Differences in life satisfaction, future orientation and locus of control between educationally active and non-active older adults

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content DIFFERENCES IN LIFE SATISFACTION, FUTURE ORIENTATION AND LOCUS OF CONTROL BETWEEN EDUCATIONALLY ACTIVE AND NON-ACTIVE OLDER ADULTS by Barbara Joan MacKenzie A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Education--Counseling Psychology) August 19 89 Copyright 1989 Barbara Joan MacKenzie UMI Number: DP25289 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. Dissertation Rubl h s n g UM! DP25289 Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 UNIVERSITY OF SOUTHERN CAUFORNIA THE GRADUATE SCHOOL c! S L UNIVERSITY PARK LOS ANGELES, CAUFORNIA 90089 £<J MISS This dissertation, written by Barbara Joan MacKenzie under the direction of hsx Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of D O C TO R OF PH ILO SO PHY Dean of Graduate Studies D a te I ? . , . . 1 9 8 9 DISSERTATION, MMITTEE Chairperson DEDICATION I dedicate this work to my parents, David and Virginia Truslow, whose twin gifts of limitless vision and support made this accomplishment possible. i i j I ACKNOWLEDGMENTS Special thanks go to Dr, David Peterson whose guidance i I I and encouragement was invaluable in the development of thisj kissertation. I wish to thank Dr. Amaury Nora for his I abundant enthusiasm, analysis, and productive criticism in; helping me reach this goal. I wish to acknowledge my deepest appreciation to Dr. Penelope Richardson for her unfailing encouragement and support as she constantly widened my' i : horizons of knowledge and confidence on this journey. TABLE OF CONTENTS Chapter Page I. STATEMENT OF THE PROBLEM......................... 1 Introduction...................................1 Purpose of the Study.......................... 5! Research Problem.................... ..........5; Statement of Hypotheses.......................l\ i Significance of the Study.....................8i | II. REVIEW OF THE LITERATURE......................... 9 j Participation of the Elderly in Education....9| [ ! | Theoretical Bases of Adult Education......... 12' ■ Behaviorism................................ 13; i i I i i > i Operant Conditioning...................... 14! \ ; l ! Social Learning Theory.................... 16> i ! Cognitive Approaches...................... 17 | Humanistic and Developmental Foundations of j Adult Education............................ 19 i ! Life Satisfaction.............................24 1 Future Orientation............................29 | Locus of Control..............................32 I ! ; Summary........................................39 III. METHODOLOGY....................................... 42 Introduction..................................42 Sample............................. 42 Educationally Active Group............... 42 Educationally Non-active Group........... 43 General Considerations about Subjects....44 Procedures and Administration of Surveys.... 45 Instrumentation.............................. 46 Criterion Variables.......................... 49 Predictor Variables.......................... 50 Data Analysis.................................50 Methodological Assumptions...................52 IV. ANALYSIS AND DISCUSSION OF FINDINGS............. 54 Review of Descriptive Statistics............ 56 Age........................................ 56 Gender.....................................56 Education..................................59 Occupational Status.......................63 Income.....................................66 Life Satisfaction.........................66 Future Orientation........................69 Internal/External Locus of Control.......72 Discriminant Analysis........................77 ______ _ _ _ _____ _ _ _ _ v Results......................... Discussion......................... SUMMARYr CONCLUSIONS, AND RECOMMENDATIONS Summary of Findings............... Conclusions and Implications..... Demographic Factors............ Attitudinal Factors............ Implications...................... Recommendations................... Summary............................ REFERENCES................................ LIST OF TABLES i j Page Table t 1. Tabulation of Responses for Age.............. 1 Table i 2. Mean Scores for Age by Category...... ....... t Table t 3. Tabulation of Responses for Gender........... .60 1 Table t 4. Education: Highest Grade Completed........... i Table 5. Mean Scores for Educational Attainment...... .62 Table i 6. Distributions of Ratings for Occupation..... .64 Table 7. Mean Scores for Occupation................... I Table 8. Distribution of Income by Category........... Table 9. Mean Scores for Income........................ Table 10. Total Scores on the Life Satisfaction Index.. .70 i Table 11. Mean Scores for Life Satisfaction Index..... .71 Table i 12. Distribution of Scores on Future Orientation. .73 Table 13. Mean Scores for Future Orientation........... Table 14. Distribution of Scores on the IE Locus of Control Scale................................. Table 15. Mean Scores for Locus of Control............. Table 16. Group Means for All Predictor Variables..... Table 17. Canonical Discriminant Functions and Pooled- Within-Groups Correlations Between Discrimin ating Variables and Canonical Discriminant Functions..................................... Table 18. Canonical Variables Evaluated at Group Means and Group Means Plotted Against Discriminant Functions..................................... I . Table 19. Pooled-Within-Groups Correlation Matrix.......87 Table 20. Classification Matrix.......................... 88 v i i i LIST OF FIGURES Page Figure 1. Plots of Group Means Against Discriminant Functions One and Two........................ 83 CHAPTER I STATEMENT OF THE PROBLEM I ‘ Introduction i j While a substantial body of literature exists describing i ■ participation rates in educational activities by older persons (Anderson & Darkenwald, 1979; Graney, 1980; Carp, Peterson &; ] ] Roelfs, 1974; Cross, 1981), not as much attention has beeni I 1 i paid to the factors associated with engagement in educational I activities by older students (Brady, 1984). In addition, 1 i insufficient data exist which can satisfactorily explain the reasons for the participation of older adults in educational I activities. ! The United States has always had a relatively young i population; m 1790, the median age of white males was under I 16 years. For the whole population, median age rose to 22 years in 1890 and to 27.9 years in 1970. By the year 2030, i ' however, the median age is expected to be within the range of, 32 to 37 years (U. S. Bureau of the Census, 1987). i j The number of persons 65 and older is projected to increase from approximately 22.9 million in 1976 to about 29.8 million in 1990. The combination of declining mortality and the aging of the post-World War II "baby boom" population is j expected to result in an elderly population of 55 million by ! 2030— a jump from 14 to 22 percent of the United States population* The average life expectancy is anticipated to increase slowly but steadily for both males and females from 69.1 to 71.8 and 77 to 81 years, respectively (Schick, 1986).! Certain social factors impact the older person which | I prompt educators and social commentators to suggest education1 i I |as an appropriate ameliorative agent to help the older person' improve and maintain quality of life and sense of satisfaction with life. The science-based, post-industrial technology of modern economics has, among other things, led to vast increases in productivity, disposable income, leisure time and educational attainment. Moreover, growing complexity and i change characterize not only technology and work, but also I social relations in marriage, family, and community. As the society changes, so too, do individuals, and education is seen i as an important vehicle for such individual change (Trent & i Trent, 1977). ! Perhaps even more important is the increase in I jeducational attainment by the population of the United States, |where the median number of years of schooling completed by ! adults 25 and over increased from 8.6 years in 1940 to 12.5 in 1979. Over the same time span, the percentage of adults 25 and older who completed 4 or more years of college rose jfrom 4.6 to 16.4 (Gross, 1982). Numerous studies have shown ithat those with more preparatory schooling are much more i i 2 likely to continue their education as adults than those with i less schooling (Cross, 1981; Peterson, 1981; Russ-Eft & Steel, 1980). In addition, there are gender differences in participation rates in adult education. Demographics of participation in adult education consistently report more females participating than males in educational activities and :he trend grows. The National Center for Education Statistics; i (1978) cites a participation rate of 12.7 percent for women| as opposed to 10.7 percent for men in adult education. U.S.j i 3ureau of Census (1987) data indicate that this has increased* :o 14.1 and 12.8 percent for women and men, respectively. j The last quarter century has seen a general increase in® disposable income and in the amount of leisure time available* to most workers in the United States. For example, median' I family income nearly doubled from $8,991 in 1950 to $17,640 i in 1978 (Gross, 1982). Median family income nearly doubled i again from 1978 to 1985 (the last year for which figures are available) reaching $31,100 (Statistical Abstract of the, i i United States, 1988). i I Total leisure time for the average urban adult increased i from 34.8 hours per week in 1965 to 38.5 hours in 1975 (Gross, . 1982). Changes in the work/leisure ratio are underscored by the continuing trend for many organizations to encourage early i retirement for persons as young as fifty-two. Indeed, most ■ people now have 25 years of life after retirement. i In summary, specific socioeconomic and demographicj factors combine to make increased participation of olderj adults in educational activities a reality. As many writers |(Cross, 1981; Gross, 1982; Peterson, 1975; 1981) point out, higher participation rates in education are demonstrated by persons with higher income and educational attainment. More; ! I and more older Americans qualify for inclusion in these! I ' categories and can therefore be expected to participate in| educational activities. \ | i j While specific socioeconomic and demographic factors1 i exist which make increased participation of older adults in i educational activities a reality, the participation rates of! i t persons 55 and older continue at about one-third that of j persons in the 35-54 age bracket. In 1981 Cross reported a i *4.5% rate of participation compared to 13.7% for the younger jadult group. In 1987, Snyder reported a 5.7% participation rate for over-55 persons compared to a 16.9% rate of1 [ i participation for persons 35-54. ! i At the same time, theorists (Peterson, 1975; 1981;j Sihvola, 1981; Trent & Trent, 1977) propose that older adults increase their sense of personal control and sense of! I l satisfaction in life through the medium of education. Most I i studies of older adults' participation in educational' \ i I i activities have attempted to correlate demographic variables! i I I with rates of participation. This attempt, however, accounts for only ten per cent of the variance associated with i i participation in education (Anderson & Darkenwald, 1979). \ l . I Researchers call for study of other, non-demographic variables such as trait and attitudinal variables. Although researchers have investigated the relationship of such variables as life satisfaction and educational' i ] i participation (Brockett, 1986; 1987; Fisher, 1986; Mizer,: j i 1975; Okun, Stock & Covey, 1982) and locus of control andj educational participation (Falconer, 1974; Hooper & Traupman, ! 1983), the results of these studies are inconclusive at best1 , and fail to address significant attitudinal variables. This; | i study attempts to address this deficiency. i Purpose of the Study ! i The purpose of the study was to advance information about’ t i important aspects of older persons' attitudes, traits, and characteristics relative to educational activity. Research i remains conflicting, inconclusive or absent relative to life! satisfaction, future orientation, and locus of control, and the relationship of these factors to the participation or non participation in educational activities of older men and' I women. This study investigated these relationships in order i to identify variables that may be used to predict ! participation in educational activities by older men and women. i 5 flesearch Problem i Although theorists propose that adults increase theiri i ; sense of personal control and their sense of satisfaction with I ■ • life over the lifespan and into old age and view education as: an agent that effects those ends, there is conflicting! ^research data to support this position. While it is known| that locus of control tends to become more external across the i lifespan, it is unknown what this tendency is in older adults who engage in educational activities, and if the tendency is‘ j f jthe same for men as women. The relationship between future orientation and educational participation is unknown., i Research is inconclusive about the relationship between j i educational activity and satisfaction with life. | The research problem was to determine whether differences in attitudes and characteristics of educationally active and educationally non-active males and females may be useful in i prediction of educational activity. The factors which were » investigated in this study included demographic and attitudinal variables. Specifically, the research 'investigated the differences in occupation, income, prior 'education, life satisfaction, future orientation, and locus i of control among educationally and non-active older men and i women. Analysis of these data yielded information about the variables identified above to discriminate between four groups of persons: educationally active males and females and educationally non-active males and females. Statement of Hypotheses Statement of Hypotheses 1. Older persons who are engaged in an educational activity will have a higher measure of occupational status than those' i persons not engaged in an educational activity. i >2. Older persons who are engaged in an educational activityj will have a higher level of education than those persons not: j engaged in an educational activity. I 1 ■3. Older persons who are engaged in an educational activity will have a higher level of income than those persons not; | i engaged in an educational activity. 1 j4. Older persons who are engaged in an educational activity will demonstrate a higher level of life satisfaction than \ those persons not engaged in an educational activity. i |5. Older persons who are engaged in an educational activity will demonstrate a higher degree of future orientation than i jthose persons not engaged in an educational activity. 6. Older persons who are engaged in an educational activity |will demonstrate a more internal locus of control than those i person not engaged in an educational activity. I 7. Low socioeconomic older adults who participate in 1 educational activities will be more similar to high l I socioeconomic older adults on measures of locus of control, l I future orientation, and life satisfaction than to low 1 i 7 socioeconomic older adults who do not participate in educational activities. t i i i Significance of the Study j i This research will serve to expand and substantiate a< l I small literature that is currently inconclusive (Brady, 1984;j Brockett, 1987) about factors related to educational activity ! ! 