. ny’qnuer-fltn? wad}, a." w . 15.3A ,." ' ' 3:3»- 13:11.33?” i: "P‘?“‘ 4‘ L ~' ~ ‘ ‘ . A “‘- '. , 3v. ’4 4.3.... "‘ ‘ 7 ' “i" 3 , -; n. u‘ .r. ft‘L"_§E‘4.:.. ' .. ‘3’. u “1 ; "r.“ 1,” . 1.! " v1.2 4. A ‘3 a! 33:23.:- N :2‘ . at guimw' 4 w x . 3km \' $439333“... JV! dub] - :1: "4:3,: wfim” $2; 'b: New aft-N 45.. my , y. -l. "(Eb . “ fifid’gp u . _, _ ~ 5: .24.! y. 3 $112... 4W; ' i :7 h ‘,;—57' .m. m m :1. 3 u, .‘..;’2‘.‘L"..‘L . . .. 4 an“) ... .m- .‘xh’L «t.- I-u. 1743.5 ’33. . N ‘ b _.‘ 3. .u :w‘xhu .r .va“ N4? ”41...; , ”an 9w“; 1,373! A » “EXP-f; gait-1E5: fig. .33.. {MM 3:" *~ i'w “hint: 327$ .X .n' «.53: «I: . 4 ‘5 ' ' i 7&5;§§§§.‘$~ )- —" "‘ “‘35 “3‘ i:§gr gt." 4- ’fiV-yf {Eu-33;. £15.} % :u :ikv‘sih‘ ‘v "*3 437:.-.“ :Lwfit’i 7.. r «1733'. J28...“ 41.x“- Mum. any 5 . ‘7: 2' ’-‘ W"? Li“! ’ ~53} V. . An". .02.” . Hrm’a'vl'ylo. .7; '1‘; }r{)yn~-uc'r'7 ' :l .J, .1... ‘Aldw J’”jk.x‘4. H‘VJE g. .‘ UNIVERSITY LIBRARIES IUHHWMHWHWUII IIHJHHJ II 31293 This is to certify that the dissertation entitled Participation and Motivation in Adult Nonformal Education: A Social Systems Approach presented by Gollapalli Prakasa Rao has been accepted towards fulfillment of the requirements for Ph . D . degreein Educational Administration &©.¢t& Major professor Dme July 16, 1991 MS U it an Affirmative Action/Equal Opportunity Institution 0-12771 (k W LIBRARY Michigan State University in , ‘— Q PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE :9”? 0 .1-7» I .. . iv‘w c 119% p .. . 0‘ r is {were 3 0 9 W MSU Is An Affirmative Action/Equal Opportunity Institution c:\cltc\ddeou.pm3-p.t Behold, I come quickly; and my reward is with me, to give every man according as his work shall be. I am Alpha and Omega, the beginning and the end, the first and the last. -Lord Jesus Christ Holy Bible, Revelation 22:12,13. PARTICIPATION AND MOTIVATION IN ADULT NONFORHAL EDUCATION: A SOCIAL SYSTEHS APPROACH By Gollapalli Prakasa Rao A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Educational Administration 1991 ABSTRACT PARTICIPATION AND MOTIVATION IN ADULT NONFORMAL EDUCATION: A SOCIAL SYSTEMS APPROACH BY Gollapalli Prakasa Rao Research and experience have shown that adult nonformal education is an effective means for achieving community development. As such, further exploration of factors that contribute to its effectiveness is important. Participation and motivation have been most enduring research concerns in adult learning. Past studies have shown that socio- demographic factors such as age, sex, and income shed little light on participation and motivation. Therefore it is appropriate to try alternative approaches to the examination of participation and motivation. The purpose of this study was threefold: To examine the aggregate and individual impact on participation and motivation of selected sociodemographic attributes like age, sex, education of adults enrolled in nonformal education; to lay out a theoretical model to understand adult participation and motivation drawing upon Merton's(1957) theory of social structure and anomie: and to examine associations between learning orientations of adults and their social systems. Gollapalli Prakasa Rao The study was conducted in North India, drawing a random sample of 261 learners from adult nonformal education programs of selected voluntary agencies. Multiple linear regression, analysis of variance, and crosstabulation procedures were used as statistical tests. The study showed that the sociodemographic variables account for 7% of variance in participation which is lower than the 10% variance in participation accounted for in American studies. The same variables accounted for 13% of the variance in motivation. The study has suggested a theoretical model that may be called the Response Pattern Model, to enable a further understanding of participation/ motivation as a function of a learner's response pattern. The Response Pattern Model is based upon a learner's acceptance of the goals and means of an educational program. The study has also suggested that the social system to which an adult learner belongs influences his/her participation and motivation. The findings of this study may help meet the challenge of further developing theoretical grounding of adult education and to plan and operate adult education programs as 'community based programs' rather than mere 'learner based,’ so that adult education can become a more effective means for achieving community development. Copyright by GOLLAPALLI PRAKASA RAO 1991 Dedicated to Gnanamma and Gollapalli Elias, my mother and father, whose faithful Christian testimony, prayers and support provided a model for my life's journey ACKNOWLEDGMENTS I wish to express my genuine gratitude to all those who have contributed to the completion of this study: First, to Lord Jesus Christ, for His never failing love and grace which enabled me to complete this study. To Dr. S. Joseph Levine, Committee Chairman, for the guidance and encouragement he provided throughout my graduate studies. To the members of my doctoral committee, Dr. Frank Fear, Dr. James E. Snoddy, and Dr. Sheldon Cherney, for their advice and assistance. To Dr. Alemu Beeftu and to Dr. Kenneth Harding for their support during my graduate studies and preparation of the dissertation proposal. To Dr. Frederick Ignatovich for his time and expert advice which helped me to develop a conceptual frame for this study. To Dr. M.v.d. Bogaert s.j. and to Major J.K. Michael for their prayers and support that I received during the time of data collection in India. To my fellow Christians at the Little Flock Christian Fellowship for their prayers. To my wife, Raj, daughter, Lisa, and sons, Esli and Ashish, for their love and support. vi TABLE OF CONTENTS LIST OF TABLES O O O O O O O O O O O LIST OF FIGURES . . . . . . . . . . . LIST OF APPENDICES . . . . . . . . . . . Chapter I. INTRODUCTION . . . . . . . . . . Need for the study . . . . . . The problem . . . . . . . . . The purpose . . . . . . . . . Theoretical foundations . . . . Hypotheses . . . . . . . . . Definitions . . . . . . . . . Organization of the dissertation II. LITERATURE REVIEW . . . . . . . . Adult Education Theory . . . . . Nonformal Education . . . . Motivation in Adult Education . Participation in Adult Education Merton's Social Structure and Anomie vii 16 17 21 22 22 23 25 29 32 III. METHODOLOGY . . . . . . . . . . . . . . . . 35 The Research Setting . . . . . . . . . . 35 The Research Design . . . . . . . . . . . . 35 The Population . . . . . . . . . . . . . . 37 The Sample . . . . . . . . . . . . . . . 39 Data Collection . . . . . . . . . . . . . . 42 Instrumentation . . . . . . . . . . . . . . 43 The Variables . . . . . . . . . . . . . . . 50 Scoring the Data . . . . . . . . . . . . . . 52 Statistical Procedures Used . . . . . . . . 54 Qualitative Study Design . . . . . . . . . . 59 IV. DATA ANALYSIS . . . . . . . . . . . . . . . . 60 Introduction . . . . . . . . . . . . . . . 60 Characteristics of the Sample . . . . . . . 61 Computed Variables . . . . . . . . . . . . 67 Descriptive Statistics . . . . . . . . . . 70 Statistical Procedures for Testing Hypotheses 83 Tests of Hypotheses . . . . . . . . . . . . 85 Hypothesis # 1 . . . . . . . . . . . . 85 Hypothesis # 2 . . . . . . . . . . . . 91 Hypotheses # 3, 4, and 5 . . . . . . . 94 Hypotheses # 6(a), 6(b) . . . . . . . . 103 Results of Additional Statistical Analyses 118 viii V. SUMMARY AND DISCUSSION . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . Findings . . . . . . . . . . . . . . . Hypothesis # 1 . . . . . . . . . . . . Hypothesis # 2 . . . . . . . . . . . . Hypotheses # 3, 4, and 5 . . . . . . . Hypotheses # 6 . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . Relationship between personal variables and participation/motivation . Relationship between participation and motivation . . . . . . . . . . . . . . A theoretical frame for explaining participation and motivation . . . . . Influence of learner's community/social system on adult learner's participation and motivation . . . . . . . . . . . Implications . . . . . . . . . . . . . . To Theory . . . . . . . . . . . . . . . To Practice . . . . . . . . . . . . . . Limitations . . . . . . . . . . . . . . . . Recommendations for Further Study APPENDICES BIBLIOGRAPHY ix 123 123 129 129 137 138 140 141 142 144 144 147 148 148 151 152 153 LIST OF TABLES Table Page 1. District-wise distribution of subjects . . . . 41 2. Reliability coefficients . . . . . . . . . . . 49 3. Distribution by district . . . . . . . . . . . 61 4. Distribution by sex . . . . . . . . . . . . . . 62 5. Distribution by caste . . . . . . . . . . . . . 63 6. Distribution by occupation . . . . . . . . . . 63 7. Distribution by formal education . . . . . . . 65 8. Distribution by type of training . . . . . . . 66 9. Distribution by age group . . . . . . . . . . . 67 10. Statistics for the variables . . . . . . . . . 71 11. Participation and motivation by economic levels. 74 12. Participation and motivation means for unmarried and married participants . . . . . . 75 13. Participation and motivation means for urban rural participants by sex . . . . . . . . . . 76 14. Participation and motivation means for urban and rural participants . . . . . . . . . . . . 77 15. Means for education, participation, and motivation by sex and by state: . . . . . . . . . . . . . . 78 16. Participation and motivation means by type of nonformal education programs . . . . . . . . . 79 17. Participation and motivation means for different occupational groups . . . . . . . . . 80 18. Participation and motivation means by caste . . 81 19. Participation and motivation means for different educational levels . . . . . . . . . . . . . . 82 20. Results of multiple regression with participation as the dependent variable: (Equation Number 1) . 86 21. Results of multiple regression with motivation as the dependent variable: (Equation Number 2) . 88 22. Analysis of variance for participation by response pattern . . . . . . . . . . . . . . . 91 23. Participation means for response patterns . . . . 91 24. Analysis of variance for motivation by response pattern 0 O O O O O O O O O O O O O O O O O O O 9 3 25. Motivation means for response patterns . . . . . 93 26. Results of cross tabulation: participation levels by response pattern . . . . . . . . . . 96 27. Results of cross tabulation: motivation by response pattern . . . . . . . . . . . . . . . 97 28. Analysis of variance for caste categories and participation . . . . . . . . . . . . . . . . . 119 29. Analysis of variance for caste categories and motivation O O O O O O O O O O O O O O O O O O 1 19 30. Analysis of variance for the five occupational categories and participation . . . . . . . . . 120 31. Analysis of Variance for the five occupational categories and motivation . . . . . . . . . . . 120 32. Results of t-tests . . . . . . . . . . . . . . 121 xi LIST OF FIGURES Figure 1. Modes of adaptation and response patterns . . 2. Assumed definitions of modes of adaptation in nonformal education context . . . . . . . . . 3. Assumed definitions of response patterns in nonformal education context. . . . . 4. Relationships between participation/motivation and goal(NFE) and means(NFE) acceptance levels 5. Diagrammatic representation of relationships between participation/motivation and the levels of goal(NFE) and means(NFE) . . . . . 6. Diagrammatic representation of relationships between community's and learner's response patters . . . . . . . . . . . . . . . . . . 7. Variables used in the study . . . . . . . . 8. Descriptors of correlations . . . . . . . . . 9. Computation of response patterns . . . . . . . 10. Dummy variables in the regression analysis . . 11. Pearson's correlation matrix for the variables Page 10 ll 12 15 51 55 68 70 72 LIST OF APPENDICES Appendix I. Letter from Michigan State University II. Letter to the Respondents . . . . . . III. Measurement of Motivation . . . . . . IV. Measurement of Goal(NFE) Acceptance . V. Measurement of Means(NFE) acceptance . VI. Personal Data . . . . . . . . . . . . VII. Data Transfer Sheet . . . . . . . . . xiii Page 155 156 157 158 160 161 163 CHAPTER I INTRODUCTION Heed for the Study Literature and experience show that rural development implies change at the community level. Acquisition of knowledge and skills by the people of the community through a participatory process are critically important if change is to occur and if the process and gains of change are to be sustained. There are two avenues for this: formal and nonformal education. Studies of Coombs (1974) have shown that nonformal education for adults is a far more effective means as it pertains to enhancement of knowledge and skills within the actual context of the community besides having the virtue of providing participatory learning possibilities. Coombs (1974) observed that, "for combating poverty, nonformal education is one of the essential weapons to be used." It is therefore imperative that developing countries give importance to nonformal education in their development plans. From the early 19705 most developing countries, particularly India, have expanded this sector of education hoping it would accelerate development. 