RECREATION BEHAVIORAL PATTERNS AND CHARACTERISTICS OF URBAN RESIDENTS IN RELATION TO BELLE ISLE PARK, DETROIT, MICHIGAN Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY NORMAN RONALDRICHARDS 1974 _ ‘ ' Wm.\.‘.‘ ‘. ABSTRACT RECREATION BEHAVIORAL PATTERNS AND CHARACTERISTICS OF URBAN RESIDENTS IN RELATION TO BELLE ISLE PARK, DETROIT, MICHIGAN BY Norman_Ronald Richards At the present rate of urbanization in the United States, it becomes of critical importance to set aside land close to our urban centers for outdoor recreation purposes. Where land is not readily available for acquisi- tion and develOpment of new outdoor recreation areas, existing ones must be efficiently redeveloped and managed to provide for the changing leisure needs of urban residents. The case study presented in this thesis focuses on the redevelOpment of Belle Isle Park, a large recreation area in the City of Detroit. The redevelopment was pro- posed by the Huron-Clinton Metropolitan Authority of southeast Michigan, however, the agency recognized that it did not have basic information on the recreation behavior, aspirations and attitudes of the inner city populations which comprise the present and potential park users. Such Norman Ronald Richards information appeared necessary for appropriate redevelop- ment and program planning. In order to assess these citizens' recreation needs and desires, the Authority entered into an agreement with the Recreation Research and Planning Unit, Department of Park and Recreation Resources, Michigan State University, to conduct a pilot research survey. The purpose of this thesis, which was a sub-project of the pilot study, was to investigate the household recreation behavioral patterns and selected characteristics of urban residents in relation to Belle Isle Park and to identify possible relationships which may assist in predicting future recreation participation. A random start, systematic cluster sampling tech- nique was employed to draw a total sample of 360 house- holds distributed throughout four geographic areas in Detroit. Difficulties in interviewing limited the number of responses to 212. The socio-economic and distance characteristics which were postulated as having an influence on partici- pation patterns were investigated in two ways: through the use of a crosstabulation method, where no allowance was made for the interdependence of such characteristics; and through the use of a multivariate technique. Results of both methods of analysis revealed that the best single determinant of recreation participation Norman Ronald Richards patterns was the distance factor. The socio-economic characteristics which were found to be significantly related to participation included: number of children, family income, occupation, length of residence, ethnic origin and ownership of recreational equipment. An attempt to predict recreation participation at the park proved to be limited in scope with only 25.7 per cent of the variance being explained in the socio- economic characteristic plus distance variable model. Although this result is comparable to other similar studies, it nevertheless follows that characteristics additional to the study variables must also be determi- nants of recreation participation. RECREATION BEHAVIORAL PATTERNS AND CHARACTERISTICS OF URBAN RESIDENTS IN RELATION TO BELLE ISLE PARK, DETROIT, MICHIGAN BY Norman Ronald Richards A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Park and Recreation Resources 1974 f \3 ACKNOWLEDGMENTS Throughout the course of this study a number of people have been extremely helpful and to them I would like to extend my appreciation. I must offer my warmest thanks to Dr. Michael Chubb, my major advisor, for his encouragement, patience and invaluable advice during the formulation and comple- tion of this thesis. Special appreciation goes to Professor and Chair- man of the Department of Park and Recreation Resources, Michigan State University, Mr. Louis F. Twardzik, who was instrumental in providing me with the opportunity to receive both practical and academic experience in the field of park and recreation resources. I would also like to thank Professors Milton Steinmueller and Clifford Humphrys for their advice and guidance during my graduate degree program. Computer programming and statistical advice from Dr. Donald Hallman and Mr. Larry Douglas, both of the Ontario Government, are gratefully acknowledged. ii Many thanks go to the secretarial staff and the staff of the Recreation Research and Planning Unit, Department of Park and Recreation Resources, particularly Mrs. Vicki Selby and Miss Anne Mills for their co-. operation. I am grateful to Mr. Robert Bryan, Development Manager and Landscape Architect for the Huron-Clinton MetrOpolitan Authority, who participated in planning the field Operation of the study and provided valuable infor- mation and suggestions. A great deal of gratitude is extended to my parents, Mr. and Mrs. Norman E. Richards, who have always supported my educational goals and career ambitions. Finally, I wish to acknowledge the effort of my wife, Carol, who inspired me to enter the park and recreation profession by encouraging me to further my education at Michigan State University. iii TABLE OF CONTENTS Page LIST OF TABLES O O O O O O O O O O O O 0 vii LIST OF FIGURES O O O O O O O O O O O O x Chapter I. INTRODUCTION . . . . . . . . . . . 1 Urban Recreation . . . . . 1 Nature of the Problem and the Need for Research . . . . . . . . . 5 Objectives and Hypotheses . . . . . . 8 Background Information on Belle ISIe Park 0 O O O O O O O O O O 9 Regional and City Setting . . . . . 9 Evolution of the Park . . . . . . . l2 Redevelopment Plan . . . . . . . . 18 Relevant Studies to be Reviewed . . . . 20 II. REVIEW OF RELATED STUDIES . . . . . . . 22 Findings of Recreation Participation Studies . . . . . 22 Methods of Analysis Used in Other Studies . . . . . . . . . . . 32 III. METHODOLOGY . . . . . . . . . . . 37 Survey Design . . . . . . . . . . 37 Pilot Survey . . . . . . . . . . 37 Sampling Techniques . . . . . . . 39 Questionnaire Design . . . . . . . 45 Interviewing . . . . . . . . . 47 Editing and Coding . . . . . . . . 50 iv Chapter Page Analysis of Data . . . . . . . . . 51 Tabulations and Statistical Techniques . . . . . . . . . 51 Computer PrOgrams . . . . 52 Dependent and Explanatory Variables . . 55 Hypotheses Testing . . . . . . . 56 IV. PRESENTATION OF RESULTS . . . . . . . 62 Recreation Participation and Aspirations Regarding the Study Area . . . . . 62 Participation in Recreational Activities . . . . . . . . . 62 General Participation Patterns . . . 68 Aspirations Regarding the RedevelOpment Plan . . . . . . . 71 Comparison of Non-Participant and Participant Households by Selected Characteristics . . . . . . . . 75 Occupation . . . . . . . . . . 76 Family Income . . . . . . . . . 78 Education . . . . . . . . . . 80 Age . . . . . . . . . . . . 80 Ethnic Origin . . . . . . . . 83 Number of Children . . . . . . . 85 Length of Residence . . . . 85 Number of Types of Recreational Equipment . . . . . . . . . . 88 Distance . . . . . . . . . . . 90 Predictive Qualities of Selected Characteristics . . . . . . . . 93 Socio- Economic Model . . . . . . 93 Socio-Economic Plus Distance Model . . 95 V. CONCLUSIONS AND RECOMMENDATIONS . . . . 99 Summary of Results . . . . . . . . 99 Limitations of the Study . . . . . . 102 Chapter Page Recommendations for Further Study . . . . 106 Home Interview Surveys . . . . . . . 106 Prediction Method . . . . . . . . 107 SELECTED BIBLIOGRAPHY . . . . . . . . . . . 109 APPENDICES Appendix A. Belle Isle Park RedevelOpment Study Family Questionnaire--Help Plan Your Park . . . 114 B. Interviewer Summary Sheet . . . . . . . 122 C. Range of Responses for Question 20 . . . . 123 D. Summary of Multiple Regression Computer Program Results: Socio-Economic Model . . 125 E. Summary of Multiple Regression Computer Program Results: Socio-Economic Plus Distance Model . . . . . . . . . . 126 vi Table 10. LIST OF TABLES Summary of sample information . . . . . Range of participation in recreational activities at Belle Isle Park . . . Range of participation in recreation activities at Belle Isle Park in person participation days per household . . . . Participation in specific activities at Belle Isle Park by participant households . . . . . . . . . . Popularity ranking of recreation activities at Belle Isle Park by participant households . . . . . . . . . . . A comparison of activity popularity ranking between the 1971 Belle Isle Study and the 1967 Detroit Study . . . . . . . General participation patterns for Belle Isle and Metro Parks by respondent households . . . . . . . . . . . Popularity ranking of individual parks by respondent households . . . . . . . . Ranking of activities suggested by respondent households as being most important in the proposed Belle Isle Park redevelopment plan 0 O O I O O O O O O O O O Ranking of activities in which respondent households aSpire to participate at Belle Isle if redeveloped according to H.C.M.A. plan . . . . . . . . . . vii Page 49 64 64 66 67 69 7O 7O 72 74 Table 11. 12. 13. 14. 15. 16. 17. 18. 19. Comparison of Belle Isle Park recreation activity non-participant and participant households by head of household occupation categories . . . . . . . . . . Comparison of Belle Isle Park recreation activity non-participant and participant households by family income categories Comparison of Belle Isle Park recreation activity non-participant and participant households by head of household education categories . . . . . . Comparison of Belle Isle Park recreation activity non-participant and participant households by head of household age categories . . . . . . . . . . Comparison of Belle Isle Park recreation activity non-participant and participant households by household ethnic origin categories . . . . . . . . . . Comparison of Belle Isle Park recreation activity non-participant and participant households by number of children in household categories . . . . . . Comparison of Belle Isle Park recreation activity non-participant and participant households by head of household length of residence categories . . . . . Comparison of Belle Isle Park recreation activity non-participant and participant households by number of types of recreational equipment owned per household categories . . . . . . Comparison of Belle Isle Park recreation activity non-participant and participant households by residence to Belle Isle Park distance categories . . . . . viii Page 77 79 81 82 84 86 87 89 91 Table Page 20. Multiple regression summary describing the relationship between household recreation participation at Belle Isle Park and combined socio-economic characteristics . . . . . . . . . . 94 21. Multiple regression summary describing the relationships between household recreation participation at Belle Isle Park and combined socio-economic characteristics plus a distance factor . . 96 ix Figure 1. 2. 3. 