THE PRGCESS 0F INDIVIBHAL MGDERNIZATION: A PATII M‘GDEL THESIS FOR THE DEGREE 01" 900702 0F PHILGSOPHY MICHIGAN STATE UNIVERSITY ARNOLD GRAY HOLDEN 1972 ____———-—- ____——-—- __’—— # ___’—- '_._——- __._.—-- # -__._——-—— __._——-- ___'——-—- ___.___——— ’ _,_'_————- '_—-—’ ___.——-— # ___—_'-—— ___————- ___,__.——— fl ____—————-- ___._————— ______——-—- ___._—-— ”4-— fl ” _-.——— _‘_.——— _____———— # _'_————— _._.__———- '__———— -_._———-— ____——— ____.-—-—- # __..——-—- _'_._.———- -_-_.——— fl ,_4_————- _.._——— LIBRARY MiCbig’T-ifln Stats: 5; University in- urn-e3 ~ V‘WJ This is to certify that the thesis entitled THE PROCESS OF INDIVIDUAL MODERNIZATION: A PATH MODEL presented by ARNOLD GRAY HOLDEN has been accepted towards fulfillment of the requirements for Ph' D’ fiegreein s°°i°1°9Y NWWZWW Mfimmmmmwr Date 57-22 3 J91, 0-7639 ABSTRACT THE PROCI'BS OF INDIVIDUAL MODERNIZATION: A PATH MODEL By Arnold Gray Holden This thesis presents a general definition of individual modernity, a processual model of individual modernization, and a test of this model using survey data from five nations. Individual modernity, as here defined, is a configuration of values, attitudes, and behaviors characterized by individual autonomy and rational decision making. Autonomy is the individual's perception of his ability to participate successfully in a variety of social systems. Rational decision making is typified by a purposive consider- ation of alternate solutions. The social systemic characteristics which facilitate individual modernity are discussed. This definition of individual modernity is contrasted and compared to earlier definitions found in this field, including relational, listing, and relativistic definitions, and more recent, abstracter, definitions. A processual model of individual modernization is presented and explicated. The salient importance of formal education is noted and discussed. The processual, or path, model states that success in formal education (which is positively related to the antecedent personal and family characteristics) contributes to an individual's power. Arnold Gray Holden Education also contributes to physical and psychic mobility, which in turn are also positively related to power. The individual's perception of his power, which is self-perceived efficacy, is positively related to the components of individual modernity—-autonomy, rational decision making, and behavioral modernity. And, behavioral modernity, assuming modern social systems, increases the individual's power, thus intensifying the process. The model was tested using data from national sample surveys of the United States, Japan, Finland, Mexico, and Costa Rica administered in 1966 and 1967. The central hypotheses were that none of the com- ponents of the path model would be negatively related, and that the relationships between components would decrease as theoretical distance between them increased. A.majority of the components use multiple indicators. The path model was transformed into a matrix, and the resulting correlation matrices were inspected. Two matrices for each of the five nations were constructedf-one for ordinal and nominal dichotomized indicators using the contingency coefficient (C) and one for those indicators measuring at the assumed interval level or higher using the product moment correlation coefficient (r). The five matrices of interval indicators support the hypotheses and the path model to a moderate degree. Of the five ordinal and nominal indicators matrices, only one clearly supports the model, but the others do not indicate any one alternate model. Finally, an analysis of attitude items which were deleted because of intra-component inconsistency, but which bore strong face validity Arnold Gray Holden as indicators of individual modernity as it was traditionally defined, was conducted. Their lack of consistent ties to other components of the model justified their omission, and cast serious question on the utility and validity of earlier listing definitions of individual modernity. THE PROCESS OF INDIVIDUAL MODERNIZATION: A PATH MODEL By Arnold Gray Holden A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1972 ACKNOWLEDGMENTS I wish to first acknowledge the contributions to this study of my guidance committee. Fred waisanen, as chairman, was instrumental in stimulating my interest in the topic of modernization, and his suggestions and criticisms were essential in its growth from tentative ideas to final version. Denton Morrison, through my graduate training, instilled an appreciation for the problems and satisfactions of research. John Useem created and nurtured an early interest in comparative studies. Jim McKee opened the tradition and excitement of sociology to me. Bo Anderson's insightful comments broadened the scope of the study. To these I must add Frank Camilleri's presentation of statistical issues, orientations, and techniques; Frank Holland's library and his arguments against the thesis of this study; and Chris vanderpool's intellectual support and cribbage matches. Included in a list far too long to present here are the invaluable people who participated and participate in my study of sociology as teachers, fellow students, friends, colleagues, and students. And Carlin, who as friend, wife, colleague, researcher, and teacher, Inade this entire project possible with her constant support in so many ways. ii TABLE OF CONTENTS Page CHAPTER I. THE CONCEPT OF INDIVIDUAL MODERNITY . . . . . . . . . 1 Historical Interest in Individual Modernity (1)-—Definitions of Individual Modernity (3)--A definition of Individual Modernity (6)—-Autonomy (6)--Rationa1 Decision Making (9)--The Tie Between Autonomy and Rational Decision Making (11)-—The Tie as Seen in Waisanen and Kumata's Scheme (12)—-This Definition Compared to the Relational Definition (18)--Individual Modernity in Pre-Modern Systems (19g--Individual Modernity and the Nature of the Social System 20 --An Excursus on the Future of Modernity as Seen in the Youth Counter-Culture (23)--This Definition Compared to Relativistic Definitions (25)-- Behavioral Aspects of Individual Modernity (26) CHAPTER II. A PATH MODEL OF INDIVIDUAL MODEENITY . . . . . . . . 28 Causes of Modernity (28)--Education and MOdernity (29)-- Education and MOdernity: A Path Model (31)-—Hypotheses and Analysis (32)--Components of the Model (34)—-Antecedent Conditions-—Persona1 Characteristics (34)—-Antecedent Condi— tions--Family Characteristics (35)--Forma1 Education (36)-- Physical Mobility (37)--Psychic Mobility (38)--The Effect of Physical and Psychic Mobility (40)--Power (41)--Education, Mobility and Power (42)--Self—Perceived Efficacy (43)-- Efficacy and Autonomy (44)--Individual Modernity: Autonomy (44)--Individua1 Modernity: Rational Decision Making (46)-- Individual Modernity: Behavioral Aspects (47)-—Behavioral Modernity and Power (48)--The Process of Individual Moderniza- tion (49)--Imp1ications (49) CHAPTER III. THE METHOD OF TESTING THE MODEL . . . . . 50 Analysis (5dL-The Data (51)-—Genera1 Plan of Analysis (53)-- Some Non-Applicable Analytic Schemes (53)—-Test Hypotheses and Rationale (55)—-An Ideal Causal Model (55)——The Test of the Model of Modernity (56)--Determining the Relationships (58)-- The Search for a Statistic (59)--Interna1 Consistency of Multiple Indicators (61)--Summary Averages (62)--Procedures Used for Calculation (62)--Levels of Statistical Significance (62) iii CHAPTER IV. A TEST OF THE MODEL: INDICATORS AT ASSUMED INTERVAL LEVEL OR ABOVE . . . . . . . . . . . . . . . . . . . 64 Review of Method (64)--Order of Presentation: Level of National DevelOpment (64)--United States Survey: Interval Indicators (65)--Indicators and their Internal Consistency (65)——Comments on Internal Consistency of Indicators: United States (67)--The Test Matrix and Summary Averages (67)-- Comments on the Test Matrix: United States (67)--Finland Survey-—Interva1 Indicators (69)--Indicators and Their Internal Consistency (70)--Comments on the Internal Consistency of Indicators (71)--The Test Matrix and Summary Averages (71)—- Comments on the Test Matrix--Finland (73)--Japan Survey-- Interval Indicators (73)--Indicators and Their Internal Consistency (73)--The Test Matrix and Summary Averages (75)-- Comments on the Test Matrix--Japan (75)--Mexico Survey-- Interval Indicators (75)--Indicators and Their Internal Con- sistency (77)--The Test Matrix and Summary Averages (78)-- Comments on the Test Matrix--Mexico (78)—-Costa Rica Survey-- Interval Indicators (80)--Indicators and Their Internal Consistency (80)—-The Test Matrix and Summary Averages (82)-- Comments on the Test Matrix--Costa Rica (82)--Comments on Interval Test Matrices of All Five Nations (82). CHAPTER V. A TEST OF THE MODEL: INDICATORS AT THE NOMINAL AND ORDINAL LEVEL . . . . . . . . . . . . . . . . . . . . 89 Review of Method (89)--United States SurveyA—Nominal Indi- cators (89)--Components and Their Indicators (90)--Finland Survey--Nominal Indicators (93)--Components and Their Indica— tors (93)-—Japan Survey--Nominal Indicators (96)--Components and Their Indicators (96)--Mexico Survey-—Nominal Indicators (99)--Components and Their Indicators (99)--Costa Rica Survey-- Nominal Indicators (102)—-Components and Their Indicators (102)—- Comments on Test Matrices Using Nominal Indicators—-All Five Nations (105)-—Comments on Indicators--Nomina1 Indicator Matrices 106 --Comments on Components-—Nomina1 Indicator Matrices 109 CHAPTER VI. RESULTS OF THE ANALYSIS OF DELETED INTERVAL INDICATORS . . . . . . . . . . . . . . . . . . . . . 111 Reasons for Further Analysis (111)-—Sca1es and Internal Con- sistency (111)--Absence of Powerlessness (112)--Interna1 Consistency (113)--Political Opinions and Tolerance (115)-- Internal Consistency 115 -—Religious Tolerance (117)-- Internal Consistency 117 --Absence of Xenophobia (119)-- Internal Consistency 120 --Preference for Individual Action (122)--Internal Consistency (122)--A Belief in Planning and iv Chapter VI (cont) Technology (125)--Interna1 Consistency (127)--Belief in Distributive Justice (127)--Positive Attitude Toward Change (129)--Interna1 Consistency (130)-—Comments on the Associa— tions of Deleted Indicators 130) CHAPTER VII. EVALUATION, COMMENTS AND IMPLICATIONS . . . . . . . 133 General Evaluation (133)—-The Definition of Individual Modernity (133)--Problems of Operationalization and Equivalence (135)--Present Evaluation of the Definition (137)--The Path Model of the Process of Individual Modernization (138)-- Selection of Components (138)--The Ordering of Components (139) --Comments on Indicators (139)--Present Evaluation of the Path Model (140)--The Method of Analysis (140)--Advantages of This Method of Analysis (140)--Requirements of the Data and Theories (140)--Its Simplicity and Flexibility (142)—-Some Disadvantages of the Method (142)—-Present Evaluation of the Method of Analysis (143) LIST OF REFERENCES . . . . . . . . . . . . . . . . . . . . . . . 144 Table 10 11 12 13 LIST OF TABLES Test Matrix for United States, Using Assumed Interval Measures or Higher . . . . . . . . . . . . . . . Test Matrix for Finland, Using Assumed Interval Measures or Higher . . . . . . . . . . . . . . . . . Test Matrix for Japan, Using Assumed Interval Measures or Higher . . . . . . . . . . . . . . . . . . . . . Test Matrix for Mexico, Using Assumed Interval Measures or Higher O O O O O O O O O O O O O O O O O O O O O 0 Test Matrix for Costa Rica, Using Assumed Interval Measures or Higher . . . . . . . . . . . . . . . Means of Cell Coefficients of All Five Nations, Using Assumed Interval Measures or Higher . . . . . . . . . Test Matrix for United States, Using Nominal and Ordinal Measures . . . . . . . . . . . . . . . . . . . . . . Test Matrix for Finland, Using Nominal and Ordinal Measures . . . . . . . . . . . . . . . . . . . Test Matrix for Ja an, Using Nominal and Ordinal Measures 0 O I C O O O O O O O O I O O O O 0 Test Matrix for Mexico, Using Nominal and Ordinal Measures . . . . . . . . . . . . . . . . . . . Test Matrix for Costa Rica, Using Nominal and Ordinal Measures . . . . . . . . . . . . . . . . . . . . . . Means of Cell Coefficients of All Five Nations, Using Nominal and Ordinal Measures . . . . . . . . . Mean Correlation Coefficients Between’Absence of Powerlessness" and Components of the Model (Interval Items) . . . . . . . . . . . . . . . . . . . vi Page 68 72 76 79 83 85 92 95 98 101 104 108 114 14 15 16 17 18 19 20 Mean Correlation Coefficients Between "Political Opinions and Tolerance" and Components of the Model (Interval Items) . . . . . . . . . . . . . . . . . . . . . . . . Mean Correlation Coefficients Between "Religious Tolerance" and Components of the Model (Interval Items) Mean Correlation Coefficients Between "Absence of Xenophobia" and Components of the Model (Interval Items) . . . . . . . . . . . . . . . . . . . . . Mean Correlation Coefficients Between "Preference for Individual Action" and Components of the Model (Interval Items) . . . . . . . . . . . . . . . . . . . Mean Correlation Coefficients Between "A Belief in Planning and Technology" and Components of the Model (Interval Items) . . . . . . . . . . . . . . . . . . Mean Correlation Coefficients Between "A Belief in Distributive Justice" and Components of the Model (Interval Items) . . . . . . . . . . . . . . . . . Mean Correlation Coefficients Between "Positive Attitude Toward Change" and Components of the Model (Interval Items) . . . . . . . . . . . . . . . . . . . . . . . . vii 116 118 121 123 126 128 131 Figure 1 LIST OF FIGURES Page The Conceptual Relationship Between Heteronomy, Alienation, And Autonomy O I O O O I O O O C O O O O O O I I O O O 7 Schematic Representation of the Model of the Process of Individual Modernization . . . . . . . . . . . . . . . 33 The Process of Individual Modernization:Conversion of Schematic Path Model (Figure 2) to Matrix Format . . . 57 viii CHAPTER I THE CONCEPT OF INDIVIIIJAL MODERNITY Historical Interest in Individual Modernity The related processes of industrialization and urbanization-— worldwide and accelerating--have had major impacts on the nature of all societies. The resultant changes have altered practically all the social systems of these societies, and restructured the pattern of relationships within and between these systems. Kahl (1959) succinctly presents the major changes occurring in modernization and their character and ramifications. Major alterations are: initial dramatic pOpulation growth from the reduced death rate; an increasingly urban nature of the society, through internal migration, efficiency of agriculture, and a cash nexus of social and economic relationships; a broadening of the person's perceptual world, from local concerns to national concerns, and to an awareness of and empathy with diverse persons; increasing horizontal and vertical occupational specialization alter social stratification systems; education becomes more formal, more universal, and closely tied to the new occupational structure; the nuclear family eclipses the extended family in significance to its members; the average material standard of living is raised; and societal cohesion is more dependent on recognized interdependence than on simi- larities of the participants (pages 54—70). (See also Hoselitz and .MOOre, 1963, and Braibanti and Spengler, 1961.) 1 Sociological interest in the process of societal modernization has been strong since the beginnings of the discipline and the Industrial Revolution. The works of Durkheim, Toennies, Marx, Weber, and others show a dominant interest in the intellectual and social prob- lems of nascent modernization in their time. However, with the notable exceptions of Durkheim's work on anomie, and Weber's study of the Protestant Ethic, strong interest in the psycho-social aspects of modernization did not appear until the advent of modernization in Asia, Africa, and Latin America in the years following WOrld war II. Survey research and public opinion polling in the developing areas asked questions concerning the psycho-social aspect of moderniza- tion (Lerner, 1958; Inkeles, 1960; Cantril, 1951; Buchanan and Cantril, 1953; and others). What kind of social psychological orientations, what attitudes and values, were required for a person to interact effectively in the changed and changing societies? How were they different from the orientations of actors in pre-industrial systems? What processes and experiences facilitated the learning of these new orientations? Answers were sought in the responses of people in exploratory surveys. While the information found in these early surveys has been crucial, the "new" concept of individual modernity tended to be more empirically based than other concepts in sociology, and lacked the abstractness and theoretical sophistication of other concepts which had their origins in the grand theories of classical sociology. Definitions of Individual Modernity Given the origin of the concept of "individual modernity" in empirical studies rather than contemplative theory-building, it is not surprising to note a moderate-to-low level of abstraction in the concept's definition. It is possible to distinguish three types of standard definitions--the relational, the listing, and the relativ- istic—-and a fourth, newer, more abstract, approach. The first of the three traditional types is what I have labeled the relational definition. In it, the writer simply states that individual modernity is that collection of attitudes, values, and actions that enable the individual to participate effectively in a modern social system (Smith and Inkeles, 1966:353). While this definition clearly states the essential tie between societal and individual modernity, it does little to specify what individual modernity, as a discrete concept, is. And by placing individual modernity in a position of total dependence upon success in modern systems, it becomes difficult to use this definition in constructing measures of individual modernity as a discrete variable, or to explore a varying relationship between the two. Another approach to a definition of individual modernity is the listing approach. These lists and discussions of the various character- istics and measures contain from seven (Inkeles, 1969:210) and nine (Inkeles, 1966:141-44)* to thirty-three "major themes" (Smith and Inkeles, 1966:354), and up to 119 indicators of "over-all modernity" *The content of this definition will be discussed below in compar- ing it to a new definition of individual modernity (see pages 13-17). (Smith and Inkeles, 1966:353). Lerner (1958:47-52) provides a discussion of a collection of more abstract components of individual modernity emphasizing a positive value on change, coming from physical, vicarious, and social mobility; rationality; and empathic capacity. Rogers (1969:51-56), in a collection of behavioral and attitudinal components of modernity, which is typical of thinking in this field, includes: literacy, mass media exposure, cosmopoliteness, empathy, achievement motivation, aspirations, lack of fatalism, innovativeness, and political knowledge. Doob (1967) in presenting measures of psychological modernization gives scales which emphasize these eight dimensions of modernity: an emphasis on the future rather than present or past; a belief in the utility of the present legal govern- ment; a feeling that one's life is pleasant and controllable; loyalty to one's country; a faith in science; a generous, trusting view of other pe0ple; approval of the nation's present leaders; and a de- emphasis of tribal ties and values (415-16). A third approach is exhibited by Stephenson (1968) who takes a strong relativistic stance. "Those values defined by the local culture as traditional comprise what may be called traditionalism; those defined as modern constitute modernism" (Stephenson, 1968:268, italics omitted). The relativistic approach presents its own set of theoretical and operational problems, as it essentially denies the possibility or comparative research or universally valid concepts. Each population's perceptions of the amount and type of change, and its perception of the nature of modernism and traditionalism, are valid and intriguing topics for study. However, using the respondents' collective definition of the amount, type, and rate of change will not yield results which are comparable to other studies, except on the most impressionistic basis (Schnaiberg, 1970:402—03). The three approaches to defining modernity all have weaknesses. The relational approach inhibits theory and research on individual modernity as a discrete variable. The listing approach tends to be too concrete, and is subject to criticism of specific items and debate on the relative importance of each item. And the relativistic approach inhibits comparative studies. A fourth type of definition of individual modernity has been evolving. This type attempts to specify the more general, abstract, nature of modernity in a variety of settings. With Lerner's approach (1958:47-52) as an apparent conceptual basis, Waisanen and Knmata (1969) present four cognitive-attitudinal correlates of modern behavioral modes. These are: information seeking; planning and investment; inter-systemic participation; and creativity and innova- tiveness (pages 6e10) Kumata and Waisanen (1969:52) note three central attitudinal aspects which are: the perception of freedom of decision making; a willingness to experiment and take risks; and innovativeness and receptivity to new ideas. The definition of individual modernity presented in this thesis (see page 6), centering on autonomy and rationality, is of this school. MOst recently, Waisanen (1971) has stated that "the essence of modernity is autonomy," with autonomy being "the perception of opportunity for decision-making on the basis of self-relevant criteria" (page 184). A Definition of Individual MOdernity At this point, I offer a new definition of individual modernity. It is abstract, in order to avoid the theoretical and methodological cul—de-sacs of the three traditional types of definitions. It is a part of the newer abstract definitional approach, and subsumes the specific dimensions and characteristics used by other scholars of the field in their definitions of individual modernity. Individual medernity is a configuration of values, attitudes,_and behaviors characterized by individual autonomy and rational decision making. Autonomy Autonomy refers to the actor's perception of the nature of the relationship between himself and social systems.