed. Whether respondents onied their homes or were renting does not influence this aspect of ad- Jinatment. It may be expected that age and depth are significantly associ- at ed. The data support this contention; the chi-square value indicates '3 significant relationship. (See Table 75.) The pattern of relation- ~135- ship seems to be curvilinear for residents who had no prior choice for fringe residence and inhabitants who are now in the area for reasons different from their original reasons for moving; respondents who gave the same reasons for moving and staying exhibit an unclear relationship between age and depth. The relationship is as follows: Less than 50 years of age: no prior choice and same reasons 50 - 59 years of age: other reasons 40 - 49 years of age: same reasons and other reasons 50 - 59 years of age: no prior choice 60 years of age and older: no prior choice. Neither education of respondents nor the presence of children are significantly associated with depth. Therefore, neither of these categories can be utilized to account for depth differentials. It is to be expected that depth is associated with family gross weekly income. Since respondents who had no prior choice for area resi- dence were farmers who are in the older age groups, it may be reasoned that low income residents are also in this category. The data tend to support this hypothesis, for the association is significant (see Table 75) , and the pattern is in the expected direction. It was hypothesized that depth of identification and the commit- ment index would be significantly associated, with inhabitants who had no Prior choice for Lansing fringe residence contributing higher than chalice values to the high investment end of the index. The significant relationship was found11 and the pattern varied in the expected direction. However, the degree of association measures, as in the case of both \ 1,. llThe.f1ndj-n8‘8 are as follows: X2 - 22.226; degrees of freedom: . p < 0001; c . eh’S?‘ Lb 3 .071e -136— integration and content, did not exhibit very high values (52 = .209; Lb - .071). Thus, to some degree the differentials of depth have been ac- counted for. The nature of the relationship indicates that respondents who have most invested in the Lansing fringe tend to be in the area through no prior choice on their part. Since the no prior choice cate- gory is primarily populated by elderly farmers who own their home, the relationship is an expected one. Since it was meaningful to examine the relationship between depth and the residential history indexes with the no prior choice category eliminated,12 it was decided to follow the same procedure in this case. Once again it was found that the elimination of this category resulted in a non-significant relationship between the two remaining aspects of depth.” This means that the commitment index may be utilized to ex- plain the grouping composed of residents who had no prior choice for fringe residence but may not be utilized to account for the difference between respondents who are in the area for reasons that are the same or different from their initial reasons for moving to the Lansing fringe 31‘an Summary: Accounting for Adjustment Patterns The prime objective in this chapter has been to account for the pa”terns of objective integration and the three subjective identifica- tion categories by utilizing demographic variables. In this section an \ 12See supra, p. 111. 3, 1)The findings are as follows: x2 s .590; degrees of freedom: ’ P > .95. -137- attempt will be made to summarize the relationships between adjustment patterns and characteristics of the population. a) Objective integration The analysis has indicated that high objective integration is as- sociated with the following variables: full-time and part-time farming, not employed in Lansing, residence location in tiers six (the area fur- thest from Lansing) and four, and owning one's home. Since general influences are hidden in an analysis that treats all variables as independent, an index of some more general processual causative factors was constructed. After much deliberation it was de- cided to construct a commitment index which would reflect the degree of investment residents have in the area. Four specific factors were util- ized in the index: a) occupation, b) tenure status, c) age, and d) the presence of school age children. It was found that the commitment index and integration are sig- nificantly associated, i.e., degree of commitment affects fringe area behavior patterns. The pattern of relationship indicated that respond- ents who are highly committed to the Lansing fringe area tend to be in- tegrated. Conversely, low commitment is associated with a paucity of W:1 thin-fringe behavior. b) Direction of identification Once again the attempt to account for direction of identifica- tion has met with failure since direction of identification is not sig- nificantly associated with any of the demographic variables considered in this chapter. Consistent with the findings in Chapters III and IV, it is suggested that this aspect of identification has little utility in the determination of fringe area adjustment patterns. -138- c) Content of identification Content is significantly associated with the following demograph- ic variables: occupation, place of work, tier of residence, tenure sta- tus, and education of respondents. The index of commitment and content of identification are signi- ficantly associated. In general, the pattern of association is such that respondents who manifest high commitment exhibit orientations which place primacy on friendship patterns and economic ties; respondents who ex- hibit a low degree of commitment to the area exhibit a personal orienta- tion based on apparent assets of the area. (1) Depth of identification Depth is significantly associated with four population charac- teristics: occupation, Lansing and non-Lansing place of work, age of respondents, and family gross weekly income. In attempting to account for the depth sub-categories, the following relationships have been found : 1. No prior choice pattern: these respondents are full and part-time farmers who do not work in Lansing. They are under thirty and over forty-nine years of age, and their family gross income is less than eighty dollars weekly. 2. Same reason as staying pattern: respondents who exhibit this depth pattern are nonfarmers and part-time farmers who do not work in Iaestnsing. These respondents are under thirty and are between forty and forty-nine years old; their family gross weekly income is between “’0 and 859 and also 8200 and over. 5. Other reasons pattern: these respondents are nonfarmers who Work in Lansing. They are between thirty and forty-nine years of - 139 - age and their family gross income ranges between 880 and 8199 weekly. A significant association was found between the index of com- mitment and depth. Once again, however, it was found that the relation- ship is a reflection of a distinction between residents who did, and did not, have a prior choice for fringe residence. Thus it has been demons strated that respondents who manifest high commitment had no prior choice in fringe residence. An attempt to account for differences be- tween respondents who prefer to remain in the area for the same reason as they initially decided on Lansing fringe residence and sample members ‘(bO gave different reasons for coming and now staying was not success- :ful. CHAPTER VI COMMUNITY ORIENTATIONS In this chapter an analysis will be made of the orientations of sample members which seem crucial to certain aspects of fringe area so- cial life. This component of fringe area social life will be analyzed by examining specific orientation patterns and determining whether dif- ferentials can be explained. A description of the nature of the orientations of Lansing fringe residents has already been presented.1 It remains now to examine the answers elicited by these questions to determine their meanings in a broader perspective. To achieve this, the specific responses to ques- tions must be examined to construct broader categories which individual queries represent. These broader categories may be derived by finding clusters of relationships and constructing from them indexes for which the clusters stand. Thus the first step in accounting for community orientations is to examine the specific meanings of the answers elicited by the ques- tions and, on the basis of common meanings, to combine certain of these into orientation indexes. Indexes of community orientation can then be examined to determine whether they can be accounted for, or explained. 1See supra, Chapter II. -140- -141- Index Construction Localite index An examination of responses indicated that three questions eli- cited responses which can be utilized in identifying sample members who may be considered localites. Such respondents perceived their place of residence as a local, independent, small area. The specific questions are as follows: 1. Does this area have a name? (See Appendix B, Tables 24 and 24a.) 2. Do you think of it as part of a larger area? (See Appendix B, Tables 25 and 25a.) 5. How far north, east, south, west does [the area around here] go? (See Appendix B, Table 50.) (On the basis of responses to these questions, localites were defined as :respondents who: 1) felt that the area has a name (all names provided Irv respondents were local in nature), 2) felt that the township, school ctistrict, or bordering town is one related local area, and 5) indicated ‘bliat the local area is small in size (8.0 miles or less), or replied to ‘tlie size-of-area question in terms of sections, the immediate vicinity, ()1? a local nearby town.2 w orientation index Another cluster was exhibited by sample members whose responses idflmiicated a positive orientation toward Lansing. Three questions which Provided respondents with an opportunity to evaluate the central city were included in the field schedule. The specific questions are: 1. Do you think that living in Lansing would have advantages over living in this area? (See Appendix B, Table 54.) 2. Do you think that the idea of starting [the Lansing Metropolitan \ 2For the index marginals see Appendix B, Table 76. .. 142 - Plasming Commission] was a good one? (See Appendix B, Table 51. 5. Would you have agreed with the Commission if it had recommended that this area be annexed by the City of Lansing? (See Appen- dix B, Table 52.) On the basis of responses to these questions, Lansing oriented sample members were defined as respondents who: I) felt that Lansing residence has advantages in comparison to fringe residence, 2) answered with an af- firmative or a "don't know" response to the formulation of the Commis- sion question, and 5) replied with an affirmative or a "don't know" re- 3 sponse to the annexation question. Definition of the area indexes To determine how respondents defined their area of residence, sample members were asked several questions designed to elicit specific definitional responses. Specifically, on the basis of replies to these questions, sample members were classified into groupings based on wheth- er they perceived the Lansing fringe area as urban or rural. The spe- C1fic questions are as follows: 1. Which [type of area--urban or rural] do you think you live in? (See Appendix B, Table 26.) 2. Do you think that there are differences in the kinds of people who live in an area like this and city people? (See Appendix B, Table 27.) 5. Do you think that there are differences in the kinds of people who live in an area like this and a much more rural area? (See Appendix B, Table 28.) l+. Does an area like this come closer to the farm or city way of thinking? (See Appendix B, Table 29a.) Two definite alternatives are evident in the above questions: resPondents have been presented with a choice of whether they perceive \ 5For the marginals see Appendix B, Table 77. -143- the Lansing fringe as an urban or a rural area. After examining the re- sponses elicited by the four questions, two community orientation indexes were constructed. These indexes may be referred to as an urban defini- tion of the area index, and a rural definition of the area index. Respondents were classified in the index as providing an urban definition of the area if they: 1) indicated that they live in an ur- ban area, 2) felt that there are no differences between local peOple and city inhabitants, and 5) indicated that their residence area comes clos- er to the city way of thinking“. In the second index, sample members were classified as providing a. rural definition of the area if they: 1) felt that they reside in a rural area, 2) indicated that no differences exist between local resi- dents and inhabitants of a rural area, and 5) felt that the local area comes closer to the farm way of thinking.5 Satisfaction index . Two of the questions elicited responses which can be utilized to identify respondents who seem to be satisfied with their area of resi- dence. These respondents felt that the area is getting better and also indicated that they are quite satisfied with the facilities present in the area. The specific questions are: 1. Do you think this area is getting better or worse? (See Appen- dix B, Table 52.) 2. What sort of facilities would you like to see established around here? (See Appendix B, Table 55.) On the basis of responses to these two questions, sample members were \ #For the marginals see Appendix B, Table 78. 5For the marginals see Appendix B, Table 79. - 144 - classified in an index as satisfied if they felt that: 1) the area is getting better, and 2) no new facilities in the area are necessary. Accounting for Community Orientations In previous chapters an attempt was made to determine how social variables affect adjustment patterns in the Lansing rural-urban fringe. In this chapter the perspective will be focused in another direction: an attempt will be made to determine how integration and identification patterns affect community orientations. This new perspective will follow the same methodological approach as in previous chapters except for one important modification: whereas the reasoning in Chapters IV and V went from social variables to adjust- ment patterns, in the present case the reverse procedure will be followed. Thus, adjustment patterns will be utilized to explain social variables. In addition to this "reversal" in procedure, one specific technique must be modified: the lambda coefficient will be utilized in the other direc- tion. This change amounts to determining the value of La’ rather than L . The specific calculation of the coefficient of contingency does not b change, but the interpretation of this measure is in the other direction. Objective integration Certainly, it would be expected that behavior patterns within an area would influence residents' perceptions of the area. Thus, it seems logical to expect that residents who participate in fringe social life would tend to perceive the area as an independent, local, small area, i.e., the orientation may be designated as a localite outlook. Similar- ly, it would be expected that within-area participants would not be 6For the marginals see Appendix B, Table 80. -145- oriented to another area of residence, i.e., they would not manifest a Lansing orientation. Another expected relationship would be that resi- dents who participate in the social life of the fringe would tend to be satisfied with the area. As far as a rural or urban definition of the fringe is concerned, it may be anticipated that integrated residents do not define the fringe as urban. The rejection of an urban definition of the area is based on a general negative evaluation of the city, an eval- uation which was reflected in the initial move of many residents from the City of Lansing. As for a rural definition of the area, the expected relationship is fairly puzzling. Since well over one-half of the resi- 7 it may be dents recognized the existence of only two areas of residence anticipated that the rejection of an urban definition for the area would necessitate a rural definition, if only by default. Thus, it would be expected that integrated residents would tend to define the Lansing fringe area as rural, rather than urban. Looking at the specific empirical relationships between integra- tion and the indexes, the only significant relationships found were be- tween participation and area definitions. (See Table 81.) Thus, three of the community orientation clusters are not significantly affected by whether or not sample members participate in the fringe area. Each of the relationships will now be discussed. Localites have been defined as those residents who perceive the local area as small and independent. The data suggest that such a per- ception is not significantly influenced by behavior patterns. Perhaps each.of the integration categories is populated by residents who hold several types of orientations. Fbr example, some of the integrated 7See Appendix B, Table 26. -146- respondents may recognize the dominance of the central city while others are not cognizant of this situation. In addition, a high degree of in- tegration in the fringe does not eliminate the possibility of behavior patterns in other areas, i.e., it is not a one-or—the-other proposition. Thus, some integrated residents may be localites who participate only in the fringe area while other integrated residents may take advantage of the facilities of more than one area simultaneously; it may be expected that the latter group would not perceive the residential area as local- ites. Unfortunately, these conjectures must remain on the level of log- ical possibility since the question in this study is not: in what areas do fringe residents participate? --rather it is: do residents partici- pate in the fringe area? TABLE 81 SUMMARY OF THE RELATIONSHIP BETWEEN OBJECTIVE INTEGRATION AND INDEXES OF COMMUNITY ORIENTATION Community Degrees orientation 2 of ._ index X freedom p C La Localite index 9.#81 8 .50 Lansing orientation index 12.