4. LIST OF FIGURES Regional Map Showing the Location of Belle Isle Park in Relation to Regional Parks . . . . . . City Map Showing the Location of Belle Isle Park in Relation to Major City Parks . . . . . . . . . Belle Isle Park RedevelOpment Plan Detroit and Vicinity Map Showing Belle Isle Park and Four Sample Areas Page 10 11 19 43 CHAPTER I INTRODUCTION Cities must continue to expand their efforts to provide the kind of recreation the people-want, when they want it, where they want it. Urban Recreation In recent years, rapid social and economic changes have exerted a significant effect upon recreation partici- pation. The nature and importance of these changes are well illustrated in the United States where increasing urbanization has brought about a much greater use of recreation resources. While urbanism has long been a characteristic of this nation's development, its emergence as a dominant cultural attribute dates back only to the early 1920's, when for the first time, urban residents exceeded half of the total population. Today, approximately 70 per cent of the United States pOpulation live in urban areas. By the year 2,000, probably 80 per cent of the 1National League of Cities, Recreation in the Nation's Cities, Problems and Approaches (WaShIngton: NaEIonal League of Cities, 1968), p. 50. total, or about 320 million peOple out of 400 million, will live in the cities and suburbs of the metrOpolitan areas. Increasing urbanism and the growth of large metro- politan centers is greatly influenced by the technological developments and social organization of the country. This gathering together of humanity is, in ecological terms, a re3ponse to an environment favorable to pOpulation concen- tration; it is easier for people to make a living and easier to provide for their needs in urban centers. For example, there can be no effective mass marketing in a predominantly rural society, and without the possibilities of a mass market, much of our present consumer oriented economy could not have been developed. So urbanism appears to be inevitable as changing social cOnditions become increasingly more favorable to further metropolitan development. Large urban centers are major users of land and water resources. These centers, with all their power and wealth, are notorious for transforming woodlots and water- sheds into commercial, industrial and residential develOp- ments. In many cases, the process of subjugation of resources to urban use has not followed any planned sequence or orderly procedure. Consequently, the result: ing patterns of resource utilization are not likely to provide for efficient satisfaction of the needs for recreation generated in a metrOpolitan society. In general, the provision of recreation facilities and the development of adequate access to recreation opportunity for the residents of densely settled urban areas have not been accorded equal consideration in the competition for urban lands. The urban transformation of American civilization has created central problems for recreation planners, developers and administrators. The sheer weight of numbers presses heavily on available recreation resources. The future offers little doubt that the provision of recreation Opportunities will become a much more difficult and complex task as existing urban centers increase in area and as new population centers of substantial size appear. While the general nationwide recreation demands Should triple by the year 2,000, it is reasonably safe to assume that the demand for recreation opportunities a half-hour from home will even be greater.1 During the past several years, an area of growing concern for professionals in the field of park and recrea- tion resource development has been the need to improve recreation facilities and services for those who live in the inner city areas. This problem has become a number lU.S., Outdoor Recreation Resources Review Commission, Outdoor Recreation for America (Washington: Government PrintIng Office, I962), pp. 25-48. one priority for many recreation administrators, particu- larly after the wave of riots which took place in cities throughout the country chiefly during the summer months of the last decade. Clawson and Knetsch have expressed their concern for this growing problem: At the same time that young white couples have moved to the suburbs, the older districts of the city have been increasingly taken over by families of racial, ethnic, or other minorities. The Negroes of the nation are leaving the farms and rural areas and settling in the cities - in northern and western cities as well as in southern ones. Housing and other restrictions often force them to live in older areas within cities - deteriorated areas, which are or often become Slums. This, in turn, vastly affects the recreation problems in such areas. The need for adequate recreation facilities is very great, yet these older parts of large cities nearly always lack park and recreation areas, and the opportunities to acquire additional suitable areas are often sharply li... - .... ..m.........-w.. .«VEM: s in m8c8s88n M 65 88» 63.83». A 4.4. .... .. .. .. , II II. ......._,..~O~ waausgmgfl. 8.88 . _ - a /Aflvl .\. >BO 0 —h0 / a ,z a I .vB...uws,v\ .F.R A \. x... .x ../ «R ...? kw E. , . / \ ... JD \ mg I =5 ,/.\A/ NON . 6v . . x m. . ...x _, \..\ .. . . . N ,, the... . .\ .2 /SB I we... . , x . woo: . . _. . 2.0 K ...\..\mo.: /\/... wasps...“ , . _ .. ... . m . I s RAGE... 90% , . . , . - - . - , - .. ...\ . .mMas 38. B8. .. .. _ I a C. . NOR‘ . m: @086 .1. .023 . .‘Tiwg.__ .nOOhO ' I . Ifilmw...» 9.". F . 685 6.3 . x .32.. 5...; 8'; 5n; _ 1 38a . 1.3% L{%§III;.. ,P g._ I \\ 5.:6 ..... 35. \88» 881.138 8.3 Jose 83 M 520 (an; _ .26 :o .28 e .- u\.I8oou.i.§ A: p k: - L\ n LuIIr: I- Wham... --I rr 44 of sample clusters in a selected geographic area. The quotient derived becomes the interval used to skip through the list of blocks for the purpose of selecting the blocks on which the sample clusters are to be located. The starting point is determined by randomly selecting any number between one and the quotient (the skip interval). 3. After the sample blocks were selected, the final stage involved selecting the actual cluster of housing units in each sample block. For the purposes of the pilot study, a cluster consisted of six housing units, that is, a group of six contiguous housing units on a block. Actual housing unit addresses were drawn from the most up-to—date city directory.1 Each address of the sample block was given a number by starting to count in the north-west corner and moving in a clockwise direction. Since only one cluster sample was required from each sample block the first address of the cluster was selected by taking from a table of random numbers one number be- tween one and the total number of addresses listed for the sample block. The final five addresses on this list were dropped for the reason that they were represented by the sixth last address considering six comprised a cluster. 2R. L. Polk, Detroit City Directory (Detroit, Mich.: R. L. Polk and Company, I970). 45 Questionnaire Design It was decided that the best method of data collec- tion for this study would be the personal interview. This method was used because the information solicited was assumed not to be ready for complete and instant recall by the respondent. Therefore, the situation re- quired one person to talk to another in order to be able to probe peOple's reSponses and reactions to particular questions and, in addition, fully explore the wide range of inquiries which might be received from the respondents. Although less expensive to conduct, the mailed questionnaire method was excluded from use since it lacked the flexibility which is needed when seeking complex in- formation. Furthermore, there are indications that there is sometimes a mail-back bias because the upper socio- economic status groups return a higher percentage of their questionnaires than the lower status groups. It is ex- tremely important to obtain the full representation of the lower socio-economic groups in the study area, eSpecially when it is known that these people comprise steadily increasing proportions of the population of Detroit's urban core. The designing of the questionnaire involved con- sultation between the Recreation Research Unit and the Huron-Clinton Metropolitan Authority. The major problems in the preparation of the questionnaire were the 46 following: which of the potentially innumerable topics of investigation should be included in the interview and in how much depth should each of these be explored? There was also a great deal of discussion about the sequence of topics and the structure which each battery of questions should take. Finally, the design team went through the process of successive refinement of the wording of each question in an attempt to clear up all ambiguities. This process was carried on through the period of field pre- testing, which consisted of interviewing 20 households, and right up to the time the questionnaire assumed its final form. The final questionnaire, a slight revision of the fourth draft, was completed in late May, 1971 and is shown in Appendix A. The questionnaire sought to obtain basic informa- tion concerning Belle Isle Park for each respondent house- hold on the following major topics: 1. general visitation patterns; 2. transportation habits; 3. degree of participation in recreational activities; 4. aspirations for potential recreation participation; 5. socio-economic characteristics; and . 6. attitudes toward the prOposed redevelOpment plan. In conjunction with the designing of the question- naire, careful consideration was given to certain questions which may appear to be outside the comprehension of the 47 respondent. For example, if a reSpondent was not aware of the proposed Belle Isle Redevelopment Plan referred to in question 17, perhaps a graphic representation of the plan could assist the person in responding to further questions. In order to obtain some common understanding of such items and concepts mentioned in particular ques- tions a flip chart of photographs and sketches was prepared for the interviewer's use. Interviewing The interviewer received instructions and full explanation of the procedures to be followed in all cases when introducing and conducting the interview. Questions were to be asked exactly as they were printed on the ques- tionnaire and permissible forms of supplementary questions were carefully prescribed. The goal was to obtain full reSponse to all the questions so the interviewer was per- mitted to make additional comments and clarifications (including the use of flip charts), but no prompting or use of leading suggestions that might plant ideas in the reSpondent's thinking was allowed. Answers to some closed questions were recorded simply by checking a box in the questionnaire. Extensive use was made in this study of Open-ended questions in which the respondent replies in his own words and the interviewer records the answer in the words of the respondent. 48 The addresses selected for the sample were listed on a form called the "Interviewer Summary Sheet" (see Appendix B). This sheet was identified by the census tract number, census block numbers and a list of the streets which bounded the blocks. The six addresses, where interviews were to be conducted, were listed down the left-hand side of the form and the balance of the space provided three sets of columns where the interviewer could record the date, time and status of his calls at each address. Interviewing commenced in early June 1971, however the interviewer, a student and part-time employee of the Huron-Clinton Metropolitan Authority, had great difficulty in completing the planned number of interviews. He faced the problems of being the only interviewer, doing the work on a part-time basis and encountering open refusals or failure to answer doors. A satisfactory number of inter- views and, hence, enough acceptable questionnaires were not completed until early 1972. A total of 2121 usable questionnaires including 118 from Area 1, 52 from Area 2, 35 from Area 3 and 7 from Area 4 were obtained from the selected sample of 360. The lThe Belle Isle Park User Pilot Study Report pre- pared by Dr. Chubb focusedion describing comparisons between black and white recreation participation patterns, therefore, his analysis did not use the 20 questionnaires where the ethnic origin was omitted. These questionnaires were used in this thesis, thereby increasing the total sample from 192 to 212. 49 summary of sample information shown in Table l is presented on a total sample basis. This pattern of presenting data continues throughout the thesis without any attempt to compare the survey results on an area basis due to the small number of usable questionnaires in some geographical areas . TABLE l.--Summary of sample information Number of Percentage Category Interviews of Total Usable Interviews 212 77.4 Unusable Interviews Not at home 33 12.1 Refusals 22 8.0 No such address or vacant 7 2.5 Interviews Attempted 274 100.0 A comparison was made between the sample results and the United States 1970 census data1 (these census data did not become available until after the commencement of the study) concerning respondent ethnic origin patterns. The percentages of the various ethnic groups in the four 1U.S. Department of Commerce, Bureau of the Census, Detroit, Michigan: 1970 Census of POpulation (Washington, D.C.: Superintendent offDocuments,*I972), pp. P-3 to P-75 0 50 geographic areas represented in the study (65.1 black and 34.9 other) are quite close to the percentages of the population in these classes listed in the 1970 census (69.2 black and 30.8 other). This result increases the confidence in the pilot study sample. Editing and Coding The first step in processing the data gathered in the survey was editing the questionnaires. This process involved checking each questionnaire for legi- bility and completeness of recorded answers. Upon reading through the questionnaires it was discovered that a com- plete set of data had not been obtained for some questions which were required for analysis. Such a situation occurred whenever a respondent refused to answer a parti- cular question, especially questions of a personal nature requesting information on socio-economic characteristics of the household. Questionnaires with missing data are a common occurrence in survey research and do present a problem. However, such questionnaires may still be in- cluded in the analysis providing the computer system being used has Special features for processing such missing data. A system of this type was used in this study and will be discussed in the next section. All responses to questions which were to be analyzed in this study were then coded before being trans- ferred to data processing cards. In essence, numbers were 51 assigned to all verbal responses. For computer analysis all data must be numerical, therefore, they were all con- verted to numbers which were coded to the actual meanings of each response. Analysis of Data Tabulations and Statistical Techniques include: 1. 