'of older adults. In addition, results may clarify the i {relationship between attitudinal sets such as locus of control' ! and future orientation and participation in education. Educational planners have long noted the failure of using jdemographic descriptors by themselves for program development, noting the need to look beyond this type of variable to {explain participation (Anderson & Darkenwald, 1979; Cross, 1981). This study, therefore, may identify attitudes critical to participation in educationally-related behavior. In addition, the study will examine gender differences associated !with locus of control that have been unstudied until now. i 8 CHAPTER II REVIEW OF THE LITERATURE As pointed out in the previous chapter, the sheer numbers jDf older persons in the United States population is increasing, and these persons have increased leisure, income and an educational history which would predict increased' participation in educational activities. i This chapter is divided into three main sections. The! first reviews and summarizes present knowledge about the 1 participation of older Americans in educational activities.; The second provides an overview of four theories which' contribute to the foundation of adult education practice— behaviorism, cognitive learning, humanistic, and developmental i lifespan theories. This section describes the basic tenents of each theory, gives examples of empirical research each theory has generated and critiques the contribution of that i research to adult education. The third main section of this, i i chapter defines three constructs that derive from the theoretical orientations described above and reviews the i l literature for each relative to older adults' participation i in educational activities. Participation of the Elderly in education I ! Older Americans are participating in educational I activities in record numbers. Data gathered in 1984 by the National Center for Educational Statistics indicate that 5.7% j of people 55 and over engaged in some form of adult education, , up from 4.5% in 1982 (Snyder, 1987). This marks the highest j I number and proportion of older people involved in adult| education ever recorded by NCES. The rate of participation! in educational activities by the elderly, while much lowerj than that of younger persons, is a figure that is growing in, | i size and importance for our society as the total number of: older adults as well as their proportion of the total I population in this country continues to increase (Cross, 1981; i I Russ-Eft & Steel, 1980). j A number of studies over the past two decades have addressed the question of the factors associated with i participation in educational activities by older adults.1 Demographic factors have been identified and studied most * i often. Researchers consistently found that adult participants in educational activities are drawn disproportionately from certain segments of the population; the better educated, those employed in professional or technical positions, persons with a higher socio-economic status and those in lower age groups (Carp, Peterson & Roelfs, 1974? Hiemstra, 1976? Johnstone & Rivera, 1965). Level of education as the principal factor i associated with participation in organized educational ; i activity has been confirmed by Perkins & Robertson-Tchabo (1981) and Heisel, Darkenwald and Anderson (1981). Cross 10 1 (1981) states "Of all the variables that have been related to educational interest and participation amount of formal j schooling has more influence than any other. Virtually all; surveys show that the more education people have, the more! interested they will be in further education" (p. 55). While Cross refers to the participation rates of all ages of adults, her statement is applicable to the elderly as well. Fisher; | . t (1986) reported twice as many college graduates among! participants in a study of older adult participation. \ In addition to prior education, other factors have been studied. Age, gender, occupation, and socioeconomic status I : i 1 are factors that have been compared to participation in education. Most studies look at more than one factor. Graney p f and Hays (1976) found that both age (inversely related) and previous educational attainment (directly related) were significantly related to interest in further education. Heisel, Darkenwald and Anderson (1981) found participants, as i compared to non-participants, tended to have higher levels of educational attainment and income, and were more likely to be I younger, white and female. In a survey of attitudes of older women toward participation in education at the university level, Papalia-Finley et al. (1981) indicated that "...in general, highly educated women over 65 are interested in i j participation in continuing education programs offered by institutions of higher learning" (p. 165). 11 Differences in the socio-economic status of participants and non-participants are reported in the literature. Participants tended to have a higher socio-economic status! than non-participants (Johnstone & Rivera, 1965).! Participation of older adults in educational activities was | i found by Hiemstra (1976) to vary according to occupation as1 well as social class and educational level. : | : The National Council on Aging (1982) summarized1 t characteristics associated with participation: "being an; older woman, being a young older person (age 65-69 rather thanj 75 or older), having a higher educational level than the! median for the total population of older persons? [and] having | i an annual income of over $20,000" (p. iv). Most research has' focused on analysis of socio-demographic factors and yields ! i a clear picture of the older adult participant: more likely to be female than male, more likely to be under 69 than over, i I a high school graduate or more, with an above average income i and occupational status. Theoretical Bases of Adult Education i | As discussed in prior sections, older adults have increased leisure, income, and prior experience with education, but still a small proportion (5.7%) (Snyder, 1987) , participate in educational activities. As Cross (1981) points i out, it is necessary to look beyond demographics to explain participation. While adult education theory draws on a number i 12 of disciplines to explain participation, it depends primarily on psychological theories to explain motivation to participate. Four psychological theories that seem most i productive of explanation for adults' participation in educational activities— behaviorism, cognitive learning theory, humanism, and developmental lifespan theory— are reviewed below. The topics to be covered in the following I section are: behaviorism,‘including classical conditioning (S-: R psychology) , operant conditioning and social learning i . ! theory? cognitive learning theory including expectancy-valencej theory? humanism? and developmental lifespan theories.j Analysis of theories introduced above suggests three constructs— life satisfaction, future orientation, and locus I of control— that may be factors predictive of older adults'1 I participation in educational activities. These constructs are i defined, and the literature describing their relationship to educational participation is reviewed and critiqued in this isection. i i i I Behaviorism i Agruso (1978) summarized the behaviorist definition of learning as "change in an organism's behavior brought about by experience and practice not otherwise attributable to i l drugs, fatigue or maturation" (p. 24). That behavioral models of learning are appropriate for adult education, and for older adult learning as well, can be seen by the fact that classical 13 conditioning techniques are systematically used to demonstratej psychophysical attributes of learning in older adults'^ cognitive functions such as short-term and long-term memory,! speed of learning, verbal ability, numerical computation, and! inductive reasoning— all functions predictive of overall! | j academic performance (Willis, 1985). j ! A more important example of classical conditioning,' however, and one overlooked by educational experimenters andj i yet indirectly recognized by educational theorists, is the- i i tendency for affective responses to be associated— often very, strongly— to physical stimuli following stimulus-response associations. This is nowhere demonstrated more clearly than, t : in the recognition that adult education suffers from the I effects of past learning behavior in a powerful and, i prestigious element in the educational process— the school. Cropley (1985) says it best: "Many people acquire [through i unintentional S-R repetitions] attitudes, values, feelings and self-image which predispose them to avoid organized learning situations once they become adults" (p. 3789). Other writers (Boshier, 1971; Long, 1983; Rubenson, 1982) cite the ■ i importance of early socialization, attitude development and i prejudices towards learning and verify the power of S-R conditioning in shaping a person's behavior overall, and behavior vis-a-vis educational participation in particular. Operant Conditioning Operant conditioning rests on research and theory whose chief proponent, B. F. Skinner, identifies as including social^ language and cognitive behavior which is modified by the] stimuli (called reinforcers) which follow behavior. This is in contrast to S-R psychology in which behavior is modified by the pairing of stimulus and response. This technique has' yielded a process unquestioned in its effectiveness as ai | i change agent for behavior called behavior modification. Itsi applicability to older adults is demonstrated in nursing home1 I feeding experiments (Agruso, 1978) as well as in the classroom, |(Davenport & Davenport, 1984). These authors cite nine principles ' of effective adult learning techniques for the classroom, five of which are specific behavior modification i i ^techniques. ! It is entirely possible that Boshier's (1973) person/environment congruence model that satisfactorily I distinguishes dropouts from persisters could be more parsimoniously explained by reinforcement theory. Most j persons for whom the classroom environment, teaching methods and teacher interaction style were positively reinforcing would have attendance/persistence behavior reinforced and would therefore demonstrate "persisting behavior". Those persons for whom these factors were not positively reinforcing would demonstrate "non-returning" or "drop-out" behavior. The i 15, point here is that reinforcement theory can explain and predict this behavior. i i i While the concepts of behaviorism, both S-R conditioning and operant conditioning, have been demonstrated to be! | i applicable to older adult learning and therefore capable of| explaining that learning, behaviorism has been criticized as! lacking in ability to explain all learning. Tolman, as earlyi as 1932, noted that the same stimuli did not evoke the same, responses in all subjects, and called for a Stimulus-Organism- i Response (S-O-R) psychology in which responses to stimuli were i i ] seen to be cognitively mediated (Bandura, 1969). Social Learning Theory Social learning theory is one theoretical outgrowth of1 i the call for recognition of cognitive mediation of stimuli. t Social learning theory has as a basic premise that behavior results from the relationship between the environment and the individual. According to a primary proponent of social learning theory (Bandura, 1977), people solve problems within I l their minds and can foresee the probable consequences of i different choices. They can choose different actions as a result of this cognitive process. People can, to some extent,. I ! I control their behavior. "Behavior comes to be regulated by ■ antecedant stimulus events that convey information about probable consequences of certain actions in given situations" (Bandura, 1969, p. 19). | i i I 16 Rotter (1975) calls social learning theory a "molar theory that attempts to integrate the stimulus-response or reinforcement theories with the cognitive and field theories' of Alfred Adler and Kurt Lewin" (p. 57). Social learning! theory recognizes that behavior is affected by the! individual's interpretation of incoming information, his or; her beliefs about the world, and the self-regulatory processes! i : by which he or she evaluates and controls his or her own i actions. Like the behavioral tendencies, the cognitive, mediations, too, are viewed as learned responses to antecedant1 conditions, with an emphasis on learning from models. ! Research demonstrating social learning theory is, generally experimental in nature and examines four classes of variables (behaviors, expectancies, reinforcements, and psychological situations). There are literally thousands of i studies that have been carried out using an experimental i framework on school children, college students, and l i institutionalized populations. One of the few experimental studies using older adults (Krantz & Stone, 1982) attempted to alter expectancies of subjects under success and failure conditions. The three theoretical approaches reviewed above share a common conceptual basis; that is, they are all, essentially, i stimulus-response psychologies. A common criticism of all behavioristic theories is that they yield experimental data j 17 about variables that are too narrow, sterile and artificial! for the understanding of the rich range of behavior of humanj i beings. Further, cognitive theorists have called for theoryj and experimentation that include complex human behaviors suet* as feeling and emotional behavior, and the creative processesj j of imagination, inventing, thinking, problem-solving andj reasoning (Bijou, 1985). ! Cognitive Approaches i Cognitive approaches view human behavior in general and motivation in particular as determined by a process of decision-making, in which an active individual, seeking meaning and control of his or her environment, considers and jselects from among alternative ways of behavior. Prevalent in this approach is the value-expectancy model which views the motivation for behavior as a function of two general factors: F the individual's perception of the value of the outcomes j expected to follow a certain behavior, and his or her i expectancy— the perception of the means and likelihood of | (achieving these outcomes. The combination of the two factors, ] I generally assumed to be multiplicative, determines the strength of the tendency to direct action towards these outcomes, as reflected in choice, intensity and persistence of behavior. Rubenson (1975) adopted expectancy-valence theory to account for motivational problems manifested by dropouts from adult education classes. He maintained that 18 •learners persist if a course or learning activity satisfies' Ln important need, that is, has positive valence, and if they! jexpect to be able to cope with and complete the course, thatj i jis, if they have positive expectancy. Although placing an ^appropriate emphasis on the importance of expectations, the; model does not specify the factors that can be manipulated to! enhance valence and expectancy. Further, the theory has' yielded no empirical data relative to adult participation in I 'educational activities and therefore concepts deriving from i this theory will not be examined and related to this research.1 Humanistic and Developmental Foundations of Adult Education ^ A primary factor affecting learning behavior left i [unaccounted for by the theories described above is the impact that maturation has on human behavior. While developmental I psychologists in the forties and fifties focused on early- and middle-childhood developmental processes, psychologists in the i ; [sixties and seventies recognized that significant i developmental change occurs across the total lifecourse (Neugarten, 1977; Willis, 1982). This period also saw a recognition of and emphasis on humanistic tenets of individual freedom and responsibility. Since most adults want to and do! assume responsibility for their own lives and learning, adult 1 education practice has experienced a very considerable influence from humanistic and developmental theories. It is, in fact, fair to say that adult education practice rests on 19 (developmental and humanistic concepts (Birren & Woodruff,| ■1973; Brookfield, 1984; Elias & Merriam, 1980; Merriam, 1977;j \ : Mezirow, 1982; Willis, 1985; Wiltshire, 1964) that view the! 