* Autonomy is the actor's perception of bgth his relative independence from any single social system, 222 his interdependence with several social systems. In this scheme, heteronomy would be the perception of complete dependence on a single social system, while ideal-typical alienation would be the perception of complete independence from all social systems. This conception of autonomy and heteronomy come from the work of Piaget (1932) on the source of rules and moral judgments of children. "For very young children, a rule is a sacred reality because it is tradi- tional; for the older ones it depends on mutual agreement. Heteronomy and autonomy are the two poles of this evolution" (page 195). This *I am using the term "social system" in a generic sense to indicate any "collectivity in organized pursuit of consensually carried goals" (Waisanen, 1963:18), and thus refers to families, communities, «Irganizations, associations, nations, societies, etc. development in the child is a result of contacts with more groups outside the family, and is analogous to the individual changes required as a society becomes increasingly specialized in its division of labor (pages 159-60). The brief, general, conceptualization of alienation here is not inconsistent with other works on the concept (for example, see Seeman (1959)). The relationship between these three perceptual states is most clearly seen as a space bounded by the three continua between the three ideal-typical end states (see Figure 1). Alienation A C Autonomy Heteronomy FIGURE 1 THE CONCEPTUAL RELATIONSHIP BETWEEN HEI‘ERONOMY, ALIENATION AND AUTONOMY Any actor may thus be seen as occupying some point with greater or less approximation to the three end states. Actor A (see Figure 1) would be a highly autonomous individual; B would be described as low in autonomy, but alienated from a source of earlier heteronomy; and C as a person between heteronomy and autonomy without significant loss of meaningful systems to participate in. (This conceptualization of these states permits easy discussion of a typical rural-urban migrant in developing countries, with a move from heteronomy to alienation to perhaps autonomy.) Also, in looking at the conceptual relationship of the three in this schema, one can easily understand why Armer and Schnaiberg (1971) found a strong negative correlation between alienation and autonomy (or "modernity") in an analysis of modernity scales which were designed to measure position on the modern—traditional (or autonomy-heterOnomy) continuum. Autonomy enables the actor to consider choosing among various social systems before making behavioral commitments. The autonomous actor perceives the possibility of participating in a variety of social systems for material gains and psychic rewards. (The heteronomous actor is able to consider participation in only one social system; while the alienated actor can conceive of no rewarding participation in any available system.) our definition of autonomy is not accidentally similar to several concepts found in theories of individual modernization. Kumata and Waisanen (1969:52) mention the individual's perception of "freedom of decision making." Inasmuch as alternate solutions to problems are often tied to different social systems, and each social system usually prescribes a preferred solution, the perception of the ability to make decisions indicates relative independence from any one, and inter- dependence with several social system(s). It is also apparent that in speaking of autonomy in this fashion, I am speaking of the perception of "intersystemic participation," or the perception of the opportunity for it (Waisanen, 1969:8; Waisanen and Kumata, 1969:5,6) . It should be noted that not all behavioral scientists regard autonomy as a characteristic particular to modernity. As an example, Cantril (1965:318) says, "Human beings have the capacity to make choices and the desire to exercise this capacity," and that this characteristic is part of "a genetically built in design that sooner or later must be accommodated" (page 315). Rational Decision Making The second characteristic of individual modernity is the ability and tendency of the actor to make rational decisions. The concept of rationality is problematic, with strong philosophical overtones and is susceptible to culture and value biases. (See Ginsberg (1965) for a philosophical discussion of rationality and social development.) To minimize to avoid these problems, I will be defining the concept in the most minimal and generic sense. A rational decision is one made only after the actor has pur- posively considered the implications of at least two alternate solu- tions to the problem. In making a rational decision, the actor attempts to collect information about possible solutions, critically evaluates the sources of this information, and informally assigns probabilities to possible outcomes, and estimates the values of alternate rewards and costs. The rational decision is made after this process. 10 This definition follows Simon's discussion of rational choice (1957:241-60). He notes that among the "givens" or constraints common to all rational decisions are: "(1) a set of alternatives open to choice, (2) the relationships that determine the pay-offs ('satis— factions,‘ 'goal attainment') as a function of the alternative that is chosen, and (3) the preference-orderings among pay-offs" (page 242). A thorough theoretical and empirical analysis of "the preference- orderings among pay—offs" is highly problematic within the boundaries of this project. I will therefore make the assumption that "purposive consideration of two or more alternative solutions" subsumes the constraints posited by Simon. (It should be noted that this definition of rational decision making does not address the question of the conservative or "progres- sive" nature of any specific solutions nor the "correctness" of any solution as evaluated by a value position or by the eventual result. However, it can and will be demonstrated that the actor who is pre- disposed to making rational decisions is also generally predisposed to new and innovative solutions. Thus, the frequent selection of new and innovative solutions is a derived characteristic of rational decision making--not a defining characteristic.) Several scholars have noted the continuation and, more interest- ingly, the revival of "traditional" traits among segments of societies which would appear to have become "modern." The continuation of some traditional traits is to be expected in any "modern" society. Modernization, while a force of the greatest magnitude, does not affect all segments of social life equally. Gusfield (1967) provides an 11 excellent statement on the nature of the complexities and segmentally differing nature of the play between tradition and modernity. Hoselitz (1961) makes the useful distinction between "traditional" norms and "traditionalistie" ideologies, the latter being a result of, though inimical to, economic development. The revival or revitalization of traditional traits, or "neotraditionalization" (Rogers, 1969:16-17) should not be seen as a denial of, or countervailing force to, modernization. Or, it can be argued that the success of modernization in some institutions (e.g., economic, politico-administrative) makes possible the renaissance of interest and activity in traditional traits in other institutions (e.g., religion, art). And the selection of traditional alternatives may well be made as a result of a rational decision. The Tie Between Autonomy and Rational Decision Making The two components of individual modernity (individual autonomy and rational decision making) are empirically interrelated, and each is logically necessary for the other to occur. If the actor is heteronomous, the opportunity for making individual decisions will not be perceived, and any problem requiring a decision will be referred to his social system for a solution (usually a traditional solution). The ideal-typical alienated person, being completely independent from all.socia1 systems, is unable to arrive at solutions which would require ‘participation (and interdependence) with any social system. In opposite direction, the actor who appears to have considerable intmumsystemic ties and thus would be thought to perceive alternate 12 Solutions, but who consistently makes decisions solely on the basis of the dictates of one social system is not truly autonomous. And the person who is unable to make decisions requiring participation in any social system is more correctly seen as alienated rather than autonomous. The Tie as Seen in Waisanen and Kumata's Scheme The interrelationship between autonomy and rational decision making is further explicated in inspecting the conceptual specification of individual modernity as presented by Waisanen and Kumata (1969:4-5). They suggest four behavioral-attitudinal modes: (1) information seeking, (2) planning and investment, (3) inter-systemic participation, and (4) creativity and innovativeness (page 4). Of these, "inter- systemic participation" is most similar to the core meaning of autonomy. "Information seeking," as the search will generally involve a variety of social systemic sources, is also exemplary of autonomy. And the search for information regarding the existence and evaluation of alternatives is also a crucial aspect of rational decision making. "Planning and investment" is the most similar of these four to the concept of rational decision making. Planning and investment are behavioral commitments to the value of rationality. The selection of alternatives for future action (planning) and the commitment of scarce resources to these selected alternatives (investment) are indicative of rational decision making, and to the extent that these will generally involve several social systems, autonomy. 13 ."Creativity and innovativeness" are examples of high degrees of autonomy and rational decision making. They indicate that the actor is willing to carry out the search for information with his own resources (experimentation) and attempt to develop previously unknown alternatives (creativity). These activities require both a commitment to rational decision making and the autonomy to inspect alternatives not prescribed (or even those which are proscribed) by particular, relevant social systems. waisanen and Kumata suggest these four modes are in a develop- mental sequence (in the order listed above). My model would suggest that autonomy is antecedent to rational decision making. The apparent discrepancies in sequence are explained by their inclusion of behavioral modes, while I offer behavioral aspects as a separate component, related to both. Thus, by utilizing the feedback loop from behavioral aspects in my model, either model can address the same phenomena with little quarrel (see Figure 2, page 33). However, the Waisanen and Kumata scheme is much more parsimonious in this conceptual area. This Definition Compared with Inkeles' Listinngefinition It is appropriate to compare these attitudinal components (i.e., autonomy and rational decision making) of individual modernity with previous definitions to note their similarities and differences. Inkeles (1966:141-44) discusses nine themes characteristic of ixuiividual modernity. Each of these themes can be generally subsumed in the two components of individual modernity presented in my definition. (As Inkeles' listing definition is very typical of listing 14 definitions, and is also one of the most extensive and adequate, a comparison with my definition is appropriate. As the concept of individual modernity is derived from, and generally thought of, in the terms presented by this and other listing definitions, it is essential that I demonstrate the consistency between those definitions and my more abstract definition.) Inkeles notes a "readiness for new experience and . . . openness to innovation and change" (page 141). This orientation can be subsumed in the process of rational decision making, and also necessitates a high degree of autonomy. This orientation is antithetical to the automatic reference to traditional solutions for problem solving, characteristic of traditionality. The central characteristic of rational decision making is the purposive consideration of at least two alternatives. Central to this characteristic is the positive orientation toward innovation, change, and new elements. While our definition stressed that any particular decision need not arrive at a "progressive" decision, we assert that in order for a person to make rational decisions, he must inspect alternate solutions. In most cases, some of these alternatives will be relatively new elements, and an awareness of these newer elements ‘will facilitate rationality. As people tend to be most aware of those things they are attracted to, a positive disposition toward innovative or "progressive" alternatives facilitates an increased number of aawtilable alternatives. To summarize, the individual who makes Tuitional decisions must be aware of several solutions (some of which 15 will be new);he must have a positive disposition toward new things. Secondly, Inkeles notes: "a disposition to form and hold opinions over a large number of problems and issues" (page 141). This theme can be derived from the nature of autonomy, as by virtue of relative independence from one system, and contact with other systems, the actor can receive alternate opinions from other systems and receive support for these opinions. Inkeles mentions a third theme regarding the modern individual's awareness of time and his tendency to be oriented toward the future and present rather than the past (page 142-43). This cognition of time as a scarce commodity and a concern for future states is an essential component of rational decision making. Implicit in the nature of decision making is a strong concern about the future and an awareness of time. One can only make decisions which have consequences in the future. The actor who avoids decisions (i.e., utilizes only traditional solutions) demonstrates either a lack of concern about the future or more likely a strong belief that fate alone decides the future. And a general belief in fatalism is usually accompanied by apprehension of the future, rather than anticipation of the future. Inkeles says that the modern man is "oriented toward and is involved in planning and organizing and believes in it as a way of handling life" (page 143). Planning or the inspection of alternate suilutions is the central aspect of rational decision making. In the fifth theme, Inkeles speaks of the attitude of modern man toward his physical environment. He states that modern man believes 16 that he can learn "to dominate his environment in order to advance his own purposes and goals, rather than being dominated entirely by that environment" (page 143). Again we see the close tie between the autonomy to hold this position and the inspection of alternatives (one of which may be the malipulation of the physical environment) of rational decision making. (If the nascent environmentalism.movement is successful, this "modern" attitude may come to be regarded as anachronistic. This illustration of possible changing criteria points out one of the flaws of the listing approach.) Inkeles says that modern man "has more confidence that his world is calculable, that other peOple and institutions around him can be relied upon to fulfill or meet their obligations and responsibilities" (page 144). The theme of calculation of outcomes is central to rational decision making, and the theme of trust in impersonal others is a product of autonomy and experience in a variety of systems. Inkeles also speaks of a lack of fatalism in this context (page 144) which, of course, is central to the premise of decision making. "The modern man has more faith in science and technology, even if in a fairly primitive wax" states Inkeles (page 141). The elementary belief in a general scientific approach to problem solving shows a desire to make decisions rationally. Technology also provides an increased array of alternatives to choose among. Inkeles notes an increased "awareness of the dignity of others" auui the greater perceived importance of the individual in many :situations (page 144). This value is closely tied to autonomy, as the acixxr,perceiving the independence and interdependence of himself as an 17 individual actor,projects this perception to other individuals. This increased tolerance of others is also tied to rational decision making, as the modern man, with a positive awareness of, and curiosity about new things, sees other people as potential sources of help and information. The more traditional man, tending to avoid new or alternate means and goals, will likely regard others, particularly unknown others, as a source of threat or competition. However, this aspect is most correctly seen as an openness to rational decision making, and thus, most closely aligned with autonomy. The modern man also "believes that rewards should be according to contributions" or believes in distributive justice" (page 144). This value orientation derives from the process and experience in assigning probabilities to rewards in the rational decision making process. And, the actor's autonomy permits him to reject those social systems and reward systems which violate his expectations generated in rational decision making. Inkeles' definition can thus be adequately subsumed in the two major themes of autonomy and rational decision making. And as Inkeles' definition is one of the most adequate and extensive of the listing definitional approaches, my definition should be an appropriate and parsimonious substitution for the listing definitions of individual modernity . 18 This Definition Compared to the Relational Definition The relational definition of individual modernity as those attitudes, values, and actions which enable effective action in a modern social system is useful in the issues which are raised by this definition. The term "modern" and its derivatives "modernity" and "moderni- zation" are mischievous in that they carry several unfortunate implications. They carry the idea that conditions of the present time are "modern," those of the past "not modern." Surely what was "modern" in the 1890's is far different from what is "modern" today. The referent of a "modern social system" is based on the time of one's writing as well as the nature of that social system. By forcing a present orientation on a phenomenon, "modern" makes it difficult to see similar patterns and characteristics in past ("pre—modern") and possible future ("post—modern") phenomena. The term also carries a Western-industrial bias. Since present Western-industrial societies have passed through conditions presently existing in the "non—modern" countries, there is a strong, but often unstated, assumption that when countries become "modern" they will be industrial, urban, secular, nations. There is also a strong technolog- ical emphasis in the idea of "modernity." (Note Inkeles' inclusion of punctuality, science and technology, and domination of the environment.) While this association of "modern" with advanced technology is empirically justified, I feel it is unnecessary, and even misleading, to associate them in its conceptualization and definition. 