2a. 2b. 3a. The methods of analysis adopted for this thesis Frequency distribution tables to present data on household recreation participation patterns, popu- larity of different activities and recreation aSpirations with respect to the proposed redevelOp- ment plan for Belle Isle Park. Crosstabulation tables to indicate possible rela- tionships between household recreation participa- tion patterns and each selected characteristic acting separately as well as to compare the differ- ences between non-participant and participant households. Statistical tests to analyze the crosstabulations by testing the significance of each relationship and measuring the strength of the relationships. Multiple regression to explain and predict rela- tionships between household recreation participation 52 patterns and selected characteristics acting interactively. 3b. Statistical tests to analyze the multiple re- gression by testing the significance of relation- ships, measuring the strength of these relation- ships and determining the predictive value of various combined characteristics. Computer Programs The SPSS1 computer system was used on an IBM 360 computer to perform the crosstabulation and multiple regression analysis. The crosstabs program permits the researcher to compute and display two-way by n-way crosstabulation tables which may be used to investigate relationships between two variables. Crosstabs produces a sequence of two-way tables showing along the vertical axis the values of one variable and along the horizontal axis the values of a second variable. This program contains a rather com- plete complement of significance tests and measures of association which are available to statistically analyze the distribution of frequency counts in the tabular displays. 1Norman Nie et al., Statistical Package for the Social Sciences (New York: McGraw-HilIfiBOOR Company, 1970). 53 The test statistic deemed most appropriate to investigate the significance of the relationship between the two variables, household recreation participation and individual selected characteristics, was the chi-square test. Since chi-square does not measure the degree of association or the strength of the relationship, an addi- tional statistic, referred to as Cramer's V, was computed for this purpose. The multiple regression program allows one to study the linear relationship between a set of explanatory or independent variables and one or more dependent vari- ables. In multiple regression analysis, the assumption is made that a dependent variable is influenced not only by a single independent variable but also by two or more independent variables. The basic goal of multiple re- gression is to produce a linear combination of independent variables which will correlate as highly as possible with the dependent variable. This linear combination can then be used to predict values of the dependent variable, and the importance of each of the independent variables in that prediction can be assessed. A variety of statistics can be produced with the use of the multiple regression program. Those used in this study include: the F statistic to test the signi- ficance of relationships between the dependent variable and several independent variables simultaneously; the 54 multiple correlation coefficient (R) which indicates the measure of association between the dependent variable and the independent variable; the multiple determination co- efficient (R2) which determines how much of the total variation in the dependent variable can be explained by all of the independent variables acting together. The actual multiple regression program utilized incorporated a stepwise technique. Stepwise regression is a variation of multiple regression which provides a means of selecting independent variables which will provide the best prediction possible with the fewest independent variables. This technique recursively constructs a pre- diction equation one independent variable at a time. The first step is to choose the single independent variable which is the best predictor. The second independent variable to be added to the regression equation is that which provides the best prediction in conjunction with the first variable. This procedure continues in a recur- sive fashion adding variables step by step until the desired number of independent variables is reached or until no other variable will make a significant contribution to the prediction equation. As mentioned in the previous section, the measure- ment of a variable is often not obtained for every case in social science research. In survey research, this is often because respondents refuse to answer certain 55 questions or respond with “don't know," or it is occasionally due to interviewer omission. For these reasons, the SPSS crosstabs and multiple regression pro- grams include a number of processing options which enable the researcher to specify missing values for each of the variables so that files containing cases with incomplete data may still be conveniently processed. Dependent and Explanatory Variables The dependent variable for this thesis, measuring the level of household recreation participation, is a binary variable which is coded "1" if the household had at least one member participate in a recreation activity at Belle Isle Park one or more times in the preceding year and "0" if the household did not. In order to project recreation demand it is neces- sary to identify those factors which affect demand. Based on the research findings cited earlier, the explanatory or independent variables selected for study and postulated as having a significant influence on household recreation participation are as follows: socio-economic characteris- tics such as occupation, family income, education, age, ethnic origin, number of children, length of residence, and recreational equipment; as well as a distance factor. 56 Hypotheses Testing After identifying the variables to be studied, the next step is to investigate sets of relationships. One or more procedures for examining relationships are selected depending upon the characteristics of the variables and the purpose of the research study. Some form of table display or correlation analysis may be chosen to identify such relationships. A statistical analysis of the significance of relationships requires that specific hypotheses be formu- lated and tested. Each of the hypotheses entertained in this thesis is stated below in the form of a research hypothesis which assumes that there is a relationship between variables. The sub-hypotheses are those which will be tested specifically as a means of proving the more general hypotheses. These general and specific hypotheses include: H1: There is a relationship between household recreation participation patterns and each selected characteristic acting separately. la: There is a relationship between household recreation participation patterns and each selected socio-economic characteristic acting separately. lb: 2a: 2b: 57 There is a relationship between household recreation participation patterns and a distance factor acting separately. There are relationships between household recreation participation patterns and selected characteristics acting together. There are relationships between household recreation participation patterns and selected socio-economic characteristics acting together. There are relationships between household recreation participation patterns and selected socio-economic characteristics plus a distance factor acting together. Hypotheses.--The chi-square statistic is em- ployed to analyze the frequency count data of the cross- tabulation tables by testing H1 hypotheses. Although socio-economic characteristics are considered under one sub-hypothesis for convenience, it is important to emphasize that a separate chi—square computation is undertaken for each hypothesized relationship between two variables. This test is commonly referred to as a test of differences, thus, as well as testing the significance of relationships between variables it also tests the significance of differences between several groups. The 58 chi-square test indicates whether or not the observed departures of frequencies between independent sample group are significantly different from those frequencies exactly prOportionate to the total number in the studied cate- gories and sample groups. In this thesis, the test of differences results in a multi-category comparison of two populations, that is, non-participant households and participant households. A difference is considered significant or real if as large or larger departures from the expected numbers could occur from chance sampling fluctuations not more than 5 per cent of the time (0.05 significance level). Only if a signi- ficant difference occurs could the two pOpulations be interpreted as different from one another on the items compared. This method of analysis and interpretation avoids the error of overlooking the effect of sample size upon the reliability of the percentages that could be calculated for the two populations. Chi-square gives the most accurate result when applied to tables with a large value of N, as chi-square distribution tables are based on large samples. Therefore, when the expected frequencies in some categories, or cells, of the table do not have at least five observations in them, it is necessary to make some correction for con- tinuity, as the possibilities of different values for chi-square are rather limited when the cell frequencies 59 are small integers. An exception exists if the degrees of freedom are greater than 1. In this case it is advisable to proceed with calculations, providing no more than 20 per cent of the expected frequencies are less than 5.1 Since none of the tables exceeded the 20 per cent level it was unnecessary to combine or collapse some cells to provide for a cell frequency of 5 in each case. H2 hypotheses.--A stepwise multiple regression technique is used to investigate possible statistical cor- relations between household recreation participation patterns and selected socio-economic and distance charac- teristics by testing H2 hypotheses. It is expected that the explanatory variables interact to significantly affect the observed differences in the dependent variable. To capture the interaction effects among the explanatory variables, eight socio- economic characteristics were hypothesized to be necessary in the first type of model. This model is expressed in the following form: S+...+bS+E P = a + b S + b 2 8 8 l l 2 where: i P - is the dependent variable of whether or not a household member- ship participated in at least one recreation activity. 1Sidney Siegel, Nonparametric Statistics for the Behavioral Sciences (New York: McGraw-Hill Book Company, 1956), p. 457 60 a, bl-b8 - are constants. 1, ..., SB - are the socio-economic charac- teristics which are combinations of explanatory variables. E - is the per cent of error deter- mined in the analysis. In the second model, the above equation is expanded by adding a distance factor to the eight socio- economic characteristics. This step is taken since the results of prediction equations of related recreation re- search indicate that when a distance variable is included with socio-economic variables the R2 seems to improve slightly.1 The form of the second type of model is: P = a + b S + b S l l 2 + ... + b S + b D + E 2 8 8 9 where: All symbols remain identical to those in the first model with the exception of the addition of D. D - is the distance factor which is combined with the other explanatory variables. In the above mentioned multiple regression models each of the independent variables is simultaneously lElwood Shafer and George Moeller, "Predicting Quantitative and Qualitative Values of Recreation Par- ticipation," Recreation Symposium Proceedings, p. 13. 