'adult person as free to make choices, capable of becoming; ! ! increasingly free, increasingly in control of his/her own. I i life, and growing and developing throughout the lifespan. I i Bloland and Walker (1981) identify humanism as being l "concerned with freedom to make choices and become responsible for oneself independent of extrinsic forces" (p. 65) J Holtzclaw (1979) writes, "... man makes himself free through free choices; he becomes what he wills" (p. 19). | Mahrer and Gervaize (1985) indicate that the humanistic approach to education is characterized by the intent to i humanize education in accord with humanistic values,! | philosophy and theory. The humanistic educator trusts the i i Rapacity of the individual for developing his own potential i and permits the individual the opportunity to choose his/her |Own way of learning. | Writers analyzing the philosophical bases of adult education return again and again to the concepts of freedom, i ^self-determination, self-control, empowerment, and to the idea jthat the person is free to choose and to bear the responsibility for that choice (Davenport & Davenport, 1984; Even, 1987). In addition, the appropriateness of these principles of education for the older adult learner is stressed (Brookfield, 1985; Cookson, 1986; Covey, 1983; Knowles, 1980; Merriam, 1977; Mezirow, 1985). ! As well as proposing that the individual experiences1 i freedom of choice, humanistic and developmental theories' assert that the development of the self continues across the lifespan. However, even though developmental theory asserts the capacity of the individual to grow and develop and thereby! change throughout life, writers on human behavior and development emphasize the tendency of the human personality! to remain stable over the lifetime. It is important to note' that while this theory proposes that "...changes occur distally, behaviorially, and interpersonally, the individual's i basic personality structure and internalized self remain fixed" (Mahrer & Gervaize, 1985, p. 2352). Back (1987) speaks \ ! of a "continuity of the self that is maintained throughout the vicissitudes of the lifespan" (p. 145). Studies of personality continuity show a valuing of the self, the j definition of the self, and self-evaluation. These global i measures show constancy, and have led to claims of i "continuity" (Thomas, 1980) verified by research (Bengston,1 Reedy & Gordon, 1985). I Further, most longitudinal studies have shown that' personality tends to be stable over the years and into old age (Britton & Britton, 1972; Moss & Sussman, 1980? Thomae, 1980). Louis Harris concluded in a 1975 study for the National i i | 21 Council on Aging that "At no point in one's life does a personj stop being himself and suddenly turn into an 'old person' with all the myths that involves. Instead, the social, economic i -and psychological factors that affect individuals when they were younger often stay with them throughout their lives" (p.; 72). ; Although there is a demonstrated and verified tendency jfor stability of personality and behavior across the life span and while this stability may derive from both psychological origins (personality traits) and social origins (socio- I economic factors), there is a small number of persons who a behave in a way that would not be predicted by knowledge of, 1 their prior conditions. Cross (1981) describes these persons < as "seriously underrepresented in organized learning i activities today: the elderly, those who failed to graduate from high school, and those with annual incomes of under i $10,000" (p. 53). However, a small percentage of persons who fit these categories in fact engage in organized educational 'activities. Cross (1981) describes them as being on the I ["...tip of the pyramid formed by adult learners..." (p. 53), i emphasizes the necessity to look beyond demographic factors I jto explain their participation in education, and suggests I 'attitudes as variables central to educational participation. | Humanistic principles underlie adult education practice and prompt adult educators to suggest the individual's sense of personal control, the person's focus on the future, and the person's sense of satisfaction with life as attitudes that may t be associated with participation. Educational theorists (Brockett, 1987? Brookfield, 1984; Mezirow, 1985) cite the importance of autonomy and control over one's life and the f f place of education in the development and maintainance of that1 control. While these authors see enhanced control overj i personal destiny as an outcome of education, it is entirely appropriate to question the direction (internal versus1 i external) of locus of control over the lifespan and its1 relationship to participation in educational activities, particularly for the aged person. The authors cited above propose that individuals become more self-directed as they i become more mature and that theoretically one might expect a trend toward an increasingly internal locus of control. i Besides a concern for locus of control, there is also an] ( ^mphasis on orientation to the future in adult education t theory. As Lindeman (1961) declared, "...all education is |always futuristic...and the adult learning process is held to constitute an effort towards self-mastery" (p. 3). Brookfield | H |(1984) reflects both the basic dependence of adult education |theory on humanistic and developmental concepts as well as i focusing on the futurity of orientation involved in education: t j"The aim of adult education is the nurturing of self-directed, empowered adults? such adults will see themselves as pro- 23 active, initiating individuals engaged in continuous re creation of their personal relationships, work worlds, and; I ] 'social circumstances, and not as reactive individuals buffeted! by circumstances. Adult education, even for the older adult, i affirms the possibility of adults to change their future" (p. 48) . Quality of life, degree of satisfaction with one's life and individual sense of well-being are all concerns for educators and other persons who interact with older persons. i Education is often cited by authorities as a medium for the jimprovement of the quality of life of older people (Crabtree, ;1965; Fisher, 1986; Gross, 1982; Hentges, 1980; Peterson, i 1981; Radcliffe, 1982; Trent & Trent, 1977), but most often this statement is made in a hortative rather than in a i normative sense. Commissions charged to study aging, such as jthe White House Conferences on Aging in 1961, 1971, and 1981, 'concluded their deliberations and presentations with recommendations for education to address problems delineated during the conferences (Peterson, 1981). Philosophically j ;oriented writers such as Crabtree (1965), Moody (1976), and j Trent and Trent (1977) cite education as the imperative link between the older person and the older person's quality of i ;life and resultant sense of satisfaction with life. In the j jfollowing sections three attitudinal variables— orientation to the future, sense of control and direction of control, and 24 sellse of sat"Tsfa ct'ionw"ith"life- -and their importance relative' to the participation of older adults in educational activities: will be delineated and discussed. ! I i Life Satisfaction | George (1979/ p. 210) has described life satisfaction as: i ! {"essentially a cognitive assessment of one's progress toward i desired goals" and by doing so, has emphasized the future time i orientation implicit in this concept. Lemon, Bengston, and I Peterson (1972) define the concept as "...the degree to which one is presently content with his general life situation" (p. 513). An important consideration that must be taken into account when conceptualizing life satisfaction is the point of reference from which the concept is measured. One approach is to look at factors that can provide an objective measure i of the phenomenon such as income, health, participation in I social activities, employment, and marital status, as Palmore i (1979) and Palmore and Luikart (1972) have done. The limitation of this approach, however, as Campbell (1981) has \ j pointed out, is that "it is not possible to understand the t psychological quality of a person's life simply from a jknowledge of the circumstances in which that person lives. jBy attempting to explain the person's sense of well-being on jthe basis of objective circumstances, we will leave’ i unaccounted for most of what we are trying to explain" (pp. An alternative to defining well-being by describing external conditions is to focus on the subjective perceptions of those persons being studied. Larson (1978, p. 378) terms well-being a "strictly internal construct" which requires an understanding of how subjects feel about themselves in terms; of satisfaction with their lives. In fact, the bulk of! j research on life satisfaction among older adults has stressed the self-reporting approach to measuring the concept i (Brockett, 1987). ! Life satisfaction can be seen, then, as a global, multi dimensional construct encompassing happiness, morale, and life jsatisfaction. As Okun, Stock and Covey (1982) point out, jthese constructs have affective, temporal and cognitive ■dimensions. For example, one major distinction between morale I jand life satisfaction may be that life satisfaction is focused Itoward the past while the aspect of life satisfaction ! identified as "morale" is oriented toward the future. All ! three constructs have affective concomitants; the cognitive i jdimension is an evaluative, reflective, thoughtful characteristic that may accompany affective response. I The relationship of life satisfaction to participation lin educational activities by older persons is uncertain partly because few researchers have looked at this relationship, Ipartly because the behavior "educational activity" has been I operationally defined differently in each study? and partly 26 'because sampling procedures are opportunistic rather than j planful. A review of studies comparing participation in j i I educational activities and life satisfaction measures follows.| Mizer (1975) compared life satisfaction measures, reading; i ability, and personality trait assessments of 75 persons age' i fifty and older who were enrolled in university classes withj those of 75 persons residing in a nearby retirement community.1 i Using the Life Satisfaction Index A (LSIA) defined as having' the following five components: Zest (versus apathy),, resolution and fortitude, congruence between desired and 'achieved goals, positive self-concept, and mood tone (Neugarten, Havighurst, & Tobin, 1961), Mizer concluded that "there was a significant difference in measures of life I satisfaction among older students attending classes and non- i students" (p. 110). She reported that "educationally active I |older people have a greater zest for living, better mood i ,tone, feel that they have achieved more of their desired jgoals, are more resolute, and have a better self-concept" than i Inon-students (p. 115). i ] Mizer's (1975) use of what might be assumed to be a primarily middle class population may limit generalizations \ ) of this study to non-middle class populations. In addition, jit is questionable whether her definition of a participant as i a person enrolled one time in a class is sufficient to define | that person as an "educationally active" person. 27 Okun, Stock and Covey (1982) reviewed seven studies f'rorci 1971 to 1981 which were designed to impact quality of lifej (and therefore life satisfaction) of older persons through! educational intervention programs. Of the seven programs,! | > three focused on retirement education and one each on ilfe; , enrichment, social support, citizen’s affairs, and health.. They reported that "three had a positive effect, one had a1 zero effect, one had a negative effect, and one had a tnix^d positive/negative effect on participants5 sense of Ilfs- l 1 i i satisfaction” (p. 525). The authors concluded that results; I " • 1 from these studies are inconclusive as to impact of education1 j : experiences on life satisfaction. Their analysis identities’ loose research design as a primary deficit of ail sove;; i studies since they all used pre- or quasi-experimental met,co>Sn! fusing Campbell & Stanley, 1966, nomenclature). The pr.ln? o.-.yi pitfall for these reviewers, as well as for the research - a5 , ! however, was the attempt to view the relationship betoc-n education and life satisfaction as a causal relatiannh- tc These are complex concepts, and as Ebell (1973) pointed out, not concepts easily explored using the experimental method nor yielding information about causal relationships. I | Fisher (1986) surveyed active older adults (active. meaning older adults who went to senior centers or other social gathering places) and compared educational attainment, sense of anomie, life satisfaction, self-directed learning, . awareness of learning sites and awareness of learning needs, A multiple regression analysis was used to explain relationships between the six variables described above and! i , j participation to assess the relative significance of each asj a predictor variable. Fisher reports that while participants Lvidenced a higher level of life satisfaction than non- t participants, the difference was not statistically significant! i ^p. 205). It is important to note that although the Fisher ! ! study produced inconclusive results relative to life! satisfaction and educational participation, that the morei jsophisticated research design used raises the level of confidence that can be placed on the results. Less confidence j i can be given the study, however, given the definition of ;"educational participant". He asked subjects one open-ended question, "List any classes, groups, workshops or conferences i you have participated in this year." Subjects who identified i i one educational activity were defined as "participants". It i i |is questionable that participation in just one class is I I sufficient to make a person an "educational participant". The ipresent study will correct what is seen as a deficiency in the ! Fisher study. Brockett (1986, 1987) compared life satisfaction measures |With engagement in self-directed learning activities of 64 ipersons. Using the Guglielmino Self-directed Learning Scale, he measured readiness to engage in a self-directed learning 29 project and compared these scores to scores on the Life ^Satisfaction Index A. He reported a "weak, but statistically! I 1 significant, positive relationship between the two variables,] suggesting that one who is high in life satisfaction is^also| llikely to be high in self-directedness" (p. 215). He! ^attempted a subsequent analysis of data using multiple1 regression techniques but conceded that "100-plus subjects are! necessary to ensure reliability of regression coefficients", j(p. 215). i Future Orientation j Underlying the emphasis in adult education on' developmental possibilities and personal achievement is the assumption of future orientation in the older as well as the younger learner. Sihvola (1985) comments that this emphasis I jon personal achievement applies not only to younger adults ibut "...continues through the life span and into old age..." I !(p. 538). Theorists propose that older adults maintain a i i [future orientational stance and that engaging in educational jactivities promotes this orientation (Brookfield, 1984; jbindeman, 1961). Willis (1982) comments that "...a lifelong view of education suggests the need for a reallocation of educational opportunity across the lifespan. The individual continues to develop across the total life course, and thus leducational opportunities must be provided to facilitate and f ! joptimize the development of older adults" (p. 10). This focus on the necessity for providing educational j opportunities for the elderly is reinforced in the opening statement on education produced by the 1971 White House Conference on Aging which stated "Our present value system implies that a person's future ends when he retires from the paid work force. Thus the society condemns its older members; ! to having no future and therefore no need to learn and to grow" (p. 217) . ; Brookfield (1984) quotes Eduard Lindeman who, forty years before the Conference on Aging, acknowledged this functional^ foreshortening of adult's lives by declaring that "Education1 ks always futuristic [and] a daring challenge to life that is | jto come" (p. 189). A future orientation is apparently healthy and conducive to productive social and personal behavior. A study of community-based male and female pre-retirees found i ! !that future-oriented older respondents were rated by psychiatrists as being more competent than non-future-oriented persons (Chiriboga, 1973). This was a clinical rather than an experimental study but demonstrated that future orientation among older persons correlated with psychiatric ratings of mental health. However, this finding was reported as data jtangential to the primary purpose of the study. The ;relationship of future orientational stance to overall mental health awaits further empirical study; the use of phenomenological or field study methods may be most productive. 