19 Individual Mbdernity in Pre-Modern Systems If we define individual modernity in terms of autonomy and rational decision making, it becomes possible to inspect this social psychological orientation in a variety of social, technological, and cultural settings. Autonomy and rational decision making Should result in behavior typified by both risk-taking and investment. (By risk-taking I do not mean a superstitious, fatalistic, or compulsive gambling, nor by investment do I mean a mere hoarding of existing resources. Nor do I wish to include in this conceptualization of risk-taking, risks taken solely for the immediate thrill received.) Behavioral modernity is typified by the calculated risking of time, material, and energy resources (investment) for some anticipated greater return in the future. The idea of some deferral of gratification is central to investment and risk—taking. Viewed in this light, one can begin to notice individual behavioral modernity in a variety of "pre-modern" societies. Early traders, merchants, and money lenders must be seen as behaviorally Inodern in their economic activities. EXplorers and military leaders attempting to discover, establish trade routes and relations, or (conquer and exploit, must be seen as risking much to receive rewards. (halonists, settlers, and traders are clearly members of several social symstems (autonomy) and facing new situations must purposively consider snaveral solutions to problems (rational decision making). (As one example, see Mander's analysis of the Spanish conquistador in the Americas, 1969:83-9L) 20 These examples from pre-industrial societies also serve to emphasize the possible limitation of modern behavior to selected institutional segments of a society. It is not unreasonable to suggest that individual behavioral modernity is most likely found first in trade and other economic activities, both in internal (e.g., engineering, public works such as irrigation, and distribution systems) and in external areas (e.g., inter-societal trade, economi- cally motivated conquest). Individual modernity is least likely to appear early in family and religious systems. Individual Medernity and the Nature of the Social System In viewing modernity as rational risk-taking rather than mere technological competence, we can begin to specify those social and ecological systemic conditions which facilitate risk-taking and modernity. A society surviving at or near a subsistence level (e.g., hunting and gathering societies, horticultural societies (Lenski, 1970:123—25)) can ill afford to take risks or to reward or condone those who do. Having no storable surplus for subsistence, the cost of failure is so absolute that it proscribes risk-taking. In addition, time, resources, and energy apart from immediate subsistence concerns are minimal, and there is little to risk or invest. With the creation of a storable surplus (surplus beyond the sub- sistence level) and the coordination of its collection, storage, and distribution (Sanders and Price, 1968), time, energy, and resources are available for investment in limited risks, and the possible failure is 21 not crucial to the social system. Trade, exploration, loans for interest are concurrent with an increase in technology and control of resources. Thus socially rewarded risk-taking increases with the amount and surety of its economic surplus. If a society is technologically advanced and sure of access to sufficient raw material, risk-taking and investment is likely to be relatively commonplace. To the extent its resources are limited and its technology is simpler, risk-taking will be more proscribed,and less modernity (qualitatively and quantitatively) will be observed. (Other variables must, of course, be considered. The society's system of stratification is one additional pervasive influence. Lenski (1970:248—50) following Childe (1951) in speaking of agrarian societies notes that the surplus was entirely controlled by the ruling classes, who lacked expertise and motivation for innovation or risk-taking. The peasant class was denied opportunity for reward for risk-taking or innovation, and thus, despite adequate surpluses, little modernity was observed.) Thus, the large, predictable economic surpluses of Western industrial societies facilitate (and reward) widespread modern behavior, and as a result, Western industrial nations have the greatest amount of individual modernity, both as measured in the proportion of their people who are modern and in its prevalence in many institutional areas. This above discussion has viewed individual modernity as a result of technological advances making a surplus available. It is equally 22 correct to look at individual modernity and its resulting innovativeness, experimentation, and risk-taking as producing the advances in technology, trade, and resources which increase the available predictable surplus. For this reason, governments of developing nations are generally anxious to increase the level and breadth of individual modernity in their people. It is interesting to look briefly at the structures of a social system which supports, regulates, and rewards risk-taking (See Moore, 1961). At the higher levels of technology, societies contain a variety of organizationswhich have the specific goal of innovation and experimentation. Research institutes and foundations, and research and development centers are common in the technologically advanced nations, rare to non-existent in undeveloped nations. The growth of laws regulating risk-taking are noticed in social systems as they increase their level of technological development. Norms which make investment legal (and less risky) (through corporations, laws for collection of debts, and the resolution of bankruptcy) encourage risk—taking in economic areas. Finally, rewards for risk-taking in technologically advanced nations are not presented solely in economic terms but in social honor as well. The degree to which achievement (vs. ascription) dictatesone's standing in the stratification system is indicative of the social rewards for risk-taking and investment. The obvious comparison of the social prestige of the Medieval money lender and a present day banker illustrates this relationship (Hoselitz, 1960:61-66). 23 An Excursus on the Future of Modernity as Seen in the Youth Counter- Culture Much has been written recently regarding the middle class adoles— cent and young adult bohemian reaction to the technologically advanced society of the present United States. In reaction to (as the counter— culture sees it) an over-rational, materialistic, impersonal culture, the new orientation emphasizes non—rationality, sensuality, and immediate gratification. This "counter-culture" or "the hippies" (Rozak, 1969; Berger, 1971; and others) can be viewed (certainly so in terms of their own rhetoric) as an anti-modern movement. While alienated from the institutions and organizations of the dominant culture, they attempt to participate solely within the systems of the sub-culture. Homogeneous youth ghettoes and attempted communal subsistence farms intensify the heteronomy of its participants. The additional rejection of rationality and the substitution of intuition, sub-cultural dogma, astrology, emotional religions, etc. in its place further add to its non-modern nature. It may be seen as an "escape from modernity." Some writers (for example, Reich, 1970) have argued that the counter—culture with its anti—materialism and non—rationality is the harbinger of a new society-wide culture which will be made in the image of the hippies. If this is likely, the concept of modernity would indeed become archaic and irrelevant for any future study. However, I feel this prediction is simplistic and fails to consider all elements of the counter-culture. I would present the following speculations. I do not see the counter—culture as uniform 24 and monolithic but see at least three trends at the present which are related to the counter-culture and which will likely continue. First, many of the artifacts and activities of the counter-culture will be ad0pted (through the commercial market-place) by broad segments of the society as leisure time items. Many of these will be mere fads or fashions, but some will be more lasting. Middle class "weekend hippies" will continue to play an important role within bureaucratic and technological systems, with no reduction but perhaps an increase in their rationality and autonomy. Secondly, some segments of the counter-culture will continue in self-imposed isolation, inhibited from participation in the larger society by deviant subcultural values ("anti—modern"), a lack of marketable skills, and discrimination from the dominant culture. It is highly unlikely that this style of life can become dominant, as it requires either a wholly non-productive role in the economic institu- tion, or because the majority of the pOpulation would not (and could not) survive on the production style (subsistence farming and handicrafts) proposed by this group. It is essentially a millenarian— style movement, although it may continue for many years as a collection of deviant cults. A third trend will likely have the largest innovative role in the future. This collectivity, presently best seen in the editors and contributors to the Whole Earth Catalog (Brand, 1971), is attempting an innovative combination of technology and counter cultural values. This group might well be considered the "most" modern, in that their autonomy permits participation in, and selection of ideas from, many 25 social systems. Their wide-ranging and critical inspection of goals and values as well as means shows an essential rationality. Their central concern with environmental and ecological concerns frequently Showsan eco-systems awareness of considerable rationality. This movement is concerned with attaining humanitarian and individualistic goals but considers what rational and appropriate new and old knowledge and technology can be used to reach these goals. As this autonomy and rationality is identhxfl.to that presented in the new definition of modernity used here, this definition should serve as a viable definition for a variety of cultural situations. Further, if this third trend proves to have a significant influence on future society, modernity, as now presented in some listing definitions, will be archaic. Technological and work aspects of future society will probably be de-emphasized (though not abolished). However, this new definition of modernity, centering on rationality and autonomy, should continue to be a viable concept for identifying and studying a class of phenomena. This Definition Compared to Relativistic Definitions Stephenson's relativistic definition (1968:268; also page 4, this paper) essentially denies the possibility of universally applicable concepts, and as such inhibits the comparison of results from different studies (see Schnaiberg, 1970:402-03). However, there is nothing in the operational definition of modernity presented in his study of Kentuckians which cannot be subsumed under the more abstract and universal dimensions of this new definition. 26 Behavioral Aspects of Individual Modernity, The attitudinal dimensions of individual modernity have been discussed above in comparison of Inkeles' definition with mine. It is now appropriate to mention some of the more concrete behavioral aspects of individual modernity. As an expected behavioral result of rational decision making, we observe a tendency to experiment purposively, to test new things and methods, and to actively search for other methods of achieving goals. We also observe the tendency to make investments in time, energy, and materials for anticipated return in the future. We observe a greater punctuality in time related behavior and the treatment of time as a Scarce commodity. And we note an approximation to the scientific method for gathering and evaluating information. As a result of the belief in distributive justice, and the related idea that rewards are distributed on the basis of actual achievement rather than on the basis of ascribed characteristics, the modern man believes in the possibility of individual achievement. Thus we see not only evidence of increased aspiration but behavior which has the intended end of returning greater rewards to the actor. The modern man invests time, energy, and goods for a return, or in other words, tries to achieve. One writer, McClelland (1961), sees "achievement" as the crucial, central, and defining characteristic of modernity. The modern man's perception of the greater value and importance (If the individual leads not only to attempts to increase his self- lesteem.by individual achievement and participation in a variety of social systems, but also creates a belief in the value of individuals 27 per se. This leads to a greater tolerance of others and results in an increased amount of egalitarian behavior. The belief in man's ability to affect change and the willingness to form and hold opinions leads to greater participation in pressure and interest groups. These would include labor unions, political parties, business organizations, and other voluntary organizations. The belief in the calculability of life and trust in impersonal others has observable consequences in the modern man's ability to participate effectively in bureaucratic organizations as a client, and in consumptive participation in the mass media to gain information as well as entertainment. CHAPTER II A PATH MODEL OF INDIVIDUAL MODERNITY Causes of Modernity Having arrived at a definition of individual modernity which appears applicable to a wide historical and cultural range of related social psychological orientations, the question of the origin of this orientation remains to be answered. This question may be approached at two levels: the societal, to explain the degree of aggregate psychological modernity in a population; or individually, to trace the factors in an actor's life leading to his own degree of individual modernity. McClelland (1961) in looking for the causes of an aggregate increase in achievement motivation (a concept closely tied to individual modernity) was content to identify the cause as an increase in the achievement imagery of the teachings of religion, stories, and parents. The most productive path toward explaining societal modernity would appear to be the tracing of the inter—relationships between ecological factors, economic surplus, and socially taught, supported, and advocated risk-taking. The conclusions of such an investigation should agree with the relationships presented above in the discussion (If the tie between individual modernity and the nature of the social 28 29 system (see pages 20-22). However, such an investigation is beyond the scope of this research. It is the purpose of this paper to identify those conditions affecting different individual orientations toward modernity, particu- larly as they exist in the Western world at present. A most salient variable, both empirically and theoretically, is formal education. Education and Modernity Many researchers have observed and noted this distinct configura- tion of values, attitudes, and behaviors that are associated with successful participation in modern and modernizing social systems (see Anderson and Bowman, 1965). The cause of this configuration (variously labeled "modernity," "modernism," or as a process, "modernization") has been attributed to a variety of factors. McClelland (1961) would attribute the cause of modernity to the acquisition of the "need to achieve." In an earlier paper, Inkeles (1960) found that employment in industry was the factor common to those ranking high in those attitudes and values. Later, Smith and Inkeles (1966) describe modernity as the product of (and requisite for) participation in modern systems. Lerner (1958) suggests that a combination of mass media participation, urban location, and literacy lead to individual modernity. However, most studies of individual modernity arrive at the conclusion that a most powerful predictor of individual modernity is success in formal education. Inkeles states that education is the pre-eminent force for modernity (1966:146). Elsewhere (1966) he notes that "Formal education 30 is clearly the most consistently powerful influence in Table 3," (a table of correlations relating education, mass media usage, urbanism, work experience, factory environment, standard of living, and other variables to "active citizenship" (page 1132)). Almond and verba note: "Among the demographic variables usually investigated--sex, place of residence, occupation, income, age, and so on--none compares with the educational variable in the extent to which it seems to determine political attitudes" (1963:379). Harbison and Myers (1964) speak of the crucial nature of education for economic development. Prediction and co-variation are, of course, not sufficient to establish causality. Formal education is often used as a control variable by those who wish to study the effects of other variables on individual modernity. As a result, formal education is often the recipient of adequate analytical but minimal theoretical attention. And when education is considered as a theoretically relevant variable, immediate causality is often attributed to it. Inkeles (1966:146-47) states that allowing for social class differences in attendance and the traditionality of some schools, We may still say that education, especially in the schools emphasizing the more modern type of curriculum, seems to be the most powerful factor in developing a population more modern in its attitudes and values. This effect depends in part on the direct instruction provided, but we assume as well that the school as a social organiza- tion serves as a model of rationality, of the importance of technical competence, of the rule of objective standards of performance, and of the principle of distributive justice reflected in the grading system. To these, I can add several other important functions of the formal cubication system as it contributes to individual modernity. Partici- Inrtion in school teaches an awareness of and respect for time, and 31 rewards punctuality. It provides experience with impersonal authority and rules. It also presents semi-modern or modern role models for emulation in form of older students, teachers, and administrators. While these observations on the significant contributions of formal education to individual modernity are all generally valid, I find that looking at education as the direct and immediate cause of individual modernity has serious shortcomings. waisanen (1971) and Waisanen and Kumata (1969) offer similar observations on the theoreti- cal and empirical ties between formal education and modernity. If education is seen as the immediate cause of individual modernity, it would suggest that we disregard the large number of very important experiences in adult life which contribute greatly to modernity. This view would imply an "educational determinism" which is unjustified on theoretical and empirical grounds. (See Waisanen's (1971) caveats regarding the conditions influencing the effectiveness of education in stimulating the growth of modernity.) Secondly, it allows no consideration of the nature of the social environment and the manner in which this environment may inhibit or stimulate the utilization of the educational experience. Thirdly, it implies an excellence of educational experience that is rarely encountered. jEducation and Modernity--A Path Model In an attempt to explain the profound effects of education on irulividual modernity (but avoiding the deficiencies inherent in zittributing immediate causality to education), I have developed a 32 processual model which considers intervening social and psychological factors. (See Figure 2, page 33) The model states that success in formal education (positively related to antecedent conditions) contributes to the individual's power. Education also contributes to physical and psychic mobility, which in turn are positively related to the actor's power. The actor's perception of his power, which is self-perceived efficacy, is positively related to the components of individual modernity. And behavioral modernity, assuming a modern social system, increases the individual's power. Hypotheses and Analysis While the test hypotheses and proposed analytic techniques will be more fully discussed in a later section (see Chapter III), it is appropriate to make a general statement at this time. My major concern is in ascertaining the degree of empirical support for the model as presented. In this sense, the model itself can be regarded as the major hypothesis or a collection of hypotheses. The model will be tested independently with data from each of five national sample surveys, from Costa Rica, Finland, Japan, Mexico, and the United States. I hypothesize that the analysis will show that: I There are no negative relationships between components in the model; and II The degree of association between components is strongest when they are theoretically adjacent, and weakest when they are theoretically distant. 33 ZOHBHQZH ho mmfloomm flue mo ammo: Hue mo ZOHesezmmfimmHm oHeHQZH m mmwam WQHmumflm Imdmm > \ J NBHZEWQQE AHQZH VBH IQHmQE UHEUWmm / wagon a? NEH IAHmcS A B----) c----> D----) E Each of the components will be positively related to all other components. ("R" indicates "relationship" with a numerical value between +1 and -1.) 56 Rab),0 Rac7 0’ Rbc7 O Rad7 0’ Rbd7 0’ Rcd>° Rae7 0’ l:Lbe7 0’ Ree? 0’ Ede? 0 The above matrix corresponds to Hypothesis I. Given: The assumptions of the causal ordering, it is reasonable to assert that the degree of association is dependent upon the proximity of the components. Thus: Rab V Rac‘Rbc V V Rad< Rbd < Rcd V ‘V \l R ae< Rbe < Rce < Rde The above matrix corresponds to Hypothesis II. By observation one notes that the strongest relationships occur along the diagonal of the matrix, and the weakest relationships are found at the extreme. This array is identical in structure to Guttman's "perfect simplex" (1954: 271) . The Test of the Medel of Modernity It is convenient at this point to place the path diagram of the model of modernity as presented on page 33 in a 10 by 10 matrix (see Figure 3, page 57). In this format, the arrows representing causal relationships remain, and the theoretical proximity of components is indicated. Those relationships between theoretically adjacent components are labeled "RI"; those of intermediate proximity (with one component 57 Antecedent Components Pers. Fam. Phys. Psych. S.