61 tested and ranked according to its influence acting to- gether with other variables and given a percentage rating as to its importance with other variables in determining the dependent variable. The stepwise multiple regression technique allows variables to be admitted into the equation even though they may be relatively insignificant in the overall effect. In the application of stepwise methods, an attempt is made to derive a regression equation containing variables that are significant at some prescribed level of confi- dence. Significant variables are those which, when included in the equation, account for sufficiently large proportions of the total variance in the dependent vari- able so that the relationship is unlikely to have resulted by chance. Because the significance of an independent variable in the equation will change with the addition of new variables, each variable already in the equation is tested for significance immediately after the addition of each new variable. The actual test of significance is accomplished by the F statistic which was discussed earlier. The minimum level of significance for testing the hypotheses was set at 5 per cent. Therefore, the confidence level was to be 95 per cent. The results of the various methods of analysis used are expressed in the succeeding chapter. CHAPTER IV PRESENTATION OF RESULTS This chapter describes the results obtained from various tabulations and statistical analyses. The des- cription includes a discussion of household recreation behavioral patterns and aspirations, a comparison of non- participant and participant household characteristics, and an examination of the predictive qualities of a combina- tion of factors which are assumed to affect recreation demand. Recreation Participation and Aspirations Regarding the Study Area Participation in Recreational Activities The findings on recreational activities are pre- sented both in terms of the number of activities and person participation days (refer to Question 6). With a large list of activities possible, the household respond- ent could have mentioned any number from 0 to 26 activi- ties and possibly more if he chose to write in others. The term "person participation days" is simply the total 62 63 number of occasions in which one or more members of a household participated in all of the activities indicated by the respondent. Obviously, households with larger numbers of members have an opportunity to generate more person participation days by going the same number of times but these values are useful in estimating a rough ordering of various segments of the population according to the degree of participation in recreational activities. The ordering is based on the number of activities and on the frequency with which each activity was undertaken. Table 2 presents the figures on participation in recreation activities as taken from the sample. The values show that 40.6 per cent of the households had one or more members who participated in at least one recrea- tional activity at Belle Isle Park during the preceding year. Almost 80 per cent (79.1) of the participant house- holds had members participating in up to four different activities, while 17.4 per cent participated in from five to eight activities and 3.5 per cent in over nine activi- ties. The mean number of activities per participant household is calculated to be 3.3. Participation rates in terms of person participa- tion days are shown in Table 3. The actual values range from totals of 0 to 432 person-participation days per household. The greatest number of participant households were in the l to 99 range. Some of the households 64 TABLE 2.--Range of participation in recreational activities at Belle Isle Park Number of Number of Percentage Activities Households of Total 0 126 59.4 1 11 5.2 2 29 13.7 3 16 7.5 4 12 5.7 5 7 3.3 6 4 1.9 7 l 0.5 8 3 1.4 9 2 0.9 10 0 0.0 11 l 0.5 Total Respondent Households 212 100.0 TABLE 3.--Range of participation in recreation activities at Belle Isle Park in person participation days per household Number of Person Number of Percentage Participation Days Households of Total 0 126 59.4 1 _ 99 75 35.4 100 - 199 3 1.4 200 - 299 5 2.4 300 - 399 2 0.9 400 - 499 l 0.5 Total ReSpondent Households 212 100.0 65 derived their person participation days by frequent par- ticipation in a few activities. In the same manner, others obtained similar totals by less frequent participa- tion in more activities. The results of Table 4 give some interesting in- formation on what participant household members did at Belle Isle Park. Picnicking, driving for pleasure, walk- ing and relaxing generated by far the largest number of person participation days for individual activities. By aggregating the person participation days and dividing by the number of households for each activity, the average number of person participation days per household is de- rived. When added together, these figures show that households with members who participated at Belle Isle Park during the preceding year generated an average of about 33 person participation days per household. Individual activities are measured for their pOpu- larity in Table 5. The top fifteen activities are ranked in order from those in which most households had members taking part to those in which household members partici- pated least. The most popular activities listed in order of importance are picnicking (75.5 per cent), driving for pleasure (60.5 per cent), walking (33.7 per cent), relax- ing (29.1 per cent), and playing ball (17.4 per cent). Recreational activities similar to those used in this study were also ranked according to their popularity 66 TABLE 4.--Participation in specific activities at Belle Isle Park by participant households (Based on a sample of 86 participant households) Number of Average . . Person Number of Person Days ACtIVltY Participation Households per Days Household Picnicking 2,195 65 33.8 Driving for Pleasure 1,807 52 34.6 Swimming 314 9 34.9 Fishing 193 10 19.3 Walking 1,155 29 39.5 Relaxing 1,684 25 67.4 Band Concert 14 3 4.7 Playing Ball 495 15 33.0 Bicycling (rental and owned) 83 7 11.9 Canoeing 81 13 6.2 Golfing (range and course) 25 3 8.3 Children's Zoo 209 12 17.4 Aquarium 13 4 3.3 Conservatory 167 5 33.4 Great Lakes Museum 28 4 7.0 Other 384 13 29.5 Totals 8,847 269 32.9 67 TABLE 5.--Popu1arity ranking of recreation activities at Belle Isle Park by participant households (Based on a sample of 86 participant households) 3:322:53. Panza? 1 Picnicking 65 75.5 2 Driving for Pleasure 52 60.5 3 Walking 29 33.7 4 Relaxing 25 29.l 5 Playing Ball 15 17.4 6 Canoeing 13 15.1 7 Children's Zoo 12 13.9 8 Fishing 10 11.6 9 Swimming 9 10.5 10 Bicycling (rented and owned) 7 8.1 11 Conservatory 5 5.8 12 Aquarium 4 4.7 13 Great Lakes Museum 4 4.7 14 Band Concerts 3 3.5 15 Golfing (range and course) 3 3.5 68 in the 1967 Detroit study of living patterns and attitudes. Table 6 allows a comparison of the results. The top three activities considered together for each study are directly comparable, with picnicking and driving for pleasure ex- changing ranks for first and second positions. Only swim- ming and golfing show any appreciable difference in ranking. General Participation Patterns In Table 7, an attempt has been made to compare general patterns of participation for Belle Isle Park with Huron-Clinton Metropolitan Authority parks (see Question 7). Slightly over half (58.9 per cent) of the 209 respondent households diSplayed in this table had at least one member participate at Belle Isle Park or one of the Metro Parks during the preceding year. Approximately 40 per cent (40.7) of the respondent households had parti- cipation at Belle Isle Park while a similar per cent (43.6) participated at one or more Metro Parks. A further comparison between Belle Isle Park and individual Metro Parks reveals in Table 8 that the most popular park of respondent households is Belle Isle. MetrOpolitan Beach and Kensington Metropolitan Park rank a close second and third respectively, whereas Stony Creek Metropolitan Park and Lower Huron MetrOpolitan Park apparently receive little participation from the respondent households. 69 TABLE 6.--A comparison of activity popularity ranking between the 1971 Belle Isle Study and the 1967 Detroit Study 1971a 1967b Percentage Percentage Activity Rank of Total Rank of Total Participant Participant Households Households Picnicking 1 76 2 67 Driving for Pleasure 2 61 l 72 Walking 3 34 3 52 Canoeing/ Boating 6 15 5 34 Fishing 8 12 5 34 Swimming 9 11 4 45 Golfing 15 4 9 12 aFrom Table 5. bJohn B. Lansing and Gary Hendricks, Livin Patterns and Attitudes in the Detroit Region, Technical Report ofTTALUS (Ann Arbor: Survey Research Center, University of Michigan, 1967), p. 14. 70 TABLE 7.--General participation patterns for Belle Isle and Metro Parks by reSpondent households Park Number of Percentage Households of Total Participated at Belle Isle and one or more Metro Parks 53 25.4 Participated at Belle Isle but not at other Metro Parks 32 15.3 Did not participate at Belle Isle but did at other Metro Parks 38 18.2 Did not participate at either 86 41.1 Total Respondent Households 209 100.0 TABLE 8.--Popularity ranking of individual parks by respondent households (Based on a sample of 209 reSpondent households) Number of Percentage Rank Park Households of Total 1 Belle Isle 85 40.7 2 Metropolitan Beach 67 32.1 3 Kensington 65 31.1 4 Stony Creek 18 8.6 5 Lower Huron 5 2.4 71 Aspirations Regarding the Redevelopment Plan One way of determining the potential need for various recreational activities is to identify those activities most desired and measure the degree of interest in these activities. Respondents were asked to indicate the most de- sired activities that should be provided in the re- developed park (Question 20). Up to three suggestions per household were recorded. When the aggregate reSponses for all respondent households were ranked in order of the ten most frequently mentioned activities or needs, the pattern in Table 9 emerged. The range of ideas represented by the activity terms in Table 9 and several other categories receiving less than 5 per cent of the mentions are des- cribed in Appendix C. The top ranked activity for all respondent house— holds was picnicking (47 per cent). When the beach and swimming categories were combined, they constituted the second most mentioned category for the respondents (32.9 per cent). Similarly, when park and recreation area re- sults were combined into one category, they represented the third most important category (27.7 per cent). Per- haps, the most interesting result of these data is the high emphasis placed on the apparent need for park serv- ices other than active recreational activities, such as adequate police protection and food concessions. 72 TABLE 9.--Ranking of activities suggested by respondent households as being most important in the. proposed Belle Isle Park redevelopment plan (In percentages based on a sample of 198 reSpondent households) a Rank ACtiVitY P§§°$2EZIQ l Picnicking 47.0 2 Beach 16.7 3 Swimming 16.2 4 Recreation Areas 14.6 5 Park Areas 13.1 6 Sports 9.1 7 Play Areas 7.1 8 Dancing 6.6 9 Food 6.0 10 Fishing 5.6 Music/Theatre 5.6 Boating/Canoeing 5.6 73 The degree of interest in various recreational activities, prOposed in the conceptual plan for the park, is shown in Table 10 (Question 21). Each activity was ranked according to the total number of households, ex- pressed as a percentage of the total sample, having at least one member aspiring to participate in that particu- lar activity. Table 10 further explores how the desire for participation in the nineteen different activities breaks down with respect to whether these aspirations come from non-participant or participant households. The results confirm those of Table 9, in that, picnicking and swimming are the top two activities. Pic- nicking, swimming, basketball, soft/hardball and canoeing are the five most significant activities where the majority of the aspirations come from non—participant households rather than participant households. By contrast, aspira- tions to go dancing, roller skating, shore fishing and indoor and outdoor ice skating were expressed more fre— quently by participant households. With reSpect to future trends, the data in both Tables 9 and 10 seem to imply large increases in parti- cipation in such activities as picnicking and swimming. Water sports, in particular, fishing, boating/canoeing and swimming were often mentioned as desired activities. In the next two sections of this chapter, the author has attempted to identify those significant factors 74 TABLE 10.