#89 9 .20 Urban definition of area index 11.929 6 .05 .512 .0 Rural definition of area index lh.015 6 .025 .556 .095 Satisfaction index 5.1+92 6 .80 _— The possible heterogeneity within the integration types may ex- ‘plain.why utilization.of within-fringe behavior patterns to account for the differentiation between localite and non-localite failed. In terms - 147 - of the study framework, a clue to the nature of the influence of adjust- ment patterns on localite orientations may be gained through an analysis of the relationship between community orientations and identification. The lack of a significant relationship between integration pat- terns and the Lansing orientation index was also unexpected. (See Table 81.) Other investigators have suggested that inhabitants of the fringe are oriented to the central city which dominates the area. Since it has been possible in this study to differentiate between fringe participants and non-participants, a hypothesis is clearly suggested: nonepartici- pants are Lansing oriented while participants are not. However, the data indicate that residents who participate in the fringe area are not significantly different in their central city orientations than inhabi- tants who do not. Of course, one can participate in the fringe area and still feel that there are advantages in the central city, but such an association cannot be anticipated in terms of the conclusions of other investigators. The possibility exists that the biased nature of the framework of others8 had so conditioned their approach that the conclusion of central city orientations is not based upon an objective approach to the problem. Clearly, the lack of a significant association between integra- tion and the Lansing orientation index suggests a re-evaluation of the expected relationship. After much deliberation on this pgg’g £22329; lev- el, the beginnings of a possible explanation have been formulated. The reasoning proceeds from the observation that people move into an area ‘with.a pre-established community orientation; previous data suggested that one of the crucial orientations held by fringe migrants is an anti- 8See Chapter I for an extended discussion of this point. .1423- city bias. This anti-city bias may or may not affect behavior patterns. Therefore, residents may reject the proposition of Lansing residential superiority and may reject the annexation idea whether or not they par- ticipate in the fringe area. unfortunately, data on pre-established community orientations are not available in this study, nor is such in- formation available in any of the studies reviewed in the first chapter. Satisfaction with the fringe area, as measured by the satisfac— tion index, was also feund to be non-significantly associated with inte- gration. (See Table 81.) As in the two cases discussed above, it was necessary to scrutinize the data to reach an understanding of why the expected relationship does not exist. Possibly, this situation.may also be explained by taking account of the fact that people move into any area with pre-established communi- ty orientations. In this case migrants may be satisfied with the fringe area since the move itself may commit them to a positive evaluation of the area. If this is the case, participation patterns may have little influence on satisfaction. In addition, inhabitants may participate in an area whether or not they are satisfied; participation may be a function of the fact that the residential area is the only one available for interaction purposes. Residents may feel that the area facilities are inadequate and that the area is getting worse but might still shop and visit friends in the area, belong to local organizations, and utilize local sources for in- formation about the area, the township, the county, and Lansing. Thus, behavior patterns may not influence inhabitants' feelings of satisfaction. The two indexes of rural and urban definition of the fringe area are significantly associated with integration. (See Table 81.) In - 149 - addition, the patterns of relationship are in the expected direction: respondents who manifest high integration define the fringe area as rur- al and reject the proposition that it is urban. The degree of associa- tion measures are not high: approximately ten percent of the variance is explained in each case (urban definitions2 - .097; rural definition '62 - .115) and knowledge of integration leads to no increase in the ef- ficiency of the prediction of urban definitions and only a ten percent increase in the prediction of rural definitions (La - .093). In spite of the low degree of association, it has been ascer- tained that integration influences definitions of the area. This may be explained by the general negative evaluation of the city held by many residents. The data suggest that if residents participate in the area, it is not defined as urban; in terms of the perceptions of residents, if the area is not urban, then it must be rural. Direction 2; subjective identification Up to this point none of the relationships between direction of identification and social variables were found to be significant. As each non-significant association was found, the possibility became clear- er that the direction variable is simply nonpdifferentiating. Neverthe- less, it was decided to examine the relationship between direction and the community orientation indexes. If significant relationships were found, the position that direction is non-differentiating could not be accepted; if non-significant associations were found once again, there would be added reason to suggest that the position is a valid one. Basing expectations on the premise that direction is sufficienthy differentiating, the problem to be investigated is the following: in what manner does direction influence community orientations? Or more -150— specifically, how do positive and negative identifications of residents influence their modes of orientation? If direction is differentiating, it may be expected that resi- dents who manifest positive identification with the fringe area would exhibit different community orientations than those who manifest nega- tive identification. Fbr example, it may be expected that the positive- ly identified would think of the area as small and independent, i.e., they would be localites. Positively identified residents would not be expected to manifest Lansing orientations since they would reject an an- nexation proposal and a statement suggesting that the central city is a superior residential area. Following a similar line of reasoning, the positively identified may be expected to define the fringe area as rural, rather than urban. Finally, it may be anticipated that those who exhib- it positive direction would be more satisfied with residence in the fringe area than those who exhibit negative direction. Hewever, an examination of the relationship between direction and community orientation indicated that all associations are non-signi- ficant. (See Table 82.) This finding, consistent with previous find- ings, suggests that the direction variable pgg s3 may be non-differenti- ating. The suspicion is now quite strong since the evidence is one- sided. Therefore, this dimension is suspect enough to warrant its re- jection as a useful determinant of adjustment patterns in the fringe area. Perhaps some other indicator of positive and negative identifica- tions can be devised which would meet the needs of an adjustment study. Unfbrtunately, it is too late to start a search at this point in the present study: it is suggested that any future research oriented in the same direction as the present study should attempt the development of a - 151 - scale to meet this need. TABLE 82 SUMMARY OF THE RELATIONSHIP BETWEEN DIRECTION OF IDENTIFICATION AND INDEXES OF COMMUNITY ORIENTATION Community Degrees orientation 2 of index X freedom p Localite index 1.554 2 .50 Lansing orientation index .782 2 .70 Urban definition of area index .712 2 .70 Rural definition of area index 5.725 2 .20 Satisfaction index 2.2#5 2 .50 Content 9; subjective identification Basically, the problem to be considered in this section may be stated as follows: in what manner do content types influence modes of orientation? It is evident that the wide range of content types makes prediction of specific relationships fairly difficult. However, it has been found in previous tests that content based on the social climate and that based on forced ties (i.e., social and coerced) tend to vary to- gether: both types are populated by residents who manifest deep social roots in the area. Since these types are of this unique character, ex- pected relationships were derived for them, with the anticipation that other content categories would vary in the opposite direction. Utilizing these categories, an anticipated association is im- mediately evident: it may be expected that respondents who focus on friendship patterns and on a subjective knowledge of the area would con- sider the community as small and independent. Therefore, a significant association between content and the localite index is anticipated, with - 152 - social and coerced contents contributing more than chance values to the localite end of the index. Likewise, it may be expected that respond- ents in these content categories would not be oriented to Lansing. Bas- ing expectations on an assumed anti-city bias, it would seem logical to expect that sample members who exhibit close ties in the fringe area, i.e., those who have been classified in the social and coerced content categories, would define the area as rural, rather than urban. Another anticipated relationship is that residents who have emphasized friend- ship patterns and a knowledge of the area would be satisfied with the area. More specifically, it is anticipated that the association between content and the satisfaction index will be significant, with the social and coerced content types related to satisfaction. However, when the empirical relationships between content and community orientation were computed, only the definition of the area in- dexes were found to be significant. (See Table 85.) Therefore, three of the community orientation categories are not significantly affected by the types of content exhibited by respondents. This finding deserves further discussion, looking at each association independently. Apparently, respondents may emphasize a knowledge of the area and its inhabitants without this viewpoint affecting whether they per- ceive the area as a small, independent, and local one. Seemingly, the Lansing fringe residents have indicated that they like the area for rea- sons other than its local nature. Central city orientations, as measured by the Lansing orienta- tion index, are not significantly associated with content. (See Table 85.) Generalizing from the relationship, it seems that respondents do not feel that Lansing domination is related to what they like about the - 153 - area. Thus, sample members probably are not cognizant of changes that would ensue if the central city was to dominate the fringe area to a greater degree than it now does. Apparently they feel that their rea- sons for liking the area are not affected by actions of the central city. TABLE 85 SUMMARY OF THE RELATIONSHIP BETWEEN CONTENT OF IDENTIFICATION AND INDEXES OF COMMUNITY ORIENTATION Community Degrees orientation 2 of _ index x freedom p 0 La Localite index 9.008 8 .50 Lansing orientation index 7.549 8 .50 Urban definition of area index ' 56.885 6 .001 .556 .061 Rural definition of area index 21*. 705 7 e 001 e “’55 o 171.. Satisfaction index 8.020 6 .50 The final non-significant association found was between content and the satisfaction index. (See Table 85.) It had been anticipated that a significant association would be found, with the social and co- erced content categories related to satisfaction with the area. The exp planation for the nonesignificant relationship may lie in pre-established community orientations. .Migrants would probably be satisfied with the area since the move itself may be perceived as a commitment which calls for a positive evaluation of the new place of residence. Since the con- tent categories do not distinguish between migration differentials, this may explain the non-significant association. Computations revealed that the rural and urban definition of the - 154 - area indexes are significantly associated with content. (See Table 85.) As expected, respondents exhibiting content based on friendships and forced ties provided a rural definition of the area and rejected the idea that the area is urban in character. The coefficient of contingen- cy values indicate that approximately twenty percent of the variance in the rural definition index is explained by content differentials (Ce - .207) while almost thirty percent of the variance in the urban defini- tion index is explained (52 a .287). However, the lambda values indi- cate a reversal in order: content may be utilized to predict rural def- inition differentials with more success that urban differentials (rural definition index La 2 .174: urban definition index La - .061). M 9_f_ subjective identification Depth is the final aspect of adjustment to be considered in the analysis of community orientations. The problem posed here is the fol- lowing: in what manner does depth of identification play a part in the formulation of community orientations? Previously, it was found that the basic differentiation made by depth is between those who did not have a prior choice for fringe area residence and those who did have a choice in the matter, i.e., it is be- tween the "no prior choice" category on the one hand and the "same rea- sons" and "other reasons" categories on the other.9 Therefore, antici- pations of relationships will take account of the dichotomy, rather than the trichotomy. Actually, the dichotomy differentiates between residents who were brought into the area at a young age, before the large influx of migrants, 9 See su re, p. 111. -155- and residents who have more or less recently chosen to move into the fringe. It may be expected that sample members who have had long and rural experiences in the area would perceive it as small and independent, i.e., as localites. Since these experiences have been within the fringe area.pgg.§g, it may be expected that respondents who had not chosen their area of residence would not be oriented to Lansing. Similarly, these respondents would be expected to define the area as rural, rather than urban. To understand the expected relationship between depth and the satisfaction index, the focus must be shifted from residents who had no prior choice in location to those who chose the area for residential purposes. At this point, pre-established community orientations seem to become an essential feature of the situation. It seems logical to inter- pret a move to the area as an initial commitment which would influence orientations. It may be anticipated that migrants would be satisfied with the area as a place of residence since the move itself may commit them to positive evaluations. An examination of the empirical relationships between depth and the orientation indexes revealed that three of the associations are sigh nificant. The data indicate that the localite index and the Lansing or- ientation index are not significantly associated with depth, while the indexes of rural definition, urban definition, and satisfaction do ex~ hibit significant associations. Each of these relationships will now be discussed. It was hypothesized that depth would be significantly associated with the localite index: this hypothesis is based on the assumption that residents who had undergone rural and longbtime experiences in the fringe - 156 — would perceive the area as small and independent. However, the analysis revealed that the association is not significant. (See Table 84.) This lack of significance suggests that localite orientations are not influ- enced by whether or not respondents' are in the area by choice. Such a negative finding serves an important function by pointing to the importance of phenomena that were not previously considered. One of the basic neglected variables up to this point is the type of reac- tion by "natives" to the ecological invasion they have witnessed. Per- haps these residents have seen the area change from what once was de- fined as a small, local, independent area. Certainly, the changes ob- served by them must have necessitated new definitions of the area; one of the results may have been an awareness of the increasing dominance of the central city. This situation.may have influenced "native" per- ceptions by emphasizing that the area is no longer small, local, and independent. Meanwhile, it is evident that the newcomers, i.e., those who chose the area as a place of residence, may quite clearly perceive the fringe as part of the larger Lansing area. Under these conditions, they would not define the fringe area as small and independent. Unfor- tunately, questions dealing with the changing perceptions of "natives" were not included in the field schedule. Perhaps this negative finding and the attempted explanation formulate the basis for another empirical study in the fringe area. It was also anticipated that depth would be significantly asso- ciated with the Lansing orientation index, with respondents who had no prior choice for residence in the area not oriented to the central city. Hewever, when the relationship between these variables was calculated, it was found to be non-significant. (See Table 84.) This suggests that - 157 - TABLE 84 SUMMARY OF THE RELATIONSHIP BETWEEN DEPTH OF IDENTIFICATION AND INDEXES OF COMMUNITY ORIENTATION Community Degrees orientation 2 of __ index X freedom p C La with "no prior choice" categggz included Localite index 6.015 5 .20 Lansing orientation index 6.565 6 .50 Urban definition of area index 7.156 2 .025 .292 .0 Rural definition of area index 15.080 4 .02 .560 .116 Satisfaction index 8.099 4 .05 .286 .0 with "no prior choice" catemry 2W Localite index 1.785 5 .70 Lansing orientation index 5.127 5 .50 Urban definition of area index .248 2 .90 Rural definition of area index .554 .80 Satisfaction index 5.275 .10 -158... depth does not significantly influence orientations to the City of Lan- sing. This non-significant relationship is not completely unexpected since the hypothesis might easily have been that both depth categories (considering the dichotomy) operate against Lansing orientations; the hypothesis of a "non-native" - Lansing orientation relationship was pre- sented because it seemed a bit more plausible. What made this hypothe- sis plausible was the assumption that the longhtime and rural experiences of residents who did not have prior choice for fringe residence would in- fluence them to reject a positive Lansing orientation. It seems logical that those natives who witnessed the expansion of the city would take a negative view of the changes forced upon them. However, the same hypoth- esis could have been stated for "non-natives" since respondents who moved into the area recently may be expected to evaluate the city in a negative manner. Host of these sample members quit the city and chose "country living" to take advantage of assets not present in Lansing. The fact that the same orientations may be held for entirely different reasons cannot be taken into account since the depth categories did not go beyond a simple division on the basis of whether respondents did, or did not, have an initial choice for fringe residence. It was expected that the definition of the area indexes and depth would be significantly associated, with respondents who had no choice for fringe residence defining the area as rural, rather than urban. The data indicate that both indexes are significantly associated with depth (see Table 84) and that the expected patterns are present. Thus, depth may be utilized to explain whether residents perceive the fringe as a rural