2 Since 'nonformal education'(NFE) for adults is so important for development, it is essential that we continue to explore factors that contribute to its effectiveness. Ih§_2£921§! Two of the variables that have, so far, been looked at for the purpose of explaining differential levels of effectiveness of adult nonformal education are participation and motivation. These two variables have been most enduring research concerns since investigations of adult learning began and continue to occupy the attention of current research. The problem is, past studies had conceptualized and measured these two variables mostly in terms of sociodemographic variables such as age, sex, income level with participants drawn “mostly from middle-class or working-class Americans. However, those studies have explained only about 10% of the variance in participation (Cross, 1983). This suggests that past research approaches are insufficient for understanding participation and motivation; also, we cannot claim generalizability for theories and concepts of adult learning derived from those studies. The real 3 explanation for participation may lie elsewhere and the sociodemographic factors relevant to adults in developing countries may explain participation and motivation differently. The following two questions arise at this point: 1. What percentage of variance in learner participation and motivation could be attributed to sociodemographic variables of adults in less developed countries like India? 2. What alternative view could better explain participation and motivation of adult learners? Who people are is sometimes less important than the cultural surroundings in which they operate. It is necessary that a conceptual and a theoretical approach is taken to understand adult participation and motivation. Merton's (1957) paradigm "social structure and anomie", which is firmly based on concepts relating to social systems and their cultures, appears to provide an alternative approach to understanding adult motivation and participation in nonformal education programs operating in social systems. We also know that man is always a part of a larger system and only in this wider context can his conduct be 4 understood (Young 1956). In other words, man in his social system, interacts with his fellow members and seeks social reinforcement for his behavior (Malinowski, 1945). For understanding participation/motivation it appears logical to look at a community as a social system and the adult learner as a part of it, sharing culturally approved goals and values of the community. Ih§_EEIDQ§§ The purpose of this study was to: 1. examine relationships of selected sociodemographic variables such as age, sex, income level and their aggregate and individual impact on participation and motivation of adult learners in nonformal education programs in North India. 2(a).1ay out a theoretical model to enable a further understanding of adult participation/motivation drawing upon Robert Merton's(1957) theory of social structure and anomie. 2(b).examine ways that the social and cultural context may influence the learning orientation of adults. 5 Thooretioal Foundations Inoozotical fouodotion relotive to oozoose (1) Most research relating to adult learning has been done in the industrialized countries. Brookfield (1986) states that: in research in adult learning, the adults who form the sampling frames are for the most part ethnically homogeneous; that is, they are Caucasian Americans. They are drawn chiefly from middle-class or upwardly mobile working-class families.... how can we then write confidently of adult learning style in any generic sense when we know little (other than anecdotally) of the cognitive operations of, for example, Asian peasants, African tribes people, or Chinese cooperative laborers? Anderson and Darkenwald (1979) found through a study of about 9,000 American adults, 17 years of age or older who were participants of organized adult education, that sociodemographic variables of adults explain only 10 percent of the variance associated with participation. They say that such variables play a relatively modest role in influencing participation. Their finding however, cannot be generalized as their sample was drawn from the 6 American learners only and therefore, studies for understanding participation/motivation in developing countries using different strategies, including the use of sociodemographic variables relevant to adults in those countries as a strategy, are necessary. e t'c l t' 'v o o Sociologists tell us that culture is a dominating factor in human life. Horton (1965) suggests that, "from before one is born until after he is dead, his life is circumscribed by his culture.... his culture defines the meaning of situations and thereby indicates the behavior expected." Therefore, as Young (1956) suggests, "it would seem completely logical that examining the cultural environment itself could make explicit the learning process .... and learning must be viewed as an aspect of the cultural milieu." However, culture covers a host of variables. Merton (1957) suggests that, "among the several elements of social and cultural structures, two are of immediate importance... the first consists of culturally defined goals held out as legitimate objectives for all or diversely located members... the second, the acceptable modes (means) of reaching out for those goals." This 7 implies that goals and means are two important factors in cultural milieu which deserve serious consideration for understanding human behavior. Based on these two factors of culture, Merton developed his theory 'Social Structure and Anomie' in an effort to explain human behavior with respect to social systems. Merton's theory suggests that individuals in a social system associate themselves to prescribed goals and available means in definite patterns which Merton (1957) calls rosoonse oattozns. He suggests that these response patterns correspond to a typology of modes of adaptation or modes of behavior. Merton's suggested modes and the corresponding response patterns are shown in Figure 1. Modes of Acceptance of Acceptance of Response Adaptation goal (G) means (M) pattern(RP) 533235112; """" 31;; """""" 112;; """"""" 5:31. Retreatism Low Low G- M- Innovation' High Low G+ M- Ritualism Low High G- M+ Figure 1. Modes of adaptation and response patterns. Conformity, according to Merton refers to the behavior of a person which is consistent with the expectations of 8 his/her society which has prescribed the goals and the means. Innovation supposes that the individual abandons institutional means while retaining the aspiration to successfully pursue the prescribed goals. In this mode his/her concentration is on finding an alternative means to attain the same goals. In the Ritualism mode, the individual abandons the goal but clings to the routines and other institutional norms in order to escape possible dangers which he/she might face by not following those norms and routines. Retreatism refers to the behavior of an individual who is in the society but not seriously involved in it's activities. Here, the individual has relinquished the prescribed goals and his/her behavior does not accord with the institutional norms. To apply Merton's paradigm to adult nonformal education we must view a nonformal education program as having two distinct components; first being the goal the program offers to its learners and second, the institution itself which serves as an available means to achieve that goal. In this study, a goal is viewed as the professional and social status or vocation an adult learner could attain by successfully participating in a nonformal education program, where participation is assumed to be the same as learner's class attendance. For example, a welding 9 training program could enable a learner who participates in it to become a welder. 'One can then say, the welding program offers a 'welder status' as a goal to its I participants. The teachers and staff who operate the program and the equipment used are assumed as a means available to the learners to attain that goal. In this study, goal(NFE) stands for the goal offered by a nonformal education program and means(NFE) stands for the institution that operates that program, as explained above. These two terms carry the same meanings wherever they appear in this study. It is assumed that Merton's modes of adaptation and the response patterns in adult nonformal education context may be defined as outlined in Figure 2. Mode Assumed definition Conformity: High levels of participation and motivation. Innovation and Medium levels of participation Ritualism: and motivation Retreatism: Low levels of participation and motivation. Figure 2. Assumed definitions of modes of adaptations in nonformal ed. context. 10 In a nonformal education setting, if learner's acceptance of goal(NFE) and means(NFE) are measured quantitatively by suitable scales, Merton's response patterns may be defined as outlined in Figure 3. Response Patterns: Definitions: ( G+ M+ ) .. Learner has high levels of acceptance of goal(NFE) and means(NFE). ( G+ M- ) .. Learner's acceptance of goal(NFE) is high and acceptance of means(NFE) is low. ( G- M+ ) .. Learner's acceptance of goal(NFE) is low and acceptance of means(NFE) is high ( G- M- ) .. Learner has low levels of acceptance of goal(NFE) and means(NFE). Figure 3. Assumed definitions of response patterns in nonformal education context: Merton's suggestion that these four response patterns correspond with the four modes of adaptation is fundamental to this study. Based on the above definitions of response patterns and modes of adaptations given in 11 Figures 2 and 3, four relationships are identified between participation /motivation and the levels of acceptance of goal(NFE) and means(NFE): these relationships are shown in Figure 4. Level of participation/ Definition of response pattern motivation in nonformal education context High level of participation/ High goal(NFE) motivation and high means(NFE) acceptance Medium levels of participation/ High goal(NFE) motivation acceptance and low means(NFE) acceptance Low goal(NFE) acceptance and high means(NFE) acceptance Low levels of participation/ Low levels of goal(NFE) motivation and means(NFE) Figure 4. Relationships between participation/motivation and goal(NFE) and means(NFE) acceptance levels. A diagrammatic representation of this construct is given in Figure 5. 12 Figure 5. Diagrammatic representation of relationships between participation/motivation and the levels of goal(NFE) and means(NFE) In Figure 5, the Circle 1 represents the goal(NFE), Circle 2 the means(NFE), and Circle 3 the learner. Through a process of examination of the goal(NFE) offered and the means(NFE) made available by a nonformal education program, the learner internal to himself/herself will form a level of acceptance for each of these factors, which in this study are termed goal(NFE) acceptance and means(NFE) acceptance. The process of examination may be 13 influenced by factors internal and external to the learner. Taken together, these two form a response pattern (Figure 5, Circle 4). If all the possible levels are categorized into two groups, low and high corresponding to Mertons 'rejection' and 'acceptance', we come up with only four different combinations. Merton suggests that each combination corresponds to typology of modes of adaptation or a mode of behavior (Figure 5, Circle 5). It is hypothesized in this study that each mode of adaptation corresponds to a low or a medium or a high participation or motivation of the learner in adult nonformal education (Figure 5, Circle 6). Theoretical foundation relativo to purpose 2(b) When a nonformal education program operates within or near a village, different culturally homogeneous communities in that village are likely to examine the goa1s(NFE) and the means(NFE) offered by that program in the light of their own culturally valued and accepted goals and means and adopt different response patterns vis- a-vis the nonformal education program. Such community response patterns, in turn, are likely to influence formation of response pattern of the learners of that 14 community because, as Horton (1965) suggests, "a young person looking for a career may feel he has a free choice, but he will choose one of the lines of work his culture has taught him to value." Young (1956) suggests that, "man is always a part of a larger system and only in this wider context can his conduct be understood." Along the same lines, Malinowski (1945) says that, "humans live not by psychological drives alone but by physiological drives molded and modified by the conditions of culture and man never deals with his difficulties alone." In other words, man in his social system, interacts with his fellow members and seeks social reinforcement for his behavior. In that process, his perception and attitude towards an educational goal and institution may be influenced by the perceptions and the attitudes of the members of that social system of which he is a part. Brookfield (1986) thinks that nonparticipation is a function of a cluster of cultural attitudes. Therefore, variations in a community's responses to a goal(NFE) and means(NFE) may, through the channel of culture, correspondingly influence and alter an individual learner's level of acceptance of those components of NFE. It is logically consistent to assume here that a learner's response to a goal(NFE) and means(NFE) may be closely related to the corresponding responses of his/her community. 15 A diagrammatic representation of this construct is shown in Figure 6. Fig. 6 Diagrammatic representation of relationships between community's and learner's response patters. In Figure 6, Circle 1 represents the goal(NFE) and Circle 2 the means(NFE). Through a process of examination of the goal(NFE) offered and the means(NFE) made available by a nonformal education program, the community (Circle 3) though a process of interaction among the members is likely to come up with a level of acceptance for each of 16 these two factors. This study has hypothesized that through the channel of culture, the response pattern of the community is likely to influence the process of formation of learner's (Circle 4) response pattern in such a way that finally the two patterns will have a close correspondence. W The following hypotheses were constructed on the basis of the above theoretical orientations. 