31 of useful "data with “this population. ~ | t The importance of future orientation is iterated by! Havighurst (19 58) who stated "What the older person j ' accomplishes is determined by his self-concept, his; aspiration, and by society around him which encourages him in certain activities and discourages him in others" (p. 47).: ;In fact, as Lindeman (1961) points out, "...the attainment of adulthood is characterized by a growing awareness of self and by a readiness to make existential changes" (p. 123). Cross and Floria (1978) add, "Goals and objectives are more; important than age" (p. 8). It is important to note that the i above comments are philosophical and idealistic in nature; there is no literature comparing future orientation between younger and older cohorts and no literature comparing future I orientation of older participants to that of non-participants i In educational activities. The present study addresses that lack. The rationale and method for assessing future orientation both derive from recent analysis of the Life Satisfaction Index A (Neugarten, Havighurst & Tobin, 1961). i ’ Okun, Stock and Covey (1982) define subjective well-being, jas a "global, multidimensional construct encompassing i happiness, morale, and life satisfaction" (p. 526) and further assert that "three underlying dimensions are important for distinguishing among happiness, morale, and life satisfaction. i These three qualifying dimensions are affective, temporal, 32 and cognitive in nature" (p. 526). They identify the temporal ^dimension as continuous/ varying from the distant past through the immediate present to the distant future. This underlying! temporal dimension produces the primary distinction between the constructs of life satisfaction and morale. Lifej satisfaction is focused towards the past while morale is | | oriented toward the future, relative to a respondent's! immediate present. Four of the items on the Life Satisfaction1 1 Index A question the repondent's anticipation of future events i (see Appendix C). It is the subject's response to and score i on these four items that will yield an operational definition' j of the construct "Future Orientation". Locus of Control i | "The adult learning process is held to constitute an: i effort toward self-mastery" (Eduard Lindeman, 1944, quoted by: Brookfield, 1984, p. 189). Educational theorists persist in ! ascribing the older adult's progress toward self-sufficiency i(Sihvola, 1985), increasing control of one's life (Mezirow, f i 1982), and increased autonomy (Brockett, 1987) to educational experience. Locus of control is a concept derived from social t •learning theory which is used in this study to operationally define and measure automony and control in the individual's j life. Locus of control refers to the extent to which an individual believes reinforcements are contingent upon his/her i I 33 Actions. Expectancy of reinforcement is an individual, not] i i i a situational, variable. The expectancy of reinforcement is' perceived differently by different individuals. Those personS| ^ho perceive that they have little or no control over their! I | environment and are subject to the actions of powerful others,j ( fate, chance, or unpredictable events have been labeled1 I ^''externals". Those individuals who believe that their actions; i have a significant effect on what happens to them are called ’ ’internals" (Rotter, 1966). Rotter (1966) states, If a person perceives a reinforcement as contingent upon his own behavior, then the occurrence of either a positive or I negative reinforcement will strengthen or > weaken potential for that behavior to | occur in the same or similar situations. I If he sees the reinforcement as being j outside his own control or not contingent, that is, depending upon chance, fate, powerful others, or unpredictable, then the preceding behavior is less likely to be strengthened or weakened, (p. 5) ! Individuals who are primarily internal, as compared to i jexternals, are more alert to the factors in the environment 'that are useful in future behavior, make efforts to improve their environmental conditions, value skill or achievement ireinforcement, and resist subtle attempts to control their i behavior. In contrast, externals have lower expectations for reinforcements following success and they tend not to generalize their experiences of success and failure to 'expectancies of future reinforcement (Rotter, 1966). Rotter (1975) noted that there is a tendency to assume that there is a value orientation associated with being more internal than external in motivation. He indicates that this; position presents difficulties and points out that in some| situations it may be desirable to be external, as id i situations where people are dealing with failing abilities.| |lt has been suggested that typical life events associated with aging may cause a decrease in the belief in personal control in the latter portion of the life cycle (Bengston, 1973; Brim i & Wheeler, 1966). While the literature on locus of control is very large, most studies target other populations; therefore, there are I ; ivery few studies of locus of control in the older adult. No! jstudies have been conducted to determine the relationship i between locus of control and participation in educational Activities by older adults. One study (Price & Lyon, 1982) examined attitudinal components of the educational i jorientations of older adults and found a significant relationship between internal locus of control and the individual's perceived ability to learn. The authors conclude that "perceived ability to learn was directly related to a i feeling of control over one's environment" (p. 478). It is important to note that this is a study of the beliefs and jattitudes toward education, not of participation in organized i i i 35 educational activities. | i < One study compared the relationship between locus of; control and voluntary proticipation in informal adult (education programs. Falconer (1974) studied 57 female and 37 male subjects, mean age, 44.5, 29 of whom were over 55. Mean! | I level of education was twelth grade; 34 subjects had not graduated from high school. Subjects were not chosen at i random and the author specified that the groups would contain substantial numbers of individuals who had low participation j rates on voluntary educational activities, low socio-economic i status, and low level of formal education. The author concluded that no significant relationship was demonstrated between general locus of control and participation in learning activities, and that persons with an internal locus of control i orientation were neither more nor less likely to engage in informal educational activities than a person with an external locus of control orientation. The inconclusiveness of this study characterizes the bulk of research conducted relative to locus of control and the i elderly. The major question that dominates the literature on locus of control and age is "Are there age differences or aging-related changes in locus of control?" Lachman (1986) i succinctly summarizes, "With regard to age differences and changes, the findings [of research] have been remarkably inconsistent" (p. 34). Nerke, Hulicka and Morganti (1980) i i summarized their study of locus of control in older adults by 36 'concluding that "virtually any directional hypothesis of locusj of control [in older adults] can be supported by the^ literature" (p. 26). j | Possible explanations for discrepant findings include; differences in sample composition study design and measurement instruments. Many research reports do not provide information about relevant demographic data about subjects.) ! When this information is given, subjects are seen to come froirl such disparate samples as low socio-economic, rural elderlyi i (Falconer, 1974), ambulatory residents in a Veteran's Administration home (Nerke, Hulicka & Morganti, 1980), and college alumni (Hale & Cochran, 1986). Almost all the data I on locus of control in the elderly seems to have been carried! ! out on opportunistic samples. I As for design, both cross-sectional and longitudinal studies have been conducted, although the majority are cross -sectional so as to provide the young-old comparison in locus i of control. No consistent patterns are demonstrated in outcomes using cross-sectional versus longitudinal designs j(Lachman, 1986). Several measurement instruments have been developed. An important distinction between locus of control instruments is the number of dimensions that are included. A multidimensional representation acknowledges that there are multiple sources of control such as chance, self, or other i i I 37 people. The survey instruments developed by Levenson (1974) and Paulhus (1983) are examples of multidimensional | I instruments. Most studies of locus of control and aging have; used unidimensional, generalized measures such as Rotter's; I (1966) IE Scale. This provides the rationale for using thisj ] j instrument in the present study since there is disagreementj among researchers as to the multi- or uni-dimensionality of, i ! the construct of the locus of control (Lachman, 1986). ! \ The relationship between age and perceived control of i reinforcements has been investigated by many researchers, but i ! there does not appear to be any agreement as to the exact1 I nature of the relationship. Several studies demonstrate that elderly adults score more "internal" than younger subjects1 (Staats, 1975; Strickland & Shaffer, 1971; Wolk & Kurtz,, 1975). Other research reports no significant differences; between elderly and younger adults (Lao, 1974; Ryckman & Malikiosi, 1975) and still other research finds the elderly to be more external than younger persons (Box & Peck, 1981;; i Bradley & Webb, 1976; Molinari & Niederke, 1984-5). j : i Locus of control has been studied in relationship to; other factors affecting the elderly. Previous research i indicates that a more internal locus of control is associated i with satisfaction with life and successful aging (Bengston, i 1973; Kuypers, 1972; Palmore & Luikart, 1972; Wolk & Kurtz, 1975), higher level of activity (Brown & Granick, 1983; 38 1 Lumpkin, 1986), and higher socio-economic status (Rotter, 1966? Cicirelli, 1980). Questions relative to gender differences in locus of| i t control are poorly addressed or ignored by research. Therei j are mixed findings indicating only a weak correlation of sex j(male) to internal locus of control (Palmore & Luikart, 1972? Cicirelli, 1980), no correlation (Kuypers, 1972), or strong' I ! correlation of older females with a more external locus of | i E control (Lachman, 1986). i j New evidence (Lumpkin, 1986), using a large national i j sample, purports to answer the age/locus of control question; 1 ' definitively. Lumpkin comments, "Theory and intuition suggest i ! the elderly should be more external. At best, prior research i is equivocal because it is based on relatively small samples( from limited locales. This study, using a national group of! i ; over 3,000 persons and including the non-elderly for comparison demonstrates that the elderly are significantly (at the .05 level or better) more external" (p. 250). This study is especially significant because of the size and characteristics of the sample. Most studies of locus of | I control in the elderly use samples of under 100? this study ! I used a large panel sample of over 4,000 which yielded 3,000j | i subjects. The panel, in general, had a higher socio-economic i level, a higher educational level, greater home ownership and a higher employment rate than the general population. [Previous citations have indicated that a more internal locus] of control is associated with higher socio-economic status. jThe significant trend towards externality of locus of control! 1 i !in this higher socio-economic status group is, therefore,] | 1 jimportant. j Not referenced in this large, definitive study is thej question of gender differences in locus of control. This, deficiency in information on this subject provides one of the jthrusts for investigation for the present study as well as i ithe need to know the relationship between direction of locus |of control and participation in educational activities by' ; older persons. i Summary [ This chapter has presented a selected resume of related research and professional literature relevant to a sound and comprehensive foundation of this study. This literature [ review has defined older adults' participation in educational activities and identified the demographic variables most often, ■studied relative to participation. It has shown that prior! ! j ieducation, income, and socio-economic status are related to 'participation in educational activities. i I In addition, four theoretical bases for adult education- i behaviorism, cognitive theory, humanistic psychology, and life jspan developmental theory— were identified and the: ! contributions of each to older adult education were 40‘ summarized. Further, three constructs— future orientation, i 1 locus of control and life satisfaction which derive from these1 i i i theories— were defined and described and the literature! | i relating each construct to older adult participation in: education was reviewed, and where appropriate, critiqued.