--Ranking of activities in which respondent households aSpire to participate at Belle Isle if redevelOped according to H.C.M.A. plan (In percentages based on a sample of 196 respondent households) Rank Activit P 59“? t Participant Percentage y ar IClpan Households of Total Households 1 Picnicking 29.1 25.2 54.3 2 Swimming 22.4 18.7 41.1 3 Indoor Swimming 14.5 15.9 30.4 4 Nature Walks 10.2 10.5 20.7 5 Dancing 6.6 13.6 20.2 6 Basketball 10.7 8.4 19.1 7 Roller Skating 7.1 11.5 18.6 8 Boat Trips 7.7 8.4 16.1 9 Soft/Hardball 9.2 6.4 15.6 10 Canoeing 8.9 5.9 14.8 11 Indoor Ice Skating 6.1 8.4 14.5 12 Children's Zoo 6.9 7.4 14.3 13 Athletics 6.4 6.9 13.3 14 Shore Fishing 4.6 7.6 12.2 15 Plays/Concerts 5.3 5.9 11.2 16 Public Boat Mooring 4.1 4.3 8.4 17 Indoor Ice Skating 2.8 4.1 6.9 18 Boat Launching 3.1 2.5 5.6 19 Par 3 Golf 2.6 2.5 5.1 75 which would affect these future recreation participation trends for Belle Isle Park. Comparison of Non-Participant and Participant Households by SelectEd Characteristics In Tables 9 and 10, an attempt has been made to determine possible future trends in recreational activi- ties at Belle Isle Park by examining frequency count data of people's wishes and aSpirations. This section will present the results of the first stage of the statistical analysis which examines the relationship between household recreational participation patterns and each selected characteristic acting separately by comparing the two populations, non-participant and participant households, with respect to a characteristic or sub-group which has multi-categories. Such comparisons will indicate whether or not differences in participation patterns exist within categories of the various sub-groups, thereby making pos- sible the identification of those sub-groups which have a significant affect on participation patterns. In cases where projections of future changes in the relative size of these sub-groups can be calculated, these comparisons can assist in estimating future needs and demands for recreational activities and facilities. The results are described according to the vari- able which is the basis of comparison. Each table shows the number and percentage of survey households falling 76 into the different sub-group categories as well as a sum- mary of the statistical comparisons of the same categories, using the chi-square test in conjunction with Cramer's V measure of association. Occupation For the purposes of this study, the occupational status of the head of the household given by the respond- ents was coded in seven separate categories which represent combined classes based on the system used by the United States Bureau of Census. Taking Table 11 (Question 12) at face value, there seem to be some differences in occupational levels between reSponse groups. For example, 14.1 per cent of the parti- cipant households fall within the highest category repre- senting professionals and managers while only 4.6 per cent of the heads of non-participant households have occupations this high. Similarly, in the lower part of the occupational status hierarchy, the categories "retired" and "other" (in- cluding students, military and unemployed) experienced a difference of 9.4 per cent and 14.5 per cent respectively; however, both categories show higher percentages under the non-participant household group. A comparative glance at the percentage of households in the two response groups falling in each occupational category indicates a differ- ence between the two. 77 .mflnmcowumamu vacuum 8 mcfiumoflocfl msam> swan m Qua: H on o Eoum oqcmu m m>mc has A>v > m.umfimuuo .mmanmwnm> cmw3umn mocmumwuwp pancamficmfim m mmumowoca msam> mumsgmlflco omuomEoo may “Om mo.v mo Amy muaaflnmnoum «n .coflummag cowummsooo 09m .oma. n > .naoo. v a .mHH.e~ u mumsvmuano “muoz Nam mm mma mzmw>umucH Hmuoa mm m 5H noncommmm oz o.ooa nma o.ooa mm o.ooa moa moaocmmsom ucmocommmm Hmuoe m.¢H mm m.m m o.H~ mm mumcuo o.m ma m.~ m m.HH ma conflumm n.oa om m.oa ma v.m n mow>umm .H 4.44 mm a.me mm m.Hv me mumuonma 0.» ma m.m m e.n m m>aumummo .cmamuom .cmsmummuo ¢.o ma H.m w «.5 m Hmoanmao .mmamm o.m SH H.4H Ha m.v m mummmcmz .Hmaonmmmmoum w # w * w * coaummsooo Hmuoe ucmmHOHuumm ucmmHOHuummucoz mmfluomoumo cowummoooo paonmmsoc mo pawn an moaocmmsoc ucmmfiowuumm 0cm ucmmaowuummucoc >us>fluom coHummnowu xumm mHmH maamm mo comHummEoouu.HH mamas 78 The crosstabulation table alone does give a good breakdown of where survey households fall with regard to occupational status, but it should not be considered as in- dicative of a statistical measure of differences in re- sponse groups. The statistical comparison in this case, however, proves that there is a significant difference between the occupations of the head of non—participant and participant households. Results of the chi-square test indicate the difference between the two response groups or essentially the relationship between the two variables, household participation and occupation is significant at the .001 level. Family Income Comparisons of the number of respondent households in the six total family income categories designated in Question 13 of the questionnaire are made in Table 12. The response rate for this particular question is by far the lowest of all questions with only 147 households reSponding out of a total sample of 212. The distribution of family income shows that the majority of respondent households (57.7 per cent) falls within two categories ranging from $5,000 to $9,999. A larger proportion of participant households are in the $5,000 or more family income range than in the case of the non-participant households. Conversely, there is a smaller 79 .c0wummsv msoocfl 09m .mo. u > .mo. v m .Hom.aa u mumswmuano "muoz mam om ems mama>uoucH Hobos mm mm ow mwmcommwm oz o.ooa sea o.ooa Hm o.ooa mm mpaonmmsom usmocommmm awuoe v.a m m.m m o.o o um>o 6cm ooo.mam m.a~ mm m.- 4H a.o~ ma mmm.ea n ooo.oam m.a~ as N.Hm as o.mm mm amm.m . coo.» m m.m~ «a ~.Hm ma H.am mm mmm.m u ooo.m w m.oa ma N.m m m.ma Ha mmm.e . ooo.m m ~.m NH ~.m N G.HH OH ooo.m m “was: a * w # w # oEoocH maflsmm Hmuoe Hmuoa ucmmaoauumm ucmmwoaunmmucoz mowuommumo meoocfl maflEMM an mpaozmmson ucmmfloauumm can ucmmfiofluummucoc muw>wuom coflummuomu xumm mHmH maamm mo c0mflumm500I1.mH mqmde 80 proportion of participant households in the $4,999 and under categories. The differences between the response groups are further substantiated with the application of the statistical test wherein the difference is found to be significant at the .05 level. Education Education was divided into the five categories listed in Table 13 (Question 14). As evidenced in other similar studies, education levels of the head of the household tend to follow the occupation and income levels to some extent but the differences between categories are not nearly as pronounced. All categories above the high school graduation level show larger proportions for parti- cipant households while all categories below this level indicate larger proportions for non-participant households. However, the statistical test of differences reveals no significant difference at the .05 level. 5% Categories for the age of the household head were structured as closely as possible to those used by the United States Bureau of Census. Perhaps the most inter- esting finding in this study is displayed in Table 14 (Question 15). Unlike the results of related studies where the differences in age between sample groups are usually striking, the results of this table show similar .cowummsv coHumosom 09m .HO. H > sMOo A m sHNHoN “ QHMDWmIflSU "0902 81 mam om mma m3mw>umucH Hmuoa mm ma mm noncommmm oz 0.004 mas 0.004 4s 0.004 mm meaosmmsom ucmocommmm Hmuoe ~.m m m.o m 4.4 4 +4 .mmmaaoo 4.44 om m.m4 on 4.04 04 mus .mmmaaoo 4.04 on «.m4 mm 4.mm mm 4 .Hoocom swam a.mm as m.mm mm 4.N4 m4 mna .Hooaom swam 4.4 a n.~ m o.m m mud .mumuamsmam a 4 w 4 w 4 coaumosom Hmuoe ucmmwowuumm ucmmfiofluummlcoz moauomwumo coflumocom paocmmsoc mo poms ma moaocmmson ucmmwoauumm can ucmmwofluummacoc >uw>fluom cOwummuomu xumm mHmH maamm mo camanmmEOUuu.ma mqmda 82 .cowummsv mmm 09m u > .mo. A m .mp4.m n wumswmufino "muoz man 00 0NH m30a>umucH Hmuoa 0H 0 m momcommmm oz 0.00H 00H 0.004 00 0.00H HNH moaocmmoom pcmocommmm Hmuoe 0.0 «H 0.~ m «.0 0H um>o can m0 0.m 0H 0.~ m 0.0 m 40 u mm m.0a mm m.ma 4H 0.0H ma 4m u m4 0.0m 40 m.H4 mm 0.4m N4 44 u mm m.m~ 0m 0.0m mm 0.0N 0m em I mm H.0 NH m.m m m.m n «N 3 0H 0 t w 0 w * mmd Hmuoa ucmmwowuumm ucmmflofluummncoz mowuoomumo omm oaosmmson mo pawn an mpaonmmooc ucmmflowuumm can ucmmflowuummncoc >ua>fluom coflummuomu xumm meH maamm mo comflHMQEounu.va mqmda 83 proportions between categories. The only noticeable dif- ference in the distribution is that fewer households with older heads tended to use the park. There are 14.8 per cent of the non-participant households in the 55 and over age categories while only 5.2 per cent of the participant households fall into these categories. Statistical testing indicates that there is no significant difference at the .05 level between non-respondent and respondent households for age of the head of the household categories. Ethnic Origin The ethnic origin information is obtained from the upper right hand of the front page of the questionnaire where "b" refers to black, "s" to Spanish speaking and "o" to other. The interviewer was instructed to check one of these categories following the interview. Table 15 shows the ethnic composition of the respondent household population. Households classified as "other" in this study are predominantly white and include Spanish speaking households. As may be expected in the areas surveyed, black citizens constituted the majority (65.1 per cent) of households interviewed. Appreciable differences exist within the categories of both response groups. Partici- pant households of black origin have a much greater pro- portion (75.4 per cent) than non-participant households (58.8 per cent). These differences are statistically 84 .camwuo vaccum 03m. m#MOflUGH OH UGHHMM H030fi>H0flGfl @390 .mo. n > .mo. v m .mp4.m u mumsmmuaao umuoz mam 00 0NH m3ww>uwucH Hobos on ma 0 mc3ocxco 0.00H ~04 0.00H m0 0.00H maa moaocmmsom unmocommmm amuoa H.m0 mma v.m0 mm 0.0m 00 xomHm 0.0m no 0.4m 0H ~.H4 04 uwcuo w 0 w # w # 00mm Hmuoa ucmmflowuumm ucmmflofluummncoz mwwnomwumo cfimfluo owccum paonmmson an moaonmmsoc ucMQAOHuumm ocm ucmmwoauummucoc muw>fiuom coHummuomu xumm meH maamm mo GOmHHmmeouun.ma mqm¢e 85 confirmed by the chi-square test which shows a significant difference at the .05 level. Number of Children The results summarized in Table 16 (Question 16) indicate that households with children under 18 years of age living at home are somewhat more likely to have members participate at Belle Isle Park than those which do not have children at home. There appears to be major differences between non-participant (51.0 per cent) and participant (28.8 per cent) households not having children. A similar difference (22.2 per cent) exists between the two reSponse groups which have one or more children. A significant difference at the .01 level is determined after statistically comparing non-participant and partici- pant households by the number of children in the household. Length of Residence Differences in recreation behavioral patterns according to length of residence are observed in Table 17 (Question 10). It is of interest to discover that 84.6 per cent of householdrheads have resided in Detroit for 11 years or more. The most obvious difference between the response groups is seen within the category "under 10 years." A difference of 17.3 per cent is noted with non- participant households having the greater proportion. 