-P. Rat. Behav. ChanlChar. Educ'LMob J Mob. Jpowe1 EfficIAuton D. M.l Mcd.j Pers. Char. Fam.; [ R Char E ‘ 3 Educ RI L r T Phys A Mob R11 R11 N I T Psych Mob. R11 R11 R11 3 f f ‘ ‘ lfipower RII RII HI ”I RI ‘ OS—P. gEffic RIII HIII R11 RII RII RI N N gAut‘m RIII RIII RIII RIII RIII RII RI / ‘ fifth). RIII RIII R111 R111 HIII RII I \ \ 3:28“ BIII RIII RII RII RII RI / RII flI FIGURES THE PROCESS OF INDIVIDUAL MODEBNIZATION: CONVERSION OF SCHFMATIC PATH MODEL (FIGURE 2) TO MATRIX FORMAT Arrows indicate causal paths, as represented in Table 1. RI indicates a relationship between theoretically adjacent components. indicates a relationship between components with intermediate theoretical proximity (i.e., one intervening component). RIII indicates a relationship between theoretically distant components (i.e., two or more intervening components). 58 intervening) are labeled "R11"; and those which are theoretically distant (two or more intervening components) are labeled "R111." The array in this case is not exactly similar in appearance to Guttman's "perfect simplex" (1954:271) as there are several instances of multiple causality and non-recursive relationships. Drawing from this matrix (Figure 3), the two test hypotheses, and the ideal model, we can state: All RI are positive, all RII are positive, and all RIII are positive, and: (the average of RI):> (the average of RII):> (the average of R111). These two statements are the test hypotheses stated in alternate form. Determining the Relationships Let us return to the hypothetical ideal causal model and assume that: Each of the components "A" through "E" is composed of a number of indicators, such that: A is composed of a1, a2, . . . an 2, . . . bn and similarly through E. B is composed of b1, b In order to determine the relationship between components, it is necessary to examine the relationships between their respective indicators. Thus the relationship between A and B (RAB) is a function of the combined relationships of their indicators (r). R.AB 1s a function of: I‘ +I‘ +ooo+r + a1b1 a1b2 albn 59 As there will be a different number of indicator relationships comprising componential relationships and the componential relationships must be compared, it is necessary to determine the arithmetic mean of the indicator relationships in each componential relationship. Thus: the sum of the indicator relationships r through r a1b1 anbn R.AB = the product of nA and 11B (where nA is the number of indicators in A and where 11B is the number of Indicators in B) (It is mathematically legitimate to average statistics of relation- ship when each of them, as in this case, is based on an equal number of cases (Edwards, 1950:132).) The Search for a Statistic Having established that the test of Hypothesis II is the varying strengths of relationships between indicators, it was necessary to select a statistical measure of association. As the majority of the indicators measure at the ordinal and nominal level, the contingency coefficient (C), computed from the Chi- square statistic, would appear appropriate. However, the contingency coefficients drawn from tables of different sizes are 233 directly comparable (Siegel, 1956:201). Thus, one would be forced to collapse all indicators to the spread of the smallest (i.e., two categories only). This is particularly costly as many of the ordinal indicators 60 (ladder and attitude items) are assumed by many to measure at the interval level. To avoid the loss of discrimination unavoidable in collapsing all indicators to dichotomies, I divided the indicators into two groups—- those which are clearly nominal and ordinal measures (the "N" group) and those which measure at the ratio, interval, or assumed interval level (the "I" group). I determined the Pearson product moment correlation coefficient (a) for each of the relationships between ratio, interval, and assumed interval indicators. (While the use of this statistic also requires the assumption of normal distributions, the required assumptions are fewer and less stringent than those required for multiple regression analysis (Hiese, 1968:57). The value of the Pearsonian g is reduced by curvilinearity, as are all other measures of association. The curvilinearity of relationships will reduce the derived Pearsonian r in several cases, and the correlation coefficient, based on the assump- tion of a straight line of regression, will be lower than it would be if the variable(s) were transformed to reduce the curvilinearity.) The coefficients of multiple indicators of components were then averaged. I computed the contingency coefficient (Q) for the nominal and ordinal indicators, collapsing all indicators, if necessary, to dichotomies. The coefficients of multiple indicators of components were then averaged. 61 Internal Consistency of Multiple Indicators As I selected the indicators on the basis of apparent relevancy to the components and not on the basis of prior knowledge of their relationships with other co-indicators, it was necessary to check their consistency with other indicators of the same component. Thus, prior to the construction of the test matrices I inspected the relationships between co-indicators. Numerous indicators which have been selected were observed to show consistently negative or very low degrees of association with the other indicators of the same component, and these indicators were dropped from analysis. Judgment of which collection to delete from the major analysis was determined on their relevance to the definitions and relationships of the components as discussed theoretically in presenting the model. No indicators were deleted solely because of a weak relationship with indicators of other components. (Most of the indicators which were deleted from the major analysis were analyzed. The relationships they show and an interpretation of them will appear below in Chapter VI.) The statistics used in this test of internal consistency were the same ones used in the respective test matrices (i.e., product moment correlation coefficient for the ratio, interval, and assumed interval measures, and the contingency coefficient for ordinal and nominal measures). 62 Summary_Averaggs Upon completion of the test matrices and average measures of relationship for each componential relationship, I combined the averages to test Hypothesis II. The scores were combined as indicated in Figure 3 (page 57) into the three categories of theoretically adjacent inter- mediate theoretical proximity and theoretically distant. Procedures Used for Calculation Data for the tests were stored in tape form. The data tape used is from the Michigan State University Political Science Department's Data Bank. variables were selected and, because of an incompatibility of the data tape format and the Computer Center's computers, were transformed to data card form. Contingency and correlation coefficients were calculated using the ACT II program of the Michigan State University Computer Institute for Social Science Research on the CDC 6500 computer. These coefficients were then averaged to determine the internal consistency of multiple indicators. After inspection of multiple indicators and the selection of those indicators to be used, test matrices were constructed and the analysis of deleted indicators was performed. Levels of Statistical Significance While not central to the test of the model, statistical levels of significance were calculated. These levels appear below each table. Because of the large samples used, many small coefficients are statistically significant. However, my concern is not with specific 63 coefficients but with the general pattern they present. Thus the levels of statistical significance are merely included to support the assumption that the results obtained were not caused by chance alone but by relationships which exist in the social world. The patterning of relationships and a possible explanation of them are of major significance here and not the statistical levels of significance. For correlation coefficients, the levels of significance are derived from the one-tailedlgrtest based on Fisher's transformation of sample and population coefficients to standardized scores, assuming a random sample, a normal distribution of coefficients, and an actual correlation coefficient of .00 in the population (Haber and Runyon, 1969:195—97). For contingency coefficients, statistically significant values of the Chi-square distribution were selected, then these were transformed to the contingency coefficients appropriate to the specific sample. Because of a concern with accuracy (particularly regarding positive and negative coefficients), all test matrices and the analysis of deleted indicators were constructed a second time. Any discrepancies were re-checked and reconciled, thus giving a high level of confidence concerning the accuracy of the calculations. CHAPTER IV A TEST OF THE MODEL: INDICATORS AT ASSUMED INTERVAL LEVEL OR ABOVE Review of Method Presented here are the results of testing the model as presented above. The methods used are discussed above. This section consists of the tests of the model using those indicators which measure at the assumed interval (i.e., ordinal) level and above. The measure of association used is the Pearsonian product moment correlation coefficient (r). Where more than one correlation coeffi- cient measuring the relationship between components is present, the unweighted arithmetic mean (3) was calculated, and the mean correlation coefficient for the two components is shown. Order of Presentation--Level of National Development The tests are presented in descending order of the five nations' level of national development. The ranking of the countries comes from Farace's (1966) article. In it the United States ranks 1, Finland ranks 11, Japan 22, Mexico 42, and Costa Rica 43. Farace's ranking is a result of a factor analysis of 54 variables in the areas of government, health and nutrition, agriculture, education and religion, economic indicators, population characteristics, and mass communications (pp. 306-08). While other studies of national development might well arrive at different ranks, the relative ranking of the five countries studied 64 65 here would be unlikely to change significantly. (Because Of the moderate urban bias in the Mexican sample, results from this particular survey may represent a population more developed than results from a survey representative of Mexico as a whole.) United States Surveye-Interval Indicators The indicators used in the test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussion cf the theoretical components and their indicators (Chapter II, this paper). Also presented here are the mean correlation coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Indicators and Their Internal Consistency Internal Consistency of Multiple Indicators as Measured by the Mean Correlation Coefficient Components and Their Indicators Personal Characteristics Age Family Characteristics Parents' Education Parents' Occupation .180* Education Years of formal education Physical Mobility No indicators *The levels of statistical significance for correlation coeffi- cients for samples of this size (N = 1528) are: p 5. .10 = .03; p 5 .05 =2 .04; and p 5.01 = .06. 66 Internal Consistency of Multiple Components and Their Indicators (cont)Indicators as Measured by the Mean Correlation Coefficient Psychic Mobility Hours of television viewed Hours of radio heard Number of newspapers read .069* Number of magazines read Place of residence Power Annual family income Occupation .259 Self-Perceived Efficacy Ladder item--influence over people Ladder item--can do much to make his life happier Attitude items-little chance to get .120 ahead unless one knows the right people Attitude item-—the feeling that other people are using him Autonomy Self related to: political party community family .261 country his work Ladder item--chances to do anything he wants to Rational Decision Making Ladder item--1iving the best possible life, five years from now Ladder item--wants to do new things Attitude item--new methods of raising children should always be .105 tried Self-rating item--how "set in his ways" is he *The levels of statistical significance for correlation coeffi- cients for samples of this size (N = 1528) are: £5: .10 = .03; p {z .05 = .04; andpé .01 = .06. 67 Internal Consistency of Multiple Components and Their Indicators (cont) Indicators as Measured by the Mean Correlation Coefficient Behavioral MOdernity No indicators Comments on the Internal Consistency of Indicators: United States The coefficients measuring internal consistency are generally at an acceptable level, with the possible exception of the indicators of psychic mobility. In the United States, participation in the various mass media appears to be unrelated as measured by zero-order correlation. Individuals who are heavy and light consumers of all media are found with individuals who vary in their own use of different media, thus erasing a simple linear relationship. With the exception of a strong (.339) relationship between newspaper and magazine usage, the other factors, including place of residence, are unrelated in this test. This low internal consistency undoubtedly detracts from the absolute value of the psychic mobility cells in the test matrix, but without other means of analysis it is an unavoidable consequence. The Test Matrix and Summary Averages Using the above indicators, the test matrix was constructed. It appears on the following page, followed by the summary averages. Comments on the Test Matrix—~United States The results, as presented in the Test Matrix (Table 1), are generally consistent with the two hypotheses of the study. 68 TABLE 1 TEST MATRIX FOR UNITED STATES, USING ASSUMED mm MEASURES OR HIGHER RESULTADT COMPONENTS Personal Characteristics Pers. Family Char. Characteristics Pam. Education Char. -020 # # Physical Educ' “320 '273 Mobility ANTPDEDPNI COMPONENTS Phys. Psychic (read down) Mob. Mobility . :3?“- .056 .041 .190 \“Power ob. ‘7 Self-Perceived Power .193 .096 #15041- .133€\\ J,Efficacy So-Po Effie. .029@ .072 .155 .038 .113# lAutonomy Auton.- .022Q..004@A .020C‘1 1’ .013j1.013 .130 {1113:3323 Making Rat. 105@ 012d 062d 1) ? Behavioral D. M. . . . .029 .039 .097 .084 Modernity Behav. Med. (# indicates theoretically adjacent cells, @lindicates theoretically distant cells. SUMMARY’AVEBAGES: Theoretically Adjacent Relationshipsw) Theoretically Intermediate Relationships Theoretically Distant Reloti.;nships.(@) Remaining cells are at a theoretically intermediate distance. RANGES: 3231*! = .212 .450 to .084 73;.“ , .075 .193 to .013 Yflmo= .032 .155 to -.022 The levels of statistical si samples of this size (N = 1 p‘ .01 I: .06 nlficance for correlation coefficients for 28)arep§.10= .03,££.05= .04 9 69 All but two correlations are positive, as hypothesized, and these two negative correlations are very small in absolute value, indicating a null relationship rather than a negative relationship. Several others are very low in absolute value and, while positive, probably indicate a near null relationship. The presence of the null relation— ships is not counter to the hypothesis that no relationships would be negative. However, as the negative coefficients are in the clear minority and occur primarily between components which are theoretically distant, and therefore a low coefficient was expected, I see support for the first hypothesis. Put another way, there is far more evidence for accepting the hypothesis than rejecting it. While not every cell coefficient is within a discrete range for relationships of its type (adjacent, intermediata.0r distant theoretical distance), the summary averages and ranges stand in the hypothesized relationship to one another, thus supporting the second hypothesis. Overlapping ranges of the three types are expected, and for this reason the use of an average as a test was prOposed. (A discussion of the cell coefficients which appear consistently within the ranges of other types of relationships appear in an evaluation of the interval level test matrices of all five nations.) Finland Survey--Interval Indicators The indicators used in the test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussion of the theoretical components and their indicators (Chapter II of this paper). 70 Also presented here are the mean correlation coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Indicators and Their Internal Consistency Internal Consistency of Multiple Components and Their Indicators Indicators as Measured by the Mean Correlation Coefficient Personal Characteristics Age Family Characteristics Parents' education Parents' occupation .283* Size of place of childhood Education Years of formal education Physical Mobility No indicators Psychic Mobility Hours of television viewed Hours of radio heard Number of newspapers read .045 Number of magazines read Power Annual family income Occupation .484 Self-Perceived Efficacy Ladder item--inf1uence over people *The levels of statistical significance for correlation coeffi- cients for samples of this size (N = 893) are: ‘p é:.10 = .04; p g .05 = .061; and p g .01 = .08. 71 Ladder item--can do much to make his life happier .195* Attitude items-little chance to get ahead unless one knows the right people Autonomy Self related to: political party community family .211 country his work Ladder items-chances to do anything he wants to Rational Decision Making Ladder items-living the best possible life, five years from now .148 Self-rating item—-how "set in his ways" is he Behavioral Mbdernity N0 indicators *The levels of statistical significance for cerrelation coefficients for samples of this size (N = 893) are: 241.10 = .04; pf .05 = .061; and p 5 .01 = .08. Comments on the Internal Consistency of Indicators The coefficients measuring internal consistency are at acceptable levels, with the exception of the indicators of psychic mobility. It appears that the same interrelationships of mass media usage patterns occur in Finland as occur in the United States and again may account for relative low values of psychic mobility cells. The Test Matrix and Summary Averages Using the above indicators, the test matrix was constructed. It appears on the following page followed by the summary averages. 72 TABLE 2 TEST MATRIX FOR FINLAND, USING ASSUMED INTERVAL MEASURES OR HIGHER RESULTANT COMPONENTS Personal Characteristics Pers. Family Char. Characteristics Fan. .039 Educ t' Char. a ton Educ # Physical - . 171 . 338 Mobil i ty ANTEDEDIENT COMPONENTS Phys. Psychic (read down) Mob. Mobility P h. “3:: .021 .062 .120? lPower Self-Perceived Pm" .080 .247 .4494 .109# 11 Efficacy s.-P. Effie. .094 .107 .169 .080 .203 lAutonomy ‘ Rational Auton.v.o42fi .031q .015+ .030@H .039 .112#* lDecision Making Rat' 181 0888 1424> 072@l 138 221 073# Beham’“ D. M. ’ ' ' ' ° ' ' Modernity Behav. Mod. J (# indicates theOretically adjacent cells, (Q indicates theoretically distant cells. Remaining cells are at a theoretically intermediate distance. SUINARY AVERAGES : RANGE : Theoretically Adjacent # g BalatiQUShips(#) XRI .196 .449 to .073 Theoretically -- Intermc diate X81 Relationships I .110 .247 to .021 Theoretically @ Distant = RelatiODShipS(@) . XIIIII 0072 .181 to "0042 The levels of statistical significance for correlation coefficients for samples of this size (N .= 893 ) are pé .10 a: .04 , pé. .05 = .06 , 0‘ .01 .-= .08. 73 Comments on the Test Matrix--Finland All coefficients save one are positive. If the one negative coefficient is interpreted, because of its low absolute value, as measuring a null rather than a negative relationship, I see reasonable support for the first hypothesis. Summary averages and ranges are in the relative positions hypothesized for adjacent, intermediate, and distant relationships, and despite the expected problem of overlapping ranges I see a pattern which supports the second hypothesis. Japan Survay-—Interval Indicators The indicators used in the test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussion of the theoretical components and their indicators (Chapter II, this paper). Also presented here are the mean correlation coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Indicators and Their Internal Consistenpy Internal Consistency of Multiple Components and Their Indicators Indicators as Measured by the Mean Correlation Coefficient Personal Characteristics Age 74 Family Characteristics Parents‘ education Parents' occupation .356* Education Years of formal education Physical Mobility No indicators Psychic Mobility Hours of television viewed Hours of radio heard Number of newspapers read .103 Number of magazines read Place of residence Power Annual family income Occupation .344 Self-Perceived Efficacy Ladder item--influence over people Ladder item--can do much to make his life happier .381 Autonomy Self related to: political party community family .360 country his work Ladder item--chances to do anything he wants to Rational Decision Making Ladder items-living the best possible life, five years from now .092 *The levels of statistical significance for correlation coefficients for samples of this size (N = 990) are: pé .10 = .04; p 5 .05 = .05; and p e. .01 = .07. 