| These reviews demonstrate conclusively the absence of I I : information or the inconclusiveness of results from prior research, clearly showing the need for the present study. CHAPTER III METHODOLOGY In chapter one, the literature presented indicated that I older adults, while growing substantially in proportion to the] i rest of the United States population, do not participate in! ‘ formal educational activities to the same extent as do young’ i 'and middle age adults. Chapter two reviewed information about( i what is known about older adults’ participation in educational I i jactivities and presented a theoretical background that suggests attitudinal orientations (sense of life satisfaction, I future orientation, and locus of control) m addition to demographic factors of (age, gender, income and socioeconomic status) may account for differences in participation in' educational activities. This chapter presents the methods and procedures utilized in examining differences among subgroups. Subsequent to a1 description of the sample, attention is given to procedures; i ]and instrumentation in this investigation including a review; ; 1 ,of the pilot study. Variables used to form groupings and predictor variables are defined and described, and an overview I of the statistical method in the data analysis is presented, ^followed by methodological assumptions underlying the study. I i i 1 42 j Study Sample Educationally Active Group Subjects in the educationally active groups were defined as males and females, age 55 or older, who have enrolled in at least three community college classes at South SeattleJ Community College, Subjects having taken or having been’ enrolled in at least three courses were identified as; ( "learnersM or "ongoing educational participants" and were differentiated from incidental or one-time enrollees who may have been taking a class at the time of the survey but who’ 'could not be called a "usual" or "ongoing" educational’ l participant. This distinction distinguishes this study from former studies of older adult participants in educational activities. Subjects were drawn from lists of interested I persons or past students at South Seattle Community College who had taken classes in the arts, humanities, and social sciences through the "Over 60" program. Educationally Non-active Group Subjects in educationally non-active groups were males and females, age 55 and older, who resided in income- 'subsidized housing in units administered by the Kitsap County i Housing Authority in Kitsap County, Washington, and older i l adults on the community college mailing list who had never attended. The selection of educationally non-participant subjects in low-income housing units was based on findings by researchers who report that the lower the older person's I income, the lower the incidence of participation in educational activities (Carp, Peterson, & Roelfs, 1974; Johnstone & Rivera, 1965; & Ventura & Worthy, 1982). Moreover, the low-income housing units in which the survey was disseminated were relatively new, and subjects for these subgroups were comparable in age to subjects who were! 1 I educational participants. Consistent with the 1982 NCOA: j summary report, the second most noted characteristic of the i j * older educational participant was that they were a "...young older person (age 65-69 rather than 75 or over)" (p. iv.). : General Considerations About Subjects * | In older adult populations, women are represented more frequently than men (Merriam & Dimmock, 1985). Moreover, females participate in education in the same proportion in the 65-69 and 75 and over categories (Hooper & March, 1978). Therefore, a larger number of females were expected in this study. j Equal numbers of subjects in each group are not a necessary requirement in discriminant analysis and therefore i no attempt was made to equalize the numbers of subjects in the I I four groups. While former studies report easy direct access to older persons as subjects (Fisher, 1986; Mizer, 1975; Price' & Lyon, 1982), both organizations contacted imposed strict 44' ^requirements on confidentiality and insisted upon preserving jthe privacy of both students and residents. This insistence jtook the form of managers or administrators receiving the i l ‘ surveys from the researcher and handing them out or mailing ! them to their residents or students themselves. Procedures and Administration of Surveys I j Because measures of demographic and attitudinal variables; ! j ^ere sought in this study, a survey format was deemed t ! appropriate. The survey format has been used extensively and ( i I f ' successfully with older populations (Flynn & DeVoss, 1986? Merriam & Dimmock, 1985; & O'Connor, 1987). While some researchers report a preference for using interview techniques' jto gather survey data (Fisher, 1986? Hiemstra, 1976? & Price * & Lyon, 1982) other researchers (Brady, 1984? & Ralston, 1981) i have employed group survey administration techniques, and i 'still other researchers report successful utilization of mail survey techniques (Daniel, Templin, & Shearon, 1977; Flynn & DeVoss, 1986? & O'Connor, 1987). i i | The survey consisted of an introductory letter, demographic questions, the Life Satisfaction Index-A, and the i 1 short version of the Locus of Control Scale. The Introductory letter sent to subjects in the educational I participant group was different from that sent to non- j participants; South Seattle Community College personnel insisted on the inclusion of the disclaimer that "prior 45 education does not affect ability to take and profit from community college courses" (see Appendix A). A preliminary survey was administered to ascertain ability of older subjects to understand and respond. A phone! interview was conducted to see if subjects understood the survey items and if they would respond the same way in a j j verbal interview as they did to the written survey. Of teni i surveys sent out, ten were returned. The researcher contacted ! | all ten persons and ascertained that verbal answers were in; fact the same as written responses. i ! Surveys to educationally active participants were mailed i at the beginning of the 1989 Winter Quarter to all persons who! had taken a class through the institution's "Over 60" program since its inception in the Spring Quarter, 1986. A total of: I 1294 persons were identified who had taken classes in one or more of five college quarters or were on a waiting list to do so. Stamped return envelopes addressed to South Seattle Community College were included in these surveys. Surveys and stamped, addressed envelopes were delivered to managers of Kitsap Housing Authority units. Residents i mailed completed surveys to the researcher. In all, 100 j surveys were handed out to low-income residents. Instrumentation Three demographic and three attitudinal variables were included in the survey. Demographic variables were: (1) 46 prior education, (2) occupation and (3) income. The three attitudinal variables were (1) life satisfaction, (2) future Lrientation, and (3) locus of control. The construct, life satisfaction, was operationalized as the score earned by a person on the Life Satisfaction Index, Form A (Neugarten, Havighurst & Tobin, 1961). The Life Satisfaction Index A is a multi-dimensional measure of the | | Construct, subjective well-being or life satisfaction. The; i t dimensions underlying subjective well-being are happiness,; zest, optimism, life satisfaction and morale. Items number1 J i J i 1, 2, 3, 4, 6, 7, 9, 10, 11, 12, 13, 15, 16, and 17 measure, I ; jthese constructs. Items are positively- and negatively-worded! statements that are randomly presented to prevent a positive i or negative response set. ! ! The construct of future orientation was operationalized! 1 i i 1 i as the score earned on four items in the Life Satisfaction i ! Index Form A that have been identified by Okun, Stock, & Coveyi ,(1982) as comprising the morale subscale of the index. These; ; i are items 5, 8, 14, and 18. ] The Life Satisfacation Index has been used extensively since its development by Neugarten, Havighurst & Tobin (1961). Formal tests of reliability and homogeneity of the Life Satisfaction Index have been conducted. Adams (1969) i ' evaluated the reliability of the LSI-A using a discrimination value (D values) and a biserial correlation between the mean i 9 47 of the affirmative-response groups for each item and the LSI- ! A mean score for the entire sample. The D-values indicate that all items except item 11 fell within the acceptable range j - from 20% to 80%. Neugarten, Havighurst & Tobin (1961) presented several :ests of validity. They correlated the LSI-A with expert I ratings called the Life Satisfaction ratings (LSR). This! I i instrument consists of five rating scales for the five i i components of satisfaction (zest, life satisfaction, optimismi j i and morale). Ratings are based not on the respondent's self- reports but on inferences drawn by the judge or rater from all i information available, including information on the i respondents' interpersonal relationships and how others reacted to the respondent. This, in addition to a comparisoni i < of scores on the Life Satisfaction Index-A to ratings made by1 a clinical psychologist, yield a correlation of .39, and constitute the evidence for content validity of the LSI-A. Lohmann (1977) examined the correlations of the LSI-A with six I : other measures of psychological well-being. Correlations of < .these measures with the other indicators of life satisfaction' i ranged from .39 to .88, constituting further evidence for content validity of the LSI-A. | The construct of locus of control was operationalized as l the score on the Rotter Internal-External Scale, short form. i Evidence exists regarding the appropriateness of use of the' I I < 48 scale with the ..elder ly_._ The_scale_may__.be used_wi.th_either] Jquestionnaires or interviews. Powers (1982) indicates that [no special training is needed for its administration, and Ldds, "as revised, it is a good scale for use with older! j samples" (p. 241). Formal tests of reliability of the IE Scale are presented by Rotter (1966). Internal consistency estimates (split-half, ! Spearman-Brown, and Kuder-Richardson) range from .65 to .79.. i ' iTest-retest reliabilities range from .49 (males, two-month | j i Interval) to .83 (females, one-month interval). ! J Rotter (1966) reports associations with measures of' ^intellectual performance that ranged from -.22 to .03, and a; i * positive correlation of .24 between the IE Scale and the’ S : I Taylor Manifest Anxiety Scale as indications of measurement: I of a construct different from either intelligence or anxiety. Criterion Variables i . | The criterion variable was participation or non-. I participation of older adults in education activities. The j i four groupings were educationally active males and females, 1 i and educationally non-active males and females. Engagement j ’ in educational activity was operationalized as past or present- i ' enrollment and participation— at least three times in the past ! ; |two years— in an educational program sponsored by a community: i 'college. Non-engagement in educational activity was' i operationalized as being socially and otherwise active but not i 1 currently engaged in an educational activity. Non-engagement in educational activity was further verified by survey items designed to determine that the subjects were not engaged in‘ any informally organized form of education such as library-,! I . ! jvolunteer-, or church-related educational activities. | I j ‘ Predictor Variables i i I j Predictor variables in the study included priori education, or highest educational attainment; occupational' i ^ status, based on the SES rating ascribed to the reported* l I f occupation using Stevens & Cho (1981) ratings; and income,; I ; coded into six categories: under $3,000, $3,000 to $5,000; $5,000 to $10,000; $10,000 to $15,000; $15,000 to $20,000; and over $20,000. These income levels reflected average incomes' i for persons who frequently had been retired for ten or more years. i The three attitudinal variables previously mentioned were also used as predictor variables: attitude toward life satisfaction, future orientation, and locus of control. Operationalization of these constructs is discussed in the ^'Instrumentation" section. Data Analysis i j Discriminant Analysis was the statistical method used to jtest the hypotheses. The major purpose of discriminant function analysis is to predict group membership on the basis of a variety of predictor variables to determine if discrimination among groups is better than chance. The 50 ^emphasis may also be on interpreting the discrimination space in terms of the variables contributing most heavily to separation of the groups in that space. Lohnes (1985) describes discriminant analysis as essentially a mathematical and geometrical modeling of data on group differences which produces a spatial model. Discriminant analysis is a special case of regression i analysis encountered when the dependent variable is nominal,j i i i and, as in this research, a classification variable. Linear \ | functions are calculated which best separate the groups. i i \ demonstrating a special case of canonical regression. j ! Discriminant analysis assumes a multivariate normal! | : independent vector variable and a nominal dependent variable. The independent variable is called the predictor variable and | I the dependent variable is termed the criterion or grouping variable. The analysis also assumes equal measurement' dispersions (variance/covariance structures) for the1 populations under study, therefore Wilks' Lambda is applicable! 1 j ; as a test of significance. i In this study, the discriminant analysis function was: used to determine which predictor variables were significant in separating groups, the canonical variables as domains that; maximally differentiated between groups, the predictor variables that loaded on the canonical variables, and a i measure of correctness of classification of groups. The discriminant functions performed by this statistical technique included setting up the model so that successive orthagonal discriminant functions which would maximally separate the groups were performed until all possible] dimensions were evaluated. Discriminant analysis produced cross-product matrices of scores on predictor variables. Determinants of the matrices were found and ratios between1 ! r :hem provided tests of the hypotheses about the ability of the predictor variables to differentiate among groups. The variables that loaded on the canonical variables were those! j i which discriminated one group from other groups. Methodological Assumptions The following methodological assumtions were considered implicit in this investigation: , J 1. The research design, subject selection procedures, and ! 1 data analysis techniques used in this study were appropriate to its intent. 2. The reliability and validity of the instruments used were; proven and sufficient for assessment purposes. 3. The subjects employed in this study constituted a I ‘ representative sample of older adults in the population in: general and in educational and non-educational settings. I |4. No problems are imposed by using unequal sample sizes in discriminant analysis techniques. ■ j 5. Discriminant analysis is robust to failure of normality of predictor variables with twenty or more cases. js. Multivariate outliers in predictor variable data will be: i ! removed, as discriminant analysis is highly sensitive to i ! multivariate outliers. i i ; ]7. Discriminant analysis assumes a linear relationship among! ! i all predictor variables within each group. I i i 8. A decision was made as to the a priori probabilities with: 1 i Which cases are assigned to groups. i i i i i i i i 53 CHAPTER IV ANALYSIS AND DISCUSSION OF FINDINGS j i Subsequent to a review of the sample and sampling! i procedures used to collect data, the predictor and criterion^ j I variables and a review of descriptive statistics for the data,! i I analysis of the data by discriminant analysis techniques is; presented. This is followed by a discussion of the findings. t The educationally active sample was drawn from lists of' i persons who had been or wanted to be enrolled in the "Over 60" program at South Seattle Community College. This sample was ] considered representative of groups in an institution of; i ■ higher education. Surveys were mailed to twelve hundred and; i 1 ninety four (1294) persons. Six hundred and eighty one (681) i surveys were returned; ten were blank, five lacked attitudinal I i ; survey responses and six were returned too late for data • < j analysis, making a total of six hundred sixty (660) surveys in the college sample. i i The educationally non-active sample was drawn from groups of residents in low income housing who were otherwise active: i i but did not participate in educational activities. This I sample was considered representative of educationally non- I active persons. One hundred (100) surveys were distributed |to persons in this sample; sixty seven (67) were returned, one (1) too late to include in data analysis, making a total of I : ! 54 sixty six (66) cases from the low income group. This study was concerned with participation or non- participation in educational activities by males and females 55 and older, which yielded four criterion (or group) variables: non-participant males and females, and participant males and females. The survey asked two questions about participation in educational activities: I Have you engaged in a learning activity through a j I library, museum, "Y", church or volunteer program , j in the past two (2) years? | i l Have you taken a class or course at a community or ' j four-year college in the past two years? I i . | The non-participant group was operationally defined as: males and females 55 and older who had answered "No" to both I : of the questions above. Sixty-three (63) males and females1 from the low income group met this criterion and were defined. ! . I as non-participants. Seventy-three (73) males and females* f from the community college sample met this criterion and were defined as non-participants. In all, the non-participant group had one hundred and thirty six (136) subjects, j The participant group was operationally defined as males and females who answered the second question "Yes"? "3 or I more". The participant group was comprised of one hundred and ! t twenty (120) males and females from the community collegej sample who met this criterion. I Altogether, both samples produced seven hundred and twenty three (723) surveys. Four hundred and sixty seven (467) cases were dropped because response categories did not jfit strict operational definition of non-participant or par ticipant groups, therefore data are based on a total sample of two hundred and fifty six (256) cases. The following i tables will present and review descriptive data for the two demographic variables, age and gender, and the six predictor i variables: education, occupation, income, and scores on life ! 