86 .cowummdv coupafino mo HmnEsc 08w .00. u > .Ho. .mmm.m4 n mumsvmuano u302 NHN mm 0NH m3ma>umucH Hmuoe H4 ma 0N mmmcommmm oz 0.00H aha 0.00H mm 0.00H mm mpaocmmoom ucmpcommwm Hobos 0.0 ma 0.0a m H.m m whoa no 0 H.Ha 0H v.0a NH N.0 n m 4.0a 0N m.mN NH N.HH Ha N «.mN 04 0.0N ma m.mN mm a m.av an 0.0N HN 0.Hm 0m 0 m * w * m * cwnpaflno mo Hmbsdz Hmuoa ucmmfluauumm ucmmflofiuummucoz mmflnommumo oaocmmsoc Ca cmupaflno mo umnEdc an moaocmmsoc HGMQNONuHmQ 0cm ucmmHOHuummncoc mufl>auom coapmmuomu xumm mHmH maamm mo conflummEooul.0a mqmde 87 .QOHummov mocmonmu mo cumcmH oem .40. u > .mo. v a .Hmm.4H u mumswmnflao ”muoz NHN om mNH mzmH>umucH Hmuoa 4N m mH mmmcommmm oz 0.00H 00H 0.00H 05 0.00H HHH moHocmmsom ucmocommwm Hmuoa N.N m m.N N 0.N m um>o pcm Hm 4.0 NH m.N N 0.0 0H 0m I H4 m.mH 0N 0.0H mH 0.0 HH 04 I Hm 4.Nm H0 H.mm 0N 0.0m 4m on I HN m.mN mm H.mm 0N N.mN 0N 0N I HH 4.mH 0N N.m 4 m.NN mN 0H Hops: w 4 w * w 4 Amummm ch mocmonmm mo camcoH Hmuoa uchHoHuumm pcmmHoHuummIcoz mmHuommumo mocmpHmmu mo cumcoH oHonmmsoc mo pawn ma moHocwmdon ucmmHoHunmm can ucmmHoHuummIcoc muH>Huom COHummuomu xumm mHmH mHHmm mo GOmHnmmEouII.0H mamme 88 When statistically analyzed, a significant difference at the .05 level is indicated. Number of Types of Recreational Eguipment Since the actual number of each type of recrea- tional equipment owned by a household is not particularly significant, each respondent household received a value of "l" for having at least one item in each equipment type category. Accordingly, with 13 specific equipment type categories and an "other" category listed in Question 9 the possible score per household could feas- ibly be higher than 13, however, all fall within the 0 to 13 range. The types of equipment most commonly owned by households included bathing suits, fishing rods, bicycles, hard/softballs and baseball bats. A comparative glance of Table 18 shows that the absence of recreational equipment in a household somewhat seems to preclude participation while ownership of equip- ment appears to be closely linked with participant house— holds. A greater prOportion of non-participant households (31.6 per cent) compared to 4.8 per cent participant households, are without recreational equipment. At the other end of the scale, households owning one or more types of recreational equipment include 95.2 per cent of the participant households in contrast to 68.4 per cent .cOHummov pwc3o ucmEmHogw HmcoHuwwuomu mo momma mo “whens 09m .mH. u > .Hoo. v m .mmm.m~ u mumswmIHno Hmuoz NHN mm 0NH m3mH>umucH Hmuoa mH m NH noncommmm oz 0.00H 00H 0.00H mm 0.00H 4HH moHocmmsom ucmocommmm Hmuoa q. 0.N 4 0.m m 0.0 H um>o cam 0H 8. N.0H 0N N.0H m 0.0 HH 0 I 0 m.Hm N0 0.0N 4N m.mm mm 0 I 4 0.0m H0 m.Hm M4 0.4N 0N m I H m.0N 04 0.4 4 0.Hm 0m 0 m 4 w 4 w 4 momma ucmEmHovm mo nonssz Hmuoa ucmmHOHuumm ucmmHOHuummIcoz mmHuommumo oHocwmaon mom UmCBO ucmEmstm HmcoHummHomu mo woman Mo Hones: an moHocOmson ucmmHoHuumm can ucmmHoHuummIcoc muH>Huom coHummuomu xumm mHmH mHHmm mo cemHquEooII.mH MHmde 90 of the non-participant households. The difference is statistically significant at the .001 level. Distance The actual distance from the residence of the respondent to Belle Isle Park was measured according to the shortest city street route. This calculation was possible since the address of the respondent household was pre-recorded on the top of the front page of each questionnaire. Table 19 summarizes these data in various mileage categories. It is evident that households situa- ted within close proximity to the park (under 3.0 miles) tend to show a greater proportion of participant house- holds (74.4 per cent) than non-participant households (37.2 per cent). Conversely, a smaller prOportion of participant households (10.5 per cent) than non-participant households (34.2 per cent) are located 5.1 miles or more from the park. The result of the chi-square test, showing a significant difference at the .001 level, is perhaps the most significant one of the study. In summary, the chi-square test showed that non- participant and participant households are significantly different in six out of eight socio-economic variables; namely, occupation, family income, ethnic origin, number of children, length of residence and recreational equip- ment. The remaining two variables, education and age, did not vary significantly between response groups. The 91 .NN. u > .Hoo. v a .mm~.44 u mumswmIHno “muoz NHN 00 wNH wzmH>HmucH Hmuoe 0.00H NHN 0.00H mm 0.00H 0NH mcHonmmsom ucmocommwm Hmuoe m.m 0 0.0 0 0.m 0 um>o ocm H.N m.n 0H m.m m m.0H mH 0.0 I H.N N.m HH 0.0 0 0.4 m 0.0 I H.0 m.m 0H 0.0 0 m.4H 0H 0.0 I H.m 0.4H Hm H.mH mH m.4H 0H 0.m I H.4 m.m 0H 0.0 0 m.4H 0H 0.4 I H.m m.mN 00 4.0m mm 4.HN 0N 0.N I H.N H.4N Hm 0.0m Hm m.mH 0N 0.N Hops: w 4 w 4 m 4 AmmHHE cHV mocmumHo Hmuoe ucmmHoHuumm ucmmHoHuummIcoz mmHuommumo mocmumHU xumm mHmH mHHmm on mocmonmH Nb mcHocmmson ucmmHOHuumm 0cm ucmmHoHuummIcoc >UH>Huom coHummuomu xnmm mHmH mHHmm mo COmHummEouII.mH MHmda 92 comparison of non-participant and participant households by a distance factor reveals a striking difference. The results tend to generally support the H1 type hypotheses which state that there is a relationship be- tween household recreation participation patterns and each selected characteristic taken separately. The majority of the relationships signify a positively directed relation- ship, in that an increase in participation is observed with an increase in the particular variable being examined. Exceptions include distance, education and age variables where a decrease in participation is noticed with an in- crease in each of these variables, thereby indicating a negative relationship. One additional comment concerning the results should be made here. Although the hypotheses testing indicates significant relationships between several pairs of variables, the findings of Cramer's V measure of association consistently show that a relatively weak relationship exists between variables. However, while these measures of the strength of the relationship be- tween variables are not overwhelming in terms of their magnitude, the overall consistent nature of the results, particularly in terms of the existence of relationships regardless of their strength, gives credibility to the hypotheses testing. 93 Predictive Qualities of Selected Characteristics This section presents the results of the multiple regression analysis, the second stage of the statistical analysis, which was used to explain and predict relation- ships between household recreation participation patterns and selected characteristics acting together in inter- correlated sets. Socio-Economic Model The first model to be discussed is that involving the socio-economic characteristics and their combined influence on the dependent variable P, household recreation participation. Table 20 summarizes the results of the multiple regression equation in which all variables ex- plaining at least .001 per cent of the variation in the dependent variable were retained for a final regression run. Four of the five independent variables which were allowed to enter the final prediction model proved statis- tically significant at the .05 level when related to the dependent variable by means of the F test. The final model contained the following variables listed in order of importance: occupation, number of children, length of residence, ethnic origin and income. The multiple correlation coefficient (R) which indicates the degree of association between the five independent variables and the dependent variable has a value of .403. The value of 94 .memHum> ucmocmmmo may CH coHuMHum> mo ucmo Mom H00. pmmmH um cHMmem ou U0HHMh44 .Hw>mH mo. um unmoHuHcmHm4 oommH. Ammo ucmHonmmoo coHumcHEHmumo mHmHuHsz omNO4. Ame ucmHonmmoo coHumHmuuoo mHmHuHsz «ammd «aucmEmHoqm HocOHummHomm «4c0Hum05pm Hnm00. 00NOH. mEoocH mHHSmm 4H04N0. mNmmH. CHOHHO UHGSum 4444mo. 0N4mH. mocmchmm mo gumcmq 404H40. 4m000. coHUHHcU mo uwQEdz «4Hmm0. 4Hmm0. COHummsooo mwcmao mm mm mHanum> moHumHumuomumno OHEocoomIOHUOm omcHnEoo cam xumm mHmH mHHmm um coHquHOHuumm cOHummuomu pHocmmson comaumn chmGOHumHmu mcu mcHnHuommp unmeasm GOHmmmummu mHmHquzII.0N mqm ucwpcmmmp may CH coHuMHum> mo ucmo mom H00. ummmH um cHMmem ou omHHmm«« .Hm>mH m0. um ucmoHMHcmHm4 ommmN. ANmV ucmHonmmoo coHumcHEumumo mHmHuHsz mmwom. Amy ucmHOHmmmoo coHumHmuuoo mHaHuHsz «4mmfl 44N00. 040mm. ucmsastm :oHummuomm mHmoo. M44mN. aoHumoswm «mNoHo. mN04N. cOHummsooo «mmHHo. mmmMN. chHHO 0chum 4NmmHo. HH4NN. mocmpHmmm mo numcmq .HomHo. mmOHN. cmupHHso mo amnesz «mmmmo. mman. msoocH NHHEME «0H40H. 0H40H. wocmumHo mocmsu Nm Nm mHanuw> nouomw mocmumHv m msHm mOHumHuouomumco GHEocooonHUOm @mcHnEoo can xumm mHmH mHHmm um GOHummHoHuumm GOHumwuomH oHonmmsoc cmo3umn mQHcmcoHUMHmu on» OCHanommU mumEEdm GOHmmmume mHmHUHDEII.HN mqmme 97 dependent variable. As expected, the addition of the distance variable assisted in improving the R2 value of the first model. This one variable alone accounted for 40.5 per cent of the variance explained in the second prediction model. More information on the output from the two multiple regression programs is presented in summary tables in Appendices D and E. Results of both multiple regression models gen- erally support the H2 type hypotheses which infer that relationships exist between the dependent variable and the selected independent variables acting together. The analysis further indicates that participation in recrea- tional activities at Belle Isle Park is somewhat in- fluenced by distance, number of children, occupation, family income, length of residence, and ethnic origin. The distance factor is by far the most important single predictor. While the predictive qualities of the ex- planatory variables taken together in the second model only yield an R2 of .26, this value does compare closely with the results of similar studies. For example, as mentioned earlier, the R2 determined by the combined effect of nine socio-economic variables in the ORRRC study1 was .30. However, that study did not utilize a lEva Mueller and Gerald Gurin, Participation in Outdoor Recreation: Factors Affecting Demand Among American Adults, p. 27. 98 distance factor which may have increased the R2. When com- paring the R2 of .26 to the evaluation scale developed by Shafer and Moeller,1 which ranges from "poor" (0 to .20) to "really great" (.81 to 1.00), this predictive value falls within the "so-so" category (.21 to .40) along with the R2 derived from the ORRRC study. The final chapter presents basic conclusions con- cerning the results and limitations of the study in con- junction with recommendations for future research. lElwood Shafer and George Moeller, "Predicting Quantitative and Qualitative Values of Recreation Parti- cipation," Recreation Symposium Proceedings, p. 9. CHAPTER V CONCLUSIONS AND RECOMMENDATIONS Summary of Results Belle Isle Park is a valuable outdoor recreation resource to many urban residents of the City of Detroit. The results of this study indicate that 40.6 per cent of the households had one or more members participating in at least one recreational activity at the park during the pre- ceding year. The most popular recreational activities en- gaged in by the members of participant households, in order of importance, were picnicking, driving for pleasure and walking. In an attempt to determine the potential needs and desires of respondent households, aspirations regarding recreational activities were examined. Picnicking and swimming appear to have broad support as the two top activities which peOple expect to undertake. There is little doubt then, that the main emphasis in redevelOping the park should be on facilities accommodating such activities. 99 100 The H1 hypotheses have been supported by the fore- going analysis. It was determined that there are signi- ficant relationships between household recreation partici- pation patterns and selected characteristics when making no allowance for the interdependence of such characteris- tics. These hypotheses may also be interpreted equiva- lently by stating that there are differences between non-participant and participant households with respect to selected characteristics. The test of differences showed that the two household populations were significantly dif- ferent at the 5 per cent level according to: socio- economic variables such as occupation, family income, ethnic origin, number of children, length of residence and recreational equipment; as well as a distance factor. The multiple regression analysis reveals that there are relationships between household recreation participation patterns and combined selected characteristics when acting together, thereby supporting the H2 hypotheses. Four of the five independent variables entering the final socio- economic prediction model were found to be valid indicators of the dependent variable at the 5 per cent level of significance. The four explanatory variables, in order of importance, include: occupation, number of children, length of residence and ethnic origin. The five variables in the final equation accounted for 16.2 per cent of the variance in the measure of recreation participation. In 101 the second model, where a distance factor was added, recreation participation proved to be most strongly in- fluenced by distance. This factor alone explained 40.5 per cent of the variance (25.7 per cent) in the second prediction model. Other significant indicators of the dependent variable in the second model, in order of in- fluence, were: family income, number of children, length of residence, ethnic origin and occupation. Contrary to past research which has concentrated on non-urban recreation areas, both the crosstabulation and multivariate analysis showed that education and age were not significant determinants of recreation partici- pation. A possible explanation of this outcome could be that when outdoor recreation resources are close to the urban center different conditions exist and certain factors assumed to affect demand no longer have a high degree of significance. The multivariate analysis used and the manner in which it has been used indicate that it offers several contributions to recreation research, especially recreation research on human behavior. Not only is it valuable as an heuristic tool for purely exploratory or descriptive purposes, but it is also valuable as a definitive and confirmatory tool as well. It is hoped that the prediction models formulated