75 Ladder item--wants to do new things Attitude item--new methods of raising children should always be tried Self-rating items—how "set in his ways" is he Behavioral Modernity No indicators The Test Matrix and Summary gyerages Using the above indicators, the test matrix was constructed. It appears on the following page, followed by the summary averages. Comments on the Test Matrix--Japan The Japanese survey and its results present the highest relation- ships in absolute values and the greatest absolute difference in summary averages. The range of coefficients in adjacent, intermediate, and distant relationships all overlap but are in the hypothesized relative position. The three negative coefficients and several low positive coefficients should best be interpreted as null relationships but are not frequent enough to cast serious doubt on the validity of the first hypothesis. Mexico Survey--Interval Indicators The indicators used in the test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussion of the theoretical components and their indicators (Chapter II, this paper). 76 TABLE 3 TEST MATRIX FOR JAPAN, USING ASSUMED INTHlVAL MEASURES‘OR HIGHER RESULTANT COMPONENTS 1 Personal Characteristics Pers. Family Char. Characteristics Fhm' 272 Eflucation Char. ° # Physical Eluc. .SSGA .379 Mobility ANTEBEDFNT ' COMPONENTS Phys. Psychic (read down) MOb. Mobility h. :3: .141 .157 .177 )Power .1 . S lf-P e'v d Power .138 .321 .3594 .198# \ hang? ‘ e Effie -.025 .039 .055 .040 .130# lAutonomy ‘ Rational Auton- —.040 .009_.017% .018@ .017 .268fi lDecision Making W" 1 > 7~ 166d O65@ 107 150 106# Behamml D. M. ' 54 '0 O ' ' ' ' ' Modernity Behav. Med. (# indicates theoretically adjacent cells, cells. SUMMARY.AVERAGES: ‘Theoretically Adjacent Relationships(#) Theoretically -— Intermtdiate XRII = Relationships Theoretically ._ Relationships(@) H ' The levels of statistical significance for corr samples of this size N = . a 04.01 g .07. ( 990 )arep§ 10 V= .272 .140 @ indicates theoretically distant Remaining cells are at a theoretically intermediate distance. RANGES: .556 to .106 .321 to .017 .166 to -.040 elation coefficients for 004 , .2"- .05 a: 005 i 77 Also presented here are the mean correlation coefficients between the multiple indicators of those components 80 measured, indicating the internal consistency of these multiple indicators. Indicators and Their Internal Consistenpy Internal Consistency of Multiple Components and Their Indicators Indicators as Measured by the Mean Correlation Coefficient Personal Characteristics Age Family Characteristics Parents' education Parents' occupation .209* Education Years of formal education Physical MObility No indicators Psychic MObility Hours of television viewed Hours of radio heard Number of newspapers read .353 Number of magazines read Place of residence Power Annual family income Occupation .131 Self-Perceived Efficacy Ladder items-influence over people *The levels of statistical significance for correlation coefficients for samples of this size (N = 1414) are: p e .10 = .03; B é .05 = 004; and R é 001 = 006. 78 Ladder item--can do much to make his life happier .329* Autonomy Self related to: political party community family .249 country his work Ladder item-~chances to do anything he wants to Rational Decision Making Ladder item--living the best possible life, five years from now Ladder item--wants to do new things .096 Attitude items-new'methods of raising children should always be tried Self-rating items-how "set in his ways" is he Behavioral Modernity No indicators *The levels of statistical significance for correlation coefficients for samples of this size (N = 1414) are: pg: .10 = .03; B 6. 005 = 004; and B é, 001 = 0060 The Test Matrix and Summapy Averagga Using the above indicators, the test matrix was constructed. It appears on the following page, followed by the summary averages. Comments on the Test Matrix--Mexico The positive relationships and the expected ranking of summary averages and overlapping ranges argue strongly for accepting both hypotheses. TABLE 4 79 TEST MATRIX FOR MEXICO, USING ASSUMED INTERVAL MEASURE OR HIGHER RESULTANT COMPONENTS 1 Personal Characterisnics Pers. [ Family Char. Characteristics Fm“. O77 Eflucation Char. ' Ed 259# # Physical ac. ° .345 Mobility ANTECEDENT COMPONENTS Phys. Psychic (read down) Mob. Mobility th- .127 .168 .372 1pc... Mob. # Self-Perceived Power .061 .195 .3894 .198 Efficacy SO-P. Effic. .062@ .115@ .186 .084 .158# lAutonomy Rational “W“ .072@ .066@l .123 .081@ .078 .199N1Decision Making Rat. B h ' l D, u, .086@ .045 .1159 .062@ .040 .203 .104 lujdggggt Behav. Med. (# indicates theoretically adjacent cells, cells. @ indicates theoretically distant Remaining cells are at a theoretically intermediate distance. SUMMARY AVERAGES: Theoretically Adjacent Relationships(#) Theoretically Intermediate Relationships Theoretiially Distant Relationships(@) xfll# g .253 x111 ‘ 7(“‘IIIE’” .083 .122 RANGES: .389 to .104 .203 to .040 .115 to .062 The levels of statis samples of t. 29.431 = .06. tical significance for iis size (N = 1414 ) are p6 corrElation coefficients for .10 g .03 ’ fig 005 - .04 , 80 Costa Rica Survey--Interval Indicators The indicators used in the test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussion of the theoretical components and their indicators (Chapter II, this paper). Also presented here are the mean correlation coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Indicators and Their Internal Consistency Internal Consistency of Multiple Components and Their Indicators Indicators as Measured by the Mean Correlation Coefficient Personal Characteristics Age Family Characteristics Parents' education Parents' occupation .480* Education Years of formal education Physical Mobility No indicators Psychic Mobility Hours of television viewed Hours of radio heard Number of newspapers read .222 Number of magazines read Place of residence *The levels of statistical significance for correlation coefficients for samples of this size (N = 1040) are: pg .10 = .04; pé.05 = .05; and pg .01 = .07. 81 Power Annual family income Occupation .459* Self-Perceived Efficacy Ladder items-influence over people Ladder items-can do much to make his life happier .169 Attitude items—little chance to get ahead unless one knows the right people Attitude items-the feeling thatother people are using him Autonomy Self related to: political party community family .274 country his work Ladder item—-chances to do anything he wants to Rational Decision Making Ladder itemr-living the best possible life, five years from now Ladder items-wants to do new things Attitude items-new methods of raising .105 children should always be tried Self-rating items-how "set in his ways" is he Behavioral Modernity No indicators *The levels of statistical significance for correlation coefficients for samples of this size (N = 1040) are: £155.10 = .04; B6— .05 = 005; and Bé 001 = 0070 82 The Test Matrix and Summary Averages Using the above indicators, the test matrix was constructed. It appears on the following page, followed by the summary averages. Comments on the Test Matrix--Costa Rica The positive relationships (though several would be better seen as null relationships) and the expected ordering of summary averages and ranges again argue for accepting both hypotheses. Comments on Interval Test Matrices of All Five Nations The observed patterns of relationships in all five studies support the two hypotheses and lend considerable support to this model of the process of individual modernity. Despite overlapping ranges and a number of null relationships, I see a convincing demonstration of support for the hypotheses. The lack of assumed interval level (or higher) indicators for physical mobility and behavioral modernity is unfortunate. However, extrapolating from the test matrices for nominal and ordinal indicators (which appear in the next chapter), it seems likely that they would have generally made the hypothesized contributions to the test of the model. Several of the cells in all tests consistently Show values which are not consistent with the appropriate theoretical distance between the respective components. In order to more easily inspect these deviations from the ideal causal model, which appear consistently (or perhaps vary with the level of national development), I will present a matrix consisting of the unweighted means of all cell coefficients (see 83 TABLE 5 TEST MATRIX FOR COSTA RICA, USINGASSUMED INTERVAL MEASURES OR HIGHER RESULTANT COMPONENTS Personal Characteristics Pers. Family Char. Characteristics Pam. 063 Education Char. ’ Physical WC- .207# .421# inability ANTECEDM ' COMPONENTS Phys. Psychic (read down) Mob. Mobility P h. "K? .016 .236 .278# (Power P # # Self-Perceived "er .040 .336 .479 .226 .. Efficacy S“P' 025d 1 102 168 Autonom Hfic. a o 29 .202 o o y @ @l # Rational Auton- .008@I .0829421 .085 .106 .124 11,90,810" Making Rat. B h - 1 D. u, .0448 .082?.157@ .109@i .109 .091 .085# 1501:3213 Behav. Mod. (# indicates theoretically adjacent cells, @lindicates theoretically distant Remaining cells are at a theoretically intermediate distance. cells. SUMMARY AVERAGES: Theoretically Adjacent‘ Relationships(#) Theoretically Intermediate Relationships Theoretically Distant Relationships(@) r = in“ .. .130 Yam" - .084 .248 RANGES: .479 to .085 .336 to .016 .202 to .008 The‘levels of statistical significance for correlation coefficients for samples of this size (N = 1040) are p 6 .10 :- 2.4.01- .07. .04 9 2.4- 005 a: .05 , 84 Table 6). (As this averaging is not wholly legitimate mathematically (see Edwards, 1950:132), it is not intended as a test of the hypotheses, and thus no summary averages or ranges are calculated. It is intended to serve only as a crude indicator of average patterns in all five matrices.) 0f the indicators used, I am least satisfied with the indicators of autonomy. The concern felt at the theoretical and indicator selection stage is supported by the weak relationships seen in the test matrices. However, the indicators of autonomy which were deleted (see Chapter VI) because of a lack of internal consistency show even weaker, more negative, and less consistent coefficients than the items retained. Thus, I am satisfied that the best available indicators of autonomy were used, although these are not as empirically dramatic as had been hoped. (It may be that the concept of autonomy, as I have defined it, is intrinsically difficult to operationally define. The tri-polar nature of the concepts and the empirical dependence on respondents' reporting of independence and involvement in social systems make it difficult to evaluate empirically. This is particularly the case as many such allegiances are prescribed by publicly endorsed values and norms of a culture, and respondents would be hesitant to express a lack of allegiance to family, community, or nation. While better indicators of autonomy may appear, or be developed, those used here are less than satisfactory. Thus, the evaluation of the utility of the concept of autonomy (as here defined) must be held until further instrumentation TABLE 6 MEANS OF CELL COEFFICIENTS OF ALL FIVE NATIONS USING ASSUMED INTERVAL MEmSURES 0R HIGHER RFSULTANT COMPONENTS Personal Characteristic-s Pers. Family Char. Characteristics Fan“ 094 am- +- Char. . . ca ion if Physical Educ ’ ‘ 303% ‘ 357 ‘Mobi 1i ty ANTEEFDENT COMPONIINTS Phys. Psychic (read down) nob. . Mobility P ' h. 3:1? .072 .133 .228# lPo-rer Power .102 .239 .4250 .173# lgfiif‘l’ermved icacy 5"“ 037 093 154 069 154# Eflfic. ° ' ' ' ° Autonomy Ant . _ b ca @ Rational on .005? .037 .002 .045 .051 .167 \ vDecision Making Hat. 0 @ . -. .114 .060@, Behdwioral D. M. 128 .067 e087 0153 0090 1M0derniby' Behav. Mod. (# indicates theoretically adjacent cells, @>indicates theoretically distant Remaining cells are at a theoretically intermediate distance. cells. 86 is attempted. However, the outlook for the concept's empirical utility must be seen as less than totally sanguine.) In the table of cell means of all five samples, the coefficient measuring the relationship between family characteristics and power (an intermediate distance relationship) is higher than many of the coefficients Of adjacent relationships. However, it is never greater than the coefficients for the relationships between family character- istics and education, and education and power (the causal path in the model). Thus it creates no need to revise the model. This relationship (family characteristics and power) can legitimately be seen as a measure of inter-generational mobility. The greater the coefficient, the less the inter—generational mobility. Judging from these correlations, the United States has had the most inter-generational mobility, Costa Rica and Japan the least. (This is due more to the amount of change (or lack of change) in the occupational structure of these countries and not likely due to individual factors such as motivation or skills.) Education shows a high correlation with power. The ranks of the values of these coefficients do not show a consistent relationship with the level of development of these five nations. Hewever, a larger sample of countries should show that as social status and economic rewards become increasingly based on achievement rather than ascribed characteristics (an assumed characteristic of developed nations), the strength of the education and power relationship should become stronger. This change in the reward system should also show a declining relationship between family characteristics and power (i.e., more 87 inter-generational mobility). However, this sample of five nations is too small to test this hypothesis. (Simpson (1970) discusses the relationships and anomalies of inter-generational mobility and aspects of alienation in the United States, Mexico, and Costa Rica. Findings are based on the same data as this study.) (Because of the small sample of nations and the lack of a clear consistent ordering of coefficients along the dimension of level of development, continued discussion of the effect of level of development on the relationships of this model is unjustified. I feel confident that the level of a nation's development does directly affect the process of individual modernization or level of individual modernity, but we have insufficient evidence here to evaluate these ideas. (For one possible pattern, see pages 20-22, this paper.) After education's strong tie to power, it continues to have strong ties with the socio-psychological components of self-perceived efficacy, autonomy, and rational decision making. These ties, which are more distant theoretically than these components' ties to power, are frequently slightly stronger than their ties to power. This is counter to the path suggested in the model. However, it does support the theoretical arguments for the supreme importance of education in the modernization process. And while personal characteristics (as indicated here by age) has a strong tie to rational decision making (as indicated here by a positive orientation toward newness and the future), education's ties are stronger than antecedent, ascribed factors. 88 The fact that power, as well as education, exercises an influence on these social psychological components indicates that the model is not entirely without support, and that adult experiences are nearly as important as education in determining modernity. (But because of the strong interdependence of the two components, it might be theoretically futile to go beyond saying that both are extremely important as determinants of modernity, with education being the more important of the two.) CHAPTER V A TEST OF THE MODEL: INDICATORS AT THE NOMINAL AND ORDINAL LEVEL Review of Method Presented here are the results of testing the model of individual modernity as it was presented in Chapter II. The methods used in testing the model were discussed in Chapter III. This section consists of the tests of the model using those indicators which measure at the nominal and ordinal level and some selected assumed interval indicators which were dichotomized. The measure of association used is the contingency coefficient (C) which is derived from the Chi-square value for each relationship. (Being derived from the Chi-square value, all values are necessarily positive.) Where more than one contingency coefficient measuring the relationship between two components is present, the arithmetic mean of the contingency coefficients was calculated and is presented in the test matrices. The tests are presented again in descending order of the five The ranking of the nations is n<':11:ions' level of national development. frOm Farace's article (1966). U§ited States Survgy--Nominal Indicators The indicators used in this test are presented here in abbreviated fOl‘m. Readers wishing to check the exact wording of the items are 89 90 referred to the previous discussion of the theoretical components and their indicators (Chapter II, this paper). Also presented here are the mean contingency coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Internal Consistency of MUltiple Components and Their Indicators Indicators as Measured by the Mean Contingency Coefficient Personal Characteristics Sex Race .011* Family Characteristics Parents' occupation Education Years of formal education Physical Mobility Has visited a foreign country Change from place of birth .088 Psychic Mobility Place of residence Power Owns own home Owns own farm Presently employed .133 Self-employed Supervised employees *The levels of statistical significance for contingency coefficients for samples of this size are: (N = 1528) £4: .10 = .033; 91 Self-Perceived Efficacy Ladder item--can do much to make his life happier Autonomy Has considered moving from present town .147* Can imagine moving from country Rational Decision Making Considered buying stock WOuld invest a large gift of money .061 Behavioral Modernity Member of work related organization Member of voluntary association .120 Presently owns stock *The levels of statistical significance for contingency coefficients for samples of this size are: (N = 1528) pg; .10 = .033; B é .05 = .042; and B é .01 = .059. 92 TABLE 7 TEST MATRIX FOR UNITED STATES, USING NOMINAL AND ORDINAL MEASURES RESULTANT COMPONENTS 1 Personal Characteristics Pers. i Family Char. Characteristics Fam. 015 Education Char. ’ Physical Educ. '114# ’032# [Mobility ANTEXJEDENT C((lMPONENTS ) Phys. # Psychic read down Mob. .107 .062 .125 lMobility Psych' 036 150 107# 111 Power Mob. ' ' ° - Power .160 .080 .048# .077# .130# J’Self-Perceived Efficacy SO-PO # Ant n? EEfic. .011 .012 .141 .044 .008 .030 Ohm) Rational Anton. .052fll .028C‘1.112@ .078@ .031 068 .037# 19901810,, Making Rat. ' @ @ @{ .. Behavioral D. M. .OOBfl .003?.147 .069 .062 004 .074 .078# lModernity 3:3?“ .106Qi.055(1.157 .087 .016 .076# .138 .034# .141# (# indicates theoretically adjacent cells, cells. Remaining cells are at a theoretically intermediate distance. SUMMARY AVERAGES: Theoretically Adjacent' Relatiorships(#) Theoretically Intermetiate Relationships Theoretically Dis tant Relatiorship800) inf 711.. -ifl.1@ II .080 .083 @ indicates theoretically distant RANGES: .141 to .034 .160 to .008 .147 to .011 .060 The levels (f statistical 5 samples of ibis size (N = = 0059 £5.01 ignificance for corral 1528) are p ‘- .10 a ation coefficients for 0033, £9 .05 == .042 p 93 Finland Survey--Nominal Indicators The indicators used in this test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussion of the theoretical components and their indicators (Chapter II, this paper). Also presented here are the mean contingency coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Internal Consistency of Multiple Components and Their Indicators Indicators as Measured by the Mean Contingency Coefficient Personal Characteristics Sex Family Characteristics Parents' occupation Education Years of formal education Physical Mobility Has visited a foreign country Psychic Mobility Number of magazines read Power Owns own home Owns own farm Presently employed .214* Self-employed Supervises employees *The levels of statistical significance for contingency coefficients for samples of this size (N = 893) are: p .4: .10 = .043; p 9 .05 = .055; and p 2.01 = .078. 94 Self-Perceived Efficacy Ladder item—-can do much to make his life happier Autonomy Has considered moving from present town Can imagine moving from country .218* Rational Decision Making Considered buying stock WOuld invest a large gift of money .178 Behavioral Modernity Member of work related organization Member of voluntary association .150 Presently owns stock *The levels of statistical significance for contingency coefficients for samples of this size (N = 893) are: [2‘5 .10 = .043; B é 005 = .055; and E £1 .01 = .0780 95 TABLE 8. TEST MATRIX FOR FINLAND, USING NOMINAL AND ORDINAL MEASURES RESULTANT COMPONENTS 1 Personal Characteristics Pers. Family Char. Characteristics Fan. 7 . Char. .006 l Educat 1011 Physical Ed“- .001# .154# \ luobmty mmmm "t— ‘ cguPONENTS ) I 1 ' read down Ph 3. Psychic “0:. 0090 .091 .326 \illobility Psych. . Mob. .146 .101 .169»).170 {Paws-1 # # \ Self-Perceived Pm" .247 .034 4276.124 .072 \ Efficacy _ -.._~. ’ SO-PO @ # . Rational Anton- .081@L.057@1 2036.176? .064@ .124 .060 1m0ision Making ' B h ' 1 3‘31, .177@ .091@' .157é.124@ .040@ .117 .091 .096# lugdggit; 3:3?