1. There is a linear relationship between participation /motivation and the set of sociodemographic variables: age, sex, caste, education, area of residence, marital status, economic status, media exposure, health, and occupation. ,2. Adult learners differ in their participation and motivation with respect to their response patterns. 3(a).High level participation in adult learning has correspondence with response pattern G+M+. 3(b).High level motivation in adult learning has correspondence with response pattern G+M+. 17 4(a).Medium level participation in adult learning has correspondence with response patterns G-M+ and G+M-. 4(b).Medium level motivation in adult learning has correspondence with response patterns G-M+ and G+M-. 5(a).Low level participation in adult learning has correspondence with response pattern G-M-. 5(b).Low level motivation in adult learning has correspondence with response pattern G-M-. 6(a).Participant's acceptance of goal(NFE) has positive correspondence with his/her community's acceptance of that goal(NFE). 6(b).Participant's acceptance of means(NFE) has positive correspondence with his/her community's acceptance of means(NFE). Qofinitions Nonforpal Eoooatiop Adopting Coomb's (1974) definition, NPR is any organized educational activity outside the graded formal system of 18 education whether operating separately or as an important part of a broad activity, that is intended to serve an identifiable clientele and objectives. Adult The definition adopted in this study is that an adult is an individual who has assumed some kind of responsibility for himself/herself and may be for his/her immediate family and or for others, and is at least 18 years old. Paraguay); A participant is an adult enrolled as a learner in a nonformal education program. 2 !' . !.O In this study participation is limited to a learner being able to attend all the classes or learning sessions organized by a nonformal education program. In this study, participation is measured in terms of class attendance. mm In this study, a goal is viewed as the professional and social status or vocation an adult learner could attain by successfully participating in a nonformal education 19 program, where participation is assumed to be the same as the learner's class attendance. M). In this study, goal(NFE) represents nonformal education goal defined above. Nonfornal education institution In this study 'nonformal education means' refers to the institution or the school that organizes and operates a nonformal education program. It includes the teaching staff, the facilities, and the equipment used. mm In this study, means(NFE) stands for the nonformal education institution or school. Acceptance of goal(NFE) Acceptance of goal(NFE) is defined as the extent to which a learner is satisfied with the goal offered by a nonformal education program where both goal and nonformal education program conform to the above definitions. Aooepranoo of means(NFE) In this study, acceptance of means(NFE) is defined as the extent to which a learner is satisfied with the means, 20 that is with the teachers, the staff, the facilities, and the equipment which play a significant role in his/her realization of the goal(NFE). Social System: A community is visualized as a social system. People of a caste or a sub-caste living is a specific geographical area is considered a community. 11211116111211 Motivation in this study is strictly intrinsic. It is defined as the learner's desire to effectively pursue a learning goal(NFE) where goal(NFE) conforms to the definition given above. Dropout: One who has discontinued attending a learning program. ociodemo a h' va ' es In this study, age, sex, caste, education, area of residence, marital status, economic status, media exposure, health and occupation are considered as socioeconomic variables relevant to adult learners participating in nonformal education programs in India. 21 Organization of the Dissertation Chapter I presents an introduction of the study. It includes need for the study, the problem, purpose, hypotheses and relevant theoretical foundations. Chapter II has a review of literature pertaining to adult nonformal education, participation, motivation, and Merton's (1957) theory about social structure and anomie. Chapter III covers methodology that includes, research design, a brief description of the population, sampling design, instrumentation, and statistical procedures used. Chapter IV presents data analyses. It covers a brief description of the sample, important descriptive statistics, and a detailed description of the statistical analyses along with interpretations. Chapter V presents summary and discussion which includes a brief description of the research problem and the research procedures used, discussion of the findings, conclusion, implications to theory and practice, and recommendations for future research. CHAPTER II LITERATURE REVIEW Adult Education Ineory Adult education reaches back to ancient times. Indeed the great social movements that produced western civilization were all founded on the conviction that adult mind could learn and change (Robinson, 1989). In fact, it is such a conviction that was responsible for all the great changes in India and most other countries.of the world. In the last twenty five years a number of respected theorists have made attempts to identify generalizable principles of adult learning in their quest to build a theory of adult learning that would aid practice (Brookfield, 1986). Boshier (1971) thinks that, "adult education is a conceptual desert." Cross (1983) seems to support Boshier's blunt statement saying that, "one of the most under-utilized vehicles for understanding various aspects of adult learning is theory .... articles dealing with theory are still rare in adult education." But, none as Cross (1983) states, "would lay claim to a fully developed theory regarding participation in adult 22 23 education." Though it looks unlikely there will ever be one single theory of adult education, there is need for adult educators to continue working towards building theories that would strengthen adult education as a conceptually rich field. Nonformal Eoncation Robinson (1989) points out three common uses of the term "adult education": (1). A field of study - An academic discipline; (2). A field of professional practice - Institutions and agencies; and (3). An adult activity - A part of every adult life. Adult education as an activity which is a part of every adult life generally includes unorganized activities, whereas the other two types of adult education listed above, represent organized educational activities. Adult nonformal education, an organized educational activity therefore, can be rightly put under the second type of adult education, 'a field of professional practice'. Adopting Coombs (1974) definition, nonformal education (NFE) is any organized educational activity outside the 24 graded formal system of education whether operating separately or as an important part of a broad activity, that is intended to serve an identifiable clientele and objectives. Adult nonformal education is considered an effective means for achieving development, particularly in underdeveloped countries. Experience has shown, as Coombs(1974) tells us that, "for combating poverty, nonformal education is one of the essential weapons to be used." As such, nonformal education is an appropriate and essential component of micro social change efforts. Nyrere (1976) talking about development says, "if adult education is to contribute to development it must be a part of life, integrated with life and inseparable from it. It is not something which can be put into a box and taken out for certain periods of the day....it cannot be imposed; every learner is ultimately a volunteer, because however much teaching is given, only he can learn." The kind of attributes that Nyrere has attached to the adult education that can contribute to development can be found only in nonformal education and hence its importance for rural development. Realizing the effectiveness of adult nonformal education, since 1970s developing countries, particularly India, have 25 expanded this sector of education with the hope of accelerating rural development. Vocational education and literacy are two main types of adult nonformal education programs that have been taken up in these countries. However, as Korten (1984) points out, "one of the clear lessons of the 1970s has been that effective participation of the rural poor in the development process is more easily mandated in programming documents than achieved in the real world of program implementation." Freire's (1983) experience at Recefe where, out of a group of five illiterate participants two dropped out on the third day of the program, also supports this view. As long as nonformal education continues to be an important mode of rural development process, efforts to understand factors that can promote it's effectiveness remain important for all those concerned with rural development. Motivation in Adult Education The three well known views on motivation are summarized by Robinson (1989) as follows: 1. Behavior impelled by inner urges (Freud); 2. behavior impelled by outer-situational forces (Povlov- Skinner); and 26 3. behavior impelled by inherent capacity for growth (Carl Rogers and other third force psychologists). An adult's motivation could conform to any one or a combination of the above three views. The adult nonformal education in the form of vocational training, literacy etc. confirms to the motivation seen as behavior impelled by outer-situational forces. What ever view one might take, it is important that we understand motivation which appears to be a dominant factor in effective adult education. Cross (1983) suggests that, " the research methods for seeking answers to the motivation of adult learners fall into four basic designs: (1) depth interviews, (2) statistical analysis of motivational scales, (3) survey questionnaires, and (4) hypothesis testing." Houle (1961), Tough (1968), Carp et al. (1974), Rubenson (1977), Vroom (1964), and Aslanian and Brickell (1983) are some of the prominent researchers who attempted to study motivation using the above methods. 0n the basis of a careful review of their studies, Cross came to the conclusion that the question of why adults choose to participate in various kinds of learning activities has not been answered definitely by any of the four major methods mentioned above. 27 There are four well known theoretical works developed as models to explain motivation of adults in learning activities. These four as explained by Cross (1983) are briefly stated below: 1. Expectancy-valency paradigm of Kjell Rubenson (1977) states that the individual always chooses the course of action that he expects will lead to the greatest degree of pleasure or which will produce the smallest degree of pain. Vroom (1964) proposed an explanation for adult motivation by drawing a comparison between "work" and "education". He suggests that, "education like work, appears to be an achievement oriented activity, meaning that people who want to get ahead will put effort into personal achievement in school or on the job." Roger Boshier (1973) proposes that, "both adult education participation and dropout can be understood to occur as a function of the magnitude of the discrepancy between the participants' self—concept and key aspects (largely people) of the educational environment...; non participants manifest self/institution incongruence and so do not enroll." 28 4. Alan Tough (1979) suggests that anticipated benefits play an important role in forming motivation. 5. Miller's Force-Field theory based on sociological research suggests that positive and negative forces existing in the individual and in the environment when combined form a resultant motivational force. While most of the other theories are based on psychological and physiological principles, Miller's theory draws from sociological research. It is to be noted that all of the above have used factors that relate to participant's environment and some psychological variables like self-concept. It is only Miller who went slightly into the area of sociology and based his analysis on compatibility of the values of the various social classes with the values of the educational system. One limitation of these studies however, is that all of them have dealt with organized school classes in which adults participate for credit or job promotion towards which either their employer or themselves had paid a fee. These dimensions build considerable pressure on the participants which develops a possibility of rendering participation more a subjective and mechanical behavior than a voluntary behavior tied up with genuine inner motivation. 29 ici on dult du tion Research points out that two most important factors that are associated with effectiveness in adult education appear to be motivation and participation. Past studies have also shown that there is a difference between those who participate and those who dropout, in terms of personal attributes or positional factors like age, sex, income, formal education etc. However, most of these studies were conducted in urban environments of industrialized countries. Brookfield (1986) suggests that, "in research in adult learning, the adults who form the sampling frames are for the most part ethnically homogeneous; that is , they are Caucasian Americans. They are drawn from chiefly middle class families.... how can we then write confidently of adult learning style in any generic sense when we know little, other than anecdotally, of the cognitive operations of, for example, Asian peasants, African tribes people, or Chinese cooperative laborers?" As such, it remains to be seen if these findings hold good for rural adults of less developed countries where people are, socially, culturally, and economically, not comparable to the people in those countries. 