in this study and refined versions thereof may be utilized 102 by the recreation planners of the Huron-Clinton Metro- politan Authority to estimate the probability of the members of a household participating in recreational activities at Belle Isle Park. This probability is condi- tional upon the selected characteristics which were determined to be statistically significant in the reSpec- tive equation. By applying such equations to Specific populations in a geographic area, the probability of participation can be determined if preferences and behavior are assumed to remain constant over time. The number of recreation participants can be estimated by multiplying this conditional probability by the number of peOple in each specific population area. If future population distributions and supply conditions are estimated, the equations can be used to forecast the future number of recreation participants. Limitations of the Study The research methods employed in this pilot project were affected by certain circumstances which may limit, to some extent, the accuracy of the final results. One of the major limitations was the errors from sampling. Although the sampling and interviewing tech- niques used appear to have resulted in a reasonably unbiased sample with respect to comparing the study data to the 1970 census data for ethnic origin, some sampling error is apparent. The sources of error include: small 103 sample size, non-response, extended period of interviewing, one person reSponding per household, and cluster sampling. As mentioned earlier, precision of sample estimates depends on sample size with larger samples yielding greater precision by reducing chance errors. Time, money and personnel constraints restricted the sample size to that within the scope of a pilot study. The limited number of responses was primarily due to difficulties encountered by the interviewer such as Open refusals, failure to answer the door, responsible adult absent and no such address or vacant houses. The problems of no such address and vacant houses indicate the shortcomings of using a city directory. These non-reSponse problems, added to the problem of one interviewer attempting to carry out all of the interviews, led to still another contributor of sampling error, that is, a prolonged period of interviewing. Originally, it had been intended that all interviews would be completed in the early spring when, it was assumed, the respondents would be more positively oriented towards the out-of-doors. Further increases in overall sampling error derived from the following two sources: by interviewing one person per housing unit assuming that the respondent was knowledgeable about the recreation behavior and desires of all the members of the household; and by sampling clusters of housing units instead of randomly sampling separate housing units. 104 In general, the sample tends to be of a purposive type especially with respect to the arbitrary selection of geographic areas and unequal distribution of the number of sampling units in each area. Consequently, although some element of randomness is used, the sample is not truly a probability sample. The prime limitation of the questionnaire design involved the respondents being required to have recall over a twelve month period. This occurred in questions seeking information on recreation participation. Most survey statisticians report that this period of time is too long a Span for the average individual. The analysis of imperfect data was also a problem. For one reason or another it seemed impossible to obtain a complete set of data for every case in the file with the exception of the distance information which was pre- recorded on each questionnaire by the author. Such a Situation occurred when a reSpondent refused or neglected to answer a question on the questionnaire. The SPSS com- puter programs utilized have special features for process- ing such missing data as explained in Chapter III, however, such incomplete data affects the rigor of the results. Furthermore, the study was hindered by the type of data used for recreation participation rates in the multiple regression analysis. These data were based on a limited range of aggregate rather than discrete responses 105 (e.g., whether or not at least one member of the household participated in one or more recreational activities at Belle Isle Park during the preceding year). The zero- binary variable was used since the questionnaire did not solicit information on individuals but total household membership participation. Therefore, total participation scores for each household visiting the park would obviously be biased by the number of persons in the household who visited the park. The above procedure resulted in a com- posite or aggregate rather than a discrete continous dependent variable. The results of the attempt to predict participation from selected characteristics, although comparable to other studies, were limited in scope. The total R2, using both the socio-economic variables plus the distance factor in the second prediction model, was only .257. It appears 2 will remain at a low level until addi- obvious that the R tional characteristics to the socio-economic and distance attributes included in this study are considered in a refined multiple regression model. AS Dr. Chubbl cautioned in his report, due to the limitations of the pilot study, the results presented must be looked upon as general indications of recreation be- havioral patterns, aSpirations, attitudes, and related characteristics of respondent household memberships. lMichael Chubb, Belle Isle Park User pilot Study, p. 30. 106 Recommendations for Further Study The data obtained from the Belle Isle Park pilot survey and analyzed in this thesis provides some insight into a previously unknown area, however, the information should only be considered preliminary. Home Interview Surveys If home interview surveys of the present type are carried out in the future by the Huron-Clinton Metropolitan Authority, it is recommended that the sample should be large enough to obtain statistical reliability for rela- tively detailed questions. Since no previous data were available, it was impossible to estimate the desirable sample Size. Previous studies conducted elsewhere have indicated that a substantial sample, perhaps as many as 200 usable responses, is needed per spatial unit of analysis if detailed questions on recreation behavior are to yield valid information. A future large scale recreation survey should be conducted by the Authority, initially and possibly at five year intervals, in order to monitor changing leisure behavior patterns and attitudes so that it may be in the best possible position to make planning and management decisions. 107 Prediction Method Refinement of the predictive models utilized, through the selection of other determining factors, would ensure a more practical and applicable tool in future recreation research and planning for Belle Isle Park. One possibility includes an expansion of the present models, which use the quantitative values of user charac- teristics and a distance factor, by adding variables repre representing relative availability or supply and relative quality of recreational areas and facilities. Shafer and Moeller make the following statement in reference to recreation participation predictions: The major reasons for forecasting the quantitative and qualitative values of outdoor recreation consumption are: to recognize possible implications of long-term recreation- management commitments made today; to prepare now for related commitments that will have to be made rapidly, economically, and with minimum disruption sometime in the future.1 Accordingly, the approach which should be taken in building predictive recreation participation models appears to be the combining of socio-economic, distance, supply and environmental quality variables in the same equation. . Elwood Shafer and George Moeller, "Predicting Quantitative and Qualitative Values of Recreation Participation," Recreation Symposium Proceedings, p. 6. 108 DeSpite certain limitations, the findings of this study hopefully provide a relatively methodical approach and practical statistical basis for the study of human behavior within the realm of recreational planning. SELECTED BIBLIOGRAPHY SELECTED BIBLIOGRAPHY Books Backstrom, Charles H., and Hursh, Gerald D. Surve Research. Minneapolis: Northwestern University Press, 1963. Blalock, Hubert M. Social Statistics. New York: McGraw- Hill Book Company, 1960. Burton, Clarence M. The City of DetEoit, Michigan 1701- 1922. Volume I. Chicago: 8. J. Clark Company, Burton, Thomas L. Recreation Research and Planning. London: George Allen and‘Unwin Ltd., 1970. Catlin, George B. Story of Detroit. Detroit: The Detroit News, 1923. Clawson, Marion, and Knetsch, Jack L. Economics ofOutdoor Recreation. Baltimore: The John HOpkins Press, 1966. Detroit Free Press. Belle Isle: Detroit's Pride. Detroit: Habbin Illustrating Company, 1890. Farmer, Silas. History of Detroit and Wayne County and Early Michi an. Detroit: Silas Farmer and Company, 18 . Goodman, William I., and Freund, Eric C. Principles and Practice of Urban Plannipg. Washington, D.C.: International City Manager's Association, 1968. Leake, Paul. History of Detroit. Volume I. New York: The Lewis PubliShing Company, 1912. Lundberg, George A.; Komarovsky, Mirra; and McInery, Alice P. Leisure: A_Suburban Study. New York: Columbia University Press, 1934. 109 110 Nie, Norman; Bent, Dale H.; and Hull, C. Hadlai. SPSS: Statistical Package for the Social Sciences. New York: McGraw-Hill Book’Company,*I970} Olmsted, Frederick L. The Park for Detroit: Belle Isle Scheme. Detroit: Richmond, Bachus and Company, Quaife, Milo M. The John Askin Papers. Volume I. Detroit: Detroit Library Commission, 1928. Siegel, Sidney. Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill Book Company, 1956. Yates, Frank. Sampling_Methods for Censuses and Surveys. New York: Hafner PuinShing Company, 1960. Reports and Documents Chubb, Michael. Belle Isle Park User Pilot Study. East Lansing: Recreation Research and Planning Unit, Department of Park and Recreation Resources, Michigan State University, 1972. Churchill, Betty, and Zazove, Lois. Prospective Demand for Outdoor Recreation. Outdoor Recreation Resources Review Commission, Study Report 26. Washington, D.C.: Government Printing Office, 1962. City of Detroit, Department of Statistics and Publications. Detroit and Wayne County. Detroit: 1935. Detroit MetrOpolitan Area Regional Planning Commission. Recreation in the Detroit Region. Part III: Home Survey of Regional Recreation Activities. Detroit: 1959. Ferriss, Abbott L. National Recreation Survey. Outdoor Recreation Resources Review CommissiOn, Study Report 19. Washington, D.C.: Government Printing Office, 1962. Huron-Clinton MetrOpolitan Authority. Belle Isle Redevelopment Plan and Report. Detroit: 1970. Lansing, John B., and Hendricks, Gary. Living_Patterns and Attitudes in the Detroit Region. Technical Report of TALUS. Ann Arbor: Survey Research Center, University of Michigan, 1967. 111 Morgan State College and Strategic Planning Corporation. A Planning Study of Urban Recreation Concepts, Behavior, Demands, FaciIities and Programs Leading to the DevelOpment of New Planning Guide- lines. Volhme I: Interpretation and Recommendi- tions. Baltimore: -The Urban Studies Institute, Morgan State College, 1970. Mueller, Eva, andGurin, Gerald. Participation in Outdoor Recreation: Factors Affecting Demand Among American AduIEs. Outdoor Recreation Resources Review Commission, Study Report 20. Washington, D.C.: Government Printing Office, 1962. National League of Cities. Recreation in the Nation's Cities, Problems and Approaches. Washington: National League of Cities, 1968. Recreation Resource Consultants. Recreation in the Lansing Model Cities Area: A Study of Spgre Time Behavior and Attitudes. East Lansing: Recreation Resource ConsuItants, Report No. l, 1972. Reid, Leslie M. Outdoor Recreation Preferences: A Nationwide Study of USer Desires. East Lansing: Department of Resource Development, Michigan State University, 1963. Saroff, Jerome R., and Levitan, Alberta Z. Survey Manual for Comprehensive Urban Planning. Anchorage: Institute ofSoc1aI, Economic and Government Research, University of Alaska, Report No. 19, 1969. University of Illinois, Department of Recreation and Park Administration. A Comprehensive Plan for Conser- vation and Recreation. Volume I: The Leisure Behavior, Attitudes and Interests of the Citizens of Lai§aIle County,Iilinois. Champaign-Urbana: I969. U.S. Department of Commerce, Bureau of the Census. Detroit, Michigan: 1960 Census of Population. WaShington, D.C.: Superintehdent of Documents, 1961. . Department of Commerce, Bureau of the Census. Detroit, Michigan: 1970 Census of Population. WaShington, D.C.: Superintendent of Documents, 1972. 