“ .1536, .1306 .150 .159 .114 .111'71 .115 .058# .201# (# indicates theoretically adjacent cells, @>indicates theoretically distant cells. Remaining cells are at a theoretically intermediate distance. SUMMARY AVERAGES: Theoretically Adjacent' Relationships(#) Theoretically Intermediate Relationships Theoretically Distant Relationshipsfl?) The levels of staiistic samples of this size (N = 2.4-. .01 = .078. RANGES: = . 120 . 326 to . 001 a .117 .247 to .006 .203 to .040 7‘R1110 a .117 893 ) a1 significance for corrcl ation coefficients for are 2‘ .10 a .043, pf. .05 z .055, 96 Japan Survey--N0minal Indicators The indicators used in this test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussion of the theoretical components and their indicators (Chapter II, this paper). Also presented here are the mean contingency coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Internal Consistency of Multiple Components and Their Indicators Indicators as Measured by the Mean Contiggency Coefficient Personal Characteristics Sex Family Characteristics Parents' occupation Education Years of formal education Physical Mobility Has visited a foreign country Change from place of birth .043* Power Owns own home Presently employed .229 Self-employed Supervises employees *The levels of statistical significance for contingency coefficients for samples of this size (N = 990) are: “Egg-.10 = .041; pg. .05 = .052; and £4; .01 = .074. ' 97 Self-Perceived Efficacy Ladder items-can do much to make his life happier Autonomy Has considered moving from present town Can imagine moving from country .225* Rational Decision Making Considered buying stock WOuld invest a large gift of money .257 Behavioral Modernity Member of work related organization Member of voluntary association .059 Presently owns stock *The levels of statistical significance for contingency coefficients for samples of this size (N = 990) are: p é, .10 = .041; RESULTANP COMPONENTS 1 Personal Characteristics 98 TABLE 9 TEST MATRIX FOR JAPAN, USING NOMINAL AND ORDINAL MEASURES Pers. Family Char. Characteristics Pam. 015 Education Char. ' Physical Educ. .008# .306# lMobility' ANTEEEDENT COMPONENTS ) Phys. Psychic read down “01). 0095 .039 .10641\ limbility P h. “2:? .025 .289 “2074.168 lPower P # # Self—Perceived ower .223 .081 .109 .106 .237 Efficacy S.—P. @ # , Effic. .049 .067 .175 .039 .000 .069 Autonomy Rational “t“- .027@ .1004 .194%.042@ .029@ .073 .068 loom-sit... 1mm Rat. B.b ' l D. M. .082@ .163fi .173é.047@ .125@ .063 .142 .084# Jrngdz:;::; Behav. Mod. .176@ .054fi .079 .057 .136 .093# .121 .062# .182# (# indicates theoretically adjacent cells, cells. SUMMARY AVEIAGES: Theoretically Adjacent Relationships(#) Theoretically Intermediate Relationships Theoretically Distant Relationzships(@) @ indicates theoretically distant Remaining cells are at a theoretically intermediate distance. XRI# 3 .126 XXII = .101 E91116) = .094 RANGES: .306 to .008 .289 to .000 .194 to .027 The levels of statistic samples of this size (N 2 5 ~01 = .074. al significance for corral 990) arepé .10 .. ation coefficients for .041 . 2.4.. .05 a.O52 99 Mexico Survey--Nominal Indicators The indicators used in this test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussions of the theoretical components and their indicators (Chapter II, this paper). Also presented here are the mean contingency coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Internal Consistency of Multiple Components and Their Indicators Indicators as Measured by the Mean Contingency Coefficient Personal Characteristics Sex Family Characteristics Parents' occupation Education Years of formal education Physical Mobility Has visited a foreign country Change from place of birth .OO7* Psychic Mobility Place of residence Power Owns own home Owns own farm *The levels of statistical significance for contingency coefficients for samples of this size (N = 1414) are: Be .10 = .034; p é .05 = .044; and pg .01 = .062. 100 Presently employed .149* Self-employed Supervise employees Self-Perceived Efficacy Ladder items-can do much to make his life happier Autonomy Has considered moving from present town .220 Can imagine moving from country Rational Decision Making Considered buying stock Wbuld invest a large gift of money .014 Behavioral Modernity Member of work related organization Member of voluntary association .115 Presently owns stock *The levels of statistical significance for contingency coefficients for samples of this size (N = 1414) are: p_ g .10 = .034; Ré .05 = .044; and Bé .01 = .062. RESULTAN? COMPONENTS 1 Personal Characteristics 101 TABLE 10 TEST MATRIX FOR MEXICO, USING NOMINAL AND ORDINAL MEASURES Pers. Family Char. Characteristics Fam. 000 Education Char. ’ Physical mm“ .088# .086# [Mobility ANTEDEDPNT ‘ COMPONst Phys. # Psychic (read down) Mob. .109 .081 .130 Mobility 5:?” .071 .013 .278# .081 JPower ‘1 . # # # Self-Perceived Pmr .129 .069 .117 .063 .180 Efficacy So-Po # t v Effie. 0000 0045 .109 0013 0025 0025 An Gnom) Rational Ant”- .005‘5’1 .0596.108@ .069@ .041@ .028 .021 lDecision Making . Bel vior 1 fit“. .132@i .040fi.117@ .073@ .034@ .087 .035 .067# luojzmit; h . _ 35;,” .2019 3706.094 .068 .049 .113# .015 .060# .130# (# indicates theoretically adjacent cells, @'indicates theoretically distant Remaining cells are at a theoretically intermediate distance. cells. SUMMARY AVERAGES: Theoretically Adjacent‘ Relationships(#) Theoretically Intermediate Relationships Theoretically Distant Relationships(@) inf it. inma " Tfie levels of statistical significance for correlation coefficients for samples of tcis size (N = 1414 ) re ~01 - .062. RANGES: .104 .278 to .021 .060 .129 to .000 .071 .201 to .000 arep‘ .10 = .034 , 23?. .05 =.044 , 102 Costa Rica Surveyh-Nominal Indicators The indicators used in this test are presented here in abbreviated form. Readers wishing to check the exact wording of the items are referred to the previous discussion of the theoretical components and their indicators (Chapter II, this paper). Also presented here are the mean contingency coefficients between the multiple indicators of those components so measured, indicating the internal consistency of these multiple indicators. Internal Consistency of Multiple Components and Their Indicators Indicators as Measured by the Mean Contingency Coefficient Personal Characteristics Sex Family Characteristics Parents' occupation Education Years of formal education Physical Mobility Has visited a foreign country Change from place of birth .047* Psychic Mobility Place of residence Power Owns own home Owns own farm *The levels of statistical significance for contingency coefficients for samples of this size (N = 1040) are: pg .10 = .040; 25.05 = .051; and p 5.01 = .072. 103 Presently employed .112* Self—employed Supervises employees Self-Perceived Efficacy Ladder item——can do much to make his life happier Autonomy Has considered moving from present town .230 Can imagine moving from country Rational Decision Making Considered buying stock fibuld invest a large gift of money .113 Behavioral Modernity Member of work related organization Member of voluntary association .188 Presently owns stock *The levels of statistical significance for contingency coefficients for samples of this size (N :2 1040) are: pg .10 = .040; 25 .05 = .051; and Bé .01 = .072. 104 TABLE 11 TEST MATRIX FOR COSTA RICA, USING NOMINAL AND ORDINAL MEASURES RESULTANT COMPONENTS Personal Characteristics Pers. Family Char. Characteristics Fam. Ed t‘ Char. .063 1 uca 101! Physical Educ. .010# .317# lilobility ANTECEDENT COMPONENTS ) Phys. Psychic read down M01). .070 .151 '153#\1M0bilit3' Ps ch. ‘ “03;, .093 .308 .268# .131 11’0"“ P # Self-Perceived “e" .167 .052 .057 .063 .142 Efficacy S 0"?0 @ @ # Autonomy Efflc- .095 .199 .190 .105 .048 .057 Rational Auton- .020@ .088@ .086 .103d.076 .055 .079# inecision Making Rat . @ @ d Behavioral D. u, .068 .098 .114 .045 .065 .067 .163 .072 Modernity Behav. @ . @ ‘ # Med. .097 .109 .146 .095 .092 .069 .157 .066 .148 (# indicates theoretically adjacent cells, cells. @lindicates theoretically distant Remaining cells are at a theoretically intermediate distance. SUMMARY AVERAGES: Theoretically Adjacent‘ Relationships(#) Theoretically Intermeciate Relatiorships Theoretically Distant Relationships(@) TEE levels of statistical significance for corro samples of this size (N = 135;.Od = .072. RANGES: inf = .115 XE“: 7' .119 7111116. .089 1040) are £5 .10 s .317 . 308 .199 to .010 to .052 to .020 :lation coefficients for 0040’ Eé- '05 3 .051 1 105 Comments on Test Matrices Using;Nomina1 Indicators-—A11 Five Nations As contingency coefficients are always positive, it is impossible to test the first hypothesis that no relationships would be negative. The inability to test this hypothesis is unfortunate, but unavoidable, given this usual measure of association of nominal indicators. (Yule's .9, a measure of association for dichotomous relationships, would have permitted this test but was unavailable in existing computer programs.) However, assuming that the majority of relationships in these matrices are positive or null (as was the case with the interval level indicators), I arrive at the very tenuous conclusion that had a suitable measure been feasible, few negative relationships would have appeared. The test of the second hypothesis, dealing with the descending value of coefficients as theoretical distance increases, is disappoint- ing and somewhat inconclusive. In only one of the five samples (Japan) do the summary averages appear in the hypothesized rank order. If I arbitrarily select an absolute difference of .05 as being a meaningful difference, I note that in two cases (United States and Costa Rica) the adjacent and intermediate associations are essentially equal, but both are greater than the mean of the theoretically distant relationships. In the Mexican sample, the mean of intermediate relationships is less than the means for the other two, although adjacent values are greater than distant values, as hypothesized. Finally, in the Finish sample, all three means are essentially equal, although the mean for adjacent associations is again greater than the mean of theoretically distant relationships. 106 Thus there is little support for the second hypothesis and the causal path presented in the model. However, there is even less support for any alternate ordering of relationships. Despite the inconsistent ordering of summary means and the almost erratic appearing values of cell coefficients, adjacent associations consistently have a higher average than do theoretically distant ones. The concluding statement is most accurately stated in the negative: I see no consistent support for any ordering of components other than the order presented in the model. Comments on Indicators-4Nominal Indicator Matrices The inconclusive pattern found across the five nations leads to further inspection of the indicators of components. I noted that a large number of the indicators deal directly with activities and plans in the economic area, with items concerning occupations, employment, stock and real estate ownership, and economic plans. This heavy emphasis on economic items was not intentional but appears because of an early decision to use as few items which appeared in the interval matrix in the nominal matrix as possible. Thus, if a component had at least one indicator at the nominal level, no additional interval level items were dichotomized and analyzed here. This decision, resulting in an over-emphasis on economic activity, was unforeseen and unfortunate. To further inspect the matrices for possible effects or causes of this economic emphasis, it may be somewhat useful to present a summary matrix constructed by averaging (without weighting) the respective cell coefficients of the five nominal level matrices. Caution is urged in 107 interpreting this table, as many cell coefficients vary widely between nations. The averages presented here reflect a centrality that, for several relationships, is more mathematical than actual. (See Table 12.) It will be noted that personal characteristics (as measured by sex, and sex and race in the United States) exercise a strong relation- ship on power (an intermediate association) and rational decision making and behavioral modernity (distant relationships). This pattern appears in all five samples. All of these last three are measured primarily by economic indicators. As the effect of sex upon economic participation is well documented and personal characteristics is not a central component of the model (although obviously empirically salient), a trial was made by excluding personal characteristics as a component. The ordering of the summary averages remained unchanged in all five nations. (However, the United States summary averages did comply with the hypothesized ranking when race was retained but sex excluded.) Thus, personal characteristics was retained, with the original indi- cators, as a component. I The raw coefficients (the relationships of indicator to indicator) were examined to determine if excluding some of the economically based indicators from the analysis would possibly result in greater conformity of results to the model. No meaningful improvement appeared likely. Further after—the-test alterations were not attempted for the practical reason that no increase in consistency 25 compliance to the model or any other ordering of components seemed likely. Also, since the only tactic Open was the deletion of indicators, and thus RISULTANT COMPONENTS 1 Personal Charac tori :tics 108 TABLE 1 2 MEANS OF CELL COEFFICIENTS OF ALL FIVE NATIONS USING NOMINAL AND ORDINAL MEASURES Pers. Family Char. Characteristics g::;. .020 l Educathnl . Physical mm“ .044# .179#\1Mobility AmmEDmT C(()\lPONl~.NTS ) Phys. # Psychic road down “0b. .094 .085 .168 \ (Mommy if?“ .074 .173 .206# .132 1PM.” Power .185 .063 .092# .087# .152# liifigfifmm \ ISM-1'11); .043d .090 .150 .064 .036 .046# lAutonomy t' 1 Anton .037@I .066(‘(.140@ .094@ .050@( .070 .051# 132.1393“ “WM 13“,} .105@ .07od.142@ .072@ .065@( .078 1101 .079# B“”""“?"91 . J'Modermty 3:3?“ .146@l .084d.1'05 .093 .081 .092 .109 .056# .160# (# indicates‘theoretically adjacent cells, @ indicates theoretically distant Remaining cells are at a theoretically intermediate distance. cells. 109 sacrificing the validity of components, it seemed appr0priate to present the matrices as they first appeared. The addition of other indicators was desired but was not feasible. Comments on Components--Nominal Indicator Matrices It will be noted that in the summary matrix (and in the five test matrices) education shows a high relationship with resultant components. While the association is generally greater than hypothe— sized for components at theoretical intermediate and distant distance, it again argues for the importance of formal education in the process of individual modernity. A great deal of variation is noted across nations in the inter- relationships of the first five components. ExPlanations for this variation undoubtedly lie within the unique occupational, educational, and stratification histories of the nations. As was the case for the interval level matrices, I see no pattern consistent with the relative level of national development. The tie between self-perceived efficacy and behavioral modernity (a relationship of intermediate distance) is high in all samples and is almost always greater than nearby, theoretically adjacent, associations. 0n the face of these coefficients alone, it would argue for revising the order of the path model. In addition, efficacy generally has a stronger tie to rational decision making (intermediate) than it does to autonomy (adjacent). (This same pattern appeared frequently in the interval level matrices also.) 110 However, I am reluctant to reorder the components of the model on the basis of this evidence alone. The weakness of the interval level indicators of autonomy was discussed, and the questionable economic emphasis of the nominal indicators is sufficient to permit me to retain efficacy in its theoretical position adjacent to power. As efficacy is essentially the perception of power, its logical location is adjacent to power. Stronger tests than these will be required to alter the theoretical location of this component. CHAPTER VI RESULTS OF THE ANALYSIS OF DELETED INTERNAL INDICATORS Reasons for Further Analysis As was noted above, a number of the indicators which measured at the assumed interval level were deleted from the major analysis. The original reason for their deletion was their lack of consistency with the other indicators of the same component. (Evidence of this inconsistency is presented here by circled coefficients in Tables 13- 20.) The possibility also existed that despite the inconsistency with the retained indicators, the "wrong" groups of indicators may have been retained and deleted. (The evidence here and previously in the interval level matrices indicatesthat the correct group was chosen.) In addition, some previously unseen pattern may have appeared, shedding further light on the process of individual modernity. As a result, the degree of association between scales of the deleted indicators and the retained indicators of the theoretical components was calculated and the results presented here. Scales and Internal Consistency The deleted items were collected into eight scales by the apparent abstract meaning of their content. (One item is analyzed by itself-~a one-item "scale.") In a general sense, they reflect some of Inkeles' (1966) major themes of individual modernity (see also pages 13—17, this 111 112 paper). Those themes reflected in these scales of deleted items include: a willingness to hold opinions and a tolerance for other opinions, a faith in planning and science and technology, a positive attitude toward change, an openness to strangers and foreign things, a belief in distributive justice, and by extrapolation, a faith in one's own individual ability. As such, these eight scales provide a tangential evaluation of these more specific themes of Inkeles' definition of individual modernity. The internal consistency of the scales is frequently low, and thus their analysis as multiple indicators of one phenomenon is often unjustified. However, as this is not the major test of the model, it is useful to continue with them as though they were internally consistent to explore the evidence of their relationship to the components of the model. (Items which show erratic internal consistency which is not readily apparent in the mean correlation coefficient will be noted.) In addition, I will briefly mention the original rationale for attempting to include them in the original test of the model, noting the theoretical component I thought they should be a part of and thus also indicating the rationale for their combination into the scales appearing here. Absence of Powerlessness (These two items were used in the major interval test matrix for the United States and Costa Rica, and the first item was used in the interval test matrix for Finland. They were deleted from the major 113 analysis of Mexican and Japanese samples, and the second item was deleted from the Finland test matrix.) These two items were felt to be indicators of self-perceived efficacy and were originally intended for use in all five nations. As efficacy is a perception of one's power, an absence of fatalism, and a faith in one's own effectiveness, these items should have been indicators of this orientation. Sometimes I have the feeling that other people are using me. (Disagreement related to modernity.) (Used here for Mexico and Japan only.) There is little chance to get ahead in this life unless a man knows the right people. (Disagreement related to modernity.) (Used for Mexico, Japan, and Finland only.) Internal Consistency: Japan: .129 Mexico: .217 The generally null relationships seen here, in a pattern which does not consistently follow the model, justify the deletion of these indicators from the interval test matrices of these countries. The fact that they were included in the test matrix of two countries and one of them in a third and that they there present a pattern which conformed to the model (see Tables 3, 4, and 7) strongly suggests that these items mean different things in the different countries. This apparent lack of a consistent relationship to a model of modernity (which appeared consistent in five separate cultural contexts) strongly 114 TABLE 13 MEAN CORRELATION COEFFICIENTS BETWEEN "ABSENCE OF POWERLESSNESS" ANDlCOMPONENTS OF THE MODEL (INTERVAL ITEMS) Nation United . . Costa States Finland Japan Mex100 Rica N = 1,528 893 990 -1 1,414 1,040 Pers. Char. .076 -.013 -.016 Fam. -0007 0001 0039 Char. Blue 0 .082 - 0030 0060 ‘— Psych. Mob. .038 .005 .068 __ Power .043 .006 .053 So-Po ' ' Effie. Anton . - . 036 .009 . 049 ' Rat. D. M. .007 .011 -.031 Levels Of pé.10 = .03 pé..10 = .04 135.10 = .04 pf; .10 = .03 pg .10 = .04 Stat. pg;.05 = .04 pg;.05 = .06 p£5.05 = .05 p5;.05 = .04 pg;.05 = .05 Sign. pg: 01 = .06 pg .01 = .08 p_4_.01 = .07 p$.01 = .06 pé.01 = .07 Circled coefficients indicate tie to original theoretical component. *Coefficient from major test matrix. 