30 Research literature in respect to adult participation in learning activities is steadily growing in volume: but a significant part of it is concerned mostly with describing the characteristics that apply to adult learners, and is in the shape of "clientele analysis" as Knox(1965) puts it. To date, most research account for only 10% of the variance associated with adult participation in organized learning activities and therefore, the real explanation for participation appears to lie elsewhere as suggested by Cross (1983). As such, it is logically consistent to assume that there is a need to look into factors from other areas of knowledge for gaining further understanding of adult motivation and participation. Roger Boshier (1972) suggests that, "it is well known to adult educators that dropout of adult learners from organized educational programs is a continuous problem." Hollenbeck (1965) points out that, "the problem is international." Dickenson, Gary et. al (1967) found through their study that, "the initial enrolment is rarely maintained." This problem is acute in adult nonformal education programs in rural India. The problem remains unsolved despite persistent efforts of the government and private organizations. 31 The theories so far put forward in an attempt to understand this problem appear to have fallen short of the need. Cross (1983) suggests that, "the answer to the question of why adults participate in learning activities will probably never be answered by any simple formula. Motives differ for different groups of learners, at different stages of life, and most individuals have not one but multiple reasons for learning." Verner and Newberry (1958) further clarify this complex issue saying that, "the decision to participate or not rests with the individual adult and is influenced by factors that are components of both the personality and social group life of the individual involved. Such personal independence in the matter of participation complicates the administrative functions, frustrates the adult educator and inhibits achievement of the maximum effective role of adult education in society." It is possible that adults' learning orientations are rooted in their cultures. Brookfield (1986) thinks that nonparticipation is a function of a cluster of cultural attitudes. As such, it may be necessary for research in adult education to direct its focus on to understanding communities and cultures of adults as a necessary step in understanding their learning orientations. 32 Merron's Theory of Social Structure and Anomio Merton (1957) suggests that, among the several elements of social and cultural structures, two are of immediate importance. These are analytically separable although they merge in concrete situations. The first consists of culturally defined GOALS/purpose, and interests held out as legitimate to objectives for all or for diversely located members. The second element of the cultural structure defines, regulates and controls the acceptable modes (or MEANS) of reaching out for these goals. Every social group invariably couples its cultural objectives with regulations, rooted in the mores or institutions, of allowable procedures for moving toward these objectives ...the cultural emphasis placed upon certain goals varies independently of the degree of emphasis placed upon institutionalized means. Merton's theory suggests that, the individuals in a social structure adapt certain role behaviors corresponding to their patters of responses explained above. His suggested adaptations are as follows: Eatrorn Moog of aoaptation (G+ M+) Conformity (6+ M-) Innovation (G- M+) Ritualism (G- M-) Retreatism (G: M1) Rebellion Where the (+) signifies "acceptance" or "internalization" and (-) signifies "rejection"; and (i) signifies "rejection of the prevailing goals/means and substitution of new goals/means". (In this study, the last mode of adaptation is not considered) 33 Merton defines the adaptation as follows: Conformity refers to the behavior of a person which is consistent with the culturally conditioned expectations of his society. Innovation supposes that the individual has been imperfectly socialized so that he/she abandons institutional means while retaining the aspiration to successfully pursue the prescribed goals. In the ritualism mode, the individual abandons the goal but clings to the routines and other institutional norms in order to escape possible dangers which he/she might face by not following those norms and routines. Retreatism refers to the behavior of an individual who is in a society but not seriously involved in it's activities. Here, the individual has relinquished the prescribed goals and his/her behavior does not accord with institutional norms. Rebellion reflects the behavior of the individual who takes an arbitrary stand in respect of the goals and the corresponding means prescribed by the society; he/she seeks to define new goals and new means. Merton thinks that when the prescribed goals and the available means are not well coordinated with respect to any member in the society, there is the possibility of his deviating from the expected behavior. In Merton's words, the reason for such a behavior is that, "disjunction between the prescribed goals and the available opportunities (or 34 means) to realize those goals exerts pressure and produces anomie, a condition of relative normlessness leading to aberrant behavior." Conversely, aberrant behavior may be regarded sociologically as a symptom of disassociation between the culturally prescribed goals and socially structured avenues (means) for realizing those aspirations. Thus anomia is centered on pressure created by the discrepancy between culturally induced goals and socially structured opportunities. But, in the case of conformity and retreatism modes, anomie does not exist. It is suggested by Merton that this paradigm is applicable to all social systems. Since nonformal education programs, particularly in developing countries, operate in clearly identifiable social systems like villages, it appears that Merton's theory could be applied to educational systems operating in them, for understanding adult learning orientations in terms of participation and motivation. CHAPTER III METHODOLOGY The s in India was chosen for this study because of the emphasis India has given to adult nonformal education in rural development programs during the past three decades. Among the several Indian states considered economically backward, Bihar, Orissa, and West Bengal had a large percentage of rural poor. Several private voluntary and governmental agencies had initiated adult nonformal education programs for development of poor communities in these states. For these reasons these three states were chosen for this study. Miss The design of the study was exploratory. Reid and Smith (1989) state that, "research has an exploratory purpose when it is used to gain preliminary understanding of phenomena or to stimulate the development of concepts, hypotheses, and technology." According to Babbie (1989), 35 36 "exploratory studies are essential whenever a researcher is breaking new ground, and they can almost always yield new inputs into a topic for research. Exploratory studies are also a source of a grounded theory." Since this study aims at exploring the possibility of finding a model that can explain participation and motivation based on Merton's (1957) paradigm in sociology, exploratory design was considered appropriate. The methodological approach the study adopted was basically quantitative but was supplemented by a qualitative approach. The rationale for this dual approach was based on Bolster's (1984) observation that, "a fundamental requirement of any inquiry (in sociology) is that peOple must be considered as both the creators and the products of the social situations in which they live...and people act on the basis of meanings events have for them." An effective way to meet this requirement was to follow a strategy that employed a quantitative approach supplemented by a qualitative approach. According to Vidich and Shaprio (1955), "such a combination will allow us to gain a more complex theoretical picture of the variables under study". Here, quantitative approach meant gathering of information relevant to the variables predetermined and measuring it in units suitable for 37 statistical treatment; and qualitative approach meant participant observation and unstructured interviews for the purpose of examining and describing social institutions and cultural events in the community. tio The target population for this study was the adults from the economically and socially backward communities engaged in organized nonformal education projects at various centers operated by nongovernmental agencies conducting rural development programs in India. For obvious difficulties in sampling from such a vast population an experimentally accessible population was taken as the focus of this study. The experimentally accessible population was comprised of adults from economically and socially backward communities engaged in organized nonformal education projects operated by nongovernmental agencies engaged in rural development programs for at least seven years, in the states of Bihar, Orissa, and West Bengal in North India. India, located in Southern Asia, is bordered by Pakistan to the northwest, China, Tibet, Nepal, and Bhutan to the north, Burma to the east, and the Indian ocean to the 38 southeast, south, and southwest. It has 22 states and 9 union territories. About 80% of the Indian population, which is about 800 million, live in rural areas. Hindus account for 83%, Muslims 11%, Christians 3%, Sikhs 2% and others 1%. Female population is about 46.7% of the total (Morgan, 1983). It is estimated that about 40% of the population live below the poverty line with most of them being rural people. Almost all of the cities of India have slums wherein live mostly the village poor who have migrated to the cities with a hope of making a living. An integral part of the Hindu religion is an elaborate network of fixed social relationships that is known as the 'caste system'. The English word caste corresponds more or less to what is locally referred to as 'verna'; a sub-caste corresponds to 'jati' (Jayaraman, 1981). There are four vernas: Brahmin, Kshatriya, Vaishya, and Sudra. Traditionally, as Jayaraman states, castes are exclusive groups, distinguished from one another by name, rules of endogamy, and association with an occupation. Castes are birth status groups (Atal, 1968). A fifth category, that of 'untouchables', has been regarded as outside the verna system altogether. Dube (1958) states that caste divisions are regarded as divinely ordained and hierarchically graded. Traditionally, this fifth category 39 of people was required to stay outside the village and do the work that was polluting or ritually unclean. This inequity, according to Hindu doctrine was considered justified by the good and bad deeds done in one's former lives, which reward or punish the doer in this life (Strayer, 1989). Gandhi, who tried hard to remove this inequity, called them Harijans, meaning 'people of God'. The fifth group includes many sub-groups; the constitution of India categorized all of them under one title, 'scheduled castes'. Though untouchability has been legally abolished, the custom continues to be a part of the Hindu caste structure. Untouchability still persists very widely (Shah et al, 1980). In the present study, Brahmins, Kshatriyas, and Vaishyas were put together under one category, 'upper castes'. The tribals and Harijans were treated as two distinct castes as they differed in their social life styles and occupations. All the others in between were grouped under the title, 'backward castes/class'. Ih§_§2E21§ The sample was drawn from the experimentally accessible population using a multistage sampling design (Babbie,89). The districts of the states of Bihar, Orissa, and West 4O Bengal were listed and a sample of one district from each state was taken. The sample of districts consisted of Ranchi district in Bihar, Bhubaneswar district in Orissa, and 24-Parganas district (including Calcutta slums) in west Bengal. Then, the adult nonformal education centers in each district were listed and a sample consisting of 10% of each group of centers was drawn. The adult learners at each center were listed and a sample of about 50% of the learners (a minimum of 10 and a maximum of 40) at each center was taken. Since the study design was exploratory in nature, a sampling process of this nature was considered acceptable. Reid and Smith (1989) state that, "research done with largely exploratory purpose in mind allows considerable flexibility in method. Samples may be selected according to which sources of data will provide the most useful information most readily, and additional sources may be added to pursue leads." The final sample consisted of 261 adult learners from 24 nonformal education centers in all the three districts. The sample of 261 adult learners in nonformal education centers was about 20% of the experimentally accessible population. 