112 U.S. Department of Commerce, Census Bureau. 1965 National Recreation Survey. Washington, D.C.: U.S. Census Bureau, 1965. . Department of the Interior, Bureau of Outdoor Recreation. The 1970 Survey of Outdoor Recreation Activities. Washington, D.CT} Government Printing Office, 1972. Articles and ngers Chubb, Michael. "Belle Isle Pilot Study: Phase 1." A Research PrOposal prepared for the Huron~Clinton Metropolitan Authority. East Lansing: Recreation Research and Planning Unit, Department of Park and Recreation Resources, January, 1971. Cicchetti, Charles Joseph. "A Review of the Empirical Analyses that have been Based upon the National Recreation Surveys." Journal of Leisure Research, IV (Spring, 1972). Clark, Alfred C. "Leisure and Occupational Prestige." American Sociological Review, XXI (June, 1956). Clawson, Marion, and Knetsch, Jack L. "Recreation Research: Some Basic Analytical Concepts and Suggested Framework for Research Programs." Proceedings of the National Conference on Ourdoor Recreation Research. Ann Arbor: Ann Arbor Publishers, 1963. Detroit Tribune. "Band Concerts Popular at Belle Isle." Newspaper article reprint. Detroit: September 8, 1903. Gerstl, Joel E. "Leisure, Taste and Occupational Milieu." Social Problems, IX (Summer, 1961). Knetsch, Jack L. "A Design for Assessing Outdoor Recrea- tion Demands in Canada." Paper prepared for the National and Historic Parks Branch, Department of Indian Affairs and Northern Development, Washington, D.C., 1967. "Assessing the Demand for Outdoor Recreation." Journal of Leisure Research, I (Winter, 1969). 113 Shafer, Elwood, and Moeller, George. "Predicting Quanti- tative and Qualitative Values of Recreation Participation." Recreation Symposium Proceedings. Paper presented at the Forest Recreation Symposium, State University of New York College of Forestry, Syracuse, October 12-14, 1971. White, R. Clyde. "Social Class Differences in the Use of Leisure." American Journal of Sociology, LXI (September, 1955). Directory Polk, R. L. Detroit City Directory. Detroit: R. L. Polk and Company,il970. Unpublished Materials Goodale, Thomas. "An Analysis of Leisure Behavior and Attitudes in Selected Minneapolis Census Tracts." Unpublished Ph.D. dissertation, University of Illinois, 1965. Lindsay, John. "Socioenvironmental Relationships Between Pineview Reservoir, Cache National Forest and the Residents of MetrOpolitan Weber County, Utah." Unpublished Ph.D. dissertation, Utah State University, 1970. APPENDICES Interview no. Tract # Date of interview APPENDIX A Block no. Interviewer ’8 name 8'0 C] b UTE-B- BELLE ISLE PARK REDEVELOPMENT STUDY FAMILY QUESTIONNAIRE-HELP PLAN YOUR PARK! I. ADDRESS-~No. & Sireef Person answering quesfions - Husband Wife Son Daughier ther Marifal sfafus - Married Separaied Divorced Single Widowed 2. DID YOU OR OTHER MEMBERS OF THIS HOUSEHOLD GO TO BELLE ISLE PARK DURING THE LAST YEAR? Yes D No D If "No," go 1'o question 7. If "Yes," ABOUT HOW MANY DAYS DID YOU AND OTHER HOUSEHOLD MEMBERS GO LAST YEAR? (year ending March 30, I97I) ADULTS TEENAGERS UNDER I2 No. x Days No. x Days No. x Days a) In The spring X = X = __ x __=.__. (April-May, I970) x = x = = Tofal Tofal Tofal b) In The summer x = x = x = (June-Aug., I970) x = x = x = Toial Tofai ‘ Tofal c) In The fall x = x = x = I (Sepf.-Nov., I970) x = x = x = Toial Tofai Toiai d) In fhe winfer x = x = X = (Dec.'70-March '7l) _ _ _ ___ x ___:____ ____x ___: ____x ___:____ Tofai Tofal Tofal TOTAL TOTAL TOTAL 115 Interview no. WHAT TYPE OF TRANSPORTATION IS AVAILABLE TO THIS HOUSEHOLD? Household owns car(s) [:3 Friend's car available [:1 D.S.R. [:1 numBer HOW DID YOU AND OTHER HOUSEHOLD MEMBERS USUALLY GET THERE? > O C r ..g (D TEENAGERS UNDER I2 a) D.S.R. bus D) Family car c) Friend's car d) Walked e) Blcycled DDDDDD DDDDDD DDDDDD f) ther (expTeih) DID YOU AND OTHER HOUSEHOLD MEMBERS USUALLY GO BY YOURSELF OR WITH OTHERS? ADULTS TEENAGERS UNDER 12 a) By self 6) This household as a family group c) Par? of a group of friends ’ (nof jusf This family) d) Organized group e) ther DDDDD UDDDU DDDDD (explain) 116 Interview no. WHAT THINGS DID YOU/YOUR HOUSEHOLD 00 AT BELLE ISLE AND HOW MANY TIMES DID YOU 00 EACH ACTIVITY? (DOHT lead respondenf) Number of Number of To+aI User Days for Parlicipanis Days Acfivify a) Picnicked b) Drove around c) Swam d) Fished e) Walked f) Relaxed ) Band concerfs h) PI ed ball I) Bi cIed (own bike) ) Renied bi Ie k) Canoed I) PIa ed If m) Used drivin ra n) Pon rides 0) Children's Zoo ) uarium ) Conservaio r) Gr. Lakes Museum 5) Casino I) Ice skafe u) Tennis v) Runnin track ' w) Handball x) Feed fhe deer ) Defroif Yachf Club z) Defroif BoaI Club Oiher XXXXXXXXXXXXXXXXXXXXXXXXXXX 117 Interview no. DID YOU AND OTHER HOUSEHOLD MEMBERS GO TO ANY OF THESE PARKS DURING THE PAST YEAR? Yes No (If "No," go on Io quesfion 9) If "Yes," ABOUT HOW MANY TIMES? ADULTS TEENAGERS UNDER I2 a) Mefropolifan Beach ::::TTmes ::::Tfifi§§ ::::TTEES b) Kensingion Me'I'ropark __Tlmes ___Times ____Times c) Sfony Creek Mefropark ____Times ____Times ____Times d) Lower Huron Mefropark Times Times Times HOW DID YOU OR OTHER HOUSEHOLD MEMBERS USUALLY GET THERE? ADULTS TEENAGERS UNDER I2 a) D.S.R. bus D) Family car c) Friend's car d) ther (explain) DOES THIS HOUSEHOLD OWN ANY EQUIPMENT FOR RECREATION 0R SPORTS? Fishing rods Yes Number___ No Bicycles Basketballs Hard/soff balls Baseball bafs FooTbaIIs Golf club seis Tennis rackefs Ice skafes Roller skaies Boafs Bafhing suifs Toboggans or sleds CICICICIDCIDDCICICIDDEJ DDDDDDDDDDDDDD ther (specify) 118 Interview no. I0. HOW LONG HAVE YOU LIVED IN DETROIT? Years Monfhs II. ARE YOU PLANNING TO STAY IN DETROIT? Yes No If no, explain I2. WHAT IS THE OCCUPATION OF THE HEAD OF YOUR HOUSEHOLD? (Type of job, no+ empldyer) I3. HERE IS A CHART SHOWING RANGES OF INCOME. PLEASE TELL WHICH BEST DESCRIBES YOUR TOTAL FAMILY INCOME BEFORE TAXES. (Show flash card #I) [:J Under $3,000 $7,ooo-+o $9,999 $3,000 Io $4,999 $I0,000 Io $I4,999 $5,000 Io $6,999 $|5,000 or more I4. WHICH OF THE ANSWERS BELOW BEST INDICATES THE GRADE IN SCHOOL COMPLETED BY THE HEAD OF THE HOUSEHOLD? 'EDDDDDDDDDDDD 4 5 6 7 8 9 I0 II l2 | 2 3 4 5 (grade school) (high school) (college) or more I5. WHAT IS THE AGE AND SEX OF THE HEAD OF THE HOUSEHOLD? Age: years Male[:] Female[:] I6. WHAT ARE THE SEXES AND AGES OF CHILDRED LIVING AT HOME? No children a‘l' homeE] Girls-~Ages Boys --Ages I7. 119 Interview no. DO YOU AND YOUR HOUSEHOLD KNOW ABOUT THE HURON-CLINTON METROPOLITAN PARK AUTHORITY'S PLANS TO REDEVELOP BELLE ISLE PARK? Yes D No D Don'f know wha1' HCMA isD (Explain as required--see flash cards 2 and 3) I9. DID YOU KNOW THAT LAST FALL THE CITIZENS OF DETROIT VOTED TO HAVE THE PARK AUTHORITY TAKE RESPONSIBILITY FOR REBUILDING AND OPERATING BELLE ISLE PARK? Yes [:I NOE] WHAT DO YOU THINK ABOUT THE IDEA? 2CL IF THE VOTERS APPROVE, THE PARK AUTHORITY PLANS TO SPEND $40 MILLION ON REBUILDING BELLE ISLE AND $2 MILLION EACH YEAR FOR OPERATIONS. WHAT DO YOU THINK ARE THE 3 MOST IMPORTANT ACTIVITIES THAT SHOULD BE PROVIDED IN THE REBUILT PART, FROM YOUR HOUSEHOLD'S POINT OF VIEW? Adulfs Teenagers Under l2 E] Cl E] El CI Cl C] CI CI 120 Interview no. 2|. HERE ARE THE MAIN ACTIVITIES THAT HAVE BEEN INCLUDED IN THE FIRST ROUGH PLANS. HOW DO YOU AND OTHER MEMBERS OF YOUR HOUSEHOLD FEEL ABOUT EACH OF THESE ACTIVITIES--WILL YOU OR MEMBERS OF YOUR HOUSEHOLD GO TO BELLE ISLE T0 (Show flash cards #4-I9) DO ANY OF THESE THINGS? I) 2) 3) 4) 5) 6) 7) 8) 9) I0) II) l2) I3) I4) I5) I6) I7) I8) I9) Picknicking Swimming af nice beach Indoor swimming Baskefball Soff or hard ball Afhlefics Indoor Ice skafing Oufdoor ice skafing Par-3 golf Canoeing Public boaf moorings Boa? Trips Boaf launching Shore fishing Dancing See plays & concerfs Nafure walks Children's Zoo Roller skafing Remarks WIII WlII mg: Adulfs Teens Under Adulfs Te§h§ Under I2 I2 DUE] CICIEI CICICI [IUD DUE] CICICI mm CICICI DUI: DUE] [IUD DUE] EJCIE] DUE] CICICI DUE] Dag DEE] [IUD CICID DUI] CIDEI DUI—.1 CIIIIEI CLUE] CIEJCI DUE] DUE] DUI] UCIU [IL—JD DUE] [:ICICI [IUD DUI] [IL—JD [IUD 121. Interview no. HAVING HEARD ABOUT THE PARK AUTHORITY'S PLANS, DO YOU HAVE ANY SUGGESTIONS FOR CHANGES? THINGS TO BE ADDED OR THINGS TO BE LEFT OUT? IF BELLE ISLE WERE OPEN 24 HOURS A DAY, LIGHTED AT NIGHT AS "GOOD AS DAYLIGHT," AND SAFE BECAUSE OF GOOD RANGER PATROLS, WOULD YOU USE IT AT NIGHT AND WHAT WOULD YOU USE IT FOR? Adu I is -- Yes No C] Teens -- Yes No I: Under I2 -- Yes No WHO DO YOU THINK SHOULD PROVIDE POLICE PROTECTION FOR CITIZENS WHO USE BELLE ISLE? (please commenf, if you wish) Defrolf Cify Police HCMA Infegrafed Park Ranger Force DO YOU HAVE ANY GENERAL COMMENTS OR SUGGESTIONS ABOUT BELLE ISLE? DO YOU HAVE ANY GENERAL COMMENTS OR SUGGESTIONS ABOUT LEISURE TIME ACTIVITIES IN DETROIT? o m 4 m N i mnumum QEHB puma maumum mSHB mama woumum «EH9 oven UHonomaom .oz HHmu cum HHmo ncN HHeO umH «0 manhood 30H>HmucH nonasz HOOHm nouns mamcoo Emmmm >m¢zzsm mm3mHDMNBZH m xHQmemd "an Swanson xoon 122 10. ll. 12. 13. 14. 15. APPENDIX C RANGE OF RESPONSES FOR QUESTION 20 Picnic - more, better, good picnic areas; picnics; picnic pavillions. Swimming — clean swimming or water; swimming. Beach - clean beach; beach. Dancing - dancing; dance area; dance hall. Play areas - children's, safe, play area or equipment. Rec. Areas - recreation area; recreation. (The exact meaning of these responses was not investigated.) Youth Areas (Again the exact meaning is unknown but it appears to mean areas which would facili- tate young people meeting and "rapping.") Sports - sports areas; track; sports. Indoor SK/SW - indoor skating (presumably ice skating) or indoor swimming. Boating - boat area; boating; boat rides; canoeing. Midway - midway; "rides." Trails - walking, nature, riding trails. Nature Center - (no other terms included). Children's Zoo - children's zoo; farm animals; new duck ponds; bird watching. Black culture - black or ethnic culture area or activities. 123 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 124 Cross-culture - cultural or cross—cultural activities or areas. MuS./Theatre - music; jazz; open air theatre. Events - events building; events. Comm. Center - community center; place to hold meetings. Sr. Citizens - senior citizens shelter or place to sit. Food - good food; eating houses; restaurants. Fishing - clean, more good fish; fishing; fishing sites. Park areas - good, clean park; park areas. Protection - good patrols; protection. Other - clean, better buildings; restrooms; better parking; roads; transportation; maintenance; black superintendent; casino; museums; etc. I 044.0 m4100.0 m4e~e.e mmwbe.o - -IIImymmzH me4.m. m4ee0.0 meemH.0 mimes.o _ lone , eHo.m.--.:- mmmoe.0 .-. mppwi.0 Hm~00.o mo nimua me~.e mm~40.0. meeem.e meemn.0 rs_xo --- - -mpmnm: - --Neeaenol- .mii4m.e .uzoem4040-. . gheeeooo i m eoeew ohm «ham a - mama_eeb. ewmmH.0 H~m00.0.. 00Ne~.0 emN04.0 . 4.. 12002_ >5—zei .mrnoz_ ~4HmH.0 104N0.0 . eNemH.0 emaom.c . _zLesoamLa no 2.0.40 oiizem . moqm momHH.0 4N4m0.0 . mm4mH.0 m4eem.o . Lezleimmx 10 140200 moxuimie nowom.o 04H4o.o 4meoe.0 emmim.0 .: ‘ , . eieenizo ic amazes. «size mHH4N.0 4me0.o 4Hem0.0 mHL4N.0 M sw>is zo_eeeaooo W040: 2peeeooo a meazcm .wozezo ewe .weeeom a a mseicsezw.- . 1544_44> msmeh.>ep_>ieoe enoiameoz mmoom ..iee<_ee> e2meymeoe a 4.-.-2---,D.-.H msm m a o L x w L. a .H. h ..-: .2 4-4 ... a...» 44.4 * 4-4.4 4.-» .1... 4.4.4..» «.42» ». LSWH .Nexomweo u whee zcihemeo. iaewm will HWQOZ UHSOZOUWIOHUOm umBHDmmm 300mm mMBDnHZOU ZOHmmmmwmm WHmHBHDE m0 ngbm D NHQmemd 125 * mem.e Noeio.e m44me.e meoee.e ..-lsew mem.~ emewo.o m.wi_.eI seeme.eI .hnuseu emm.w ememo.e -mimes.e. smnme.e x»-niooo m-.~ 4m4we.e N40~H.0 mm4-.0 moan mH~.N mmnee.e me4-.e m~4ee.o mn.nwmwe mee.m Nc~4e.0 memes.e 44040.0 .sixm. 44H.4; pesee.o eemm~.e Nomee.e m:uoz. 044.4H mmmee.e emoem.eI. m4wee.eI mmuunm_e i a 40441 ohm . able .e. msuuiue> mp4eo.0 ~4Noo.0 ooemN.0 meoomre p2mseieowluo mma>h.uo.emm::z. w4nusom edemH.e meoe.e. m44mm.oi H44om.e sm>ms zoip4oeem me4l: quesee mHH4N.o 4N050.0 eme4m.e pmew4.e Si>us zoseaeeeoe meal: annnoom. ~4Hm4.o mesio.o memm~.0 ekme4.e heiszoamme 10 210140 0_2:»m moqe momHH.e. Nmmse.e isemm.e e4m~4.e . mozwe_mae 10 leezms mornimme HeweN.o. Hemflo.o.l emeswwo.,- eemm4.e amees_lo 10 410202 ms_io eemeH.o emmo0.0 empes.o em444.0 4. - _ 020020 >3_;41 : Nunez. e-~m.oI 4140H.e 4H4es.e wkwmm.e M eeee o» m0410_mme. mo/4_m_e m wseiim...mozezo 0mm m msmniee> I zim w m.m u 4,0 w a (I I I In) .mimqp >eFH>~eoq enormmeoi HWQOZ WUZ¢BmHD mDHm UHSOZOUMIOHUOm Laws mwoum 1N>\0m\eo u when zoihemmo. umBHDmmm Edmwomm mMBDmZOU ZOHmmmmUmm MHmHBHDS m0 wm name meme 0 ,._ “134...: z- 4 .4 4.... 4144-44 4-4.4. 4 .4 4 4 «.4 4 4 n 4 a. 453mm maxi 126