or urban area. The differential perceptions of - 159 - the area seem to be based on the within-area experiences of residents. Respondents who had no prior choice for the fringe as a place of resi- dence were in the area when it was strictly rural, before the large-scale migration commenced. Therefore, it may be anticipated that their reac- tion to recent developments would be influenced by what the area once was, i.e., strictly rural and not urban. It is suspected that this group would be the last to recognize that the local area is undergoing changes which might eventually result in completely urban existence. Elimination of the "no prior choice" category from the table re- sults in a non-significant chi-square value. (See Table 84.) Thus, once again it is evident that the functional utility of the depth vari- able lies in the differentiation between respondents who did, and re- spondents who did not, have an initial choice for fringe residence. The final orientation variable to be considered here is the sat- isfaction index. It may be expected that depth is significantly asso- ciated with satisfaction differentials. The influence of depth can most clearly be seen by treating the two categories which are composed of re- spondents who had an initial choice for fringe area residence. The cru- cial factor is the pre-established community orientations held by mi- grants. It is assumed that migrants moved into the area with relatively high degrees of satisfaction since a choice for the area had to be made befbre the move. Furthermore, this initial orientation may be congealed by the fact that respondents are somewhat aware of the commitment re- flected in the move. Therefore, it is expected that respondents who made a choice for the area would tend to exhibit fairly high degrees of satisfaction. The data indicate that a significant association does exist (see - 160 - Table 84) and that the pattern of relationship is as expected. However, the degree of association measures are quite low: less than ten percent of the variance in the satisfaction index is explained by depth (52 - .082) and the utilization of depth differentials does not increase the efficiency of predicting satisfaction. Summary: Accounting for Community Orientations The object of this chapter was to detenline the influence of ad- justment patterns on community orientations. Five community orientation indexes were constructed, as follows: localite, Lansing orientation, urban definition of the area, rural definition of the area, and satisfac- tion. An attempt was made to account for each of these orientations by utilizing the objective integration variable and the three aspects of subjective identification. In this section the empirical relationships between indexes and the adjustment patterns will be summarized. a) Localite index Localites were defined as respondents who perceive the fringe area as small, independent, and local. NOne of the adjustment categories were found to be significantly associated with this index. Therefore, adjustment differentials do not significantly influence localite orien- tations. Some of the suggested reasons for the failure of adjustment patterns to differentiate between localites and non-localites are as fOllows: l. Irrespective of adjustment patterns, some respondents may recognize the dominance of the central city while others do not. 2. Local participants may also participate in other areas; if this is the case, local participation may not lead to local orientations. - 161 - 5. Residents may be positively inclined toward the local area for reasons other than localite reasons. 4. "Natives" have witnessed an ecological invasion that may have, by their definitions, modified the area to such an extent that they no longer perceive it as a small, local, independent area. "Non- natives" would not be expected to view the area as a local, independent one o b) Lansing orientation index Respondents were classified as oriented to Lansing if they felt that living in the city has advantages over fringe residence, if they endorsed the formulation of the Lansing Metropolitan Planning Commission, and if they indicated a favorable attitude toward annexation. None of the adjustment categories were found to be significantly associated with the Lansing orientation index. Various explanations for this negative finding were discussed. Two of these reasons are as follows: 1. An anti-city bias may influence all respondents by producing a negative evaluation of Lansing; this influence might produce similar orientations among "natives" and "non-natives." 2. Possibly, there is a lack of understanding on the part of residents of the consequences of fruther domination by Lansing. c) Rural and urban definition of the area indexes Respondents were categorized on the basis of whether they felt that the local area is urban or rural. The indexes took account of how respondents defined the area, opinions of whether area residents were rural or urban people, and opinions of whether the area comes closer to the farm or city way of thinking. Direction was found to be the only -162- adjustment variable not significantly associated with the indexes; on the basis of this and previous findings, the direction variable was dis- missed as non-differentiating. The remaining adjustment variables were utilized to explain how integration and identification influence community orientations. It was feund that sample members who define the fringe area as rural, and not urban, are respondents who participate in the area, are positively in- clined toward the area because of friendship patterns and a knowledge of the community, and may be considered "natives." d) Satisfaction index An.index of satisfaction was constructed by taking account of 'whether respondents thought that the area is getting better or worse, and whether they were satisfied with area facilities. Depth alone was found to be significantly associated with the satisfaction index. The lack of a significant association between integration and satisfaction was explained by pointing out that residents may partici- IP£1te in an area whether or not they are satisfied. The lack of a sig- nificant association between content and satisfaction was explained by E”Juggesting that content does not take migration differentials into ac- C“mint. This latter position received support when it was demonstrated that depth and the satisfaction index are significantly associated. As anticipated, respondents who had an initial choice for residence in the area exhibit relatively high degrees of satisfaction. This relationship hadi 'been hypothesized since it was expected that pre-established commun- ity' (arientations, and a commitment to the area of residence which was re- fle<=llzed in the move, would influence respondents to manifest a satisfac- tim’l with the area. .. ...._._".~.-__ ._..- _-——_.. . . - 165 - Two-thirds of the respondents were raised within the study counties or contiguous counties. However, since this total area contains both rural and urban settlements, approximately equal numbers were raised in rural and urban areas. Mobility patterns were also examined and the data in- dicate that Lansing fringe residents as a whole have not experienced many changes of residence since 1940, other than the move to the area. 2. Demographic characteristics of the sample were examined to A determine the composition of the population. As expected, residents ex- hibit mixed occupational characteristics; almost one-quarter are full- time farmers, twelve percent are part-time farmers, more than one-third are employed in factory work, and one-quarter are engaged in a diversity of other urban occupations. Two-thirds of the sample members work in Lansing. A density ratio tier-of—residence gradient pattern was not found. (>verwhelmingly, fringe area residents are owners, rather than renters, (if their homes. The age structure of the area suggests that the popula- izion is fairly young: two-thirds of the respondents are between thirty taund forty-nine years of age. Since a large proportion of respondents snare in the child-bearing ages it was expected that a large number of C=liildren would be found. Hewever, this was not the case: less than one- QLqurter of the respondents have more than two children under seventeen and almost forty percent have none. The income distribution is quite high, with one out of five families grossing more than 8200 a week; the median gross family income is 8107. 5. Definitions of the area were patterned to some degree. When 86m131e members were asked why they chose the particular area for resi- dential purposes only a few indicated positive choice for the area; -166- responses suggesting an economic motivation were quite prevalent. Inde- pendence of the area, as measured by two questions, was not perceived: more than two-thirds of the respondents could not provide a local name and one-third felt that the fringe is part of a larger area. More than two-thirds reported that the fringe area is rural in nature although a majority indicated that it is occupied by city people. However, two— thirds of those who did not classify the area as rural (i.e., those who classified the area as urban or "in-between") felt that local inhabitants are rural people. In addition, of those respondents who felt that rural and urban people think differently, three-quarters reported that local inhabitants think like rural people. Almost two-thirds of the sample members defined their neighborhood boundaries as an area encompassing a total of eight miles or less. 4. A battery of questions designed to determine respondents' evaluations of the fringe area was included in the field schedule. An examination of responses revealed an overwhelming positive orientation toward the local area with a simultaneous negative evaluation of the city as a place to live. Over ninety percent indicated that if they had a choice they would prefer to remain in the area. Respondents were asked to compare the advantages of fringe residence with city and rural life; an overwhelming positive evaluation of the local area was found. In ad- dition, a majority felt that the area is improving although one-half felt that it lacked some facilities. 5. Social participation differentials were also measured. In addition to determining the behavior patterns of sample members, respond- ents were asked for their opinions on how well people in the area know each other; a five item checklist ranging from "very well" to "not at - 167 - all" was utilized. The extreme categories on the checklist were not se- lected very frequently while the middle category ("fairly well") was chosen by one-half of the sample. A checklist was also provided to de- termine how many residents respondents know by name; the most frequent choice was "a few." An examination of family interaction patterns revealed that al- most one-third of the sample rarely come into contact with any families in the area, forty percent never visit local families, and two-thirds do not lend anything out. The influence of Lansing is evident when friend- ship patterns are examined: friends are visited in the city more fre- quently than they are in the area, and Lansing was the most frequently mentioned place when respondents were asked for the addresses of their three best friends. One-half of the sample reported membership in formal organiza- tions; the most frequently mentioned meeting place was local towns. As for local informal activities, it was found that these are almost nonex- istent: over ninety percent have not participated in any. The influ- ence of Lansing is evident when shopping patterns are examined: forty Percent reported that all their shopping was done in city stores and mother twenty percent indicated that Lansing was utilized for more than oz1e—half of their shopping. Respondents were also asked to report the 8Cbuqrces from which they obtain local and township news; one-quarter re- Ported that they did not keep up with news of the immediate area. Among 1”Kiss who do keep up with local and township news, friends and neighbors wer e the most frequently mentioned sources for this information. Watch- ing- ‘television was the most frequently mentioned leisure time activity. 6. Approximately two years before field work started an attempt ~168- had been made to organize a commission which had as its goal the formu- lation of plans for the Lansing Metropolitan Area, including the study site. One of the important factors leading to the eventual rejection of the idea was the suspicion on the part of individuals outside of the cen- tral city that this was an initial manuever which would eventually lead to political domination and annexation by Lansing. Under these condi- tions, it was felt that many respondents would interpret the proposed commissionwithinacontext of a "threat to independence." Therefore, a question calling for either endorsement or rejection of the commission was included in the field schedule. In addition, since respondents were thus provided with an "independence" frame of reference, it was decided to include a question asking for opinions on annexation at this time. One-third of the sample members indicated opposition to the formulation of the commission, one-half endorsed the organization, and the remainder could not decide; those respondents who were in opposition did interpret the commission as a threat. Over eighty percent were opposed to annexa- tion but the most prevalent reason given for this position was a fear that an increased tax rate would result; a threat to independence was mentioned, but not as frequently as the tax rate. Construction _q_f_ the scales Two batteries of questions, one designed to determine integration Patterns and the other oriented to modes of identification, were includ— ed in the field schedule. Scalogram analysis of responses was utilized to derive categories of integration and identification, with such cate- 8°Pi es differentiating among adjustment types. The scalogram procedure was successful in the case of objective intag-ration; respondents were classified into five scale types, on the -149 - basis of their degree of integration. However, the analysis of identi- fication responses did not meet scalogram criteria. As an alternative procedure, identification responses were subjected to another type of analysis which might provide a basis for classifying sample members into scale types. However the second method, known as the method of summated ratings, also failed to produce a scale. Since the two methods suggested that the individual battery items could not be utilized together, a totally new approach to the increasing- ly complex identification dimension was sought. Therefore, responses to all identification-related questions were examined, to determine whether any of the specific queries were particularly differentiating. This ex- amination indicated that two of the questions had elicited responses which seemed to be most relevant to the question. On the basis of re- sponses to these questions, sample members were categorized into subjec- The complexity of the identification dimen- di- tive identification types. Sion necessitated a classification of respondents on three levels: rection, content, and depth. The three levels were found to be too com- plex to combine into one measure of identification. Results relevant _tg fringe area social life 1- Extent of social life Marginals of the objective integration scale distribution indi- cfilmed that over forty percent of the respondents fall into the "high" and 0f the scale and approximately the same proportion occupy the "low" end. Such a distribution questions the validity of the conclusion that the fringe area is an "institutional desert" since approximately the same In ad- number of respondents are area participants as non-participants. dition, over ninety percent of the sample members may be‘classified as - 170 - positively identified with the area. Thus emotional bonds do exist in the fringe area, and to an overwhelming degree. The existence of integration and identification suggests that the descriptions of fringe area social life presented by other investi- gators cannot be generalized to the Lansing fringe. This criticism of previous research has further implications: it is suggested here that the conclusions of previous investigators are not relevant to rural-ur- ban fringe areas in general. Basically, this criticism is based upon two factors which supplement the findings of this study: 1) previous in- vestigators conducted their research in areas which are more clearly suburban than fringe, and 2) sociologists working in areas surrounding central cities have usually concentrated on recent arrivals, while ig- noring long-time residents. These two biases have created a false impression of fringe area social life. It is true that a large proportion of the residents do not participate in the area, but the proportion is not large enough to jus- tify the conclusion that the fringe lacks within-area participation. In addition, the position that residents of the area lack emotional bonds cannot be accepted. 2. Accounting for the adjustment categories After finding that integration and identification do exist in the Lansing fringe area, the focus of the study was shifted to variables which may be utilized to account for modes of adjustment. In general terms, it was ascertained that long-time, relatively stable respondents who have rural, noneLansi g backgrounds and who are committed to the area for various reasons tend to participate in the fringe area more fre- quently then sample members who exhibit converse characteristics. - 171 - Specifically, the following variables are significantly associated with a high degree of fringe area participation: lifetime residence; pre- WOrld War Two residence; migration from areas other than Lansing, East Lansing, or other Michigan cities of over 25,000 population; rural up- bringing; migration from a rural area; upbringing in the area itself or in the study counties (with the exception of Lansing or East Lansing); less than four changes in residence since l9#0; full or part-time farm- ing; non-Lansing employment; residence in study tiers six (the area fur- thest from Lansing) and four; and home ownership. Attempts to account for subjective identification met with fail- ure in the case of direction and success for both content and depth. An- alysis of the data revealed that content is significantly associated with length of residence, stabilitybmobility, rural-urban experience, Lansing- non-Lansing experience, and commitment to the area. The specific vari- ables affecting content of identification are as follows: year moved, life or non-life residence, place moved from, rural or urban previous residence, moving experience since 1940, rural or urban.upbringing, oc- cupation, place of work, tier of residence, tenure status, and education. Depth is significantly associated with the same general variables as content and with the following specific variables: length of resi- dence, moving experience since 1940, place of previous residence, rural or urban previous residence, rural or urban upbringing, occupation, place of work, age of respondents, and family gross weekly income. 