1 satisfaction, future orientation and internal-external locus; of control scales. Mean scores for all groups (N = 256) will; t i be contrasted in the next eight tables with mean scores for! 1 the entire sample (N = 723) to show generalizability of^ | ! ■findings. I i i REVIEW OF DESCRIPTIVE STATISTICS ; i Age , i To report age on the demographic portion of the survey, respondents checked one of four categories. The values represented by each category were: (1) 55-59, (2) 60-64, (3) | 65-69, (4) 70 or older. It can be seen from Table 2 that the average age of respondents was age 65-69. This is true for j male participants and male and female non-participants.1 'Female participants, on average, tended to fall in the 60-64 jage group. Gender I | To report gender on the demographic portion of the survey, respondents checked category (1) male or category (2) ■Table 1. Tabulation of Responses for Age N = 256 N = 723 Age # % # % 55-59 13 5.1 42 5.8 60-64 60 23.4 168 23.3 65-69 67 26.2 199 27.6 70 + 116 45.3 311 43.2 i -Table 1 j Means 2. Mean Scores for Age (by category) of Non-participant Males/Females and Participant Males/Females Total Total i NP Males NP Female P Male P Female N = 256 N = 723 3.16 3.16 3.20 2.96 3.11 3.08 i (65-69) (65-69) (65-69) (60-64) ! 58 female. The values for these response categories are represented in Table 3. Data are reported for the responses | i from the entire sample (N = 723), the total sample (persons j whose responses fit the operational definitions of1 participants and non-participants, (N = 256), and the distribution of males and females in the participant and non participant groups. | The congruence between the two samples is further; Remonstrated by the consistent two to one proportion of j females to males in both samples. This two to one ratio of1 | I females to males in both groups provided the rationale forj I I setting prior probabilities at 2:1 for data analysis. I i i i I I i Education , ^ i ! Educational attainment ranged from the fifth grade levelj i : to twenty three years in the entire sample (N = 723). In the; non-participant/participant sample (N = 256) the range was five to twenty years of education. Table 4 shows distribution' i * pf prior education for all groups together. ! I It can be seen that non-participants as well as participants had, on the average, a twelth grade education, i with male and female participants demonstrating a higher ; i ' average level of education and non-participant males and females demonstrating a lower level of prior education (see Table 5). When data for all four groups were averaged, it was1 59 ‘ Table 3. Tabulation of Responses for Gender in Entire Sample j and Participant and Non-participant Groups j Entire Total NP p i [ Sample Sample Group Group N = 723 N = 256 N = 136 N = 120 # 36 84 I t Category # % # % # i 1 (male) 213 30.0 89 34.8 53 2 (female) 497 70.0 167 65.2 83 60 Table 4. Education: Highest Grade Completed Years # % 5 1 .4 7 1 .4 8 5 2.0 9 3 1.2 10 9 3.5 11 3 1.2 12 95 37.1 13 21 8.2 14 28 10.9 15 7 2.7 16 46 18.0 17 10 3.9 18 22 8.6 19 2 .8 20 3 1.2 N = 256 Mean = 13.71 S.D. = 2.37 i Table 5. Mean Scores for Education for Non-participant Males/Females and Participant Males/Females NP Male NP Female P Male Mean ! s.d . 13.81 3.03 12.39 2.04 15.22 2.46 Total Total P Female N = 256 N = 723 14.32 2.29 13.71 2.41 13.71 2.37 62 found that mean level for all four groups, participant and non-participant, was 13.71 years. This was exactly the same* I i as the mean number of years of education for all persons inj 4 | the entire sample (N = 723). : ! Occupational Status * Occupational status was determined by rating thej i occupation of each respondent. The socioeconomic ratingsj reported by Stevens and Cho (1985) which provide socioeconomic scores for the 1980 census occupational classification schemei were used. These ratings are based on the total work forcej I ! in the 1980 census and range from a low of 14.83 (personal1 ' i service - domestic) to a high of 90.45 (professional - physics i . ! researcher). These ratings are reported in Table 6. The decimals were not used in this table as they provided no, further discrimination between cases than the whole digits.j Occupational status of subjects in this study ranged from 14' i i to 89. It can be seen from Table 7 that participant males and j females had the highest occupational status, that non-| participant males had an occupational status mean higher than i that of the sample as a whole, and that non-participant females had a mean occupational status measure that was below: I f the average for the sample and below that of the entire- sample. i : The data for occupational status yielded a trimodal | | i curve. The modal values were 79, 52, and 30. The occupation: Table 6. Distribution of Ratings for Occupational Status PERCENTS PERCENTS VALUE COUNT CELL CUM VALUE COUNT CELL CUM 14 . 1 .4 . 4 3 1 . 9 3.5 34.0 17 . 2 .8 1 . 2 32 . 5 2.0 35.9 18 . 5 2.0 3 . 1 33 . 1 . 4 36 . 3 19 . 3 1 . 2 4 . 3 34 . 5 2 . O 38 . 3 20. 1 1 4 . 3 8 . 6 35 . 1 . 4 38 . 7 2 1 . 7 2 . 7 11.3 36 . 3 1 . 2 39 . 8 22 . 12 4 . 7 16.0 37 . 10 3.9 43. 8 23 . 1 1 4 . 3 20. 3 38 . 2 . 8 44 . 5 25 . 5 2.0 22 . 3 39 . 1 . 4 44 . 9 26 . 4 1 . 6 23 . 8 4 1 . 6 2 . 3 47 . 3 27 . 3 1 . 2 25 .0 45 . 2 . 8 48.0 28 . 3 1 . 2 26 . 2 46 . 7 2 . 7 50. 8 29 . 1 . 4 26 . 6 47 . 7 2 . 7 53 . 5 30. 10 3 . 9 30. 5 48 . 5 2.0 55 . 5 PERCENTS PERCENTS VALUE COUNT CELL CUM VALUE COUNT CELL CUM 49 . 1 . 4 55 . 9 67 . 1 . 4 8 1.3 50. 1 .4 56. 3 68. 2 .8 82.0 5 1 . 3 1 .2 57 . 4 70. 2 . 8 82 . 8 52 . 12 4 . 7 62 . 1 74 . 1 . 4 83 . 2 53 . 6 2.3 64.5 75. 1 . 4 83.6 54 . 12 4 . 7 69 . 1 76 . 4 1 . 6 85 . 2 55 . 2 .8 69.9 77 , 4 1 . 6 86. 7 57 . 4 1 .6 7 1.5 79 . 21 8 . 2 94 . 9 58 . 2 . 8 72 . 3 8 1 . 3 1 . 2 96 . 1 59 . 3 1 . 2 73 . 4 83 . 5 2.0 98 .0 6 1 . 1 . 4 73.8 86. 1 . 4 98 . 4 64 . 9 3 . 5 77 . 3 88 . 2 . 8 99. 2 65 . 8 3 . 1 80. 5 89 . 2 . 8 100.0 66 . 1 . 4 80. 9 Mean = 46.31 S.D. = 20.84 64 JTable 7. Mean Scores for Occupational Status for Non participant Males/Females and Participant Males/Females Hean S.D. NP NP Males Females 49.18 33.86 22.22 14.92 P P Males Females 54.55 53.26 19.58 20.13 Total Total N = 256 N = 723 46.31 47.08 20.84 20.63 represented most frequently (21 cases) had an SES rating of 79 — engineer. Income Income was reported categorically in this study. Six categories were used which corresponded to the following j/alues: (1) under $3,000; (2) $3,000 to $5,000; (3) $5,000 to $10,000; (4) $10,000 to $15,000; (5) $15,000 to $20,000; and 6) over $20,000. The range of these figures reflected median income figures of $8,991 in 1950 and $17,640 in 1978 (Gross, 1982). Some subjects wrote in the year of their retirement beside their income response category. It is apparent from inspection of Table 9 that males had higher average income than females whether they participated in educational activities or not. Female participants had slightly higher average income than the average for the entire group. Statistical analysis in a later section will emphasize the importance of these differences. Life Satisfaction The Life Satisfaction Index-A was the instrument used to measure subjects' attitude toward life satisfaction. Consisting of eighteen items, it is a multi-dimensional measure of the construct, subjective well-being, or life satisfaction. Four of the items were temporal in nature and! measured the respondent's orientation to the future. Response Table 8 i \ i i Distribution of Income by Category of Response for Non-participants and Participants i Category # % Under $3,000 6 2.3 $3,000-$5,000 16 6.3 $5,000-$10,000 36 14.1 $10,000-$15,000 32 12.5 $15,000—$20,000 39 15.2 Over $20,000 127 49.6 67 Table 9. Mean Scores for Income for Non-Participant Males/Females and Participant Males/Females Total Total NP Males NP Females P Males P Females N = 256 N = 723 Mean 5.32 4.08 5.88 4.73 4.80 4.89 Salary $15,000- $10,000- $15,000- $10,000- $10,000 $10,000 Range 20,000 15,000 20,000 15,000 15,000 15,000 to these four items is discussed in the next section on future orientation. Subjects' responses to the fourteen items which | ! measured life satisfaction (happiness, zest, and optimism) are reflected in Table 10. The Life Satisfaction Index-A is a series of positively- and negatively-worded statements about past, present and future achievements, feelings and expectations which are | i randomly presented in order to avoid a response set. In thisj | i study, the Index was scored to measure life satisfaction.! | i There are fourteen items on the index, therefore the maximum I i score that a person could earn was fourteen. Although two persons had a score of zero on the index, most respondents had! a positive score. Over fifty percent of respondents checked! ten or more items, indicating a high life satisfaction score. Differences in life satisfaction between males and females and i i non-participants and participants are shown in Table 11. I Future Orientation ; One of the dimensions in the life satisfaction scale was morale, which has a distinct temporal orientation. This time orientation is towards the future. Four items of the Life' Satisfaction Index-A measured morale, and therefore future orientation. These were items 5, 8, 14, and 18. Scores on i I 1 J See Appendix C for the response pattern demonstrating a life satisfaction score of fourteen. i ! i 1 69 ■ Table 10. Total Scores on Life Satisfaction Index Total Number of Checked Number Percent 1 7 2.7 2 3 1.2 3 9 3.5 4 9 3.5 5 6 2.3 6 16 6.3 7 12 4.7 8 16 6.3 9 25 9.8 10 29 11.3 11 26 10.2 12 37 14. 5 13 26 10.2 14 33 12.9* * (Does not total to 100%; 2 persons had a zero score). Table 11. Mean Scores for Life Satisfaction for Non- Participant Males/Females and Participant Males/Females Total I NP Males NP Females P Males P Females N = 256 £ i i i Mean 8.52 8.50 10.83 11.07 9.67 S.D. 4.10 3.66 3.22 2.84 3.65 Total = 723 10.06 3.32 71 'future orientation are recorded cumulatively; that is, some persons responded positively to all four items, some to three' jof the items, some to two, and some responded positively to; ‘ none of the items. Table 12 details the responses to future! orientation items. Inspection of Table 12 shows that nearly fen percent of respondents were not oriented to the future at i jail, but that half of the respondents responded positively to1 'three or four of four possible items. Inspection of Table 13; i j reveals that participant males and females had group meanj scores larger than those of the total group and larger than; I i mean scores for non-participant males and females. | Internal/External Locus of Control ! i | The third attitudinal variable considered as a predictor i : variable was the individual's locus of control which varied; i from internal to external. The construct was operationally! I defined as the score on the Rotter Internal/External Scale,, short form, which consists of eleven forced-choice items. 1 Each pair of items has one choice that expresses a belief in; personal control over life activities (internal locus of control) and one choice that expresses a belief in luck, jchance, or the behavior of others as the effective source of Icontrol over a person's life activities (external locus of control). The scale is scored by counting the number of items 2 j See Appendix C for the response pattern indicating a positive score on the measure of future orientation. 1 i i 1 72' Table 12. Distribution of Scores on Future Orientation Items Scores # % 0 24 9.4 1 34 13.3 2 72 28.1 3 89 34.8 4 37 14.5 Table 13. Mean Scores for Future Orientation for Non- Participant Males/Females and Participant Males/Females NP Males NP Females P Males P Females N = 256 N Mean 2.00 1.74 2.63 2.94 2.31 i i S.D. 1.28 1.06 1.01 .82 1.04 i = 723 2.49 1.10 iTable 14. Distribution of Scores on the IE Locus of Control Scale Number of Items Checked # 1 3 l.: 2 6 2.3 3 19 7.4 4 19 7.4 5 24 9.4 6 33 12.9 7 52 20.3 8 44 17.2 9 32 12.5 10 20 7.8 11 4 1.6 75 Table 15. Mean Scores for Locus of Control for Non- Participant Males/Females and Participant Males/Females Total NP Males NP Females P Males P Females N = 256 I Means 6.90 6.42 7.38 6.53 6.69 S.D. 2.22 2.32 2.49 1.92 2.20 Total = 723 6.72 2.22 76 that indicate a belief in personal (internal) control over one's life. A maximum internal score was eleven points. The numbers in Table 14 indicate the total number of items! subjects checked indicating an internal locus of control. Inspection of Table 14 shows that half of the respondents] in the sample had an internal locus of control? internal locus I of control was operationally defined as a score of six or more j pn the Rotter IE Scale, short form. The median score for this! , i ^sample was seven, as was the modal score. i DISCRIMINANT analysis ! I ] A direct discriminant analysis (Marascuilo & Levin, 1983?; i iTabachnick & Fidell, 1983) was performed using three1 demographic and three attitudinal variables as predictors of, j membership in four groups. Predictor variables were i educational attainment, occupation, income, and attitude itoward life satisfaction, future orientation and locus of ! [control. Groups were males and females who participated in: educational activities and males and females who did not; I participate in educational activities. j Of the original 723 cases, 102 were dropped due to- I i missing data. An additional 365 cases had grouping values \ ithat did not fit the criteria established by the operational i definition of participation used. Non-participants were j defined as those who responded "No" to both questions about 77 educational activity on the survey; participants were defined J i as persons who checked "3 or more" on the second question (See| | f Appendix A for items). For the remaining 256 cases,| evaluation of assumptions of linearity, normality, multi-j | I collinearity and homogeneity of variance-covariance matrices! revealed no threat to multivariate analysis. ! Prior probabilities for group membership were set using I data from the entire sample (N = 723); probabilities were set i i at 2:1 (see Table 3). Direct discriminant analysis was used, to identify the variables that separated the groups described above, and to identify canonical functions as domains that' distinguished one group from another. j | The analysis that follows presents information about predictor variables first, then data to show how predictor' i variables load on canonical functions, and finally, a■ comparison of group means plotted against canonical functions. Table 16 summarizes group means for the six predictor variables for the four membership groups and all groups1 ^together, and will be used as a referent point in later | •discussion. ! Results Two canonical discriminant functions were identified in this analysis of participation of older adults in educational i activities with a combined Chi-squared (18) = 148.99, p. < 1001. After the first discriminant function was removed, the i i i i 78 1 Table 16. Group Means for Predictor Variables Group EDUC OCCUP INCOM LSAT INTRL 13.81 49.18 5.32 8.52 6.90 12.39 33.86 4.08 8. 