41 The district-wise breakup of this number is given in the Table 1. Table 1. District wise distribution of subjects. District Number of NFE Number of learners Centers in the sample Ranchi district, Bihar 9 97 Bhubaneswar district, Orissa 6 55 24-Parganas 9 109 Total: 24 261 The selected adult nonformal education centers were organized and managed by organizations supported either by private funding agencies or by the government or by both. _Their main objective was to provide training considered necessary for rural adults in order to raise their income levels and social status. Training was provided free or at a nominal charge. All the learners were from communities living around the centers. A few centers had hostels where food and shelter were provided to students at a small charge. Class schedules and duration of 42 nonformal education programs varied from center to center. Learners included both men and women. Most of the learners were from economically and socially lower strata of people and represented almost all the major castes, occupations, and religions in the selected states. Data Collection At the selected nonformal education centers, learners who were identified for participation in this study were taken to separate rooms and four data collection instruments, one at a time, were administered to them. In the case of those who were functionally illiterate, the researcher interviewed them and completed the instruments for them. A letter introducing the study, it's purpose, and the participants' right to discontinue participation if desired, was read (Appendix-II). The participants were also told that their responses would remain anonymous. Teachers and staff members of the center were not present at the time data collection instruments were administered to the participants. Staff at each center provided the researcher with information about participants' class attendance. Instrumentation for this study had four parts: 1. Measurement of Motivation (Appendix-III) This questionnaire had five open-ended questions. The questions were based on Knowle's (1978) paradigm on adult motivation that adult learners are: - self directed, - concerned with development/life transition, - rich in experience, - desire immediate application, and - problem centered. These questions were reviewed and judged as appropriate for the purpose they were constructed by three Indian scholars. C 2. y‘or- ‘u‘! 0 O ; ‘9 a! ‘ 1‘9"!!- This measure had twenty four Likert-type items and the responses were recorded on a scale of 1-5. The items were developed on the basis of studies by Houle 44 (1961) and Burgess (1971) which showed that adults have reasons to participate in learning activities. One of the implications of Houle and Burgess' paradigm is that, given freedom, adults will desire to take up learning activities in which they see a fulfillment of their reasons. In other words, the extent to which an adult learner desires or accepts a learning activity is related to the extent he/she sees it fulfilling his/her reasons. Burgess(l971) had classified the reasons he identified into the following seven categories: - the desire to know, - the desire to reach a personal goal, - the desire to reach a social goal, - the desire to reach a religious goal, - the desire to escape undesirable situations, - the desire to take part in an activity, and - the desire to comply with formal requirements. The original three categories of Houle(1961) are: the activity-oriented, the learning-oriented, and the goal-oriented. The first two of these three categories are already included in the above seven 45 reasons delineated by Burgess. Therefore, with the addition of the third category, the goal-oriented of Houle's set, we have eight categories of reasons for adult learning. This instrument was constructed on the basis of these above eight factors. Three Likert-type items were developed for each factor making a total of twenty four items. The researcher compiled this list in consultation with 15 Indian scholars at Michigan State University who had the experience of interacting with rural adults of India. Subsequently, the list was judged by three Indian experts in rural development as representative of all the eight factors of Houle and Burgess. e su ent o 5 NF c e tance A end' -V This measure had eight Likert-type items which were developed on the basis of the following eight institutional factors responsible for student satisfaction suggested by Apps (1981): - administrative procedures, - structure (schedule of classes), 4. 46 - support systems, - curriculum rationale (rationale for what is taught), - financial aid, - teaching apparatus, - faculty attitude (towards learners), and - faculty development (abilities of the faculty). Apps (1981) says, the list was based on his interviews with eighteen professors including Malcolm Knowles and Ralph Tyler, review of literature, and his own personal experience. These questions were also reviewed and judged appropriate for the purpose they were constructed, by three Indian experts. This questionnaire required respondents to provide information on the selected personal variables. Respondents were asked to answer all the items except for Media Exposure. The researcher personally interviewed the respondents individually and completed this item for them. Media exposure was 47 determined on the basis of the time one spent on reading newspapers/periodicals, hearing radio news, and viewing television programs. The variable had four levels: Very poor, Poor, Good, and Excellent. Valldipy of the instruments Validity refers to the extent to which an empirical measure adequately reflects the real meaning of the concept under consideration (Babbie, 1990). There are three widely recognized types of validity; content, criterion, and construct(Reid, 1989). Construct validity as Babbie(1989) suggests, is based on the way a measure relates to other variables within a system of theoretical relationships. Since all the items on the first three instruments were based on certain well recognized theories as explained above, the instruments were considered validated. The fourth questionnaire being a mere demographic and personal data collection instrument did not require validation. 48 Rel'ab'lit Reliability may be defined as the level of stability of the measuring device over time (Borg & Gall, 1989). Reliability analysis was conducted on each of the above three instruments using Cronbach's alpha method. For testing the reliability of the instrument Measurement of Motivation, individual scores of each respondent on all the five items on this measure were entered into the computer using the SPSS PC+ (version 3.0) program and Cronbach's alpha was computed. The reliability of the instrument 'Measurement of Goal(NFE) Acceptance' was tested for reliability by entering into the computer the individual scores of each respondent on all the 24 items on this instrument. Similarly, the instrument 'Measurement of Means(NFE) _Acceptance' was tested for reliability by entering the individual scores of each respondent on all the 8 items on the instrument. Reliability coefficients for the instruments are given in Table 2. 49 Table 2. Reliability coefficients: Instrument Alpha Motivation measure : .64 Goal(NFE) acceptance measure : .74 Means(NFE) acceptance measure: .74 The alpha value .64 for the motivation measure was some what low: as such, reliability of this measure was further investigated using test-retest procedure on a randomly chosen sample of 12 participants. A time gap of 30 days was allowed between the test and retest. The coefficient of correlation between test and retest scores was .81 and was significant at the .001 Alpha level. Reliability of the three measures used in this study were considered established on the basis of the above reliability tests. In addition to establishing validity and reliability, the instruments were pretested by administering them to 25 adult students at the vocational training center, Mason, Michigan. A second pretest was conducted in India (Bihar). The purpose of these tests was to ensure that the questionnaires were clear, understandable, and 50 unambiguous. The purpose was fulfilled by these pretests. All the four instruments were translated in India, into Hindi, Oria, and Bengali languages. The average literacy level of the participants was considered for choosing an appropriate writing style in each language. The translations were done by professional experts in the areas of sociology and rural development. The variables In this study, motivation and participation were used as dependent variables. The researcher in consultation with adult education experts working for rural development in India, selected the sociodemographic variables sex, age, marital status, caste, occupation, economic status, health, education, area, and media exposure as the independent variables. The variables, media exposure and health, not used in American studies were included in this study as the researcher and the Indian experts whom he consulted believed that these variables should be included in the sociodemographic variables as they very much related to the social and personal lives of the adult learners. These variables, their operational definitions, and levels of measurement are presented in Figure 7. 51 Variable Operational Levels of IV/DV* name: definition: measurement: Motivation Mean Score Interval DV Participation Percent of days attended out of total number of school days. Ratio DV Sex 1. Male Nominal IV 2. Female Age Number of years Ratio IV Marital status 1. Unmarried Nominal IV 2. Married/divorced/ widowed Caste 1. Harijan Nominal IV 2. Tribal 3. Backward caste/caste 4. Upper caste Occupation 1. Laborer Nominal IV - 2. Small scale farmer 3. Large scale farmer 4. Small scale business 5. Service Economic Value of property and Ratio IV Status annual income in rupees Health 1. Excellent Ordinal IV 2. Good 3. Poor Education No. of years of formal Ratio IV education Area of 0. Urban area Nominal IV Residence 1. Rural area Figure 7. Variables used in this study. * DV: Dependent variable, IV: Independent variable. 52 Fig. 7. continued: Variable Operational Levels of IV/DV* name: definition: measurement: Media Exposure 1. Very poor Ordinal IV 2. Poor 3. Good 4. Excellent Goal(NFE) Acceptance Mean score Interval IV Means(NFE) Acceptance Mean score Interval IV Mother's ed. No. of years of formal education Ratio IV Father's ed. No. of years of formal Ratio IV education Figure 7. Variables used in the study. * DV: Dependent variable, IV: Independent variable. W The completed questionnaires were given identification numbers. All four questionnaires completed by one individual received one number and were kept together. Scoring was done as follows: 1. 53 e o 'va '0 endix- II This instrument had five open ended questions. A respondent could get a lowest score of 1 and a highest score of 5. To begin with, 10 questionnaires of each measure were scored independently by the researcher and another scorer trained by the researcher. No significant differences were found in their scores. As such, the researcher scored the rest of the questionnaires and mean scores were computed for each individual. A mean score for a respondent was computed by adding all the five scores and dividing the sum by 5. V-tr- ‘11-! ' .2 , i '9 t! ‘ '99‘!! This instrument had 24 Likert-type items. The lowest and the highest possible scores were 24x5=120, and 24x1=24. A mean score was calculated by dividing the sum of all the 24 scores by 24. Using the scores checked by the respondents, mean scores were computed by using a calculator. However, mean scores were later computed on a computer also. The calculator computed means were used as a check for accuracy of data entered into the computer: 54 3. Measurement of Means(NFE) Acceptance (Appendix V). This instrument had 8 Likert-type items. The lowest and the highest possible scores were 8x5=40, and 8x1=8. A mean score was calculated by dividing the sum of all the eight scores by 8. Using the scores checked by the respondents, mean scores were computed by using a calculator. However, mean scores were later computed on a computer also. The calculator computed means were used as a check for accuracy of data entered into the computer. Data from the four questionnaires were posted on the Transfer Sheet (Appendix-VII) and double checked for accuracy of the data entered. These sheets were then used for keying the data into a computer file using the SPSS PC+ (version 3.0) program. WW Mean, mode, and Median were used as measures of central tendency. Standard Deviation and Standard Error were used as measures of dispersion for the quantitative data pertaining to variables measured as interval and ratio levels. 