0n closer analysis it was ascertained that all of the depth associations are based on a differentiation between respondents who had an initial choice for residence in the Lansing fringe area and those who did not. This means that the depth patterns are a function of whether or not - 172 - residents were coerced into residence in the area. 3. The affect of adjustment patterns Up to this point emphasis has been placed upon isolating the so- cial variables which affect adjustment patterns. In the following sec- tion the perspective was focused in the other direction, to determine the influence of adjustment categories on community orientations. The orientations considered were the following: localite perspectives, Lan- sing orientation, satisfaction with the area, and rural and urban defi- nition of the area. Attempts to account for the first two orientations, i.e., local- ite perspectives and Lansing orientation, were not successful. A signi- tficant association between satisfaction with the area and depth was :found, with respondents who were not coerced into fringe area residence «exhibiting the highest degree of satisfaction. This relationship sugb gests that pre-established community orientations significantly influ- ence satisfaction patterns; probably migrants moved into the area with positive inclinations toward it. . Rural and urban definitions of the area were found to be signi- ficantly associated with objective integration and with both content and and depth of identification. The data suggest that respondents who par- ticipate in the area, who identify with the area because of friendship patterns and.familiarity with it, and who had no prior choice for living in.the area, define their place of residence as rural, rather than urban. Generalizations In terms of the objectives of this study the most important find- ing'is that social life in the Lansing fringe area does not fit the -l'73 - stereotyped pattern described by previous investigators. It is suggested that the discrepancy is due to a conceptual confusion among sociologists since most "fringe studies" which preceded the present undertaking were not conducted in rural-urban fringe areas. Previous studies were car- ried out in suburbs, or subdivisions, or in other types of residential areas although the titles and universe of generalization indicate that the study site was a rural-urban fringe area. Concentration of previous investigators in non-fringe areas has led to the neglect of approximately one-half of the rural-urban fringe residents; this group is the farmers. Tonnies and Durkheim have pointed out that urban social formations necessitate a different order of inter- action than their rural counterparts. A co-existence of these divergent life styles is the unique characteristic of the fringe area. In such a unique situation of residential contact, what pattern of life emerges? Is there a tendency toward side-by-side existence, simple toleration, or a fusion of rural and urban ways of life? Generalizing from the data on the Lansing fringe, it seems that the first pattern has developed. The data indicate, for instance, that it is the lifetime-resident farmers who participate in the area. Then where do the factory workers who have recently moved to the area Partici- pate? The data suggest that these residents participate in functions located in their former area of residence. These, then, are the resi- dents who are in, but not 3f, the area. Therefore, the findings of the present study are in agreement with the conclusions of previous investigators only if the farmers are not considered. But one cannot ignore a large cement of residents and then generalize to the total population; the present study has pointed -174- to the importance of taking farmers into account in any study of social life in the rural-urban fringe. Like many research projects, the present study has brought into focus questions which should receive attention in future research. First of all, sociologists possess very little information about the adjust- ments forced upon farmers by the ecological invasion of urban migrants. How do the farmers perceive the situation, and how do they react to it? A longitudinal study of farmers' perceptions and reactions seems neces- sary. Perhaps much may also be gained by starting at the inner boundary of a given fringe area and working outward, with comparisons made by distance; this technique may allow a comparison along a continuum of ur- 'ban to rural predominance. In addition, the data suggest the possibility of conflict be- ‘tween farmers and nonfarmers in the area. For instance, one can visua- lize a conflict of interests in a town meeting called to decide upon the curriculum, or even the location, of a fringe area school. An import- ant question is: why was no open conflict found in the Lansing fringe area? Or, more generally, under what conditions might open conflict oc- cur? These are important questions which may be answered by comparing fringe areas, or by studying the same area over a period of time. Thirdly, the findings of the present study have important impli- cations for an understanding of social life in the city pg; g3, since migrants from the central city tend to return to their area of previous residence to carry on patterns of social interaction. This finding im- plies that if the Simmel-Wirth construct is accepted too literally, im- portant factors in urban life styles will be ignored. However, the pos- sibility does exist that some urban patterns of integration and identi- -175- fication are formulated by hindsight; such a possibility must be taken into account. Research oriented to this problem is needed in order to obtain an understanding of social life in the fringe area and the city. APPENDIX A RESIDENT ADJUSTMENT PATTERNS IN THE RURAL-URBAN FRINGE STUDY Interview Schedule Township Section lumber 1. When did.you.nove to this area? Year 2. Has your last address in either Lansing or East Lansing? Yes: What was it? No: where was it? (NAME or PLACE) (NAME or com (runs or sun) About how many people lived there? (GET A NUMBER) Did.you live in town or in open country? Town Open country 3. In what year did you move to this address? Year #. How many different homes have you lived in since l9#O? 5. How did you happen to move around here rather than some other part of the Lansing area? 6. _I_§_ m g your choice, would you continue living around here? Yes: What do you like about it? No: what do you dislike about it? Where would you.love? -l76 - -177- (IF A DEFINITE LOCATION IS GIVEN)What do you like about that place? Don't know: Do you have any reasons for wanting to stay? DO you have any reasons for wanting to love? (IF REASONS FOR MOVING ARE GIVEN) where would you want to move? (IF A DEFINITE PLACE IS GIVEN) who: do you like about that place? 7. Does this area have a name? Yes: What is it? No 8. Do you think of it as a part of a larger area? Yes: What is the name of this larger area? No 9. How well do you think that the people in this area know each other? (READ) _Not at all _Bot so well _Pairly well ___Quite well ___Very well (DON'T READ) Don’t know - 178 - 100 About how many of them would you say that you know by name? .____None ' ____A few .___‘About half Most All 11. About how many families around here do you come in contact with fer at least a few minutes every day or so? (NUMBER) 12. About how many do you spend a whole afternoon or evening with every now and then? (NUMBER) 13. a) In the previous questions we've been talking about people in this area. Now I'd like you to think about your best friends-- (pg,matter where they live. How, will you think fer a moment of your three best friends-~we'll call them 1, 2, and 5 because we don't want their names. Could.you please tell me where each of them lives? gamg g; plagg §3gggt Nearest Egg crossing streets 1. 2. 5. b) Now I'm interested in finding out the occupation of each of these people and whether or not any member of the family works with them. Could you please tell me about this? 1. 2. 3. - 14. How do you and.your (husband/wife) spend your (his/her) time when not working? Husband: Wife: Together: -179- 15. Do you have a T.V. set? No Yes: Can you receive the college station (WKAR-TV) on it? ___No: Are you planning to have your set converted to receive it? Yes No __;Yes: How often do you tune in the college station? What are your favorite programs? 16. What are your favorite radio programs? 17. Do you read any newspapers regularly? No ___Yes: Which ones? 18. Do you, or does any member of your family, read any magazines reg- ularly? NO ‘__;Yes: Could.you please give me the names of these magazines? 19. Do you and.your (husband/wife) belong to any groups or organiza- tions? _No (SKIP TO QUESTION 23) Yes :2. be A.. a » 2822 5 .55 .«o 6533 3338 as soofio so» on A3832 4 no means .203 .5 alalslmlml 2...»... z...»| zl»... slain... Galena: alalzlmlal ml»... 2...»... ml»... zlola... 2913...“: zlo mood when: decodes: s so» a». gases ooamo mom c.“ some?» a we: 50» 0.33 no 3583 we cowpuomonm eon) seen 50» one: as: o no so» so» .35 ...“an e .Hso -oH o «a oH A2. game can a 433 use same 235%; Baggage .35 gm was: 5583:» m» .383: 4 mo magma .203 ... .5 to ~33... o Aooazegooov as» one so» one 3333.33 ass: .8 21. 22. 23. -181- How did you come to join: Name Circumstances (INTERVIEJER: ASK THIS QUESTION ONLY IF THE RESPONDENT HAS INDI- CATED MEMBERSHIP IN MORE THAN ONE ORGANIZATION IN QUESTION 20) If you had to drop out of all the organizations you belong to ex- cept one, which would you choose to remain in? (NAME OP ORGANIZATION What are your reasons for choosing this one? What type of people belong to this organization? Is there any particular organization you'd like to join? _No (SKIP THE REST OP THIS qumTION) _Yes (GET THE POLLowINc INFORMATION ON EAOR) Name Location Why would you like to join? Are any of your friends members? _Yes _No How would you describe the people who belong? Are these people persons who are important? —182- No .__;Yes: In what way? Why haven't you joined this organization? 2#. Do you feel that membership in some of these organizations will help you, or has helped you to ”get ahead" in the world? No ___Yes: Which organizations and in what way? Name In what way? 25. Have you or your (husband/wife) taken a part in any voluntary com- munity groups such as the volunteer fire department or civil de- fense? No ___Yes: Which ones? 26. Do you know of any cases in which the people around here got to- gether to do something about an issue such as bond issues, de- cisions concerned with schools or the fire department, and so forth? No .__;Yes: What was the issue? Did you or your (husband/wife) participate? Yes No What were your (his, her) reasons for (not) participating? A} -183— 27. a) Are there any places around here where people get together in- formally or hang out? ‘ No .__;Yes: What? Where? Do you go there? ._;Y__N __Y_;N __Y;_N b) Do you go to any similar places outside of this area? No ___Yes: What? Where? 28. How do you keep up with the news of what is happening: right around here? in this township? in this county? in Lansing or E. Lansing? 29. We've been referring to the "area around here" for some time now, and I'd like to get an idea of what area you've been referring to. (INTERVIEWER: PRESENT A MAP TO THE RESPONDENT AND TAKE DOWN SPE- OIFIC STREET NAMES) HOw far north does it go? How far east does it go? How far south does it go? How far west does it go? How did you decide on these boundaries? -l8.’.- 50. Here are two ways of classifying the t es of areas people live in (INTERVIEWER: HAND CARD TO RESPONDENT . Which do you think makes more sense in terms of the way things actually are? _Rural-urban: Which do you think you live in? (IF "RURAL" IS SELECTED SKIP QUESTIONS 3]. and 33b) _Rural-urban-in between: Which do you think you live in? What would you call this third area? 31. a) (DO NOT ASK IP "RURAL" WAS SELECTED IN QUESTION 30) Do you think that living in an outlying farm area would have advantages over the area in which you are now living? ‘ No _Yes: What are some of these? b) (DO NOT ASN IP "RURAL" WAS SELECTED IN QUESTION 30) Do you think that the area in which you are now living has advantages over living in the outlying farm area? No _Yes: What are some of these? 32. a) Do you think that living in Lansing would have advantages over living in this area? No Yes: What are some of these? b) Do you think that the area in which you are now living has ad- vantages over living in Lansing? NO Yes: What are some of these? 35- 35- ~185- a) Do you think that there are differences in the kinds of people who live in an area like this and city people? No ___Xes: What are some of these? b) (DO NOT ASK IF "RURAL" WAS SELECTED IN QUESTION )0) Do you think that there are differences in the kinds of people who live in an area like this and a much more rural area? No Yes: What are some of these? Some people say that farm people think differently from city dwell-1 ere. Do you agree? ___No: Why do you feel this way? _Don' t know: ___Ies: Does an area like this come closer to the farm or city way of thinking? .__;Fanm: In what ways? ___JCity: a) Does where a person lives have anything to do with "getting ahead" in the world? No .__;Yes: In what ways and areas? b) In.your own case, do you think that living in this area has helped you to "get ahead"? No .__;Yes: In what ways? 56. 57. 39. - 186 - a) What kinds of people would you like to see move into this area and become your neighbors? b) What kinds of people who live in this area now would you like to see move out? About two years ago an attempt was made to organize a group called the LansingVMetropolitan Planning Commission. The purpose of the group was to help coordinate the development of the City of Lansing and Ingham, Eaton, and Clinton Counties by collecting information and formulating plans for the total.Metropolitan Area. The group was nonsprofit and had no political power. This area would have been one of those covered by the commission's work. Do you think that the idea of starting such an organization was a good one? .__;Yes: What are your reasons for feeling this way? No: .___Don't know: Why do you have doubts about it? Do you think this area is getting better or worse? Don't know .___Getting better: .__JGetting worse: What has been happening' Can.you give me to make you feel this some examples? way? What sort of facilities would you like to see established around here? Anonpo no enmov nonundm uobnasaon anon» 65nd 33> genome nne>na no poaoagmns -137- nounen hunnon no nonunonndn onoanmsnu m munnuoao new menu cube: d can manaonnmkaau nonnooonm new menu weioufllno an «83 ~23 3 23.3 flown 131332 .3 WIHIIIIS 8 23).... .3 «.55: E3 Agar. .5 ago 3 33 98». $383 3 on 825 snowmen and ones» and - 188 - (QUESTIONS 41.-#3 APPLY ONLY To THOSE we DO NOT MARKET A FARM PRODUCT) #1. Do you raise or produce any food of your own? _No (SKIP To QUESTION M) ___Ies: What prOportion of your family food cost would.you say your own production covers? _Less than #1: _About 1/2 _More than 3/# _About 1/1: ___Be tween V2 _Don' t know and 3/4 ___Detween.yh ___WOn't answer and 12’2 _About 3/4 #2. Do you ever find it necessary to ask the advice of other people about your garden? No ___Ies: Can you please give me the location and occupations of these people? Location (SPECIFIC STREETS) Person's Occupation 45. Do you ever find it necessary to borrow tools or equipment for use in your garden? No ___Ies: Can you please tell me about these people from whom you borrow? Where do they live and what are their occupations? Location (SPECIFIC STREETS) Person's Occupation _ 139 - hk. Do you or your (husband/wife) ever lend out anything? No ‘__;Ies: Can you please tell me about these people to whom you lend. Where do they live and what are their occupations? Location(SPECIPIC STREETS) Person's Occupation #5. Are there any types of persons you wouldn't want to lend things to? I wouldn't want to lend out anything at all. No ___Xes: What types of persons are these? (PROBE FOR OCCUPATIONAL, LOCATIONAL, ETHNIC GROUP, ETC. REASONS RATHER THAN AC- CEPT§RG ANSWERS SUCH AS "PEOPLE WHO WON'T RETURN THEM," ETC. #6. (DO NOT ASK OF PART OR FULL-TIME FARMERS) What is the occupation of the head of this household? Where does he work? (GET PLACE AND LOCATION) #7. What kinds of paying Jobs has the major family wage earner had since 19%? (RECORD IN ORDER OR SEQUENCE) #8. #9. 50. - 190 - (ASK ONLY IF RESPONDENT IS NOT A EARNER AND IF FARMING WAS NOT MEN- TIONED IN QUESTION #7) Have you (or, refer to the major wage earn- er) ever worked on a farm? .___No ___Yes: FOr how long a period of time? As a farm operator or a hired hand? I___Parm.operator Hired hand (ASK ONLY OF NON-FULL TIME FARMERS) How do you usually get to work? own car .__;bus picked up by another person .__;walk ‘___other (WRITE IN) HOw long does it take you to get to work? (IF VARIES, GET AVERAGE FOR LAST WEEK) What would.you do by way of an occupation if you were Just start- ing out now? Why that? (ASK UNLESS PRESENT OCCUPATION IS GIVEN) What are the things that make this preferable to what you're doing now? (ASK IF NO CLEAR-CUT OCCUPATIONAL CHOICE OR SET OF REASONS IS GIV- EN) Do you ever think of getting out of (PUT IN PRESENT OCCUPATION) and starting over again? ‘__;Yes: Why do you feel this way? No: (ASK ONLY OF THOSE WHO SAY "YES") What would you rather do? 51. 52. -191- Why that? Could you please tell me why you are a (PUT IN OCCUPATION)? (ASK ONLY OP FARMERS) I'd like you to try to describe what seems to you to be the most important features of farming as a way of life and an occupation. Here is a list of phrases that might be used to describe farming. (INTERVIEWER: HAND CARD TO RESPONDENT) Could you pick out those that best describe your feelings? _ _"dirty work" _ _"poor Job" _ _"long hours" __ _"clean work" _ _"seasonal Job" _ ___"free time when- __ "easy work" __ _"busy all year ever you want it" _ _"hard work" round" __ _"takes brains" __ _"creative work" __ _"boring work" _ _"don't have to __ _"healthy work" __ _"always different be smart" _ _"unhealthy work" jobs involved" _ _"don't need a _ _"outdoor wor " __ _"a man is his own big formal edu- __ "takes special boss" cation" skill that not _ _"a man is always __ _"poor paying everyone has" tied down by his work" __ _"takes a lot farm" __ _"good paying of money" _ _"have to be a work" mechanic" _ _(OTHER) Which one phrase fits best? (PLACE A "1" IN THE SECOND SPACE) Which phrase fits second best?(PLACE A "2" IN THE SECOND SPACE) Which phrase does the poorest job of describing taming? (PLACE A "P" IN THE SECOND SPACE) 53. 