50 6.42 15.22 54.55 5.88 10.83 7.38 14.32 53.26 4.73 11.07 6.53 FUTOR NP Male 13.81 49.18 5.32 8.52 6.90 2.00 1.74 P Male 15.22 54.55 5.88 10.83 7.38 2.63 P Female 14.32 53.26 4.73 11.07 6.53 2.94 i t TOTAL 13.71 46.31 4.80 9.67 6.69 2.31 i i i 79 Table 17. Canonical Discriminant Functions and Pooled-Within-i Groups Correlations Between Discriminating Variables and Canonical Discriminant Functions Eigenvalues Canonical R Percent Variance Wilks 1 Lambda P Function 1 .45247 .5581396 64.96 .5510276 .0000 2 .21952 .4242709 31.52 .8003536 .0000 3 .02454 .1547646 3.52 .9760479 .1946 1 j Function 1 Function 2 Function 3 1 FUT OR i .70892* -.33015 .14553 OCCUP p .67243* .18367 -.47203 1 EDUC .59461* .24815 .24021 INCOME .48198* .75433* .09596 i INTERNAL | .10709 .26992 .26050 LIFE SAT .49056 -.25320 .57899* * Values must be .40 or above to be considered. 80 remaining discriminant function was also significant, Chi- squared (10) = 55.67, p. < .001. The between group variance accounted for by the two discriminant functions was 64.96 for! ! the first discriminant function/ and 31.52 for the second: function (see Table 17). Lohnes (1985) describes discriminant analysis asj i essentially a mathematical and geometrical modeling of data I on group differences which produces a spatial model. The data( in Table 18 relates the canonical variables to the grouping’ jvariableS/ non-participant males and females and participant males and females, and plots the means for these groups against the discriminant functions. Since a primary benefit: i of discriminant analysis is its capability for spatial discrimination and display of variables, the data in Table 18 are displayed graphically in Figure 1. 1 A plot of the centroids (see Figure 1) reveals that the first discriminant function separated participant males and1 females from non-participant females, with non-participant! i males found to resemble the other three groups in the study. Examination of the classification matrix (see Table 20) reveals that the non-participant male group resembles the other groups in the study. There were fifty three (N = 53) j non-participant males in the study (Table 3). The classification matrix correctly classified 18.9 percent or ten of these as non-participant males and incorrectly classified i : 81 Table 18. Canonical Variables Evaluated at Group Means and Group Means Plotted Against Discriminant Functions First Second Canonical Canonical Group Variable Variable NP Male -.00912 .60943 NP Female -.87256 -.18356 P Male .67140 .62679 P Female .63697 -.47177 Symbol Mean Group For Mean Coordinates P Male 1 -.67 .63 P Female 2 -.64 -.47 NP Male 3 .10 .61 NPFemale 4 00 rH • i 00 • \ i 82 z a z > n Figure 1. Plots of Group Means Against Discriminant Functions One and Two CANONICAL DISCRIMINANT FUNCTION 1 - 1 O 1 -2 -1 Group 1 Group 2 Group 3 Group 4 Participant Males Participant Females Non-Participant Males Non-Participant Females 83 82 percent (N = 43), 18 as non-participant females, 7 as participant males, and 18 as participant females. Because jthe non-participant group consisted of subjects drawn from^ both the community college and low-income samples, the group: was diverse, not homogeneous. The variables that loaded onj the first discriminate function were not able to discriminate! ! I between this group and the others in the study. The second i idiscriminant function separated males, participant and non-; □articipant, from females, participant and non-participant. ' i An examination of the structure matrix, the pooled- within-groups correlations between discriminating variables and canonical discriminant functions (see Table 17) suggests that three variables, future orientation, occupation and i i (education, load on the first discriminant function which discriminates participant males and females from non-i participating females. The primary variable in the first discriminant function was the attitudinal variable, future orientation. Participant males and females (means = 2.63, 2.94) had higher future orientation scores than non-participant females. These persons either have plans for or expectations about events j which will take place in the short- or long-term future. If ! !they have plans or expectations, the expectations are for events that are positive in nature. See Table 16 for a j •display of means for all demographic and attitudinal variables 84 for the four groups. I The second variable in the first discriminant function :hat distinguished participant males and females from non-, participant females was occupation. Occupation is a single-! number representation of socio-economic status which rangesj | i from 14 to 95 in the general population and from 14 to 89 | > (Mean = 46.31) in this sample. Participant males (Mean =' i i 54.55) and participant females (Mean = 53.26) had higher mean] i i i scores than non-participant females (Mean =33.86). ! I 1 l j The third variable in the first discriminant function that distinguished groups one and two from group four was t education. Participant males and females (Mean = 15.22, i i 14.32) had higher education scores than non-participant i females (Mean = 12.39). No other variables had loadings in I i excess of .40. | j The second discriminant function distinguished groups one’ and three (males, participant and non-participant) from groups two and four (females, participant and non-participant). Only one variable, income, had a loading in excess of .40 on the second discriminant function. Participant males (Mean = 5.88) | i and non-participant males (Mean = 5.32) had higher income than participant females (Mean =4.73) or non-participant females I (Mean = 4.08). i The pooled-within-groups correlations among predictor i variables are displayed in Table 19. Seven of the fifteen ! 85 bivariate correlation coefficients would show statistical j significance at p. <.01 if they were tested individually. i i ! tWith 254 degrees of freedom, any biserial correlation value i above .164 is significant. j The next question discriminant analysis answers is hovj |Well the variables classify cases into the four grouping i jvariables. The classification scheme in Table 20, using Isample proportions as prior probabilities (2:1 females to ! males) classified 55.9 percent of the cases correctly using i i ithe predictor variables that were identified. It is females,! i • participant or non-participant, who are classified most1 I t correctly. As a check against the correctness of this ; ! classification, a jackknifed classification procedure was I ;used. Using a jackknifed classification, 52.3 percent or 134 out of 256 cases were classified correctly. I DISCUSSION Findings from discriminant analysis suggest that the i i ifirst discriminant function separates participant males andj ;females as groups from non-participant females. The variables^ l ithat load on the first discriminant function are future orientation, occupation and education. Participant males and females have a strong orientation to the future and have both F P expectations and plans for the future. Further, these plans and expectations are positive in nature. Non-participant 86 Table 19. Pooled-Within-Groups Correlation Matrix 254 DF EDUC OCCUP INCOME LISAT INTRNL FUTOR EDUC 1.0 occu 1 .51 o • i —1 1 INCOM i .33 .27 1.0 LSAT .03 .07 .19 1.0 INTRN i -.10 o 0 • 1 .09 .26 1.0 FUTOR .04 .04 .29 .59 .15 Table 20. Classification Matrix Group | NP Male NP Female P Male I P Female I I i i Total Percent Correct Number of Cases Classified into Group NP Male NP Female P Male P Female 18.9 10 18 7 18 77.1 3 64 2 14 19.4 3 4 7 22 73.8 0 17 5 62 55.9 16 103 21 116 88 females do not demonstrate this attitudinal orientation to the I , future. i A second variable that loaded on this discriminant1 i function was occupation. Participant males and females had, higher occupation or socio-economic ratings than non participant females. A third variable that loaded on the ^first discriminant function was education, and participant; I males and females had a higher educational level than non participant females. J i i The second discriminant function separated males, participant and non-participant, from females, participant [from non-participant. Income was the only variable that I ! loaded on this function. From inspection of the means of the* four groups, it is apparent that males had higher incomes than [females. Participant males had higher mean income than non-1 participant males, and participant females had higher mean | income than non-participant females; however, the discriminant function separated males as a group from females. i i The first hypothesis proposes that a high level of' occupational status is associated with participation in educational activities. Discriminant function analysis does yield data that support this hypothesis. Participant males and females have higher socio-economic status than non participant females. Hypothesis one, as stated, is i substantiated by this research. Hypothesis number two stated that older persons engaged * in an educational activity will demonstrate a higher level of I life satisfaction than those persons not so engaged. The evidence from this study indicated that life satisfaction| | i measures, higher for educational participants than non-! participants (group means for male and female participants, j respectively, are 10.83 and 11.07, and for non-participant i i females, 8.50) loaded on canonical variables as a domain which 4 separated the groups. Hypothesis two supports the literature! I ! in this area. 1 i j Hypothesis number three states that older persons engaged in an education activity will demonstrate a higher degree of i future orientation than those persons not so engaged.- I Referring to Table 16, it can be seen that the group means for future orientation measures for participant males and females are higher than those for non-participant females. The third hypothesis is also supported in this study. | Hypothesis number four states that older persons engaged; i in educational activities have a more internal locus of i control than those who do not. Males have a more internal locus of control as measured by group means, 6.90 for non participant and 7.38 for participant males, than females. The group mean for female participants was 6.42, and the group mean for female non-participants was 6.53. In discriminant j analysis, locus of control was not a predictor variable that i 90 discriminated between any of the groups or loaded on a canonical variable. Hypothesis number four is therefore unsubstantiated by this research. Hypotheses 5, 6, and 7 werej similarly not supported by the findings. I | In summary, the relationship of three demographic 1 > variables (education, occupation and income), and three! * i ; i attitudinal variables (attitude toward life satisfaction, future orientation, and locus of control) to participation of' i i i older adults in educational activities was examined. Using a discriminant function analysis, it was demonstrated that i ■ i locus of control and life satisfaction as variables had no functional relationship with participation in educational I activities. Future orientation as an attitudinal variable,' \ and occupation, education and income were variables that i loaded on discriminant functions that separated groups from; 1 each other. i The groups that were separated by the first discriminant function were participant males and females from non-I participant females. The second discriminant function distinguished males, participant and non-participant, from females, participant and non-participant. I i 91 CHAPTER V ! SUMMARY, CONCLUSIONS AND RECOMMENDATIONS f i i | The number of older persons in the United States I i population is increasing. Concomitantly, the proportion ofi t t i lolder persons in the total population is also increasing. This i population of older adults possess characteristics (higher ! levels of socioeconomic status, income and educational; preparation) , that have been found to be associated with; i . i increased participation in educational activities (Carp,; I i Peterson, & Roelfs, 1974; Heisel, Darkenwald & Anderson, 1981; Hiemstra, 1976; Johnstone & Rivera, 1965). However, prior research (Cross, 1981; Anderson & Darkenwald, 1979) indicates i that demographic variables such as income and educational! i attainment account for just ten per cent of the variance in1 participation figures, suggesting the need to examine other kinds of variables, (i.e., psychological and motivational). The purpose of this investigation was to identify both i . I demographic and attitudinal variables which differentiated i between males and females who participated in educational activities and those who did not. A discriminant function ! analysis was utilized to determine which of six variables- education, occupation, income, attitude towards life: satisfaction, future orientation and locus of control - would 92 be useful in the identification of factors critical to i participative behavior. ! Summary of Findings i Discriminant function analysis yielded two canonical1 ; i ■functions which separated the four groups. Three predictorj variables (future orientation, socio-economic status, andl l I l | ’ education) loaded on the first discriminant function. This function distinguished participant males and females and non participant males from non-participant females. One predictor' ^variable, income, loaded on the second discriminant function and separated males from females, regardless of participation I : behavior. The remainder of this chapter will discuss these1 i i findings, implications of the discriminant functions for older; adults, adult education and administrators of educational i institutions, and recommendations for future policy and1 action. Conclusions and Implications i Demographic Factors I 1 Numerous studies have shown that those persons with more' i i i preparatory schooling are much more likely to continue their I i education as adults than those with less schooling (Cross,; !l981; Graney & Hays, 1976; Gross, 1982; Heisel, Darkenwald &! I Anderson, 1981; Perkins & Robertson-Tchabo, 1981; Peterson, 1975; Russ-Eft & Steel, 1980). Differences in the socio-i economic status of participants and non-participants are reported in the literature? participants tended to have a i higher socio-economic status than non-participants (Carp, i ; Peterson & Roelfs, 1974? Johnstone & Rivera, 1965). i I j Prior research has shown that participant males andi i females have a higher level of prior education and a higher level of socio-economic status. The findings of this studyl ; j were consistent with prior research. For non-participant females, lower levels of socio-economic status and education' were consistent with other findings in the literature. ! i j Review of means (Table 16) for educational attainment and socioeconomic rating for participant females shows them below, i participant males but above both non-participant males and < females. It is important to note the age of participant! females (Table 2)? this was the youngest age group in the i ' 3 study. Perhaps this reflected the beginning of the trend, now! established, of encouraging females equitably with males to! persue educational and occupational goals. ' Differences in income among participants and non-; | i participants have been reported in prior research which! I . . . . ! ^indicate that participants have higher income level than non- . . I participants (Cross, 1981? Gross, 1982? Peterson, 1975? 