55 Co a ions To study linear associations between two variables, Pearson's coefficients of correlation were calculated. Such calculations were made only for the variables measured as ordinal, interval and ratio levels. The strength of association between two variables was determined on the basis of the descriptors proposed by Davis (1971) (Figure 8). Coefficient(r) Description .70 or higher Very strong association .50 to .69 Substantial association .30 to .49 Moderate association .10 to .29 Low association .01 to .09 Negligible association Figure 8. Descriptors of correlations. fiypothesls f l was tested using the multiple linear regression procedure at an alpha level of 0.05. Two regression models were constructed taking the 10 personal variables listed in the hypothesis as independent 56 variables and motivation and participation as the dependent variables. The categorical variables, sex, caste, area of residence, marital status, health, and occupation were transformed into dummy variables before entering into the regression equations. Dichotomous variables such as sex can be used as independent variables if they are numerically coded (Glass and Hopkins, 1984). Ott (1988) suggests that, "the presence of dummy variables present no major problem for ascertaining the adequacy of the fit of a polynomial model. The important thing to remember is that when quantitative and dummy variables are included in the same regression model, by substituting any combination of ls and Os for dummy variables, we obtain a regression in the quantitative variables". For each model an F-test of homogeneity of variances was performed as multiple regression analysis works best when the variables entered into the equation have homogeneity of variances. The results of F-tests are presented along with the regression analyses. A correlation matrix for all the independent variables was calculated. It was noted that there were no 57 multicollinear variables among those entered into the regression models. The variables were tested for normality. The method used in the regression analysis is known as 'ENTER' in the SPSS PC+ statistical program. Under this method all the specified independent variables gain entry into the equation at one time. The F value in the analysis of variance, a part of the results of regression, is a test of the hypothesis that Beta(sex) = Beta(age) = ....... = 0. The T significance given for each variable in the equation is for the hypothesis that the Beta value in the population is zero. Beta values are standardized regression coefficients or regression weights. Each Beta value represents the effect that a standard deviation in the corresponding independent variable would have on the standard deviation in the dependent variable. R-Square is the portion of the variance in the dependent variable associated with the variance in the independent variables and indicates the explanatory power of the 58 regression model (Hedderson, 1987). The goodness of fit of the linear model is also indicated by the R-Square value. The adjusted R is a measure of goodness of fit of the model in the population. The Multiple R is the correlation between the dependent variable and the entire set of the independent variables used in the model. fiypothesls i 2 was subjected to the analysis of variance procedure to test for equality of means of participation/ motivation for all the four levels: G+M+, G+M-, G-M+, and G-M- of the variable 'response pattern'. Scheffe's multiple range test procedure was used to identify groups significantly different from others. For both the procedures the significant level was kept at 0.05 alpha. Barlett-Box F test was also performed for testing homogeneity of variances, an assumption needed for applying analysis of variance procedure. Hyporhooiorfl 3. 4. ano_§ were tested using the cross tabulation procedure. A chi-square statistic was calculated to test independence of the variables. 59 HYEQ§h§§i§_£§ was verified using a qualitative approach. The researcher spent 20 days with two communities in the sample area collecting field notes. The field notes were analyzed and studied for drawing inferences which were then related to the hypothesis. opalitative Study Design The qualitative or the field study focussed on investigation of social interaction and the cultural activities of three communities in Ranchi district in Bihar and 24-Parganas district in West Bengal. This was done by the researcher participating in the daily lives of the people, watching what was happening, listening to what was said, asking questions, and in fact, by collecting what ever data were available to throw light on the issues with which the study was concerned. As Mead (1977) stated, "every thing that happens will become data." The researcher looked for answers to questions like, "How does what is happening here compare with what happens in other places?" "How is culture passed on to the next generation?" CHAPTER IV DATA ANALKSIS Introduction This chapter presents the analyses of data obtained from the questionnaires and the results of statistical testing of all the six hypotheses. In so doing, the relevant data, the results of statistical analyses and the statistical interpretations are arranged in a logical sequence. The results of the analyses and statistical interpretations are organized around the hypotheses tested. The tables and figures contain the relevant information needed for interpretation of the test results. The entire analysis of data in this report has been done using the statistical program called SPSS PC+ version 3.0. The chapter is divided into the following six sections: 1. Characteristics of the sample 2. Computed variables 3. Descriptive statistics 4. Statistical procedures for testing the hypotheses 5. Tests of hypotheses 6. Results of additional statistical analyses 60 61 Characteristics of the Sample For better understanding of the analytical results of this study, certain demographic characteristics of the sample in terms of sex ratio, economic levels, caste system, regional groups etc. considered relevant for the study are presented in this section. The sample, composed of 261 participants, was drawn from the Ranchi district of Bihar, the Bhubaneswar district of Orissa, and the 24-Parganas district (including the Calcutta region) of West Bengal. The distribution of participants by district is presented in Table 3. Table 3. Distribution by district: District/State No. of (%) participants E§£2£IZEI£§§ """""""" ;§"'"""""Z3§TS§" Bhubaneswar/Orissa 55 (21.0) 24—Parganas/W.B. 109 (41.8) Total 261 (100.0) Women accounted for 30.7% and men for 69.3% of the distribution in the sample (Table 4). 62 Table 4. Distribution by sex: Sex Number of (%) participants Men 181 (69.3) Women 80 (30.7) Total 261 (100.0) Hindu society, which constitutes about 80% of the Indian population, is highly caste-structured as explained in chapter III. Harijans and Tribals sometimes known as Untouchables, are at the low end of the line while the brahmins are at the upper end. In between are Kshatriayas, Vaishyas, and Sudras. In this study, Brahmins, Kshatriyas, and Vaishyas and their sub-castes were grouped together under one category 'upper caste.’ Sudras with all their sub-castes were categorized as "backward castes.' Even though Muslims do not come under the Hindu caste system, on the basis of local understanding they were included in the backward castes/class category. All the tribal communities, irrespective of their religion, were put under the category 'tribals.' Table 5 gives the caste distribution of the participants. The highest number of participants 63 (45.2%) were from the upper castes and the lowest (6.9%) from Harijans. Table 5. Distribution by caste: Caste Number of (%) participants Harijans 18 (6.9) Tribals 74 (28.4) Backward caste 51 (19.5) Upper caste 118 (45.2) Total 261 (100.0) The occupational patterns of the participants in the sample are given in Table 6. Table 6. Distribution by occupation: Occupation Number of (%) participants Daily-wage laborers 22 (8.4) Small scale farming 81 (31.0) Large scale farming 8 (3.1) Small scale business 40 (15.3) Service 110 (42.2) Total 261 (100.0) 64 Daily laborers included in the sample were mostly illiterates and lived on wages they earned on a daily basis. Most of them could find employment only for about 8 months in a year. During the remaining period of 4 months, known as the 'off season,' they either migrated to nearby cities or lived on savings, on charity, on loans or on a combination of these. The daily wage for men ranged from Indian rupees 8.00 to 15.00 depending on the type of work they did and the employer (One US dollar = Rs. 16.40 in April 1989). Wages for women were 50 to 80 percent of men's wages depending on the nature of the work done and the employer. Small farmers owned agricultural lands of 0.5 acres to 3.0 acres. Most of those lands were rain fed or had inadequate irrigation facilities and therefore were generally cultivatable only once a year. Large farmers included in the sample were not from the landlords communities but were those who had lands ranging from 4 acres to 10 acres in area. The business class in the sample were those who worked as hawkers, grocery shop owners, and other small scale businesses. The service holders were those who worked for the government of India, for commercial agencies, for religious organizations, and 65 for agencies of the state government. Their average salary was estimated at Rs. 900.00 a month. It was supposed in this study that high school graduation would take 12 years of formal education, college graduation 14 years, and master's degree 16 years. Education levels of the subjects varied. Illiterates accounted for 4.2% and college graduates 20.3% of the sample. Distribution by formal education is given in Table 7. Table 7. Distribution by formal education: Ed. level Frequency (%) Illiterates 11 (4.2) Below 8 grade 47 (18.0) 8 - 12 grade 146 (56.0) College graduates 53 (20.3) Graduates (Masters level) 4 (1.5) Totals 261 (100.0) The training programs covered by this study were classified under 8 different categories. These categories are listed in Table 8. In the sample, the 66 highest number of participants (24.2%) were in typing and shorthand training programs. The least (3.1%) were in literacy programs. Table 8. Distribution by type of training: Type of NFE Frequency (%) Literacy 8 (3.1) Carpentry 9 (3.4) Driving 16 (6.1) Tailoring & Embroidery 29 (11.1) Electrical jobs 34 (13.0) Handicraft 47 (18.0) Mechanical works 55 (21.1) Typing and short-hand etc. 63 (24.2) Totals 261 (100.0) An age breakdown of the sample is given in Table 9. A majority of the participants (72%) were in the interval 20-25 years of age; only 28% were 26 years of age, or older. This indicates that fewer adults take up learning projects after attaining the age of 25 years, which may be their culturally based attitude towards adult learning. 67 Table 9. Distribution by age group: Age group No. (%) Between 20 - 25 188 (72.0) Between 26 - 50 73 (28.0) Total 261 (100.0) Computog variaples The following variables were computed for the purpose of testing the hypotheses: Rooponse Pattern Response pattern was computed using the variables 'goal(NFE) acceptance' and 'means(NFE) acceptance' in different combinations. Scores above and below the mean (3.26) of 'goal(NFE) acceptance' were denoted by the symbols G+ and G- respectively. Similarly, the scores above and below the mean (3.85) of 'means(NFE) acceptance' were denoted by the symbols M+ and M- respectively. A response pattern was a combination of 6+, G-, M+, and M- taken two at a time. Thus four response patterns were computed as shown in Figure 9. 68 Response pattern Computation formulae: (G+M+) Goal(NFE) acceptance scores 2 3.26 and Means(NFE) acceptance scores 2 3.85. (G+M-) Goal(NFE) acceptance scores 5 3.26 and Means(NFE) acceptance scores <3.85. (G-M+) Goal(NFE) acceptance scores < 3.26 and Means(NFE) acceptance scores 3 3.85. (G-M-) Goal(NFE) acceptance scores < 3.26 and Means(NFE) acceptance scores < 3.85. Figure 9. Computation of response patterns: 'c t' eve Participation was categorized into three groups or levels; high participation, medium participation, and low participation. These levels were computed using the following intervals: Low level: Scores < 74.0 Medium level: Scores 2 74.0 and < 90.0 High level: Scores 2 90.0 Scores of 74 and 90 were scores 10% below and 10% above the participation mean (82). Based on the discussions 69 with some nonformal educators in India, it was assumed that 10% below the mean could be considered as low and 10% above the mean as high participation. Motivation was categorized into three levels: high, medium and low. The rationale and the procedure used to compute these levels were exactly the same as those used for computing participation levels. The intervals used in this computation were as follows: Low level: Scores < 2.91 Medium level: Scores ; 2.91 and < 3.56 High level: Scores ; 3.56 W All the dichotomous variables such as sex, are numerically coded (0,1) for use as independent variables in the regression equations. These variables which may be called dummy variables, are given in Figure 10. 7O Variable Code-0 Code-1 £2; """"""" {15; """""""" £115; """"" Caste Harijans and Tribals Backward and upper castes Marital status Unmarried Married Occupation Laborers and small Others farmers Health Good health Poor health Area Urban Rural Figure 10. Dummy variables in the regression analysis: D . l' E! l' I' Table 10 gives the basic statistics for the variables measured in ratio levels listed in Figure 7. These statistics indicate that except for economic status all the variables closely conformed to normality. The standard deviation for the variable economic status is 161,800 which reflects the nature of variation in the distribution of economic levels of the participants. The large standard deviation indicates that the distribution has scores which are a long way from the mean and suggests large economic disparities among participants. 71 Table 10. Statistics for the variables: Variable Min. Max. Mean Median Mode S.D. Skewness Participa. 40.0 99.0 82.0 85.0 90.0 12.0 -0.95 Motivation 1.6 4.6 3.2 3.2 3.0 0.6 -0.13 Age 20.0 49.0 24.2 23.0 20.0 4.4 1.79 Education 00.0 17.0 10.1 10.0 10.0 3.3 -1.13 Economic 5.0 190.0 75.9 25.0 10.0 161.8 6.99 status(Rs.) x1000 x1000 x1000 x1000 x1000 x1000 Mother's 00.0 14.0 4.9 5.0 00.0 4.1 0.09 education Father's 00.0 22.0 7.4 9.0 10.0 5.0 -0.19 education Gaol acceptance 1.8 4.7 3.3 3.3 2.9 0.6 0.15 Means acceptance 1.6 5.0 3.8 4.0 4.3 0.7 -0.78 Pearson's correlational matrix for all the variables other than categorical variables used in the data analyses is given in Figure 11. 72 83353 05 8.. 555: coma—028 9.588.. “: 23E Ssvo sadism . co.— 8§< as: . 8o 8; 8568a... .80 So So 8.“ 3.838 8.3. So So . So 8.“ 68828 £9.82 So to . so . so 8." .8099 58: 8o moo . moo . omo . to o3 as» 2928» So 8.? so . so . 3o . 8o 8." 8.3.3 So. . 9.? moo So So :o . 2o 8." of . ...vo . $8 So So . «so So . So Bo 8.2 5.282 . ono . was So. moo. moo So. So So. . So 8; 88985 ocean—08a. 088903. 