5#. -l92 - Here is a list of phrases that might be used to describe (USE RE- SPONDENT‘S OCCUPATION. IF RESPONDENT IS A FARMER, USE "FACTORY WORK." IF THE RESPONDENT IS A PART-TIME FARMER, USE THE RESPOND- ENT'S NON-FARM OCCUPATION). (INTERVIEWER: HAND CARD TO RESPONDENT) Could you pick out those that best describe your feelings? _ _"dirty work" __ _"takes special __ _"have to be a __ "clean work" skill that not mechanic" _ _"easy wor " everyone has" __ _"long hours" __ _"hard work" __ _"poor Job" _ _"takes brains" __ _"creative work" __ _"seasonal job" __ _"don't have to __ _"healthy work" __ _"busy all year be smart" __ _"unhealthy work" round" __ _"don't need a big __ _"takes a lot of __ _"boring work" formal education" money" __ _"poor paying _ _(OTHER) _ _"good paying work" wor " Which phrase fits best? (PLACE A "1" IN THE SECOND SPACE) Which phrase fits second best? (PLACE A "2" IN THE SECOND SPACE) Which phrase does the poorest job of description? (PLACE A "P" IN THE SECOND SPACE) Now I'd like to get some background information from you which we will need for interpreting parts of the study. First of all, would you mind telling me where you grew up? _Lansing or East Lansing Elsewhere : Where? (TOWN) (COUNTYT (STATE)w Did you live in town or open country? Open country _Town: About how many people lived there then? - 193 - 55. We'd appreciate knowing who else lives with you and what their ap- proximate ages and education are. Last grade Relationship Approximate completed Marital to respondent age in school Sex status Respondent ._JN_;P __}1_;F __}{_;P __}{_;F N F 56. Do you own or rent this place? Own ___Rent 57. Fbr the purposes of our survey, we need to have a rough idea of your family's total weekly income. Would you mind telling me into which of these classes it falls?(INTERVIEWER: HAND CARD TO RE- SPONDENT) (INTERVIEWER: ENTER CODE NUMBER) *I-i‘l’fl’fl'l'i’l’i‘i-H‘I” Thank you very much for your cooperation. If you.are interested in see- ing the results of this research, we'd be glad to send them to you, but we'd need your name and address. Name Address INTERVIEWER FILL OUT THE FOLLOWING: PLACE OF INTERVIEW: County Township Date of interview Interviewer's name APPENDIX B TABLES TABLE 2 PERCENTAGE DISTRIBUTION OF MALE WORKERS BY INDUSTRY GROUPS FOR LANSING, 1950 CENSUS Industry;group Percent Manufacturing ##.2 Construction ‘ 7.0 Wholesale and retail trade 19.1 Transportation, communication, public utilities, government 13.1 Personal services, business and repair services, other 6.0 Other services, protection, and nonsdomestic personal 10.6 TOtal 100e 0 TABLE 3 PERCENTAGE DISTRIBUTION OF MALE WORKERS BY MAJOR OCCUPATIONAL GROUPS FOR LANSING, 1950 CENSUS Occupational group Percent Professional and semi- professional 10.9 Sales, clerical and related 17.7 Skilled, semi-skilled, unskilled workers and foremen 53.} Domestic and service workers 8.# Self-employed, retired, and others 9.7 Total 100.0 — 194 - -195- TABLE # PERCENTAGE DISTRIBUTION OF HOUSEHOLD HEADS BY AGE FOR LANSING SAMPLE Age of household heads Percent Under 50 21.2 30 - 59 23st} 1:0 - 49 21.3 50 ‘ 59 1908 60 and over 1#.5 TOtal 100.0 TABLE 5 PERCENTAGE DISTRIBUTION OFIMALE WORKERS BY EDUCATIONAL ATTAINMENT FOR LANSING SAMPLE Educational attainment Percent None and l - 6th 11.2 7th and 8th 21.0 Some high school--did not graduate 2#.5 High school graduate 57.0 Some college and college graduate 6.5 Total 100.0 TABLE 6 PERCENTAGE DISTRIBUTION OF 19#9 FAMILY INCOME LANSING, 1950 CENSUS Family income in 19#9 Percent Under 32000 10.7 32000 - 82999 13.1 3#OOO - 84999 18.8 85000 - 85999 13.7 86000 - 86999 7.5 87000 and over 12.3 Total 100.0 Median: 8#097 - 196 - TABLE 7 PERCENTAGE DISTRIBUTION OF BIRTHPLACE OF WORKERS FOR LANSING SAMPLE Place of birth Percent Lansing 20.6 Elsewhere in.Michigan--urban 19.0 Elsewhere in.Michigan--rura1 2#.0 Adjacent states 12.6 Southern states 11.# All other states 6.} Foreign born 6.1 Total 100.0 TABLE 8 YEAR MOVED TO THE AREA Year moved? Number Percent Before 1950 #7 2#.9 1930 - 193# 8 4.2 1935 - 1939 10 5-3 19#O - l9#2 6 5.2 1943 - 19h5 10 5.3 l9#6 - 1948 24.5 13.0 19u9 - 1951 “ZeS 2205 1952 - 195“ 34.5 18.2 .After 1954 6 3.2 No answer .5 .5 Total 189 100.0 57 respondents, or 19.6%, were born in the area; in these cases the "year moved" category is actually the year of birth. - 197 - TABLE 9 DIFFERENCE BETWEEN YEAR.MOVED TO THE AREA AND YEAR MOVED TO PRESENT ADDRESS Difference in years Number Percent Same l#5 75.7 One to five 12.5 6.6 Six and over 52.5 17.2 No answer 1 .5 Total 189 100.0 TABLE 10 NUMBER OF RESIDENCES OCCUPIED SINCE 19#O Number of residences Number Percent One “A 2303 Two 52.5 27.8 Three #8 25.# Fbur 13 9-5 Five 6 5.2 Six and over 1# 7.# No answer 6.5 5.# Total 189 100.0 TABLE 11 PLACE OF PREVIOUS RESIDENCE a Previous residence Number Percent Lansing or East Lansing 87.5 57.6 Other Michigan city of 25,000 or more 8.5 5.6 Study counties (excluding Lansing and East Lansing) 33.5 22.0 County contiguous to study counties 8 5.5 Other Michigan area 1.5 1.0 Midwestern state other than Michigan 6 5.9 Southern state # 2.6 Eastern state 2 1.5 No answer 1 . Total 152 100.0 The 57 respondents who were born in the area are excluded from this table. -198- TABLE 12 RURAL AND URBAN NATURE OF PREVIOUS RESIDENTIAL AREA * Type of area Number Percent Rural 26 17.1 City 125 82.2 No answer 1 .7 Total 152 100.0 * The 57 respondents who were born in the area are excluded from this table. TABLE 15 AREA IN WHICH HEAD OF HOUSEHOLD WAS RAISED Area raised Number Percent In the area 57.5 19.8 Lansing or East Lansing #7.5 25.1 Other.Michigan city of 25,000 or more 1#.5 7.7 Study counties (excluding Lansing, East Lansing and in the area) 27.5 1#.5 County contiguous to study counties 1#.5 7.7 Other Michigan area 13 9-5 Other state 25.5 15.5 Foreign country 5.5 1.8 No answer .5 .5 Total 189 99.9 TABLE 1# RURAL AND URBAN NATURE OF AREA IN WHICH HEAD OF HOUSEHOLD WAS RAISED Type of area Number Percent Rural 87 #6.0 No answer .5 .5 Total 189 100.0 - 199 - TABLE 15 OCCUPATIONS 0F SAMPLE MEMBERS Occupation Number Percent Factory work 68 56.0 Farmer #5 25.8 Farmer and factory worker 15 7.9 Farmer and other 8 #.2 Urban retail, store 15 6.9 Has own business 10 5.5 Professional 9 #.8 Sales 6 5.2 Government employee # 2.1 Construction work 5 1.6 Retired 6 5.2 Other 2 1.1 Total 189 100.1 TABLE 16 LANSING AND NON-LANSING PLACE OF WORK Place of work Number Percent WOrks in Lansing 117.5 62.2 Does not work in Lansing 71.5 57.8 Total 189 100.0 TABLE 17 TIER 0F RESIDENCE Mean number of respond- ents in each Tier Number Percent section One (the 28 sections closest to Lansing) 9.5 5‘0 ~54 Two (56 sections) 2# 12.7 .67 Three (## sections) #0 21.2 .91 FOur (52 sections) 2# 12.7 .#6 Five (60 sections) 55 17.5 .55 Six (the 68 sections furthest from Lansing) 58.5 50.9 .86 Total 189 100.0 - 200 - TABLE 18 TENURE STATUS Tenure status Number Percent Own 167 88.# Rent 22 11.6 Total 189 100.0 TABLE 19 AGE OF HEAD OF HOUSEHOLD Age Number Percent Under 50 years of age 16 8.5 50 - 3“ 1905 1003 35 - 39 ' 35.5 18.8 ['0 - 41* 50 1509 A5 - A9 22 11.6 55 - 59 15 7-9 60 - 6# 15 6.9 65 years of age and over 17 9.0 Total 189 100.0 TABLE 20 SCHOOLING 0F HEAD OF HOUSEHOLD Last school grade completed Number Percent let to 5th grade 5 1.6 Grade school graduate 50 26.5 let to 5rd year of high school . 5#.5 18.2 High school graduate 5#.5 28.8 Some college 10.5 5.6 College graduate and over 1# 7.# No answer 11 5.8 Total 189 100.0 Median: 5.15 years of high school. - 201 - TABLE 21 NUMBER OF CHILDREN UNDER SEVENTEEN Number under seventeen Number Percent None 73-5 58.9 One 33 17.5 Two 59.5 20.9 Three 2# 12.7 Four 12 6.5 Five and over 7 5.7 Total 189 100.0 TABLE 22 FAMILY GROSS WEEKLY INCOME Income Number Percent Less than 8#O 7.5 #.O ‘ “O ’ ’ 59 1505 802 8 6O - 8 79 19.5 10.5 ‘ 80 ' T 99 #1 2107 8100 - 812# 55.5 17.7 8125 - 81#9 19.5 10.5 8150 - 8199 13.5 7.1 8200 and over 57.5 19.8 No answer 1.5 .8 Total 189 99.9 I Median: 8107.65 - 202 - TABLE 25 WHY THE RESIDENTIAL AREA WAS CHOSEN* Reason for choosing the area Number Percent Indication of a positive choice for the neighborhood 15.5 8.2 Indication that respondent would have moved anywhere for what was wanted (best buy, this was available, etc.) 98 51.8 Wanted to live in an unsettled area 50 15.9 Always lived here 27.5 1#.5 Relatives or friends in the area 22 11.6 Closer to work, business reasons 18 9.5 Convenient for living in the country while near town 17 9.0 Inherited, belonged to relatives 10.5 5.6 Familiar with the area, lived here before 6.5 5.# Other 7 3-7 No answer 1 .8 ‘I Number and percent are above the sample base because some respondents gave multiple answers. In subsequent tables this situation will be in- dicated by the phrase "multiple answers." TABLE 2# OPINIONS 0F WHETHER THE AREA HAS A NAME Opinion Number Percent No, the area does not have a name 155 70.# Yes, the area does have a name 56 29.6 Total 189 100.0 TABLE 2#a NAMEGIVENTOTHEAREA* Name given Number Percent Local name 19.5 5#.8 Bordering town 15 26.8 Township name 1#.5 25.9 School district 7 12.5 Total 56 100.0 * This table is based on the 56 respondents who indicated that the area has a name. -203 - TABLE 25 OPINIONS OF WHETHER THE AREA IS PART OF A LARGER AREA Opinion Number Percent No, the area is not part of a larger area 126 66.7 Yes, the area is part of a larger area 65 55.5 Total 189 100.0 TABLE 258 NMME OF THE LARGER AREA* Name of the larger area Number Percent Township name 20 51.7 Bordering town 20 51.7 Lansing 15 20.6 Local name A 6.5 School district 2 5.2 Other name 4 6.5 Total 63 99.8 * - This table is based on the 65 respondents who indicated that the area is part of a larger area. TABLE 26 OPINIONS OF WHETHER RURAL, URBAN, AND "IN-BETWEEN" AREAS EXIST, AND THE TYPE CHOSEN T0 DESCRIBE THE FRINGE Areas indicated and chosen Number Percent Rural-urban, chose rural 109.5 57.9 Rural-urban, chose urban 8.5 #.5 Rural-urban-in—between, chose rural 15.5 8.2 Rural-urban-in-between, chose urban .5 .5 Rural-urban-in-between, chose in-between 54.5 28.8 NC answer .5 o 3 Total 189 100.0 - 204 - TABLE 27 OPINIONS OF WHETHER THERE ARE DIFFERENCES BETWEEN PEOPLE IN THE RESIDENTIAL AREA AND CITY PEOPLE Opinion Number Percent There are no differences between residen- tial area people and city people 120 66.7 There are differences between residential area peOple and city people 57 51.7 No answer 5 1.7 Total 180 100.1 ‘— w The 9 respondents who indicated that they are living in an urban area (see Table 26) have been eliminated from this table. TABLE 27a STATED DIFFERENCES BETWEEN RESIDENTIAL AREA PEOPLE AND CITY PEOPLE Stated differences Number Percent More friendly here, more neighborly 55 57.9 Different social life 5 8.8 Occupational difference 5 8.8 City people think only of themselves A 7.0 City people are always in a hurry, rushing 5 5.5 Fanm people are harder workers 2.5 4.# Farmers are behind the times 1.5 2.6 City people think that farmers are ignorant 1.5 2.6 Other 6.5 11.# * This table is based on the 57 respondents who indicated that there are differences between residential area people and city people. Multiple answers. - 205 - TABLE 28 OPINIONS 0F WHETHER THERE ARE DIFFERENCES BETWEEN PEOPLE IN THE RESIDENTIAL AREA AND RURAL PEOPLE Opinion Number Percent There are no differences between residen- tial area people and rural people 45.5 68.5 There are differences between residential area people and rural people 18.5 29.1 No answer 1.5 2.# Total 65.5 100.0 * The 125 respondents who indicated that they are living in a rural area (see Table 26) have been eliminated from this table. TABLE 28a STATED DIFFERENCES BETWEEN RESIDENTIAL AREA PEOPLE AND RURAL PEOPLE Stated differences Number Percent Rural people are friendlier, more cooperative 6 52.# Rural people are more provincial 5.5 29.7 People here are more interested in the city 2.5 15.5 Farmers have one-track minds 2 10.8 People here have more money 1 5.# Other 5 16.2 * This table is based on the 18.5 respondents who indicated that there are differences between residential area people and rural people. Mul- tiple answers. TABLE 29 OPINIONS 0F WHETHER FARM.PEOPLE THINK DIFFERENTLY FROM CITY DWELLERS Opinion . Number Percent Yes, farm and city people think differently 92 #8.? No, farm and city people think the same 92 48.7 No opinion 4 2.1 No answer 1 .5 Total 189 100.0 - 206 - TABLE 29a OPINIONS OF WHETHER PEOPLE IN THE AREA COME CLOSER TO THE FARM 0R CITY WAY OF THINKING Opinion Number Percent Closer to the farm way of thinking 70 76.1 Closer to the city way of thinking 19 20.7 Both ways of thinking are represented 5 5.5 Total 92 100.1 <5 This table is based on the 92 respondents who indicated that farm and city people do not think alike. TABLE 50 OPINIONS OF THE SIZE OF THE LOCAL AREA '5 Size of local area Number Percent Less than 4 miles 52 16.9 h miles (usually "one square mile") 5h 28.6 #01 - 8 miles 5805 200‘} 8.1 - 12 miles 22 11.6 12.1 - 20 miles 17 9.0 2001 - 40 miles 8 #02 More than #0 miles 4 2.1 Name of nearby town given 5 2.6 "Immediate vicinity" 5 2.6 Answered in terms of sections 2 1.1 Other criteria 1.5 .8 Total 189 99.9 * Mileage refers to number of miles, not to square miles. - 207 - TABLE 50a HOW "AREA AROUND HERE" BOUNDARIES WERE DECIDED UPON Criteria utilized for delimiting "area around here" boundaries Number Percent People I know 80 #2.5 Friends live within this area 55 17.5 "Built-up" in this area 2% 12.7 School district 25.5 12.9 Meetings, groups, or organizations within this area 15 6.9 Shopping area 10.5 5.6 These are town or city limits 8 4.2 Roads frequently travelled 5 1.6 Houses I can see 2.5 1.5 Other 1 5 6 e 9 No answer 1.5 .8 * Multiple answers. TABLE 51 OPINIONS 0F WHETHER RESPONDENTS WOULD LIKE TO CONTINUE LIVING IN THE AREA Opinion Number Percent Yes, would like to continue living in the area 175 91.5 No, would not like to continue living in the area 15.5 7.1 Cannot make a choice 2.5 1.5 Total 189 99-9 - 208 - TABLE 51a REASONS FOR wANTINc TO REMAIN IN THE AREA* Reasons for wanting to remain Number Percent Not crowded #5 26.0 Quiet, peaceful, freedom, country life, fresh air 56 20.8 Friendly neighbors 28.5 16.5 Good farm land, near farm market 25 15.5 Good for children 19.5 11.5 Born or raised here 18.5 10.7 Close to town and yet outside 16 9.2 Likes the facilities 15.5 9.0 Close to work 11.5 6.6 Can garden 11 6.# Friends are here 11 6.# Good neighborhood 10.5 6.1 Relatives here, near relatives 10 5.8 Likes the weather, scenery 7.5 h.5 It feels like home 5.5 5.2 Other 10 5.8 * This table is based on the 175 respondents who indicated that they would like to continue living in the area. Multiple answers. TABLE 51b REASONS FOR WANTING TO LEAVE THE AREA* Reasons for wanting to move Number Percent Complaint about facilities 9 29.6 Too far from relatives A 29.6 Weather 4 29.6 Too far from Lansing 5 22.2 Taxes are going up 1.5 1.1 Becoming crowded 1 7.4 Farming is suffering because of growth 1 7.# Net suitable for farming 1 7.# Other 5 22.2 * This table is based on the 15.5 respondents who indicated that they would not like to continue living in the area. Multiple answers. - 209 - TABLE 52 OPINIONS 0F WHETHER THE AREA IS GETTING BETTER 0R WORSE Opinion Number Percent Getting better 107 56.6 Getting worse 8.5 4.5 Staying the same 25 12.2 Can't make a decision 50.5 26.7 Total 189 100.0 TABLE 52a * OPINIONS OF WHY THE AREA IS GETTING BETTER Opinion Number Percent Nice houses are being built 52.5 49.1 More and/or better facilities 25.5 25.8 It's growing fast 22 20.6 Property values are going up 10.5 9.8 Business or occupational improvement 9.5 8.9 People are interested in the area 5.5 5.1 In terms of the types of people moving in 4.5 4.2 Other 1.5 1.4 * This table is based on the 107 respondents who indicated that the area is getting better. Multiple answers. — 210 - TABLE 33 FACILITIES DESIRED IN AREA* Facilities desired Number Percent None 85.5 44.2 Roads: better, improved, paved 41 21.7 Sewage 21 11.1 Community hall, recreation hall, recreation facilities 18.5 9.8 Bus service 9.5 5-0 Schools: better 10 5.5 Schools: closer 5 2.6 Fire protection, police protection 5 2.6 Industry 4 2.1 Zoning regulations: instituted or enforced 2.5 1.5 Other 16 8.5 No answer .5 .5 i ‘ ‘ L w’ ‘—“““ .— Multiple answers. TABLE 54 OPINIONS 0F WHETHER LIVING IN LANSING,HAS ADVANTAGES OVER LIVING IN THE AREA Opinion Number Percent Present residence area has advantages 182 96.5 Lansing has advantages 58.5 50-9 * Multiple answers. - 211 - TABLE 54a STATED ADVANTAGES OF LANSING AND PRESENT RESIDENTIAL AREA * Lansing advantages Number Percent Closer to work 22.5 58.5 More facilities, better facilities 22 57.6 Closer to stores 10 17.1 Reference to advantages of city occupa- tions (work hours, money, easier work) 7 12.0 Friends in Lansing 5.5 6.0 Business reasons 2.5 4.5 More clubs -1 1.7 * This portion of the table is based on the 58.5 respondents who indi- cated that Lansing has advantages. Multiple answers. * Present residential area advantages Number Percent Not congested 85.5 47.0 Healthier, quiet, fresh air, clean 74.5 40.9 Room for children to play 50.5 16.8 More freedom, privacy 27.5 15.1 Can garden or farm here 27 14.8 Friendlier, more neighborly 15 7.1 Cheaper living 11.5 6.5 Business reasons, closer to work 9 4.9 Other 1105 603 * This portion of the table is based on the 182 respondents who indi- cated that their present residential area has advantages. Multiple answers. TABLE 55 OPINIONS OF WHETHER LIVING IN THE OUTLYING FARM* AREA HAS ADVANTAGES OVER PRESENT RESIDENCE AREA Opinion Number Percent Present residence area has advantages 50.5 79.5 Outlying farm area has advantages 12.5 19.7 No answer .5 .8 Total 65.5 100.0 * The 125 respondents who indicated that they are living in a rural area (see Table 26) have been eliminated from this table. _ 212 - TABLE 55a STATED ADVANTAGES OF OUTLYING FARM AREA AND OF PRESENT RESIDENTIAL AREA * Outlying farm area advantages Number Percent Cheaper living 5.5 44.0 More freedom, more privacy 5 40.0 More cooperative, people are friendlier 4 52.0 Similar interests 2.5 20.0 * This portion of the table is based on the 12.5 respondents who in- dicated that the outlying farm area has advantages. Multiple answers. Present residential area advantages* Number Percent More facilities, better facilities 22.5 42.1 Closer to town 19 55.5 Closer to work, business reasons 12.5 25.4 Closer to shapping places 5 9.5 Friends are here 5.5 6.5 Other 2.5 4.7 i This portion of the table is based on the 55.5 respondents who indi- Multiple cated that their present residential area has advantages. answers. TABLE 56 OPINIONS OF HOW WELL PEOPLE IN THE AREA KNOW EACH OTHER Opinion Number Percent Not at all 1 .5 NOt 30 Well 2405 1500 Fairly well 95-5 50-5 Quite well 56 19.0 Very well 25-5 15-5 Cannot make a choice 6.5 5.4 Total 189 99-9 - 213 - TABLE 57 HOW MANY PEOPLE IN THE AREA RESPONDENTS KNOW BY NAME Number known by name Number Percent None 7 5.7 A few 78 “105 About half 53 17-5 Most 47.5 25.1 All 25.5 12.4 Total 189 100.0 TABLE 58 NUMBER OF FAMILIES RESPONDENTS COME INTO CONTACT WITH FREQUENTLY Number of families Number Percent None 55-5 29-4 One 44.5 25.5 Two 58.5 20.4 Three 21.5 11.4 Fbur 15.5 8.2 Five 4 2.1 Six and over 9 4.8 No answer .5 .5 Total 189 100.1 TABLE 59 NUMBER OF FIMILIES WITH WHICH RESPONDENTS SPEND AFTERNOONS OR EVENINGS Number of families Number Percent None 75 59-7 One 60 51.7 TWO 5#.5 1802 Three 9.5 5.0 Four and over 9.5 5.0 No answer .5 .5 1 Total 189 99-9 -214- TABLE 40 WITHIN-AREA LENDING PATTERNS Indication of whether anything is lent Number Percent No, nothing is lent out in the area 125 65.1 Yes, things are lent out 66 54.9 Total 189 100.0 TABLE 41 AREAS IN WHICH FRIENDS ARE VISITED* Area Number Percent Never visit friends 5.5 2.9 Lansing 110 58.2 Local area 95-5 50-5 Other 52.5 17.2 * Multiple answers. TABLE 42 RESIDENTIAL AREA OF THREE BEST FRIENDS Residential area Number Percent All in present neighborhood 54 18.0 All in Lansing 45.5 24.1 Mixture of present neighborhood and Lansing 35 18.5 All in towns within the study counties 9 4.8 Mixture of present neighborhood and study county towns 7.5 4.0 Mixture of Lansing and study county towns 14 7.4 All in other area 15 6.9 Mixture of other area and present neighborhood 12 . 