115 suggests that the phenomenal equivalence of these items is low across the five nations. Political Opinions and Tolerance Inkeles (1966) notes that a modern man has opinions on a wide variety of topics but is also tolerant of the opinions of others. I saw this dimension as an aspect of autonomy, as the man who is relatively independent of any one system and interdependent with several social systems is likely to have opinions about the preferable means and ends of a variety of activities. Also because of a widely based participation, he must be tolerant of the opinions of others. I always try to keep my political beliefs to myself. (Disagreement related to modernity.) Political beliefs should have nothing to do with a person's work. (Agreement related to modernity.) Internal Consistency: United States: .192 Finland: .108 Japan: .193 Mexico: .200 Costa Rica: .283 Despite the acceptable level of internal consistency between the two items, they present an inconsistent pattern when compared with the components of the model. In the two most developed countries, they show null and negative relationships, while in the other three generally present null associations. Again, the phenomenal equivalence of these 116 TABLE 14 MEAN CORRELATION COEFFICIENTS BETWEEN "POLITICAL OPINIONS AND TOLERANCE" AND COMPONENTS OF THE MODEL (INTERVAL ITEMS) Nation United . ,. Costa mates Finland Japan Mexico Rica N = 1,528 893 990 1,414 1,040 Pers' —.050 -.006 .035 .008 -.046 Char. Fam‘ -.069 ...O49 .068 .053 .004 Char. Educ. -.127 -.107 .059 .057 .009 Psych. Mob. .009 -.036 .031 .041 .049 Power -.065 -.082 .055 .012 .050 S.—P. Effie. -.025 -.026 -.017 .040 -.022 Auton- .038 .014 .025 .032 62) Rat. 4., ‘7 " D M .001 -.011 .000 .045 .019 Levels of pé.10 = .03 pet-“10 = .04 pé.10 = .04 pg.10 = .03 pé.10 = .04 Stat. pé.O-5 = .04 pé...05 = .06 pg.05 = .05 peg-.05 = .04 pé.05 = .05 Sign. pé.01 = .06 pg .01 = .08 p£.01 = .07 p$.01 = .06 pé.01 = .07 Circled coefficients indicate tie to original theoretical component. *Cocfficient from major test matrix. 117 two items must be questioned. While I am satisfied with their nominal equivahyme (because of the translation, back-translation process in the preparation of interview schedules), people's response to these two items seems to be calling forth different orientations to the) general issue of political opinion. Religious Tolerance Continuing with Inkeles' theme of tolerance and my inclusion of tolerance as a likely indicator of autonomy, here are two items dealing with the person's orientation to the religious beliefs of others. It really doesn't matter what an individual believes about religion as long as he is happy with it. (Agreement related to modernity.) I believe the world would be a better place if more people had the religious beliefs which I have. (Disagreement related to modernity.) Internal Consistency: United States: .118 Finland: -.O58 Japan: -.326 Mexico: .034 Costa Rica: .014 This curious inconsistency between the relationships between the two indicators for the five nations is probably due to a lack of phenomenal equivalence. The dimension most likely involved here is the second item, which could evoke a response based on the person's faith in his own religion, or the person's 118 TABLE 15 MEAN CORRELATION COEFFICIENTS RENEE) "RELIGIOUS TOLERANCE" . AND COMPONENTS OF THE MODEL (INTERVAL ITEMS) p53.01 Nation United . ‘ . Costa States Finland Japan MeXIco Rica N = 1,528 893 990 1,414 1,040 Pers' 014 090 015 023 - 026 Char. -' ° ° ° ' Fam’ 003 041 004 .069 .100 Char. ‘ ' ' Educ. .039 .038 .010 .144 .125 PSYCh' .041 .033 .010 .067 .089 Mnb. Power .038 .076 .007 .104 .111 So-Po o o "o 4 0042 0071 Effie. 000 003 O 8 6) 6) C) 6) Rat. ' VF D. M. 0009 .031 "' 017 0042 0031 Levels Stat. p4,..05 = .04» pé.05 = .06 p5...05 = .05 p5..05 = .04 pé.05 = .05 Sign. = 006 IJé .01 = 0.08. Pg .01 = .07 p5001 = 006 135.01 = 007 Circled coefficients indicate tie to original theoretical component. *Coefficient from major test matrix. 119 tolerance of other religions. Apparently (particularly in Japan), faith and tolerance are not mutually exclusive. The lack of consistent phenomenal equivalence is again apparent. The highest coefficients are found in the Mexican and Costa Rican data, and it is possible to guess that respondents in these two Roman Catholic countries are responding to the dimension of faith in their own religion, rather than on tolerance for others. The low coefficients with the other indicators of autonomy suggest that this might be the case, and the guess is also based on the low correlation between the two items used here in the "Religious Tolerance" scale. Absence of Xenophobia By extending Inkeles' theme of tolerance in modern man and this paper's theme of autonomy, it appears that the modern person should be tolerant of, in fact desirous of, contacts with strangers and foreign people and institutions. Wm should be as helpful to people we don't know as we are to our friends. (Agreement related to modernity.) It is a good thing for companies and business firms from other countries to do business and have factories in our country. (Agreement related to modernity.) It is a good thing for our young people to marry people from other countries. (Agreement related to modernity.) 120 Internal Consistency: United States: .149 Finland: .123 Japan: .119 Mexico: .090 Costa Rica: .221 (It should be noted that these mean coefficients reflect a strong relationship between the last two items and a relatively weak tie between the first and the last two.) The lack of conclusive findings regarding the relationship of this scale and the components can be viewed in light of the various dimensions of the items. Responses to the items could have been based on cultural proscriptions ("help strangers") without actual intention or orientation, nationalism, or racism ("marry foreigners"), and a variety of orientations to national-foreign economic policy. A.response to the item regarding foreign business should have greatly differing meanings and intensity depending on the respondent's citizenship in a country where foreign competition is feared, encouraged, prohibited, or negligible. Perhaps because of this variety of dimensions, the scale fails to develop any consistent pattern or even a relationship when compared to the components of the model. The vast majority of coefficients should be seen as representing null relationships. 121 TABLE 16 MEAN CORRELATION COEFFICIENTS BETWEEN "ABSENCE OF XENOPHOBIA." . AND THE COMPONENTS OF THE MODEL (INTERVAL ITEMS) Nation United . ' . Costa States Finland Japan MOXICO Rica N = 1,528 893 990 1,414 1,040 Pers ' .050 - .039 .040 - .029 .000 Char. mm -.015 -.058 .036 .028 .068 Char. Mlle o 0040 -' 0046 0067 o 023 0091 PSYCh' .039 -.016 .022 .013 .052 Mob. .— Power .004 -.065 .051 .020 .076 SO‘PO 031 O Effie. . -. 05 .071 .031 .023 Anton. .028 .037 038 .013 .022 Rat. D. M. .056 —.002 .089 .019 .053 Levels . of pea-.10 = .03 pé.10 = .04 péJO = .04 pé.10 = .03 p540 = .04 Stat. p4,...05 = .04 pé.05 = .06 pg.05 = .05 p5..05 = .04 pé.05 = .05 Sign. pg .01 = .00 pg .01 = .08 135.01 = .07 p.13 .01 = .06 P5 .01 = .07 Circled coefficients indicate tie to original theoretical component. *Coefficient from major test matrix. 122 Preference for Individual Action As a final aspect of the concept of autonomy, a scale is presented which measures the person's preference for individual action. This aspect is central to the idea of autonomy, as it is the individual who must move from system to system, inspecting ends and means and choosing between them. The autonomous man has a preference for and propensity to individual action. Everyone should think the same about what is right and what is wrong. (Disagreement related to modernity.) I find it easier to follow rules than to do things my way. (Disagreement related to modernity.) Whatever we do, it is necessary that our leaders outline carefully what is to be done and exactly how to go about doing it. (Disagreement related to modernity.) Children should be taught that there is only one correct way to do things. (Disagreement is related to modernity.) Internal Consistency: United States: .278 Finland: .361 Japan: .320 Mexico: .205 Costa Rica: .239 (These high levels of internal consistency reflect a relatively consistent pattern among the four items.) 123 TABLE 17 MEAN CORRELATION COEFFICIENTS BETWEEN "PREFERENCE FOR INDIVIDUAL ACTION" AND COMPONENTS OF THE MODEL (INTERVAL ITEMS) Nation __ United . j . Costa States Finland Japan Mexico Rica N = 1,528 893 990 1,414 1,040 Pers' .115 .137 .041 .024 .030 Char. , Fam- .106 .136 .016 —.004 .068 Char. Educ. .291 .275 .081 -.O3O .139 Psych. 07 Mob. . 5 .058 .023 —.026 .025 Power .201 .157 .007 -.000 .136 80-1,. Effic. .073 .051 -.081 -.022 .052 6) 6) @ e 6) Rate ‘ D. M. .011, .082 —.009 -.033 -.009 Levels of p540 = .03 115.10 = .04- p5.10 = .04 pé.10 = .03 pé.10 = .(4 Stat. 11.5.05 = .04 1)é._.05 = .06 p5...05 = .05 pé..05 = .04 pé.05 r. .05 Sign. p_.r_.:._.01 -- .06 pg .01 = .08 pé.Ol =: .07 p5.01 = .06 p_/___.01 = .07 Circled coefficients indicate tie to original theoretical component. *Coefficient from major test matrix. 124 Because of the strong internal consistency of these four components, their clear theoretical relevance to the concept of autonomy, their deletion from the model did not occur automatically. However, their negative and null relationship to the retained indicators of autonomy would mean that those other indicators would have to be discarded if these four were to be used in the main test of the model as indicators of autonomy. The issue then is which of the two collections of items provide the best indicator of the concept, on theoretical and empirical grounds. The rationale for including the items pertaining to the individuars ties to social systems is more circuitous than the one arguing for including the items related to a reference for individual action. However, both appear equally valid on theoretical grounds. (Problems encountered in operationalizing the concept of autonomy were discussed above, pages 84 and 86.) The decision to use the items which were used in the major test was made after considering the following points. Since autonomy is a product of both independence and interdependence, those items which best reflect this duality were chosen. The items actually used include an item relating to independence-~the ladder item in which a respondent evaluates his chances to do anything he wants to--and this item related reasonably well to the social systemic ties (interdependence) items. These four deleted items would only reflect the independence aspect of the dual nature of autonomy. Secondly, the phenomenal equivalence of this type of Opinion item has seemed to lack the level of phenomenal 125 equivalence of the ladder type of item, and for this added reason the choice was made. The decision seems to be a correct one. In inspecting the relation- ship of this scale of "Preference for Individual Action" to the components (Table 17) and comparing it to the association of the retained indicators to other components (Tables 1-6), one notes that while neither yields totally consistent or large coefficients, the retained indicators present results more in keeping with the pattern of the model. The deleted indicators tend to show a higher relation- ship with the early antecedent components but a weaker relationship with those components which are theoretically closer. A Belief in Planningyand Technolcgy_ Two of the central characteristics of modern peOple, as seen by many writers, and by this model, is a positive attitude toward science and technology. Modern men also plan and do not leave the future to fate. These two items, both subsumed under the model's concept of rational decision making attempt to measure the respondent's attitude toward planning for the future and the use of technology to reach those desired goals. Health experts say adding chemicals to drinking water results in less decay in people's teeth. If you could add these chemicals to your water, with little cost to you, would you be willing to have the chemicals added? (Affirmative answer related to modernity.) 126 TABLE 18 MEAN CORRELATION COEFFICIENTS BETWEEN "A BELIEF IN PLANNING AND TECHNOLOGY" AND THE COMPONENTS OF THE MODEL (INTERVAL ITEMS) Nation United . . Costa States FInland Japan Mex1co Rica N = 1,528 893 990 1,414 1,040 1’9"“ .137 - -.050 .144 .023 .064 Char. Farm .033 -.012 .051 .101 .120 Char. Educ . .143 - .099 ' .110 .027 .178 5?“ .030 .076 .037 .026 .108 Power .095 .020 . .019 .022 .141 So-Po Effie. .018 .003 -.041 -.003 .073 Auton . .000 - .012 - . 012 .001 .085 Rat. ‘ Levels Sign. 135.01 = .06 pg .01 = .08. 1.5.01 = .07 p5,.01 = .06 pé.Ol = .07 Circled coefficients indicate tie to original theoretical component. *Coefficient from major test matrix. 127 Family planning by birth control has been discussed by many people. What is your feeling about a married couple practicing birth control? (Believing it acceptable related to modernity.) Internal Consistency: United States: -.O76 Finland: -.097 Japan: .059 Mexico: .032 Costa Rica: .065 In light of the low internal consistency between the two items, and the welter of political and religious issues raised by fluoridation and birth control, it seems surprising to achieve coefficients as great as these. However, they are most strongly related to the antecedent items, and generally Show a null association with those components theoretically close to rational decision making. In addition, the items appear to lack phenomenal equivalence, with mean- ingful differenCe between Finland and Mexico and the remainder. (The meaningful difference in coefficients here between Mexico and Costa Rica is one of several in this paper. The difference between two countries at the same level of development and with the same general cultural heritage invite speculation and further research.) Belief in Distributive Justice The effort and cost of making rational decisions, as opposed to the relative ease of automatic referral to traditional solutions, must be offset by the anticipated reward; the man who makes rational 128 TABLE 19 MEAN CORRELATION COEFFICIENTS BETWEEN "A BELIEF IN DISTRIBUTIVE JUSTICE" _ AND COMPONENTS OF THE MODEL (INTERVAL ITEMS) Nation United . . Costa States Finland Japan Mex1co Rica N = 1,528 893 990 1,414 1,040 Pers' -.057 .067 .011 .003 -.006 Char. __ Fam' -.066 .006 .065 .038 .051 Char. Efluc. -.134 -.016 .063 .092 .018 PSYC" - - .013 - .001 .046 .075 .031 Mnh. .— So-Po . "'0 1 0037 0009 0032 0020 Effic. 03 Auton. .051. .001 .017 .037 .033 “‘5 ® ® @ D. M. Levels of pé.10 = .03 pé.10 = .04 pé.10 = .04 139.10 = .03 pé.10 = .04 Stat. pé.05 = .04 pé.05 = .06 pg.05 = .05 psi-.05 = .04 pé.05 = .05 Sign. pé.01 = .06 pg .01 = .08 p_4_.01 = .07 p5,.01 = .06 pé.01 = .07 Circled coefficients indicate tie to original theoretical component. *Coefficient from major test matrix. 129 decisions has a general faith that rewards will be commensurate with his contributions. Therefore, one derivative of rational decision making is a belief in "distributive justice." A man who works hard and as well as he can deserves a comfortable standard of living. (Agreement related to modernity.) It is not difficult to explain the consistently low coefficients associated with this item. The item is stated as a general principle, and over 94 percent agreed with the item in every nation. At the least, we can assume the item is phenomenally equivalent. Positive Attitude Toward Changg As rational decision making is predicated on the inspection of a variety of solutions, and the inspection of alternatives implies a partial dissatisfaction with conditions at present, modern men would be expected to be optimistic about change and the future. People's ideas change so much that I wonder if we'll ever have anything to depend upon. (Disagreement related to modernity.) Running a city or village or any governmental organization is an important job. What is your feeling on this statement? "Political leaders should be changed regularly, even if they are doing a good job." (Agreement related to modernity.) I like the kind of work that lets me do things about the same from one week to the next. (Disagreement related to modernity.) Some people feel that in bringing up children new ways and methods should be tried whenever possible. Others feel that trying new methods is dangerous. What is your feeling on this statement: "New methods of raising children should always be tried out." 130 (Agreement related to modernity.) (This item analyzed here for Finland only; it is included in the main analysis for the other four nations.) Internal Consistency: United States: .077 Finland: .045 H? Japan: -.000 Mexico: .048 Costa Rica: -.008 These coefficients do not show a particularly strong pattern. The ' majority of them indicate a null association. While the highest coefficients appear in the United States sample, they are not divergent enough to seriously question the phenomenal equivalence of the items. Perhaps it is sufficient to say that they are apparently unrelated with the major components of the model. This seems strange, as the retained indicators of rational decision making (generally dealing with atti— tudes about the future) do show a reasonable relationship with other components of the model.* This perhaps suggests that there is a significant difference between looking optimistically at the future and looking optimistically at change. Perhaps future rewards are defined very much in present terms, and the means to those rewards are defined in terms of the means available in the present. Thus, change may actually threaten one's hopes for the future. Comments on the Associations of Deleted Indicators The results of this analysis clearly show that the deletion of these indicators was justified on empirical grounds, for if they had 131 TABLE 20 MEAN CORRELATION COEFFICIENTS BETWEEN "POSITIVE ATTITUDE TOWARD CHANGE" AND COMPONENTS OF THE MODEL (INTERVAL ITEMS) Nation United . . Costa States Finland Japan Mex1co Rica N = 1,528 893 990 1,414 1,040 Pers' .064 .091 .057 .025 .034 Char. Fam' .060 .050 .012 .003 .029 Char. Educ. .175 -.012 .073 .016 .063 Psy°h° .055 .022 .023 -.023 -.006 Mob. Power .144 .033 .017 .014 .050 So-Po Effic. .086 .042 .017 .021 .038 Rat. ' , D. M. Levels of pé.10 = .03 pé.10 = .04 p§.10 = .04 p540 = .03 pé.10 = .04 Stat. pé..05 = 004 pé.05 = 006 pg.05 = .05 p9005 = 004 pé.05 = 005 Sign. pé.01 = .06 pg .01 = .08 p._¢_.01 = .07 p$.01 = .06 pé.01 = .07 Circled coefficients indicate tie to original theoretical component. *Coefficient from major test matrix. lift] 132 been included, they would have threatened the pattern which did appear. Their exclusion can also be justified, in most cases, by a lack of phenomenal equivalence, a lack of discrimination, or a lack of total validity due to multi-dimensionality. In themselves, fley do not test Inkeles' themes of his listing definition adequately enough to question its validity. However, they do underline the extreme importance of valid, phenomenally equivalent indicators. The opinion item with specific content must be regarded with a great deal of suspicion in cross-cultural survey research. "'7 CHAPTER VII EVALUATION, COMMENTS AND IMPLICATIONS General Evaluation [a This section will present a restatement of the themes of the first three chapters in light of the results of the testing of them. The three themes are the definition of individual modernity, the path I model of the process of individual modernization, and the method of analysis similar to Guttman's "simplex" (1954:271). The results of the analysis of the data (Chapters IVLVI) establish no conclusive or definitive support for the three main aspects of the thesis. However, the results do strongly indicate that further study of all three areas is warranted. The Definition of Individual Modernity The findings of this study and the findings and experiences of many social scientists indicate that there is a social psychological orientation toward the social and physical world that characterizes those who are socially successful in urban, technological, and secular social settings. This orientation toward life is significantly different from that of those in rural, technologically primitive, and ascriptive societies. This orientation has been labeled "individual modernity." 