83833 com-030m 0.5398 388 882 .80 8.3 {£82 £8: 3588a 33%» «9.. 8582 c8335; «mags, 73 Most of the correlation coefficients in Figure 11 are below 0.45. As Borg and Gall (1990) has suggested, correlations in the range of 0.2 to 0.4 may be all that should be expected for relationships between variables studied in educational research; the reason being that many factors influence most of the behavioral patterns and personal characteristics. For interpreting the correlations in Figure 11, Davis' (1971) descriptors given in Figure 8 are followed. The correlation between participation and mean(NFE) acceptance (r=.30) indicates moderate association; but the association between participation and goal(NFE) acceptance is low as indicated by the low correlation (r=.25). This observation suggests that there is some possibility of adult learner's participation being associated with his/her means(NFE) acceptance while no association with _goal(NFE) acceptance may be expected. The correlation between participation and formal education (0.04) is negligible. This suggests that formal education has no relationship with adult participation. 74 The association between motivation and participation is moderate (r=.37) indicating that a variance in motivation predicts only 1.4% (r square=0.14) of variance in participation. This also suggests that motivation is not a good predictor of participation in rural adult learning situations, which in turn suggests that participation for the most part may depend on non-psychological factors. Means and standard deviations of participation and motivation by economic level are given in Table 11. Table 11. Participation and motivation by economic levels: Economic level Participation Motivation Annual income plus value of property: I. Less than 79.6 11.6 3.2 0.69 Rs. 10,000 ($610) II. Between Rs.10,000 - 83.1 11.1 3.3 0.57 50,000 ($610 - 3,500) III. Between Rs. 50,000 82.6 10.2 3.4 0.60 100,000 ($3,500-6,100) IV. Above Rs. 100,000 79.1 16.3 3.1 0.69 ($6.100) Total 81.7 12.1 3.3 0.62 75 The mean values noted in Table 11 suggest that participation of the economically very poor participants (79.6) and that of the comparatively rich (79.1) are almost the same. Participation level of the middle groups is some what higher (83.1) than other economic groups. However, the motivation means of different economic groups (3.2, 3.3, 3.4, and 3.1) are almost equal. This suggests that while motivation may not be affected by adult learner's economic status, participation might. Table 12 gives the mean and standard deviation for participation and motivation by marital status. Table 12. Participation and motivation means for unmarried and married participants: Marital status Participation Motivation Mean S.D. Mean S D Unmarried 81.9 12.0 3.2 0.63 Married 81.3 12.3 3.4 0.55 From Table 12 it is noted that unmarried participants' motivation (3.2) is slightly lower than that of married 76 participants (3.4); but their participation means are almost equal (81.9 and 81.3). The variation in participation (S.D = 12.0) is almost the same for both the married and unmarried participants. In the case of motivation, the variance is noticeable; the standard deviation in motivation for unmarried participants is .63 and that for married is .55. Table 13 presents the mean and standard deviation for participation and motivation for men and women, and Table 14 for urban and rural participants. Table 13. Participation and motivation means for urban and rural participants by sex: Category Participation Motivation Mean S.D. Mean S.D. Men 80.6 12.0 3.2 0.62 Urban men 77.8 15.0 3.1 0.72 Rural men 82.4 9.0 3.2 0.54 Women 84.0 12.3 3.4 0.61 Urban women 83.2 12.3 3.4 0.58 Rural women 87.2 11.9 3.1 0.71 77 Table 14. Participation and motivation means for urban and rural participants: Category Count Education Participation Motivation ------------------- Mean--S;D. Mean S.D. Mean S.D. Urban 136 11.3 2.5 80.4 14.0 3.3 0.67 Rural 125 8.8 3.6 83.1 9.5 3.2 0.56 Total 261 10.1 3.3 81.7 12.1 3.2 0.62 From Tables 13 and 14 it is noted that rural adults' participation (83.1) is higher than that of urban adults. Within the rural group, the participation of rural women (87.2) is higher than that of rural men (82.4). Within the urban group, the participation of urban women (83.2) is higher than that of urban men (77.8). Between the rural and urban participants, the rural women and men had higher participation means (87.2; 82.4) than those of their urban counterparts (83.2; 77.8). One possible reason for this could be that, because fewer learning alternatives are available for rural adults than what is available for urban adults, the rural adults tend to stick to the available training programs. 78 Table 15 gives the means and standard deviations of participation and motivation for men and women of Bihar, Orissa, and West Bengal states. Table 15. Means for education, participation, and motivation, by sex and by state: Participants Count Education Participation Motivation Bihar 97 8.8 3.9 80.3 11.2 3.1 0.55 Men 71 8.7 3.7 80.5 10.2 3.2 0.51 Women 26 9.0 4.7 79.8 13.8 3.0 0.63 Orissa 55 10.0 2.4 84.3 10.3 3.6 0.60 Men 41 9.8 2.5 84.5 9.1 3.6 0.60 Women 14 10.7 2.1 83.9 13.6 3.4 0.58 West Bengal 109 11.3 2.6 81.5 13.6 3.2 0.64 Men 69 11.3 2.6 78.4 14.4 2.9 0.60 Women 40 11.8 2.7 86.8 10.1 3.6 0.50 Total 261 10.1 3.3 81.7 12.1 3.2 0.62 Men 181 9.8 3.1 80.6 11.9 3.2 0.6 Women 80 10.7 3.6 84.0 12.3 3.6 0.6 Bihar adult learners are at the bottom of the list in terms of education, participation, and motivation (Table 15) . 79 Table 16 contains means and standard deviations of participation and motivation by type of nonformal education program. Table 16. Participation and motivation means by type of nonformal education programs: Participants Count Education Participation Motivation Mean S.D. Mean S.D. Mean S.D. Carpentry 9 7.9 3.3 75.3 12.6 3.0 0.4 Driving 16 9.1 2.7 77.7 9.1 3.7 0.6 Elec. works 34 11.4 2.5 82.8 12.1 3.1 0.5 Handicrafts 47 9.7 3.1 81.8 15.5 3.0 0.7 Literacy 8 0.0 0.0 70.1 4.2 2.9 0.3 Mechani. jobs 55 9.8 1.2 83.0 8.1 3.1 0.6 Typing & shorthand 63 10.1 3.3 82.5 11.3 3.3 0.6 Tailoring 29 13.3 1.4 83.1 15.1 3.6 0.6 Total 521' 131'; 31'3" SIT—3 III ”STE” 3?? From Table 16 it is seen that both participation and motivation levels are the lowest for adult learners in literacy programs. are from rural areas and most of them belonged to the lowest economic category. All the participants in this program Tailoring, mechanical, and 80 electrical training programs have higher participation and motivation levels when compared with other types of training. Table 17 presents the mean and the standard deviation of participation and motivation for each occupational group of participants. Table 17. Participation and motivation means for different occupational groups: Participants Count Participation Motivation Mean S.D. Mean S.D. Labor/min.wage 22 80.4 13.5 3.5 0.7 Small farming 81 83.3 10.0 3.2 0.5 Large farming 8 81.4 13.4 3.0 0.7 Small business 40 79.0 14.8 3.3 0.6 Service 110 81.7 12.1 3.3 0.6 Total 261 81.7 12.1 3.2 0.6 From Table 17 it is seen that the participation of the laborers group (80.4) is lower and their motivation (3.5) is higher compared to most other occupational groups. The small farmars group has higher participation. 81 Means and standard deviations of participation and motivation, by caste are given in Table 18. Table 18. Participation and motivation means by caste: Participants Count Education Participation Motivation Harijans 18 9.5 1.7 82.4 11.4 3.3 0.6 Tribals 74 8.2 3.8 81.8 11.4 3.2 0.6 Backward ca. 51 10.3 3.1 80.5 10.7 3.1 0.6 Upper caste 118 11.3 2.5 82.0 13.3 3.3 0.7 Total 261 10.1 3.3 81.7 12.1 3.2 0.6 Participation and motivation slightly varied by caste. The means for these two variables are high for Harijans and upper caste groups compared to Tribals and backward castes (Table 18). Mean education of tribal participants is the lowest (8.2) in the sample. Table 19 presents the mean and standard deviation of participation and motivation for each educational group. 82 Table 19. Participation and motivation means for different educational levels: Education Count Participation Motivation levels Mean S.D. Mean S.D. Illiterates 11 74.4 8.1 2.8 .39 Below 8 grade 47 83.1 11.8 3.1 .67 8 - 12 grade 146 81.5 12.2 3.2 .59 College grads. 53 81.6 12.6 3.4 .62 (B.A/B.Sc/B.Com) Masters level 4 93.3 5.0 4.0 .50 (M.A/M.Sc/M.Com) Total 261 81.7 12.1 3.2 .62 From Table 19 it is noted that illiterate participants have lower participation and motivation than others. Those with higher education appear to have higher motivation, but the correlation between these two variables (r = 0.2; sig. p < 0.01)(Figure 11), though statistically significant, may not be practically significant. The group 'below 8 grade' seems to have higher participation than 'college graduates'. These observations suggest that learners from the illiterate group tend to be less involved with learning compared to the educated. 83 Sratistical Procedures ror Testing Hypotheses Statistical procedures used for testing hypotheses are described in chapter III; A few key steps are noted here. HYEQLh§§i§_i_l was tested using a multiple linear regression procedure at an alpha level of 0.05. Two regression models, one with participation as the dependent variable and the other with motivation as the dependent variable, were constructed taking all the 10 sociodemographic variables as independent variables. For each model, an F-test of homogeneity of variances was performed. The F value in the analysis of variance, a part of the results of regression, is a test of the hypothesis that Beta(sex) = Beta(age) = ....... = 0. The T significance given for each variable in the equation is for the hypothesis that the Beta value in the population is zero. R-Square is the portion of the variance in the dependent variable associated with the variance in the independent variables, and indicates the explanatory power of the 84 regression model (Hedderson, 1987). The adjusted R-Square is a measure of goodness of fit of the model in the population. Betas are standardized values; they can be used for comparing relative importance of different independent variables. The Multiple R is the correlation between the dependent variable and the entire set of the independent variables used in the model. HYDQID§§1§_£_Z was subjected to the analysis of variance procedure to test for equality of means of participation and motivation for all the four levels: G+M+, G+M-, G-M+, and G-M- of the variable 'response pattern'. Scheffe's multiple range test procedure was used to identify groups significantly different from others. For both the procedures, the significance level was kept at 0.05. The Barlett-Box F test was also performed for testing homogeneity of variances, an assumption needed for applying the analysis of variance procedure. Hyporhoooo # 3. 4. ano_§ were tested using a cross tabulation procedure. A chi-square statistic was calculated to test independence of the variables. 85 Hypornooi§_£o was verified using a qualitative approach. The researcher spent three weeks with three different communities. Field notes were analyzed and studied for drawing inferences which were then related to the hypothesis. Tests of Hypotheoes Hypothesis #1 states that there is a linear relationship between participation/motivation and the sociodemographic variables, age, sex, caste, education, area of residence, marital status, economic level, media exposure, health, and occupation. Null hypothesis: Beta(age) = Beta(sex) = ....... = 0. Taking participation and motivation as dependent variables and the sociodemographic variables as independent variables, the hypothesis was subjected to a least squares multiple regression procedure. The results are shown in Tables 20 and 21. 86 Table 20. Results of multiple regression with participation as the dependent variable: (Equation Number 1) Multiple R .2634 R Square .0694 Adjusted R Square .0322 Standard Error 11.9237 F = 1.864 Signif. F = .050 Variable B SE B Beta T Sig T Area of resi. 5.5433 2.2165 .2289 2.501 .0130 Health .7301 2.4229 .0190 .301 .7634 Marital status -2.5185 1.9864 -.0871 -1.268 .2060 Eco.status -5.3579 4.9864 -.0715 -1.075 .2836 Education .3208 .2610 .0874 1.229 .2203 Age .0239 .1862 .0086 .128 .8982 Sex 5.8788 1.8488 .2241 3.180 .0017 Media exposure .4312 .8263 .0390 .522 .6023 Caste .8572 1.9179 .0338 .447 .6553 Occupation -.7588 2.0093 -.0307 -.378 .7060 (Constant) 72.9795 5.8879 12.395 .0000 It is noted from Table 20 that the value 1.864 of the test statistic F is significant at p < 0.05. Therefore, the null hypothesis that Beta(area) = Beta(Marital) = ... = 0 or that no linear relationship exists between participation, the dependent variable, and the independent variables is rejected in favor of the alternative hypothesis that a linear relationship does exist with p <0.05. The R-Square value 0.0694 suggests that the goodness of fit of a linear model is rather weak. The adjusted R of 87 square 0.0322, a measure of goodness of fit of the linear model in the population, is also low indicating a weak fit. The R-Square value of 0.0694 suggests that only 6.90% of the variance in participation is predictable from a combination of all the 10 independent variables in the equation. In equation # 1, the beta value 0.2289 of area of residence is significant with p < 0.0130 and 0.2241 of sex is significant with p < 0.005. The beta values of marital status, economic status, and education are significant only with p < 0.300. These results indicate that among the above set of variables, area of residence appears to be the best predictor of participation. Rural adults coded 1 under the variable 'area' appear to have higher motivation than the urban adults (coded 0) as indicated by the positive beta value. Also, the participation mean for rural adults (83.1) is higher than that of urban adults (80.4) as shown in Table 14 indicating higher motivation for rural participants, as above. Sex appears to be the next best predictor of Participation; women's participation (coded 1) appears to Ibe.higher than men's participation (coded 0) as suggested 3by’the positive beta value 0.2241. This observation concurs with the observation that the participation mean 88 for men (80.6) is lower than the mean for women (84.0) as shown in Table 13. Impact of education on an individual's participation level is very small (Beta = .0874) and statistically insignificant (p = 0.22). This finding indicates that there is no difference in participation based on one's level of education. Beta for economic status is -0.0715 and is statistically insignificant (p = 0.28), suggesting that no relationship can be expected between these variables. Table 21. Results of multiple regression with motivation as the dependent variable: (Equation Number 2). Multiple R .3575 R Square _ .1278 Adjusted R Square .0929 Standard Error .5916 F = 3.663 Signif. F = .0001 Variable B SE B Beta T Sig T Area of resi. .0936 .1100 .0754 .851 .3955 Health -.0950 .1202 -.0481 -.790 .4303 Marital status .2109 .0986 .1423 2.140 .0333 Eco.status -7.8672 2.4742 -.0205 -.318 .7508 Education .0368 .0129 .1957 2.844 .0048 Age 2.9539 9.2404 .0209 .320 .7495 Sex .0723 .0917 .0538 .788 .4315 Media exposure .1471 .0410 .2596 3.587 .0004 Caste -.0674 .0952 -.0519 -.708 .4794 Occupation -.1116 .0997 -.0880 -1.120 .2640 (Constant) 2.39356 .29216 8.193 .0000 89 From Table 21 it is noted that the value 3.663 of the test statistic F is significant at p < 0.0001. Therefore, the null hypothesis that Beta(area) = Beta(Marital) =.... ...