5 6 . 6 Mixture of other area and Lansing 15 6.9 Mixture of other area and study county towns 4.5 2.4 No answer 1 .5 Total 189 100.1 - 215 - TABLE 45 NUMBER OF ORGANIZATIONS BELONGED TO Number of organizations Number Percent None 95 50-3 One 42.5 22.5 Two 20 10.6 Three 15.5 7.1 Fbur 6.5 5.4 Five and over 11.5 6.1 Total 189 100.1 TABLE 45a MEETING PLACES OF ORGANIZATIONS BELONGED TO* Organization meeting place Number Percent Local town 50 55.2 Neighborhood 59 41-5 Lansing 52.5 54.6 Other location 6.5 6.9 * This table is based on the 94 respondents who indicated that they belong to organizations. Multiple answers. TABLE 45b TYPE OF ORGANIZATIONS BELONGED TO* Organization type Number Percent School 28.5 50.5 Farm 28 29.8 Church 28 29.8 Fraternal 26.5 28.2 Local (sewing club, bridge, etc.) 9.5 10.1 Business 7 7-4 Veteran 6 6.4 Professional 5.5 5.9 Other 9 9.6 No answer 1 1.1 * This table is based on the 94 respondents who indicated that they belong to organizations. Multiple answers. - 216 - TABLE 44 PARTICIPATION IN LOCAL INFORMAL FACILITIES Participation Number Percent Does not participate 172 91.0 Does participate 17 9.0 Total 189 100.0 TABLE 45 PARTICIPATION IN VOLUNTARY COMMUNITY GROUPS Participation Number Percent Has not participated 144.5 76.5 Has participated 44.5 25.5 Total 189 100.0 TABLE 45a *- TYPE OF VOLUNTARY GROUPS IN WHICH PARTICIPATED Type of group Number Percent Civil defense 21 47.2 Community fund raising campaign (Red Cross, Community Chest, etc.) 20 44.9 Volunteer fire department 8.5 19.1 Other 1.5 5.4 * This table is based on the 44.5 respondents who indicated that they have participated in volunteer community groups. TABLE 46 PARTICIPATION IN LOCAL ISSUES Multiple answers. Participation Number Percent Did not participate 156 72.0 ‘Participated 55 28.0 Total 189 100.0 - 217 - TABLE 46a TYPE OF ISSUE IN WHICH PARTICIPATED* Type of issue in which participated ' Number Percent School 55 100.0 Zoning 5.5 10.4 Fire department 7 15.2 Hospital 105 208 *- This table is based on the 55 respondents who indicated that they have participated in local issues. Multiple answers. TABLE 46b * REASONS FOR PARTICIPATING IN LOCAL ISSUES Reasons for participating Number Percent Would benefit the children 25.5 44.5 Tax rate affected 12.5 25.6 To improve the community 11 20.8 Felt it was needed, interested 8.5 16.0 Was board member, school teacher, fire department member 5 5.7 * This table is based on the 55 respondents who indicated that they have participated in local issues. Multiple answers. TABLE 460 REASONS FOR NOT PARTICIPATING IN LOCAL ISSUES* Reasons for not participating Number Percent No knowledge of the issue 122 89.7 "Beyond control" reason (working at the time, not around at the time, health reasons) 7-5 5-5 No interest, didn't concern me 5.5 2.6 Other reasons 5 2.2 Total 156 100 O I- This table is based on the 156 respondents who indicated that they have not participated in local issues. - 218 - TABLE 47 AREAS UTILIZED FOR SHOPPING Shopping area Number Percent All in Lansing 76 40.2 More than half in Lansing but not all 58 20.1 All in area (including surrounding towns) 10 5.5 More than half in area but not all 48 25.4 50-50, Lansing and area 17 9.0 Total 189 100.0 - 219 - TABLE 48 SOURCES UTILIZED FOR NEWS ABOUT THE AREA, THE TOWNSHIP, THE COUNTY, AND LANSING Immediate area news Number Percent Doesn't keep up 46.5 24.6 Neighbors, friends 100.5 55.2 Local source other than neighbors and friends 55.5 18.8 Paper, radio, television 28.5 15.1 No answer 2 1.1 Township news Doesn't keep up 46 24.5 Neighbors, friends 54 28.6 Local source other than neighbors and friends 55 28.0 Paper, radio, television 59.5 51.5 No answer 2 1.1 County news Doesn't keep up 19.5 10.5 Neighbors, friends 25.5 12.4 Local source other than neighbors and friends 49 25.9 Paper, radio, television 116 61.4 Other .5 .5 No answer 2 1.1 Lansing'news Doesn't keep up 8 4.2 ”Giglbors, friends 1505 802 Local source other than neighbors and friends 5.5 2.9 Paper, radio, television 175.5 91.8 Other 305 1.8 No answer 5 1.6 * Multiple answers. - 220 - TABLE 49 * LEISURE TIME ACTIVITIES Activity Number Percent Watch television 92.5 48.9 Sports 65.5 55.6 Goes to movies 41 21.7 Garden 40.5 21.4 Listens to the radio 55 17.5 Visit friends 52 16.9 Read 25 15.2 Attend meetings 25.5 12.4 Sew, knit 15.5 8.2 WOrk on house 11.5 6.1 Visit relatives 5.5 2.9 Other 5 206 1 Multiple answers. TABLE 50 FOOD RAISING BY NONPARMERS* Indication of whether or not food is raised Number Percent Food is not raised 51 42.1 Food is raised 70 57.9 Total 121 100.0 N This table is based on the 121 nonfarmer respondents. TABLE 50a * PROPORTION OF FAMILY FOOD COST THE PRODUCTION COVERS Proportion Number Percent Less than.y4 47 67.1 About 1/4 13 18.6 Between V4 and 1I2 7 10.0 About 92 1 1.1+ Don't know 2 2.9 Total 70 100.0 * This table is based on the 70 nonfarmer respondents who raise food. - 221 - TABLE 51 ATTITUDES TOWARDS THE FORMULATION OF THE LANSING METROPOLITAN PLANNING COMMISSION Attitude Number Percent In favor of formulation 95.5 49.5 Opposed to formulation 65 54.4 Cannot decide 50.5 16.1 Total 189 100.0 TABLE 513 REASONS FOR ENDORSING THE FORMULATION OF THE COMMISSION* Reason for endorsement Number Percent It's good to plan 28.5 50.5 For orderly growth 27 28.9 Better facilities would result 21 22.5 It would made for a closer understanding With the City 9 e 5 10. 2 To see where we‘re heading 8 8.6 Lansing is growing in this direction 8 8.6 A better community would result 7 7.5 Other 2.5 2.7 * This table is based on the 95.5 respondents who indicated that the idea of starting the organization was a good one. Multiple answers. TABLE 51b REASONS FOR OPPOSING THE FORMULATION OF THE COMMISSION Reason for opposition Number Percent The city doesn't understand the problems of a rural area 18 27.7 We're too far from Lansing 10.5 16.2 We want to stay rural 10.5 16.2 It would lead to annexation 9 15.8 We can handle our own problems 8.5 15.1 They'd try to dictate to us 6.5 10.0 It would lead to higher taxes 6 9.2 We're not part of Lansing or the Metropolitan Area 5 4.6 Other 1.5 2.5 *- This table is based on the 65 respondents who indicated that the idea of starting the organization was not a good one. Multiple answers. TABLE 52 ATTITUDES TOWARD A COMMISSION PROPOSAL FOR ANNEXATION Attitude Number Percent In favor of an annexation proposal 22.5 11.9 Opposed to an annexation proposal 157.5 85.5 Cannot decide 8 4.2 No answer 1 -5 Total 139 99-9 TABLE 52a REASONS FOR ENDORSING AN ANNEXATION PROPOSAL* Reason for endorsement Number Percent Improvement in facilities 15.5 68.9 Lansing has to expand somewhere 5 15.5 The area would benefit 1.5 6.7 Lansing is well run 1 4.4 This area is part of Lansing anyway 1 4.4 Would provide more money to the area .5 2.2 T0138]. 2205 9909 * This table is based on the 22.5 respondents who indicated that they would agree with an annexation proposal by the Commission. TABLE 52b * REASONS FOR OPPOSING AN ANNEXATION PROPOSAL Reason for opposition Number Percent Taxes would go up 80 50.8 We want to stay rural 51 19.7 We're too far from Lansing 22.5 14.5 We don't want the city to tell us what to do 15.5 9.8 Our problems are different from city problems 12 7.6 I moved here to get out of the city 8.5 5.4 It would provide no advantage 7.5 4.8 It would become too crowded 6 5.8 We can handle our own problems 5.5 2.2 Other 5.5 2.2 * a. This table is based on the 157.5 respondents who indicated that they would not agree with an annexation proposal by the Commission. Multi- ple answers. - 223 - TABLE 76 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG LOCALITE INDEX TYPES Index type Number Percent High localism 9.5 5-0 Semi-high localism 54 28.6 Semi-low localism 102 54.0 Low localism 23.5 12.4 Total 189 100.0 TABLE 77 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG LANSING ORIENTATION INDEX TYPES Index type Number Percent High Lansing orientation 16 8.5 Semi-high Lansing orientation 59.5 20.9 Semi-low Lansing orientation 83.5 44.2 Low Lansing orientation 48 25.4 Not determinable 2 1.1 Total 4189 100.1 TABLE 78 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG URBAN DEFINITION OF THE AREA INDEX TYPES Index type Number Percent Urban definition 5 2.6 Mixed-urban definition 14.5 7.7 ‘Mixed-non urban definition 114.5 60.6 Non urban definition 55 28.0 Not determinable 2 1.1 Total 189 100.0 - 224 - TABLE 79 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG RURAL DEFINITION OF THE AREA INDEX TYPES Index type Number Percent Rural definition 69 56.5 Medial rural definition 101.5 55.7 Non rural definition 17 9.0 Not determinable 1.5 .8 Total 189 100.0 TABLE 80 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG SATISFACTION INDEX TYPES Index type Number Percent High satisfaction 59.5 20.9 Medial satisfaction 111.5 59.0 Low satisfaction 57.5 19.8 Not determinable .5 .5 Total 189 100.0 TABLE 85 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG STABILITY-MOBILITY INDEX TYPES Index type Number Percent High stability 55 18.5 Medial stability-mobility 67.5 55.7 High mobility 85 45.9 Not determinable 5.5 1.8 Total 189 99-9 - 225 - TABLE 86 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG RURAL-URBAN EXPERIENCE INDEX TYPES Index type Number Percent High rural experience 56.5 29.9 Mixed rural-urban experience 56 19.0 High urban experience 95.5 50.5 Not determinable l .5 TOtal 189 990 9 TABLE 87 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG LANSING EXPERIENCE INDEX TYPES Index type Number Percent No Lansing experience 95.5 49.5 Medial Lansing experience 55 28.0 High Lansing experience 41.5 22.0 Not determinable 1 .5 Total 189 100.0 TABLE 88 EREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG URBAN AND LANSING EXPERIENCE INDEX TYPES Index type Number Percent No urban and Lansing experience 55.5 29.4 Some urban and Lansing experience 10 5.5 JMedial urban and Lansing experience 55.5 29.4 Much.urban and Lansing experience 25 15.2 High urban and Lansing experience 42.5 22.5 Not determinable .5 .5 Total 189 100.1 — 226 - TABLE 89 FREQUENCY DISTRIBUTION OF SAMPLE MEMBERS AMONG COMMITMENT INDEX TYPES Index type Number Percent Low commitment I 8.5 4.5 Low-medial commitment 27 14.5 Medial commitment 102 54.0 High-medial commitment 46.5 24.6 High commitment 5 2.6 Total 189 100.0 APPENDIX C THE MEASUREMENT OF ADJUS'IMENT Two separate sets of questions were included in the field sched- ule in an attempt to classify respondents in terms of subjective identi- fication and objective integration.1 After the data had been collected each set was analyzed to determine whether the battery questions were from the same unidimensional universe. The technique chosen for this analysis is specifically designed to test unidimensionality and is gen- erally referred to as either the Cornell technique, the Guttman tech- nique, or scalogram analysis.2 The Theory and Method of Scalogram Analysis Scaling techniques are designed to order a group of respondents into a gradient pattern such as more - less, or light - dark. In addi- tion to determining the gradient pattern, the Guttman technique imposes an important limitation on the items which compose the contimum: they must be unidimensional. Unidimensionality is a unique attribute of a series of items which gives those items a special character relative to one another. The determination of this dimension allows for ranking 1Por convenience the specific questions will be presented in a later section. 2Extensive treatments of the scalcgram technique can be found in Louis Guttman, ”The Cornell Technique for Scale and Intensit Analysis," Educational §_n_d_ Psychological Measurement, VII (Summer, 1947 , pp. 241- 2793 Samuel A. Stouffer, _e_t 9;. ,Measurement and Prediction, Studies .i_n_ World Leg ;_I_, Vol. 4 (Princeton, New Jersey: Princeton University Press, 1950;; and Matilda White Riley, John W. Riley, Jr., and Jackson Toby, Sociological Studies in Scale Anal sis (New Brunswick, New Jersey: Rutgers University Press, 1954;. -227- - 228 - respondents in such a manner that after they are placed on a continuum their specific responses can be reproduced merely by knowing their ranks. In.more precise terms, a measure of reproducibility is derived which in- dicates the percentage of responses which can be reproduced without er- ror. Thus a coefficient of reproducibility of .90 indicates that ninety percent of the responses can be reproduced perfectly, merely by knowing the assigned ranks or, conversely, a ten percent error will result if each respondent is categorized as an error-free rank type. The coeffi- cient is used in reference to res uses, not respgndents. Therefore, re- producibility refers to a.measure of total scale reproduction, rather than respgndent reproduction. The ranking procedure is based upon an empirical determination of the way in which the scale items were responded to by individuals, and is based upon the logical expectation that questions from the same universe are responded to in a consistent manner. To illustrate the point by utilizing the now classic example provided by Stouffer, suppose that the following four questions are asked:5 1. Are you over 6 feet tall? 2. Are you over 5 feet 6 inches tall? 5. Are you over 5 feet tall? 4. Are you.over 4 feet 6 inches tall? If a respondent answers Item 1 in the affirmative, he must answer Items 2, 5, and 4 in the affirmative also, unless he commits an error. Simi- larly, if a respondent answers Item 4 in the negative, he cannot give affirmative responses to Items 1, 2, and 5, unless he commits an error. The pattern of responses is illustrated in the following chart, in which 3This is a modified version of the example provided by Samuel A. Stouffer in Stouffer,‘gt‘§;.,‘gp.‘git., p. 10. The example has been modified by the addition of a fourth question. - 229 - "X" signifies an affirmative response and "0" signifies a negative re- Spouse. CHART 2 PERFECT SCALE TYPES FOR A FOUR-ITEM GUTTMANrTYPE SCALE Respondent Item Scale Type 1 2 3 4 A X X .X X V B O X X X IV C O 0 X X III D 0 0 0 X 11 E 0 0 0 O I Respondent A has replied to Item 1 affirmatively and has commit- ted no errors; respondent E has replied to Item 4 in the negative and has committed no errors. The other three respondents are also error- free and therefore, along'with respondents A and E, may be classified as "perfect" scale types, i.e., scale types that are error-free. The chart also illustrates a very important attribute of the Guttmanptype scale: the items which compose the scale must possess a special cumulative property} if the items do not possess this property, they cannot be utilized in soalogram.ana1ysis. For example, the follow- ing battery of questions cannot be utilized in scalogram analysis:“ 1. Are you over 6 feet tall? 2. Are you between 5 feet 6 inches and 6 feet tall? 5. Are you between 5 feet and 5 feet 6 inches tall? 4. Are you under 5 feet tall? .Although nonpcumulative items cannot be formulated into a.Guttmanptype scale, they may be utilized in scales which do not require the special —-—— “Ibide, p. 11o _230 .. cumulative property of Guttman scale items.5 To this point the discussion has considered perfect scale types exclusively. However, some non-consistency among respondents is to be expected, especially if attitudes are being'analyzed. To illustrate the point, suppose that a respondent's replies to the four height questions produces a pattern of responses of X X X 0 or X X.O X, rather than one of the error-free patterns. Obviously, this respondent has not been consistent. In following the Guttman technique, this respondent is as- signed to Scale Type V since this placement minimises error; assignment to any other scale type would contribute at least two errors to the scale while placement in Scale Type V would contribute only one error. If the responses were of the order X 0 X X, the respondent may be class- ified as either Scale Type V or Scale Type III since, in either case, he would contribute one error. Cases in which assignment to more than.one scale type is possible, without the introduction of more error, are ran- domly'assigned.6 This discussion of scale error logically raises the question of scale acceptability, since scales which contain many errors are obvious- 1y not as precise as scales which have few errors. To determine the ex- tent cf error in a scale, Guttman has introduced a measure referred to as the "coefficient of reproducibility." This coefficient is calculated by dividing the number of nonperrors by the total number of responses or, more simply, by dividing the number of errors by the number of 5For a scaling'method that does not require cumulative items, see the discussion of the Likert scale, infra. 6Andrew F. Henry, "A Method of Classifying Nonrscale Response Patterns in a Guttman Scale," Public Opinion.QE§gterly, 16 (Spring, 1952), ;pp. 94-106, suggests that such cases should be assigned to the most pop- ‘ular scale type. -231 .- responses, and subtracting the derived figure from 1.0. A non-arbitrary "acceptance" or "rejection" coefficient of reproducibility does not ex- ist; however, it is a fairly standardized procedure today to accept .90 as the criterion. If a coefficient of reproducibility is as high as .90, and the scale meets all other criteria, the items are accepted as scal- able; if the coefficient value is below .90, but the scale meets all other criteria, a quasi-scale is said to exist. To this point, two criteria of scalogram analysis have been dis- cussed, viz.: items must be cumulative, and reproducibility must be sufficiently high. A third condition.must be met: errors must be ran- dom. This means that the number of ”error types" should not exceed "per- fect types" in any given scale type. Thus, if the analysis of responses indicates that Scale Type V contains more cases of the pattern X X 0 X than the pattern.X X X X, errors in this category are not random. This criterion is related to another: no item should contain more error cases than nonperror cases.