133 134 The issue of the proper definition for this phenomenon is unresolved. The type of definition which is presently most popular in thinking and research is the listing definition of individual modernity. Inkeles' definitions (see Inkeles, 1966 as an example) are typical, salient, and important ideas. Despite their success in leading to empirically reliable scales for measuring modernity (Smith and Inkeles, 1966), they are subject to the serious criticism that they are bound to the time and culture of Twentieth Century middle class North America for their ideal-typical end state. If the people of the world continue to strive for this style of life and middle class Americans remain unchanged, this listing definition would be adequate for theoretical and empirical purposes. However, the degree to which this striving and stability is likely is debatable and, in any event, a questionable assumption. Thus, I attempted to construct a definition of this orientation toward life which would be less time and culture bound, leading to a greater conceptual equivalence in a variety of cultural settings. From the existing descriptions and definitions of modernity, I abstracted two characteristics which seemed centra1--autonomy and rational decision making. USing these concepts, we can visualize a man modern rather than traditional in his outlook on life but who can choose to be tardy, can question the efficacy of technology, can evaluate the desirability of change, and can celebrate many traditional and non— utilitarian aspects of life. All of these characteristics would violate the letter of Inkeles' listing definition. These two central 135 characteristics seem to be a more accurate conceptualization of the phenomenon known as modernity. Problems of Operationalization and Equivalence My satisfaction with the new definition at the conceptual level does not carry to the operational level. The tests of the so—defined orientation (and the model of its antecedent stages) were less than satisfactory or conclusive. Rational decision making and particularly autonomy showed themselves to be difficult concepts to operationalize. Part of the difficulty is undoubtedly caused by attempting to use existing items in a secondary analysis of collected data. However, had I constructed items before data collection, they would have been similar and likely would have met with the same lack of success. Operational definitions of the two concepts should, in any future research, attempt to get at the essential characteristics of the two components. For Operationalization, autonomy would likely have to be divided into sub-components of independence and multiple participation and items which would measure a person's evaluation of his standing relative to social systems devised. The "self-anchoring scale" items (Kilpatrick and Cantril, 1960) (here called "ladder items")seem to be a more valid and phenomenally equivalent ordinal measure than the typical opinion item having more specific content. Rational decision making also remains difficult to operationalize but should not be impossible. Items directly dealing with the process of decision making as well as the derived characteristics of innovativeness, planning, and risk-taking could be valid indicators. An attempt should also be made 136 to broaden the scope of this component from merely an economic context. The listing definition, with its more specific characteristics, is easier to operationalize. This is not to suggest that the more specific characteristics are simple to operationalize. As seen in my analysis of the deleted indicators here (Chapter VI) and in Smith and Inkeles' (1966) use of 119 items to find reliable indicators for much shorter scales, face validity does not always, or frequently, result in reliability. However, operational validity and reliability at the expense of conceptual validity and equivalence is a dubious exchange. The problems and questions of reliability and validity are intensified owing to the inter—societal nature of the data and the test of the model and definition. The issue of conceptual validity—-or in the comparative case, conceptual equivalence--is one which must be raised regarding the general definition of individual modernity and the Operationalization of the ten components of the model of modernization. The definition of modernity presented here, by being an abstract definition,has greater potential for inter-societal equivalence than the listing definitions. In some cases (see Doob's and.Rogers' listing definitions, page 4, this paper) listing definitions address and emphasize clearly different aspects of the concept of modernity. The more abstract definition here does have the potential for greater conceptual generality. "The theoretical significance of concepts defined in terms of specific measuring operations is limited. The 137 generality of such concepts is low and often specific to a social system" (Przeworski and Teune, 1970:95). While the more general definition here has more potential for intersocietal validity, its conceptual validity is not yet established. Przeworski and Teune (100-103) present the general logic of establishing discriminant validity of general concepts, and this procedure was used in this research (see pages 58-59 and 61, this paper). The establish- ment or approximation of discriminant validity does not yet establish equivalency across social systems. Przeworski and Teune suggest that equivalence, using common indicators, is established (in probabilistic terms) if "the indicators behave the same way in all systems" (page 121) and if one examines and finds similar "correlations among the indicators in the pooled population" (page 122). In this test of the model, the behavior of indicators (internal consistency) and similarity of correlations was not identical in each of the ten tests of the model. However, a repetitive pattern was observed in the retained indicators and provides evidence that equivalence is approached, if not established. Present EValuation of the Definition In spite of the problems of Operationalization, I believe that this definition of individual modernity, emphasizing rationality and autonomy, is extremely useful at the conceptual and theoretical level. Its utility in research must await valid and reliable indicators. The rewards of efforts to create good indicators should outweigh the costs and difficulty of developing them. 138 The Path Model of the Process of Individual Modernization The model of the process of individual modernization, as presented in Chapter II, was tested with data from surveys in five nations. The items were divided into two groups—-those which measured at the assumed interval level and higher and those which measured at the nominal and ordinal level. Thus a total of ten tests of the model were conducted. All of the tests utilizing interval items gave positive results. While the absolute degree of support was not great, a consistent pattern of coefficients did appear, decreasing as theoretical distance in the model increased. I interpret these results as supporting the path presented in the model. One of the tests using nominal indicators gave positive results, and the other four did show that components which were theoretically adjacent had a higher degree of association than those which were theoretically distant. I interpret this as partial support for the selection of components and the theorized ordering as presented in the model. Selection of Components As the components of the model reflect a general consensus in the field regarding possible antecedents to modernity, it seems an adequate collection of possible and likely causes. The three components used to collectively identify individual modernity were discussed above as issues relating to the definition of modernity. With the possible exception of the division of autonomy into its independence and participation dimensions and the use of these as two separate components, 139 I can see no change in the components which would present a sounder theoretical statement or a more accurate empirical picture. The Orderipg of Components Again the ordering of components reflects the thinking and writing of those in this field, and also mirror to a large extent the events and stages in a person's life. Educational achievement generally is a result of personal and family characteristics, economic power is a resultant factor of education, etc. The results of six of the ten tests indicate that this is a fairly adequate ordering of components, and the remaining four tests indicate no single alternate ordering. Comments on Indicators The most efficacious way to improve the empirical correspondence of the model would be, as often stated, the availability of more valid and representative indicators. ‘Generally I was able to test only a few dimensions of each component. Among those dimensions missing and not tested were: the type and content of one's education, non—economic aspects of power, fuller information on domestic travel, more indicators of non-economic behavioral modernity, etc. However, it must be noted that for this purpose of an early and feasible test of the ideas, the quality of indicators and data collection were quite adequate. As much of the model and its empirical test hinge on the opera- tional definition of individual modernity, the problems discussed above in connection with the definition of modernity could be restated here. 140 Present Evaluation of the Path Model Taking into account the problems with indicators and the partially positive tests, I see much to recommend this model of the process of individual modernity for further study. It is the only time ordered, processual, model of individual modernization and as such warrants further study to more conclusively test and refine it. The Method of Analysis This method of testing the path model of modernity was originally selected by default, as the traditional techniques of path analysis and other multivariate techniques made too many unmet requirements on the data and model. Through the use of this method, I now find much to recommend its further inspection and possible use in a variety of situations testing path models which do not meet the rigorous requirements of the more advanced techniques. Advantages of This Method of Analysis The_main advantage of this method of analysis lies in its ability to utilize a variety of data, its Openness to inspection and alteration at all stages, and essentially its intrinsic simplicity. Requirements of the Data and Theories The more rigorous multivariate techniques of analysis (and specifically path analysis) generally require interval data, no reciprocal relationships or feedback loops, the inclusion of all 141 relevant variables, large samples, and certainty of the correct causal ordering before analysis. (See Hiese, 1968 and 1969.) None of these assumptions need be met with the technique used here. As a result of its simplicity and flexibility, its results cannot be interpreted as precisely or finally as those of other techniques, but as an initial, exploratory device it would be very useful. The technique can utilize any of a number of measures of association and thus any type of data. (However, all calculations must consist of the measure of association appropriate to the lowest level of measurement present.) The technique can utilize the contingency coefficient (C), Yule's Q, rank order correlation (rrho), and product moment correlation (r). (However, as these measures of association are not directly comparable, one measure must be used consistently.) The technique can be used as a method of proof and in the frequent combined practice of testing an original hypothesis while looking for alternate hypotheses. It can also be used as a method of discovery if the indicators of components are known, but the order of components is not known. The method also does not require the prior development of index values for multiply indicated components prior to analysis, and items in multiply indicated components can be evaluated during the process. The method's ability to incorporate feedback IOOps and reciprocal relationships is another advantage, and one can present these socially common relationships in theory and then test them. 142 Its Simplicity and Flexibility The fact that the construction of final matrices is done with paper and pencil is also an advantage. (The construction could be handled by computer with a simple program.) This somewhat menial task does permit added insight into the workings of the various relationships and may lead to alteration of the model or indicators before final presentation. In addition it permits the development of alternate hypotheses and the possibility of quick tentative testing of them at the time. Some Disadvantages of the Method The method is crude and does not permit precise generalizations of relationships between components. This crudeness is primarily a result of the difficulty of control for multiple effects. Control is impossible for more than a few (one or two) variables and would require the construction of a complete matrix for each value of the variable controlled. It also does not easily permit examination of curvilinear relationships, although this is dependent upon the measure of association used. Correlation coefficients and Yule's Q (based on a four-cell table only) would disguise any curvilinearity. The contingency coefficient, used with tables with more than one degree of freedom, would present some indication of a relationship which might be curvilinear. However, as the special characteristics of these common measures of association are well known, the researcher can easily evaluate what results are possibly artifacts of the measure of 143 association and inspect the relationship separately. The lack of control may mean that one or several variables exercise undue influence on variables far removed from them in theory. In this study, education showed such a tendency. However, it is frequently of great concern to inspect these relationships, reflecting such a strong relationship in reality, before they are controlled. Present Evaluation of the Method of Analysis While a scholar with greater statistical expertise than myself must evaluate this method of analysis, its simplicity and flexibility seem to offer a great deal to the researcher tentatively testing a causal model which violates the theoretical or empirical assumptions of more rigorous multivariate techniques. It strongly calls for further consideration, as it appears to fill a need for an analysis technique between the cumbersome two, three, and four variable analytic schemes and the more rigorous multivariate designs. LIST OF REFERENCES LIST OF REFERENCES Adams, Richard N 1967 The Second Sowing: Power and Secondary Development in Latin America. San Francisco: Chandler. Almond, Gabriel A., and Sidney Verba 1963 The Civic Culture: Political Attitudes and Democracy in Five Nations. Princeton, N. J.: Princeton University Press. Anderson, C. Arnold, and Mary Jean Bowman, eds. 1965 Education and Economic Development. Chicago: Aldine. Armer, Michael, and Allan Schnaiberg 1971 "Measuring modernity: myth or reality?" Paper presented at the annual meeting of the American Sociological Association, Denver, Colorado. Barndt, Deborah 1969 "Changing time-conceptions." Summation I (February): 17-27. Berger, Bennett M. 1971 Looking For America. Englewood Cliffs, New Jersey: Prentice Hall. Blalock, Hubert M. 1964 Causal Inference in Non-experimental Research. Chapel Hill: University of North Carolina Press. Braibanti, Ralph, and Joseph J. Spengler, eds. 1961 Tradition, values, and Socio—Economic Development. Durham, North Carolina: Duke University Press. Brand, Stewart, ed. 1971 The Last Whole Earth Catalog. New York: Random House. Buchanan, William and Hadley Cantril 1953 How Nations See Each Other. Urbana: University of Illinois Press. Cantril, Hadley (ed.) 1951 Public Opinion, 1935-1946. Princeton, N. J.: Princeton University Press. 1965 The Pattern of Human Concerns. New Brunswick, N. J.: Rutgers University Press. 144 145 Childe, V. Gordon 1951 Man Makes Himself. New York: Mentor. Doob, Leonard W. 1967 "Scales for assaying psychological modernization in Africa." Public Opinion Quarterly 31 (Fall):414—421. Duncan, Otis D. 1966 "Path analysis: sociological examples." American Journal of Sociology 72 (July):1-16. Edwards, Allen L. 1950 Ekperimental Design in Psychological Research. New York: Rinehart. Feldman, Arnold S. and Christopher Hurn 1966 "The experience of modernity." Sociometry 29 (December). Ginsberg, Morris 1965 "Towards a theory of social development: the growth of rationality." Pages 27-66 in Raymond Aron and Bert F. Hoselitz, conductors, Social Development. Paris: Mouton and Co. Gusfield, Joseph R. 1967 "Tradition and modernity: misplace polarities in the study of social change." American Journal of Sociology 72 (January):531-562. Guttman, Louis 1954 "A new approach to factor analysis: the radex." Pages 258- 348 in Paul F. Lazarsfeld (ed.) Mathematical Thinking in the Social Sciences. Glencoe, Illinois: Free Press. Haber, Audrey and Richard P. Runyon 1969 General Statistics: Reading, Massachusetts: Addison-Wesley. Harbison, Frederick and Myers, Charles A. 1964 Education, Manpower, and Economic Growth: Strategies of Human Resource Development. New York: McGrawaHill. Hiese, David R. 1968 "Problems in path analysis and causal inference." Pages 38- 73 in Edgar F. Borgatta (ed.) Sociological Methodology 1969. San Francisco: Jossey—Bass. 1969 "Progress in Multivariate Causal Analysis." Paper presented at the meetings of the Midwestern Sociological Association, Indianapolis, Indiana, May, 1969. Hoselitz, Bert F. 1960 Sociological Aspects of Economic Growth. New York: Free Press of Glencoe. 146 Hoselitz, Bert F. 1961 "Tradition and economic growth." In Ralph Braibanti and Joseph Spengler, eds., Tradition, values, and Socio-Economic Development. Durham, N. C.: Duke University Press. Pp. 83-113. Hoselitz, Bert F. and Wilbert E. Moore, eds. 1963 Industrialization and Society. Paris: UNESCO and Mouton. Inkeles, Alex 1960 "Industrial man: The relation of status to experience, perception, and value." American Journal of Sociology 66 (July):1-31. 1966 "The modernization of man." Pages 138-150 in Myron Weiner (ed.) Modernization: The Dynamics of Growth. New YOrk: Basic Books. 1969a "Making men modern." American Journal of Sociology, 75 (September):208-225. 1969b "Participant citizenship in six developing countries." American Political Science Review 62 (December):1120-1141. Kahl, Joseph A. 1959 "Some social concomitants of industrialization and urbanization." Human Organization 18 (Summer):53—74. Kilpatrick, Franklin P. and Hadley Cantril 1960 "Self-anchoring scaling, a measure of individual's uni ue reality worlds." Journal of Individual Psychology 16 Novemben:158—173. Kumata, Kideya and Frederick B. Waisanen 1969 "Personality, mass media and development." Papers in International and WOrld Affairs (International Programs, Michigan State University) 1969 Series (April):51-72. Lazarsfeld, Paul F. 1955 "Interpretation of statistical relations as a research operation." Pages 115-125 in Paul F. Lazarsfeld and Morris Rosenberg (eds.), The Language of Social Research. Glencoe, Illinois: Free Press. Lenski, Gerhard 1970 Human Societies. New York: McGraweHill. Lerner, Daniel 1958 The Passing of Traditional Society: Modernizing the Middle East. New York: Free Press. Mander, John 1969 The Unrevolutionary Society: The Power of Latin American Conservatism in a Changing WOrld. New York: Knopf. 147 Mead, George Herbert 1934 "Self," in Anselm Strauss, ed., George Herbert Mead on Social Psychology. Chicago: University of Chicago Press, 1964:199-246. McClelland, David C. 1961 The Achieving Society. New York: D. van Nostrand. Moore, Wilbert E. 1961 "Social framework of economic development." In Ralph Braibanti and Joseph Spengler, eds., Tradition, values, and Socio-Economic Development. Durham, N. C.: Duke University Press. Pp. 57-82. Piaget, Jean 1932 "Social factors in moral development." From The Moral Judgment of the Child. New York: Harcourt, Brace, and Co. .Reprinted in: Theodore M. Newcomb and Eugene L. Hartley, eds., Readings in Social Psychology. New York: Henry Holt and Co. 1947:158—168. Przeworski, Adam and Henry Teune 1970 The Logic of Comparative Social Inquiry. New York: John Wiley and Sons. Reich, Charles A. 1971 The Greening of America. New York: Random House. Rogers, EWerett M with Lynne Svenning 1969 Modernization Among Peasants: The Impact of Communication. New York: Holt, Rinehart, and Winston. Roy, Prodipto, F. B. Waisanen, and E. M. Rogers 1969 The Impact of Communication on Rural Development: An Investigation in Costa Rica and India. Paris: UNESCO and Hyderabad: NICD. Rozak, Theodore 1969 The Making of a Counter Culture. Garden City, New York: Anchor Books. Sanders, William T. and Barbara Price 1968 Meso-America: The Evolution of a Civilization. New York: Random House. Schnaiberg, Allan 1970 "Measuring modernism: theoretical and empirical explorations," American Journal of Sociology, 76:399-425. 148 Seeman, Melvin 1959 "On the meaning of alienation." American Sociological Review 24 (December):783-791. Siegel, Sidney 1956 Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill. Simon, Herbert A. 1957 Models of Man. New York: John Wiley. Smith, David H., and Alex Inkeles 1966 "The OM Scale: a comparative socio-psychological measure of individual modernity." Sociometry 29 (December):353—377. Stephenson, John B. 1968 "Is everyone going modern? A critique and a suggestion for measuring modernity." American Journal of Sociology 74 (November):265—275. Thurstone, L. L. 1947 Multiple-Factor Analysis. Chicago: University of Chicago Press. Waisanen, Frederick B. 1963 "Stability, alienation, and change." The Sociological Quarterly 4 (Winter):18-31. 1964 "Educational aspirations, modernization and urban integra- tion." Translated from.America Latina 10 (4):3—21. 1969a Actors, Social Systems, and the Modernization Process. Bloomington: Carnegie Seminar, Department of Government, Indiana University, and Guillermo Briones. 1969b "Education, functional literacy and participation in development." A paper presented to the Annual Meeting of the Society for Applied Anthropology, Mexico City, April, 1969. 1971 "Education and participation in development." Acta Sociologica 14:183—192. Wright, Sewall 1934 "The method of path coefficients." Annals of Mathematical Statistics 5:161-215. HICHIGRN STRTE UNIV. I Him )1; IIIWINII ll HR 312 31049095 RARIES [WWI 22?