= 0 or that no linear relationship exists between the dependent variable motivation and the set of independent variables is rejected in favor of the alternative hypothesis that a linear relationship exists with p < 0.0001. The sample R-Square value of 0.1278 suggests that the goodness of fit of the linear model for motivation as the dependent variable, is rather weak but slightly stronger than what it was with participation as the dependent variable, shown in Table 20. The adjusted R-square value 0.0929 suggests that the goodness of fit of the linear model in the population with motivation as the dependent variable though not very strong, is higher in strength than what it was with participation as the dependent variable. The R-Square value 0.1278 suggests that 12.8% of variance in motivation may be explained by the 10 independent variables taken together. In the equation #2 (Table 21), the beta value 0.1423 of marital status is significant with p < 0.05; beta value 90 of 0.1957 of education is significant with p < 0.005 and 0.2596 of media exposure is significant with p < 0.0005. These results suggest that media exposure, education, and marital status impact adult motivation more than the other variables in the equation. It appears, married adults (coded 1 ) have higher motivation than unmarried adults (code 0) as indicated by the positive beta value 0.1423. The positive beta value 0.1957 for education suggests that adults with higher education appear to have slightly higher motivation than those having comparatively lower education. This concurs with the motivation means for different education levels given in Table 19. Media exposure seems to impact motivation more than any other variable in the equation; higher exposure appears to be related to higher motivation as indicated by the positive beta value 0.2596. Media exposure, education, and marital status which have high statistical significance in the prediction equation for motivation have no statistical significance in the prediction equation for participation. These results suggest the possibility that the variables which impact participation may not impact motivation. 91 EYDQth§i§_£_Z Hypothesis #2 states that adult learners differ in their participation and motivation with respect to their response patterns G+M+, G+M-, G-M+, and G-M-. Analysis of variance (ANOVA) procedure was used to test the null hypothesis that u(G+M+) = p(G+M-) = p(G-M+) = p(G-M-). The results are given in Tables 22 and 23. Table 22. Analysis of variance for participation by response pattern: Source df. Sum of Mean F F Squares Squares Ratio Prob. Between groups 3 4603.128 1534.376 11.739 0.0000 Within groups 257 33590.872 130.704 Total 260 38194.000 Barlett-Box F = 5.276, p <0.001 Table 23. Participation means for response patterns: Group No. Mean S. Dev. Standard Error $3.217"'E;""""§ZT;Z """ ET?"""""'IT3Z """ (G+M-) 40 80.95 10.4 1.65 (G-M+) 66 84.68 10.3 1.26 (G-M-) 68 74.97 14.7 1.78 Total 261 81.67 12.1 0.75 92 The Scheffe a priori comparison indicated that the response pattern G-M— significantly differed from patterns G+M+ and G-M+ at an alpha value of 0.05. Since the F value 11.739 (Table 22) is significant at p < 0.0000, the null hypothesis that 0(G+M+)=p(G+M-)= u(G-M+)=u(G+M-)=0 is rejected with p <0.0000. It is therefore unlikely that adult learners of different response patterns have the same participation mean. As such, the hypothesis that adult learners differ in participation with respect to their response patterns is retained with p <0.0000. The mean 84.9411.04 of group G+M+ is the highest and the mean 74.97i1.78 of the group G-M- is the lowest which is also below the average mean 81.6710.75 (Table 23). These means also indicate that adult learners with response patterns G+M+, G-M-, and G+M- differ in terms of participation levels. The means of the G-M+ and G+M+ groups (84.94 and 84.68) are almost equal and are higher than those of other groups. It appears, the factor M+ which is common to these two groups may be responsible for this equal and higher participation. 93 Table 24. Analysis of variance for motivation by response pattern: Source df. Sum of Mean F F Squares Squares Ratio Prob. Between groups 3 23.71 7.90 26.51 0000 Within groups 257 76.62 0.30 Total 260 100.34 Barlett—Box = 5.072, p < 0.005 Table 25. Motivation means for response patterns: Group No. Mean S.Dev. Standard Error (G+M+) 87 3.62 .41 0.04 (G+M-) 40 3.22 .66 0.10 (G-M+) 66 3.16 .59 0.07 (G-M-) 68 2.84 .57 0.07 Total‘ 261 3.24 .62 0.04 The Scheffe a priori comparison indicated a statistically significant difference between G-M- and the other three response patterns; between G-M+ and G-M-; and between G+M- and G+M+ groups. 94 The F value 26.51 is significant at p < 0.0000. Therefore, the null hypothesis that u(G+M+) = u(G+M-) = u(G-M+) = u(G+M-) = 0 is rejected with p <0.0000. It is therefore unlikely that adult learners of different response patterns have the same motivation mean. Therefore, the hypothesis that adult learners differ in motivation with respect to their response patterns is retained with p <0.0000. The mean 3.62 i 0.04 of group G+M+ is the highest and the mean 2.84 i 0.07 of the group G-M- is the lowest and is also below the average mean 3.24 i 0.04 (Table 25). These differences though numerically small, are considered important in view of the very small standard errors. These observations suggest that learners with different response patterns may have different motivation levels. Witmtm Hypothesis # 3(a) states that high level participation has correspondence with response pattern G+M+. Hypothesis # 3(b) states that high level motivation has correspondence with response pattern G+M+. 95 Hypothesis # 4(a) states that medium level participation has correspondence with response patterns G+M- and G-M+. Hypothesis # 4(b) states that medium level motivation has correspondence with response patterns G+M- and G—M+. Hypothesis # 5(a) states that low level participation has correspondence with response pattern G—M-. Hypothesis # 5(b) states that low level motivation has correspondence with response pattern G-M-. Cross tabulation and Chi-square test procedures were applied to test these hypotheses. The results of these procedures are presented in Tables 26 and 27. Table 26 has the results of crosstabulation of participation by response pattern and Table 27 of motivation by response pattern. In these tables, the first entry in each cell is the number of cases or frequency; the second entry is the row percentage of that frequency; and the third entry is the column percentage. Row percentages indicate the distribution of response patterns for low, medium, and high participation/motivation. Column percentages 96 indicate the distribution of participation/motivation levels for each of the response patterns. Table 26. Results of cross tabulation: participation levels by response pattern: Count Row Pct (G+M+) (G+ M-) (G-M+) (G-M-) Col Pct Row Total 11 11 10 30 62 Low 17.7 17.7 16.1 48.4 23.8 Participa. 12.6 27.5 15.2 44.1 52 24 37 28 141 Medium 36.9 17.0 26.2 19.9 54.0 Participa. 59.8 60.0 56.1 41.2 24 5 19 10 58 High 41.4 8.6 32.8 17.2 22.2 Participa. 27.6 12.5 28.8 14.7 Column 87 40 66 68 261 Total 33.3 15.3 25.3 26.1 100.0 Chi-Square D.F. Significance 27.44508 6 .0001 In Table 26, the Chi-Square value 27.445 significant with p < 0.0001 indicates that the two variables, participation and response pattern, are not independent of each other. 97 Table 27. Results of cross tabulation: response pattern: motivation by Count Row Pct (G+M+) (0+ M-) (G-M+) (G-M-) Col Pct Row Total 2 11 21 38 72 Low 2.8 15.3 29.2 52.8 27.6 Motivation 2.3 27.5 31.8 55.9 29 17 24 22 92 Medium 31.5 18.5 26.1 23.9 35.2 Motivation 33.3 42.5 36.4 32.4 56 12 21 8 97 High 57.7 12.4 21.6 8.2 37.2 Motivation 64.4 30.0 31.8 11.8 Column Tot. 87 40 66 68 261 and pct. 33.3 15.3 25.3 26.1 100.0 Chi-Square D.F. Significance .0000 71.39174 6 In Table 27 the Chi-Square value 71.39 with p < 0.0001 indicates that the two variables, motivation and response pattern, are not independent of each other. Expgthssis_iilal It is noted from the high participation row in Table 26 that 41.4% of all the 58 high participation participants are from the G+M+ response pattern group, while the 98 percentages of those under the other three response patterns are 8.6, 32.8, and 17.2 only. From the column G+M+, it is noted that the highest percent (59.8%) of adult learners are medium level participants and only the next highest percent (27.6) are the high level participants. These results indicate that the largest percentage of adults with high participation belong to G+M+ group; but most adults with the response pattern G+M+ had a medium level of participation. Taken together, the medium and high participation groups account for 97.4% of all the 87 participants in column G+M+, suggesting that the group G+M+ has a high degree of correspondence with this combined group. From the table of means (Table 23) it is seen that the participation of the group G+M+ (84.94i9.7) is higher than the means of the other three groups, which suggests correspondence between group G+M+ and higher participation. The above results provide some evidence in support of the hypothesis # 3(a) that high participation has 99 correspondence with the response pattern G+M+. However, there appears a strong evidence of correspondence between the response pattern G+M+ and a combination of medium and high level participants. Hypornesls 3(a) From the high motivation row in Table 27, it is noted that the highest number (57.7% ) of the 97 high motivation participants are from the G+M+ group. Also, looking at the column G+M+, it is noted that the highest percent (64.4%) of all the 87 learners of the G+M+ response pattern have high level motivation. Also, from the means table (Table 25) it is seen that the group G+M+ has the highest mean (3.62). These results indicate that most of the high participants belong to G+M+ group. Therefore, the hypothesis # 3(b) that high motivation has correspondence with the response pattern G+M+ is retained with p < 0.0000. mm In Table 26, looking at the column G+M- which gives the distribution of all the 40 participants of G+M- group 100 participation levels, it is observed that the highest percent (60.0%) are medium level participants. In the column G-M+, out of the 66 participants of the group G-M+, 56.1% are medium level participants, 15.2% low participants, and 28.8% of high participants. Looking at the 'medium participation' row (Table 26), it is observed that the combined percent of groups G+M- and G-M+ is 43.2, which is higher than the percentages of the other two groups. Also, the participation means (Table 23) of groups G+M— and G-M+ being 80.95 and 84.68 are between the highest (84.94) and the lowest (74.97). These results indicate that the response patterns G+M— and G-M+ are associated (p < .0000) with medium level motivation. Therefore, the hypothesis # 4(a) that there is an association between medium level participation and the response patterns G+M- and G-M+ is retained with p = 0.0000. sis b From Table 27 it is noted that of the 40 participants of G+M- group across motivation levels, the highest percent of participants (42.5%) have medium level motivation. In 101 the column G-M+ also, out of the 66 participants of the group G-M+ the highest percent (36.4%) of participants have medium level motivation. Looking at the 'medium motivation' row (Table 27), it is observed that the combined percent of groups G+M- and G-M+ is 44.6, which is higher than the percentages of the other two groups. Also, the motivation means (Table 25) of groups G+M- and G-M+ being 3.22 and 3.16 are between the highest (3.62) and the lowest (2.84). These results indicate that the response patterns G+M- and G-M+ are associated (p < .0000) with medium level motivation. Therefore, the hypothesis # 4(a) that there is an association between medium level motivation and the response patterns G+M- and G-M+ is retained with p