- This fourth criterion necessi- tates an examination of the patterning which is obtained by ordering all respondents from high to low on the basis of assigning weights to each affirmative and negative response to individual items. "Cutting points" are imposed upon this ordering by categorizing respondents into their appropriate scale types; if each item represents a different pro- portion of affirmative and negative answers (it will be pointed out let- er that this is a necessary condition), each item will have a different cutting point. Each item is examined independently to determine wheth- er the number of non-error responses exceeds the number of error respon- see on both sides of the cutting point. A fifth criterion is that each item should split between a 20-80 - 232 - or 80-20 range.7 This range is utilized as a criterion because the co- efficient of reproducibility will be artificially high if any item is split more extremely. For example, a 100-0 split on any item allows no error at all, while a 90-10 split assures a minimum of ninety percent non-errors in that particular item. Thus, if a totally unrelated item which splits 100-0 or 90-10 is imposed upon a scale of items, this item will raise the reproducibility of the scale, while defeating the objec- tive of testing unidimensionality. Since the reproducibility criterion is set at .90, as an item approaches a 90-10 split, reproducibility ap- proaches artificiality. It is possible, of course, that an item which splits 90-10 may be from the same universe as the others in the scale, but will be rejected at the outset of scalogram analysis because of its effect on reproducibility. Ford suggests that if an item outside the 80-20 range is retained in the scale for special reasons, the scale should be tested for reproducibility without the item.8 The discussion of the range of individual items brings a sixth criterion into focus: the marginals of the items in the scale should represent more than a five percent spacing. Ford presents two reasons for maintaining that a five percent spacing is necessary: "1) so that subsequent samples will not show reversals in the scale order of ques- tions, and 2) to help spread the respondents out into fairly equally sized groups, from high to low."9 In addition, another important reason is evident: if two items 7A discussion of this point is presented by Robert N. Ford, in Riley, Riley, and TOby, 93. Me, Do 281+, no #e 81bid. 91bide, Do 2850 -233 .. possess precisely the same marginals the use of both in the scale would be superfluous. In other words, the use of an item possessing the same marginals as another would add nothing to the utility of the derived scale since it would be redundant. It would be best to have both ex- tremes of the 80-20 range represented, with a wide spacing of items be- tween them. Thus, in a seven item scale the goal should be a set of distributions with approximately the fellowing spacing: 80-70-60-50-40— 50-20; in a five item scale the following spacing should be approximated: 80-65-50-55-20. Guttman presented another criterion of scalability (this is the seventh, and the last, to be discussed here): The more items included in a scale, the greater is the assur- ance that the entire universe of which these items are a sample is scalable. If the items are dichotomous (or dichotomized from.more than two categories as a result of the scale analysis), it is prob- ably desirable that at least ten items be used, with perhaps a les- ser number being satisfactory if the marginal frequencies of sever- al items are in the range of 50 per cent to 70 per cent. Just four or five items, with marginal frequencies outside such a range, would not give much assurance as to how scalabia the universe was, no matter how scalable the sample might be. In actual practive, however, this criterion has been both ignored and modified by others. Ford, for instance, illustrates a technique of scalogram analysis by which he derives a scale consisting of four dicho- tomous items with the following marginals: 74-58-44-20.11 In addition, James A. Davis comments on this criterion as fellows: “It is our impres- sion that the rule of thumb of Guttman scaling which requires a large number of items is wasteful as well as vague."12 Ferd and Davis, however, loLouis Guttman, in Stouffer, _e_t_ 21,, p. 79. llFord, 22s 93:2" ppe 276-2960 12James A. Davis, "0n Criteria for Scale Relationships," Ameri- ggn Journal 9; Sociology, LXIII (January, 1958), p. 579. - 234 - neglect to point out that few-item scales artificially approach the .90 reproducibility criterion by chance alone. To borrow from the findings which will be presented in the next section, a four-item scale with mar- ginals of 29-45-59-67 possesses a‘ghangg reproducibility of .86. In general, the more items in the scale, the lower the chance reproducibil- ity. Chance reproducibility is also a function of the distribution of responses. Green points out that a set of four items with distributions of 20-40-60-80 possesses a chance reproducibility of .90 while a set of nine items distributed 10-20-50-40-50-60-70-80-90 possesses a chance re- producibility of .85.15 However, both examples presented by Green.vio— late the criterion discussed above; the first example presents a four- item scale with two items outside of the 50-70 range suggested by Gutt- man for scales with few items, while the second example presents a nine item scale with two items outside of the suggested 80-20 range. For the sake of simplicity the previous discussion considered dichotomous items exclusively. However, it is obvious that many ques- tions on interview schedules are not simply of the "yes-no" variety but, instead, present the respondent with a wide range of possible responses. Questions which present respondents with more than two alternatives can also be handled by scalogram analysis techniques, but the construction of the scale becomes more complex. In actual practice many Guttman scales deal with dichotomies exp clusively, even if the original interview questions presented respondents 13Bert F. Green, "Attitude Measurement," Gardner Lindzey (ed.), Handbook 2;,Social Ps cholo , Vol. 1 (Cambridge, Massachusetts: Addi- soanesley Publishing Company, Inc., 1954), p. 556. - 235 - with several alternative answers. Dichotomization is accomplished by combining actual responses into two categories. Guttman outlines the procedure and justifies dichotomization in the following manner: It has seldom been found that an item with four or five cat- egories will be sufficiently reproducible if the categories are re- garded as distinct. One reason for this is the verbal hibits of people. Some people may say "Strongly Agree" where others may say "Agree," whereas they have essentially the same position on the bas- ic continuum but differ on an extraneous factor of verbal habits. By combining categories, minor extraneous variables of this kind can be minimized. by examining . . . each question, it can be de- termined how best to combine the categories so as to minimize the error of reproducibility for the combinations. . . . The way I8 combine categories is determined for each question separately. The usual procedure is to present questions to respondents which allow several responses, and to dichotomize during scale construction. This procedure may seem like operating in reverse since, if items are to be dichotomized during the later analysis, of what utility is the inclup sion of schedule questions which present the respondent with more than two alternatives? Ford justifies this procedure with three answers to the question: 1) A third group, the ”no opinion" or "no answer" group, will still arise, calling for a decision in dichotomization. 2) The simple yes-no type of question can be quite frustrating to the respondents; selecting an answer from multiple responses can be much less of a task. 5) We generally cannot tell in advance how people will split propor- tionately over the responses of a question. Thereforf5we have no clue as to how dichotomous responses could be worded. It is possible to construct a Guttmanrtype scale which contains a combination of dichotomous and trichotomous items, but only if the tri- chotomous items meet the same criteria as the dichotomous items. The prime advantage of introducing trichotomous items into a scale is the 1“Guttman, ”The Cornell Technique for Scale and Intensity Analyr 813’" flog-gs, pe 256s 15Nora, pp. 313., pp. 285f. - 236 - refinement obtained; for each trichotomous item used, one more scale type emerges. However, in many situations, further refinements by the introduction of additional scale types would be a superfluous precedure. Most empirically derived Guttman-type scales consist of a set of dichot- omous items. Construction of the Scales Objective integgation: scaloggam analysis As was indicated in the first chapter, a battery of questions, designed to determine the nature of the social relationships of fringe dwellers in their community of residence, was included in the field schedule. Objective integration is the summary term which refers to the specific nature and the degree of these relationships. The questions which were designed to measure respondent objective integration were similar since all were oriented toward determing the location of activi- ties. The specific questions included in the battery all reduce to the following general question: "What is the nature of your fringe area ac- tivity pattern?" Respondents who indicated that certain activities are carried out in the fringe area are considered integrated; those who in- dicated that these activities are carried out elsewhere are considered nonrintegrated. Of course, it was expected that these would not be ex- clusive categories, i.e., they represent the extreme positions of a gradient pattern of objective integration. The specific questions designed to measure the degree and na- ture of objective integration are as follows: 1. About how many [people in this areaJ would you say that you know by name? .. u.— _— _,—_.,—... . -237 - 2. About how many.families around here do you come in contact with for at least a few minutes every day or so? 5. About how many [families in this area] do you spend a whole af- ternoon or evening with every now and than? 4. In the previous questions we've been.talking about people in this area. Now I'd like you to think about your best friends-- ‘ng,matter where they live. Now, will you think for a moment of your three best friends--we'll call them 1, 2, and 5 because we don't want their names. Cculd.you please tell me where each of them lives? 5. Do you and your (husband/wife) belong to any groups or organiza- tions? Where does it meet? 6. Have you or your (husband/wife) taken a part in any volunteer community groups such as the volunteer fire department or civil defense? 7. Do you know of any cases in which the people around here got to- gether to do something about an issue such as bond issues, de- cisions concerned with schools or the fire department, and so forth? Did you or your (husband/wife) participate? 8. Are there any places around here where people get together in- . formally or hang out? Do you go there? 9. How do you keep up with the news of what is happening: right around here? in this township? in this county? in Lansing or East Lansing? 10. Where do you go to: shop for groceries; buy gasoline: shop for clothing; a drugstore; barbershop or beauty parlor; nightclub or tavern; church? 11. Where do you go to visit friends? There were two steps in testing whether the eleven questions, or a specific sub-group of questions, could be utilized in the study as a Guttman-type scale. First, the responses had to be analyzed to deter- mine whether the questions were unidimensional and from the same uni- verse.16 Second, if unidimensionality was exhibited, the responses had to be classified so that respondents could be placed on an objective 16Responses by all sample members were utilized for this analy- sis rather than those obtained by deflating the sample: see an re, p. 27. - 238 - integration continuum. As was previously indicated, scalogram analysis is specifically designed to do both tasks simultaneously. The responses to each of the questions were examined to deter- mine whether they meet the seven criteria discussed. The first tests fOr scalability were performed with a one-third sample of respondents because of the cumbersome nature of the initial procedures; after scala- bility was determined for the sample, all the respondents were analyzed. First, each response was assigned a weight based on judgments of "most positive and most negative," with intervening gradations if the question allowed more than two alternatives. Each respondent was then assigned a weight which was the sum of the weights of all the alternatives he en- dorsed. 0n the basis of these total scores, respondents were ordered on a chart which indicated their precise replies to each question. Exami- nation of this chart indicated that it was necessary to combine certain response categories in order to raise the reproducibility of certain items. In addition, it was necessary to eliminate certain of the items from the scale because they were not differentiating respondents with high weights from respondents with low weights. Through a series of successive approthations, and an analysis of each derived ordering, the necessary combinations and eliminations became evident. Question 8 eas eliminated before the analysis was undertaken be- cause of the extreme nature of the response distribution; the split was 9-91, eleven percent outside of the 20-80 criterion. Three items were eliminated because they exhibited more error than non-error; these items (numbers 2, 1+, and 6) were rejected on the grounds that they did not fit into the unidimensional continuum pattern. Items 1, 7, and 11 were eliminated from consideration because their distributions were too close _ 239 - to those exhibited by other items; the criterion of choice utilized was error contribution, i.e., those items which contributed most errors were eliminated. In this manner a four-item scale was derived, composed of ques— tions 5, 5, 9, and 10. Each of the four was eventually dichotomized in order to raise the coefficient of reproducibility to an acceptable level. The four questions, and their affirmative and negative responses, are as follows: 5. About how many [families in this area] do you spend a whole af- ternoon or evening with every now and then? 7b% "Negative" responses: none, one. 29% "Positive" responses: two or more. 5. Do you and.your (husband/wife) belong to any groups or organiza- tions? Where does it meet? 51% "Negative" responses: belongs to none, belongs to no or- ganization which meets in the neigh- borhood or a local small town. 45% "Positive" responses: belongs to an organization which meets in the neighborhood or a lo- cal small town. 9. How do you keep up with the news of what is happening: right around here? in this township? in this county? in Lansing or East Lansing? 4b% "Negative" responses: does not use local sources, uses one local source. 59% "Positive" responses: uses two or more local sources. 10. Where do you go to: shop for groceries; buy gasoline; shop for clothing: a drugstore: barbershop or beauty parlor; nightclub or tavern: church? 51% "Negative" responses: all in Lansing. 6T% "Positive" responses: some or all in the local area. Table 90 summarizes the scale obtained through an analysis of the four questions. An examination of the chart indicates that all the criteria are fulfilled with the exception of one minor deviation: the Item 5 marginals are 71-29, whereas Guttman suggested that few-item scales should fall within a 70-50 distribution range. mam. « huaaaowosboumoa no pooaoauuooo cosponno mm a muoaao mo Mensa: oeshcmno arm. n muaaanwoseoumon we «cowoauuooo voodoo mn.m:a n cheese «0 Hones: common me sm.hm~ mooo.a fleece I a. 23m $8. R. 3. R. E. H o o o 2m H m Hm.oH made. mm. as. me. a». H o o x 0 NH mm.o~ maho. mm. mm. pm. as. a o x o o m: 8.1: «$0. mm . 3. are. . 2. . o o o o o E N 3.3 :96. S. 3. R. mm. a x o o x mm mm.om NHHH. pm. a:. pm. a». o x o o o HHH mm NH.H¢ coma. em. mm. pm. as. o x x o 0 ma mm.HN ammo. pm. as. m:. H». H x o N o 5 r em .3 :30. R . mm. 3. .2. . H o x x c an «0.3 82. 5. mm. 3. :3 o x x x o m mm.m mmno. mm. mm. hm. om. N o x o x 0 an... $8. mm. 3. 3. mm. m. o o x x b s m~.m nsmo. mm. mm. ms. mm. a o x x x s oo.m omno. pm. as. me. mm. a x o x K S 233 38. S. mm. em. mw. a x x o x mm >w.~a nmso. so. mm. no. mm. o x x x N wonky doshomoo doaooowm wonky mo 0H m m m eschew msmpH mason nomso no moose Ho heaaansnoum no Henson hopes: mucosa: souH choppmm MAHsomnmo om mummy - 241 - In addition to presenting the results of the scalogram analysis, Table 90 reports the chance coefficient of reproducibility, given the distributions of responses observed. This information is presented be- cause of the dangers inherent in few-item scales. As was indicated, the chance reproducibility of the scale is .861 while the empirical repro- ducibility is .915. No statistical test of significance between the ex- pected and the observed reproducibility has been devised as yet. How- ever, since the empirical scale meets the necessary criteria, it is ac- ceptable by all standards. One final note: an examination of Table 90 indicates that fifty- four cases do not fit perfect scale types. Therefore the possibility exists that they have been misplaced. However, the method of placement, discussed above, provides certain assurances that their placement is satisfactory. Subjective identification: scaloggam analysis The subjective identification battery was designed to determine the nature of the basic orientation of fringe dwellers in their communi- ty of residence. Respondents were provided with opportunities to indi- cate their evaluation of the fringe community, to define the area on a subjective level, and.to compare fringe area residence to other types of residence areas. The questions were designed to determine the degree to which respondents possess positive or negative attitudes toward their community. Basically, the battery of questions can be reduced to the fol- lowing general query: "How do you feel about the fringe area as a place in which to live?" Respondents who indicated that their evaluation was positive were to be considered identified: those who indicated a nega- ..242 _ tive orientation were to be classified as non-identified. Again, "iden- tified" and "non-identified" are not exclusive categories, i.e., they are extremes in a continuum of identification patterns. The subjective identification battery consisted of the follow- ing ten questions: 1. If y_o_g had mm; M, would you continue living around here? 2. Does this area have a name? 5. Do you think of it as part of a larger area? 4. If you had to dr0p out of all the organizations you belong to ex- cept one, which would you choose to remain in? (The location and nature of these organizations had been ascertained previous- 15'.) 5. We've been referring to the "area around here" for some time now, and I'd like to get an idea of what area you've been referring to. How far north, east, south, west, does it go? 6. Do you think that living in Lansing would have advantages over living in this area? 7. Do you think that there are differences in the kinds of people who live in an area like this and city people? About two years ago an attempt was made to organize a group called the Lansing Metropolitan Planning Commission. The purpose of the group was to help coordinate the development of the City of Lansing and Ingham, Eaton, and.Clinton Counties by collecting information and formulating plans for the total Metropolitan Area. The group was nonprofit and had no political power; This area would have been one of those covered by the Commission's work. 8. Do you think that the idea of starting such an organization was a good one? 9. Wbuld you have agreed with the Commission if it had recomp mended that this area be annexed by the City of Lansing? 10. Do you think that this area is getting better or worse? As in the case of objective integration, the responses to the questions in the subjective identification battery were subjected to an analysis to determine whether they met the criteria of a unidimensional scale. However, the analysis of the subjective identification questions —243 '- failed to yield a series of items which were scalable. Thus, it is not possible to take the position that the questions in the subjective iden- tification battery are unidimensional. The Method of Summated Ratings Since the analysis indicated that the items deviated from the scalogram model, it was decided to subject the battery to another type of analysis which might provide a basis for the ranking of respondents on the Oasis of subjective identification. The method of summated rat- ings (frequently referred to as the "Likert scale" method) was chosen.17 In this method a systematic item analysis is utilized to select discrim- inating items and respondents are assigned scores on the basis of how they responded to these items. In addition, the Likert-type scale indi- cates whether specific items are related to the total scale. Cumulative items are not necessary for a Likert-type scale. The initial step in the construction of a Likert-type scale is to assign weights to each item, e.g., a weight of 5 to the "most posi- tive" possible response and a weight of l to the "most negative" possi- ble response, with intermediate weights assigned to responses between these extremes. After the weighting process each respondent is assigned a total score, based on a summation of his responses to all items; these are the scale scores. Respondents are ordered by scale scores to facili- 17Extensive treatments of the method of summated ratings are found in Rensis Likert, "A Technique for the Measurement of Attitudes," Archives 2; Ps cholo , 140 (June, 1952), pp. 5-55; Marie Jahoda, Morton Deutsch, and Stuart A. Cook, Research.Methods in Social Relations, Part One (New York: The Dryden Press, 19515, pp. 1517-197; William J. Goods and Paul K. Hatt, Methods ig Social Research.(New York: McGraw—Hill Book Co., Inc., 1952), pp. 270-285: and Green, 92. gi§., pp. 551-555. -244- tate separation into two groups: the upper quartile ("most positive") and the lower quartile ("most negative").18 The discriminating power of each item is determined by comparing weighted totals, as illustrated in Table 91.19 Through this procedure the discriminating power of each item in the scale is derived. Obviously, as the discriminating power of an item approaches 0, the utility of the item is brought into question. USually, items with a discriminating power of less than 1.0 are elimi- nated from the scale. Since items are handled independently, each is actually a scale in itself and the derived total scale is a "scale of scales." Thus, a Likert-type scale may actually contain several dimensions, instead of a single dimension. This factor has led to one of the most severe criti- cisms of the procedure since ". . . such a scale can give only a crude measure, for the definition of what it actually measures is always open to question."20 A second criticism is that knowledge of a respondent's total score provides no indication of the specific items endorsed, ex- cept in the case of extreme scale scores. The Likert-type scale is particularly suited to a reliability test which is referred to as the "split-half" method. The calculation is very simple: the scale is divided randomly in half, into two equiva- lent groups of items; the usual procedure is to divide the items into odd-numbered and evenrnumbered groups. After the random division is made, a correlation coefficient is calculated between the halves. Since 18The upper and lower deciles are sometimes used. 19Adapted from Gardner Murphy and Rensis Likert, Public Opinion ing _t_r_:_e Individual (New York: Harper and Brothers, 1938), p. 289. 20Goods and Hatt, pp. 2g... p. 285. -245 - $4 5 o o o m H m 30A oo.~ SA R m n N H o m seem Asses assesses Anaconda Aoeoou m s n N a £5.83 sienna 30H : some some 5 van» 5 assumes: n35 hopes: 933030 oeoom 53H Henson nosom + Hopes. Hones: wcapggnema depsmaozv x ouoemv coo: H309 oneness: corneas: mm 9.549 madam mEIBmmMHA 4 2H EH me $58 efiadzgmvan a ho KOHHHEHHQ - 246 - the two groups are treated as independent scales, both of which are as- sumed to measure the same phenomenon, the correlation coefficient should approach +1.0. If the correlation is low, the scale cannot be accepted as reliable. Subjective identification: the method 9; summated‘gggiggg The initial step in the attempt to construct a Likert-type scale was to assign weights to the question responses in the subjective iden- tification ten—question battery. Fbr each item a weight of 4 was as- signed to the most positive response, a weight Of 0 was assigned to the negative extreme, and intervening responses were assigned intervening weights. Respondents were assigned total scores, based on the sum of their individual item weights, and were ordered from high to low. The upper and lower deciles were drawn from this array. By following the precedure indicated in Table 91, an index of the discriminating power of each item was obtained. Six of the ten items met the criterion of a discriminating power above 1.0. The items which qualified, and their indexes of discrimination, are the following: 1. If yggflhgg your choice, would you continue living around here? Iadscriminating power--1.67) 2. Does this area have a name? (discriminating power--2.55) 6. Do you think that living in Lansing would have advantages over living in this area? (discriminating power--4.00) 7. Do you think that there are differences in the kinds of people who live in an area like this and city people? (discriminating power-- 3 e 56) 8. Do you think that the idea of starting such an organization [the Lansing Metropolitan Planning Commission] was a good one? (dis- criminating power-~2.44) 9. Would.you have agreed with the Commission if it had recommended that this area be annexed by the City of Lansing? (discriminat- -247 .. ing power-~2 . 00) The next step was to divide the six items into two equivalent groups and to calculate the correlation coefficient between them. All respondents were assigned two scale scores, one on the basis of items 1, 6, and 8, and the other on the basis of items 2, 7, and 9. The correla- tion coefficient obtained by the split-half method was an unacceptable +.l61. Since the split-half correlation did not meet the reliability criterion, the derived Likert—type scale had to be abandoned as non-re- liablee Subjective Identification: .Modes of Orientation The failure of the scaling methods to indicate that the items could be utilized together created a need for a totally new approach to the increasingly complex subjective identification dimension. The re- sponses to each of the ten questions were re-examined to determine the degree to which they were individually performing the task for which they were constructed. On this basis it was decided that question num- ber 1 ("if 193 224. Lqu_r M, would you continue living around here?") had elicited the most relevant responses since the question was carried beyond a simple "yes" or "no" endorsement by the following probes: If yes: "What do you like about it?" If no: "What do you dislike about it?" If don't know: "Do you have any reasons for wanting to stay? Do you have any reasons for wanting to move?" Question 1 had brought direction and content responses which were clear indications of the modes of orientatiOns of respondents. In addition, examination of the responses to a previous question ("How did you happen to move around here rather than some other part of the Lan- sing area?"), added depth to the direction and content of these responses. -248 - By utilizing these two questions in conjunction, the mode of a respond- ent's orientation to the fringe in terms of direction, content, and de th, could be estimated. Respondents were classified in each of these re- spects by a careful scrutiny of their responses. The classification utilized was as follows: 1) direction a positive b negative c unclear 2) content a) social: friends here, knows people here, good neighborhood, friendly, good neighbors, feel at home, near friends, never lonesome b) ecolo cal: near work, near facilities, good location, close to town and yet outside c) economic: good land, making good living, got money tied up here, best buy, lower taxes d) coerced: close to family, too old to move, all I know, raised here, know the area e) personal: good for raising children, freedom, good place to retire, anti-city, quiet, peaceful, fresh air, good facili- ties, can garden, not crowded, room for children f) aesthetic: good weather, scenic, good view 5) depth a no prior choice b moved for same reason staying c moved for other reasons The classification into mode of orientation typologies was thus undertaken on the basis of the responses to what was judged the most relevant subjective identification questions. Each respondent was as- signed to a typology on the basis of his responses to these questions and probes. It is evident that the derived categories are not mutually exclusive; respondents who indicated multiple orientations were assigned to combined categories. The formulation of typologies seems a logical alternative after the failure to develop a subjective identification scale. It became -249 .. increasingly evident as the development of scales was attempted, that subjective identification is a much more complex dimension than was ori- ginally assumed. Rather than a simple positive or negative identifica— tion pattern, it became evident that there existed patterns of identifi- cation. In addition, there is no reason to expect that a given individ- ual possesses a single mode of orientation; the definition of the situa- tion by an individual is frequently quite complex. Therefore, it was expected that many multiple orientations would be found when respondents were assigned to subjective identification typologies. BIBLIOGRAPHY Angell, Robert C. "The Moral Integration okamerican Cities," American Journal 9; Sociolo , LVIII, Part 2 (July, 1951), 1-140. Beegle, J. Allen, and Schroeder, Widick. Social Orgapization'ipwppg NOrth Lansing Fripg . iMichigan State University A.E.S. Technical Bulletin 251, September, 1955.- Blizzard, Samuel W., and Anderson, William P., II. Problems ip,Rural- Urban Fringe Research: Conceptualization ppd Delineation. Pennsyl- vania State College A.E.S. Progress Report No. 89, NOvember, 1952. Davis, James A. "On Criteria for Scale Relationships," American Journal ‘2; Sociolo , LXIII (January, 1958), 571-80. Dobriner, William M. (ed.). ‘Thg Suburban Communi y. New York: G. P. Putnam's Sons, 1958. Duncan, Otis Dudley. "Community Size and the Rural-Urban Continuum," in Hatt, Paul K., and Reiss, Albert J., Jr. (eds.). Cities ppd Society. Glencoe, Illinois: The Free Press, 1957, 55-45. . "Optimum Size of Cities," in Hatt, Paul K., and Reiss, Albert J., Jr. (eds.). Cities ppg Society. Glencoe, Illinois: The Free Press: 1957. 759-72. Durkheim, Emile. ,Thg Division pf Labor i3 Society. Glencoe, Illinois: The Free Press, 19 7e Firey, walter. Social Aspects 39 Land ggp'Plannipg‘ipnppg’Countpy-City Frgge: _T_I;_e_ Case 9}; Flint, Michigag. East Lansing: Michigan State College A.E.S. Special Bulletin 559, June, 1946. Fbrm, William H.,‘gpflgl. "The Compatibility of Alternative Approaches to the Delimitation of Urban Sub-areas," American Sociological R2: view, 19 (August, 1954), 434-40. Gist, Noel P. "Ecological Decentralization and Rural-urban Relation- ships," Rural Sociol , 17 (December, 1952), 528-55. Gist, Neel P., and Halbert, L. A. Urban Society. New York: Thomas Y. Crowell COe, 1956s Geode, William J., and Hatt, Paul K. Methods 3p Social Research. New York: McGraw-Hill Book Co., Inc., 1952. Goodman, Lee A., and.Kruska1, William H. "Measures of Association for Cross Classifications," Journal of ppp.American Statistical Associa- tion, 49 (December, 1954), 732-657 - 250 - ~25]. - Green, Bert F. "Attitude Measurement," in Lindzey, Gardner (ed.). Handbook‘pf Social Psychology. Vol. 1. Cambridge, Massachusetts: Addison-Wesley Publishing Co., Inc., 1954, 555-69. Guttman, Louis. "An Outline of the Statistical Theory of Prediction," in Horst, Paul, 33 _s_.l_. (eds.). T_hg Prediction pg Personal Adjust- ment. Social Science Research Council, Bulletin 48, 1941, 258- 5. . "The Cornell Technique for Scale and Intensity Analysis," Educational _a__n_d_ Psychological Measurement, VII (Summer, 1947), 2 7-79. Hagood, Margaret J ., and Price, Daniel 0. Statistics f2; Sociologists. New York: Henry Holt and Co., 1952. Hawley, Amos H. Human Ecolog: .A_ Theopy 23: Community Structure. New York: The Ronald Press 00., 1950. . _'I_'_h_e _(m_a_.ng_i_ng Shape 9_i_‘ Metromlitan America: Deconcentration Since 122 . Glencoe, Illinois: The Free Press, 195 . Henry, Andrew F. "A Method of Classifying Non-scale Response Patterns in a Guttman Scale," Public Opinion Ql_a.t_'ter1y, 16 (Spring, 1952), 94-106. Hillery, George A., Jr. "Definitions of Community: Areas of Agreement," Rural Sociolo , 20 (June, 1955), 111-25. Jahoda, Marie, Deutsch, Morton, and Cook, Stuart A. Research Methods in Social Relations. 2 vols. New York: The Dryden Press, 1951. Janowitz, Morris. The; Community Press _ipn _ap Urban Settipg. Glencoe, Illinois: The Free Press, 1952. Kimball, Solon T. The New Social Frontier: 1133 Fr e. East Lansing: Michigan State College A.E.S.. Special Bulletin 5 0, June, 1949. Kurtz, Richard A., and Eicher, Joanne B. "Fringe and Suburb: A Confu- sion of Concepts," Social Forces, 57 (October, 1958), 52-7. Likert, Rensis. "A Technique for the Measurement of Attitudes," Archives 9_f_ Ps cholo , 140 (June, 1952), 5-55. Loomis, Charles P., and Beegle, J. Allen. Rural Social Systems. New York: Prentice-Hall, Inc., 1950. MacIver, Robert M., and Page, Charles H. Society: 53 Introductogy Anal- ysis. New York: Rinehart and Co., 19 9. Martin, Walter T. The Rural-Urban Fripge: _A_ Study 9_i_‘ Adjustment ‘_t_<_>_ Residence Location. Eugene: University of Oregon Press, 1955. -252- McKain, Walter 0., Jr., and Burnight, Robert G. "The Sociological Sig- nificance of the Rural-Urban Fringe: From the Rural Point of View," MW: 18 (June. 1953). 108-14. Mercer, Blaine E. ‘Thg American Community. New York: Random House, 1956. Peters, Charles C., and Van Voorhis, Walter R. Statistical Procedures and their*Mathematical Bases. New York: McGraw-Hill Book Co., Inc., 1940. Queen, Stuart A., and Carpenter, David B. The American City. New York: MCGI‘BW‘Hill 300k COe, Ines, 1953-. . "The Sociological Significance of the Rural-Urban Fringe: From the Urban Point of View," Rural Sociology, 18 (June, 1955), 102‘8e Quinn, James A. Human Ecology. New York: Prentice-Hall, Inc., 1950. Reiss, Albert J., Jr. "Some Logical and Methodological Problems in Com- munity Research," Social Forces, 55 (October, 1954), 51-7. . "The Community and the Corporate Area," Universit 2; Penn- sylvania Law Review, 105 (February, 1957), 445-65. Riley, Matilda White, Riley, John W., Jr., and Toby, Jackson. Sociologi- ppl Studies pp Scale Anal sis. New Brunswick, New Jersey: Rutgers University Press, 195 . Rhyne, J. J. "Social Man and His Community," Peterson, Elmer T. (ed.). Cities App Abnormal. Norman: University of Oklahoma Press, 1946, 136’57e Rossi, Peter H. Why Families Move. Glencoe, Illinois: The Free Press, 1955- Sanders, Irwin T. The Community: An Introduction to a Social System. New York: The Ronald Press Co., 1958. Siegel, Sidney. Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill Book Co., Inc., 1956. Simmel, Georg. "The Metropolis and.Mental Life," in Hatt, Paul K., and Reiss, Albert J., Jr. (eds.). Cities Egg Society. Glencoe, Illi- nois: The Free Press, 1957, 655-46. Sjoberg, Gideon. "The Preindustrial City," in Hatt, Paul K., and Reiss, Albert J., Jr. (eds.). Cities and Society. Glencoe, Illinois: The Free Press, 1957, 179-88. Smith, Joel, and Form, William H. "Urban Identification and Dis-identi- fication." (Typewritten.) - 253 - . "Urban Identification: Orientations and Mechanisms." (Type- written.) Smith, Joel, Form, William H., and Stone, Gregory P. "Local Intimacy in a Middle-sized City," American Journal‘pf Sociology, LX (November, Spykman, Nicholas J. "A social Philosophy of the City," in Burgess, Ernest W. (ed.). ‘Thg‘Urban Communi y. Chicago: The University of Chicago Press, 1926, 55-64. Stone, Gregory P. "City Shoppers and Urban Identification: Observations on the Social Psychology of Urban Life," American Journal pf Sociol- 2m. Lx (July. 195A). 36-45. Stouffer, Samuel A., 21; 31. Studies ip Social Psychology in World War II, VOl. 4: Measurement and Prediction. Princeton, New Jersey: Princeton University Press, 1950. Tate, Merle W., and Clelland, Richard C. NOnparametric ppd Shortcut Statistics. Danville, Illinois: Interstate Printers and Publishers, Inc., 1957. Thompson, Warren S. Population Problems. New York: McGraw-Hill Book Co., Inc., 1955. Tomars, Adolph S. "Rural Survivals in American Urban Life," in Wilson, Logan, and Kolb, William L. (eds.). Sociological Analysis. New York: Harcourt, Brace and Co., 1949, 571-8. T6nnies, Ferdinand. Fundamental Concepts‘pf Sociolo (Gemeinschaft pp; Gesellschaft). New York: American Book Co., 19 O. U. S. Bureau of the Census. U.S. Census‘pf Population: 125 . Vol. II, Characteristics pfmppp,Popplation, Part 22, Michigan, Chapter B, P-B22. Whetten, Nathan L. "Suburbanization as a Field for Sociological Re- search," Rural Sociolo , 16 (December, 1951), 519-50. Wirth, Louis. "Urbanism as a Way of Life," in Hatt, Paul K., and Reiss, Albert J., Jr. (eds.). Cities ppg Society. Glencoe, Illinois: The Free Press, 1957, 46-65. H afar .- ’40.; «or. a. -.‘e S"— ta- 1.! 2' “ 1; {fl-*- {QT E! inAR 1’2 1962 1,5, 5' '. 2...» 1041an SW: UNIV. LlsReRIEs \IHIWllWIIHIIWWI}INNVIWIIHIHUWIWI 31293008538872