1981). In this study, differences in income were not found betweeni participants and non-participants, but between males and' !females, regardless of whether they participated in. educational endeavors or not. Findings from this research did 94- not support those in the literature. j Age and gender-related factors are important in' i understanding the findings. Difference in gender and societalj J ; norms for males and females in the study population could have; affected attitudes and values that greatly influenced male andj female employment and income patterns. Future studies could' possibly find a more even distribution of income between males and females. This finding could be anticipated because by | ; 1965, females were entering the workforce in almost equal; | numbers to men, as early as men did, and staying as long. Consequently, at the end of their worklife, female income1 levels would be closer to that of males. Participant females ■in this study population had comparable levels of education and occupation as participant males, and higher than either1 i i non-participant males or females. The life course for these women who had managed to break from norms that restricted other women was quite different from the usual or average i female work and life history, and one which perhaps established productive personal examples for females in later decades. I ! Attitudinal Factors j Educational theorists (Brookfield, 1984; Lindeman, 1961) ' i emphasize a future orientation extant in all adult education. Brookfield (1984) asserts the basic dependence of adult education theory on developmental concepts and emphasizes the concept of future orientation involved in education: "The aim i of adult education is the nurturing of self-directed, empowered adults [who] will see themselves as pro-activef ! initiating individuals engaged in continuous recreation of | their personal relationships, work worlds, and social circumstances, and not as reactive individuals buffeted by circumstances. Adult education, even for the older adult, affirms the possibility of adults to change their future" (p. j 48) . I | Future orientation discriminated among groups in the I study population in a manner which substantiated prior research. Participant males and females scored high on measures of future orientation. Non-participant females, as expected, had lower measures on future orientation scores. Social learning theory (Bandura, 1977) suggests that individuals who have experienced success in the past will expect success in the future. Participant males and females had high levels of education and high socio-economic standing. These are positive experiences that would allow a self judgment of "success" and would allow a person to anticipate I future positive experiences and events. The non-participant i female group in this study would not have experienced high levels of education, employment or income, and would be unwilling or attitudinally unable to look forward in time with positive perspectives. It appears, then, that gender differences underlie the differences in future orientation observed in this study.j Indeed, it may be safe to say that males, encouraged by values! I 1 and attitudes prevalent in their youth, prepared by adequate! education, and seasoned by four decades of competence in work! ! . . ! and income have a more positive assessment of their present! and, therefore, their future opportunity than their female coevals. Females, targets of opposite values and attitudes, i would not have this same positive anticipation of future occurences. j Since most of the persons in this study were in their i ! formative years almost fifty years ago, it can be concluded’ 1 J that societal values and attitudes favoring education of males I above females, a male work force and male income providers j shaped the differential behaviors, experience and attitudes between educationally participating males and females in this i study. Values and attitudes, no less than family and personal effort, affect opportunity structures, and the societal values I could have constituted very real barriers to the optimal development of educationally non-participant females in this age group. [ I Implications | ■ The implications of the findings and conclusions of this study for adult educators and administrators center on the j identification of certain societal values and attitudes as I I I 97 ' barriers to personal development for a female segment of the population who were in their formative years before World War ■II. These values and attitudes served to prescribe ! differential life courses for males and females resulting in' i ja tremendous late-life gap in educational attainment, income, and socio-economic standing among males and females. Those| females, however, who by chance or personal determination- 1 i secured early educational attainment and occupational! competency demonstrate a late-life pattern much closer to that I l of participant males. This discrepancy among female subgroups1 produces two problems for educators who provide programs for i older adults: they must not only address current needs produced by a rapidly changing environment, but must redress past omissions for opportunity in older females' lives. Developers of programs to meet current needs of older adults must be cognizant that change is not rapid, but rampant in i this era, and must provide programs for all females alike. ! Income was a factor that separated males from females,' i 1 whether participant or non-participant. Lack of money is a1 i ' female issue, as shown by the data in this study, and public I ; policy makers as well as senior adult educators must be fully, lapprised of this factor and its impact on the lives of i ^females. The societal values and attitudes extant during this group's formative years consitituted an unbridgeable social i barrier to education, occupation and especially income | 98' barrier to education, occupation and especially income ^opportunity. A social ill perpetrated informally through1 I i mores and cultural value systems must be redressed through (formal social policy and action: low-cost or no-cost 1 j jeducational benefits and opportunities must be made available^ ito all older persons, but especially to older females, j As plaintive testimony to this need, one female in the! i low-income sample wrote regarding the number of classes taken:. t ;"There is no money for classes". When a person is living on ^social security benefits, survival needs must be met first.; i A compassionate society and responsive policy makers and I ; ^educators must plan a system to make educational opportunities possible for a low-income group that comprises two-thirds of the largest growth in the population in the nation. i If Lindeman's (1961) brave challenge that "all education is always futuristic and the adult learning process is held to constitute an effort toward self-mastery" is true, then the .implications for educational policy and action are clear. Two-thirds of the fastest-growing segment of the total (population have experienced social and attitudinal barriers I large enough in their lives to diminish their sense of future .orientation; they may have given up that very spark that i 'comprises the motivational core of education activity. A society cognizant of its pluralism must acknowledge the pressing need of such a large segment of that society to have 99 access to educational opportunities that provide for personal mastery and increased control over life events. j i Access must be made available for all older persons, l regardless of personal income, to educational programs; r i I 1 ^necessary in an increasingly complex society. These findingsj prompt the following recommendations for political policy,] ! i educational action, and further research. ! l \ , Recommendations 1 I Higher education administrators must educate policy makers, law makers and members of legislative bodies empowered i to designate public funds for older adults. Public monies; 1 must be made available to educational institutions to offer 1 \ j no-cost or extremely low-cost programs to the over 65 female i segment of the population. Adult educators and administrators must maintain a keen social awareness in program planning so that curriculum, program and class offerings target and continue to target the very real and pressing needs of the older, low-income females in the population. Health, money management, information about changing public resources as well as recreation and i opportunities to meet and affiliate with other people are all* important content areas for classes to help elderly females maintain optimal control and effectiveness in their lives. The marketing tactics used to target this older low- income female population must take into account how little 100 jnoney there is for some women to purchase things often taken i for granted: newspapers, for example. Marketing strategiesj must include direct targeted mailings, posters inside low- income housing units, and dissemination of educational I opportunity information through churches, television or radio I that carry no cost to the user. J \ i I Last, further research needs to be conducted of older j adults who are active in non-educational activities such as | senior center activities and other group social events. It : ( would be valuable to know if the same factors studied in this research are associated with participation in social and: ; : recreational activities, since these types of activities would also influence adult future orientation. Summary | The findings in this study served to validate educators' conclusions that education is opportunity (Gross, 1982; Peterson, 1975; 1981; Sihvola, 1985). Furthermore, when persons are denied access to education by custom or barriers, such as income or geography, their entire life course and quality of life is affected. Males in this study had higher average amount of education and socioeconomic standing than females whether they participated in educational activities or not, and clearly had higher income than females. Poverty and opportunity become gender issues and all the more important for our society to face in view of the anticipated increase in the proportion of older persons in the total population and the consistent two to one ratio of females to males in this population. Amelioration is possible. As time progresses, it may be that the expectation of equity in opportunity for education and occupational preparation for males and females becomesj jstandard. As that occurs, the differences now observed among! females and males may decrease, to the benefit of all. j i Quality of life over an ever-lengthening life span is an Important societal issue and one which must be addressed at' I * fhe macro-level by policy makers and at the action level by i providers of educational programs for the elderly. Programs must be carefully planned to redress past omissions of opportunity and address current needs of all elderly persons. I ( ] By providing greater opportunity for educational activity for1 elderly females, participants' evaluation of the quality of; their lives and their future may continue to increase. Perhaps their sense of control over their life circumstances may become closer to that shown by males. "The aim of adult ^ducation is the nurturing of self-directed empowered adults i 1 j[who] see themselves as pro-active, initiating individuals! engaged in continuous recreation of their personal jrelationships, work worlds, and social circumstances, and not as reactive individuals buffeted by circumstance" (Brookfield, 1984, p. 48). Institutions of higher education, reinforced 102 by national policy promoting equity in educational Jopportunity, can and must provide the education necessary to make Brookfield's statement a reality. The ameliorative! i . i ^outcomes of education— quality of life and empowerment in' jliving— must be made available to all segments of the growing i group of elderly persons who comprise nearly one in every five l persons in the United States today, and who, as Willis (1982) i points out, continue to develop and grow across the life span^ and throughout old age. REFERENCES Adams, D. , (1969). Analysis of a life satisfaction index. Journal of Gerontology, 24, 470-474. Agruso, V. , Jr. (1978). Learning in the later years: Principles of educational psychology. New York: Academic Press. Anderson, R. , & Darkenwald, G. (1979). Participation and I persistence in American adult education. New York:I College Entrance Examination Board. App, J. (1973). Toward a working philosophy of adult education. Syracuse: Syracuse University publications! in continuing education. I Back, K. (1987). Continuity theory. In G. Maddox, (Ed.), The j j Encyclopedia of aging (pp. 144-145). New York: Springer.! Bandura, A. (1969). Principles of behavior modification. New i York: Holt, Rinehart, & Winston. | I I Bandura, A. (1977). Social learning theory. Englewood! ; Cliffs, NJ: Prentice-Hall. Bass, S. (1986). Matching educational opportunities with the able elderly. Lifelong Learning, 9^(5), 4-7. 1 ^ Baur, P., & Okun, M. (1983). Stability of life satisfaction j in late life. The Gerontologist, ^3(3), 261-265. I I Bengston, V. (1973). Self determination: A social and I psychological perspective on helping the aged, j Geriatrics, 33, 301-307. j i Bengston, V. , Reedy, M., & Gordon, C. (1985). Aging and self-' conceptions: Personality processes and social contexts. In J. Birren, & K. Schaie (Eds.), Handbook of the psychology of aging (2nd ed.) (pp. 544-593). New York: Van Nostrand Reinhold. 104 Bijou, S. (1985). Behaviorism: History and educational j applications. In T. Husen and T. Postlethwaite (Eds.),! I International encyclopedia of education (pp. 444-451) . ) 1 Oxford: Pergamon. I Birren, J., & Woodruff, R. (1973). Human development over thej | life span through education. In P. Baltes & K. Schaiei I (Eds.), Life span developmental psychology: Personality! and socialization (PP« 305-337). New York: Vanj Nostrand Reinhold. | Bloland, P. & Walker, B. (1981). A humanistic existential' perspective on career counseling. The Vocational ! Guidance Quarterly, 29, 197-204. BMDP Statistical Software. (1983). Los Angeles: University of California Press. t < Bolton, C. (1985). Lifestyle management, pro-action, and , educational efficiency. Gerontology, 11, 181-190. Boshier, R. (1971). Motivational orientations of adult education participants: A factor analysis exploration oft Houle's typology. Adult Education, 21(2), 2-26. i Boshier, R. (1973). Educational participation and dropout: ; A theoretical model. Adult Education, 23(4), 255-282. i | Boshier, R. (1985). Adult education: Motivation of participants. In T. Husen & T. Postlethwaite (Eds),! International encyclopedia of education (pp. 149-153). Oxford: Pergamon Press. i j Box, T. , & Peck, D. (1981). Perceived low self-actualization I and external fatalistic determinism as a function of aging. The Journal of Genetic Psychology, 139, 295-304. I Bradley, R., & Webb, R. (1976). Age-related differences in ! locus of control orientation in three behavior domains.; Human Development, 19, 49-55. Brady, E. (1984). Demographic and educational correlates of self-reported learning among older students. ■ Gerontology, 10, 25-38. Brim, O., & Wheeler, S. (1966). Socialization after childhood: Two essays. New York: Wiley. 105 Britton, J., & Britton, J. (1972). Personality changes ini aging: A longitudinal study of community residents. New York: Springer. Brockett, R. (1986). Methodological and substantive issues - in the measurement of self-directed learning readiness. Adult Education Quarterly, 3j5(l), 15-24. i Brockett, R. (1987). Life satisfaction and learner self- j direction: Enhancing the quality of life during the later years. Educational Gerontologist, 13, 225-237. f i i Brookfield, S. (1984). The contribution of Eduard Lindeman| i to the development of theory and philosophy in adultj j education. Adult Education Quarterly, 2£(4)* 185-196. i Brookfield, S. (1985). A critical definition of adultj i education. Adult Education Quarterly, ^6(1)* 44-49. \ J I Brown, B., & Granick, S. (1983). Cognitive and psychological differences between I & E locus of control for aged 1 persons. Experimental Aging Research, £, 107-110. Campbell, A. (1981). The sense of well-being in America. New I York: McGraw-Hill. i ( Campbell, D. T. , & Stanley, J. C. (1966). Experimental and' i quasi-experimental designs for research. Chicago: Rand McNally. Carp, A., Peterson, R., & Roelfs, P. (1974). Adult learning ; interests and experiences. In K. P. Cross, R. Valley,, & Associates (Eds.), Planning non-traditional programs:, j An analysis of the issues for postsecondary education' (pp. 11-52). San Francisco: Jossey-Bass. ! ; Chiriboga, D. (1973). Time and the life course. San j Francisco: Langley Porter Institute. Cicirelli, V. (1980). Relationship of family background: ! variables to locus of control in the elderly. Journal : of Gerontology, 35, 108-114. Cohen, J., & Cohen, J. (1975). Applied multiple-regression. correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, New Jersey: Erlbaum. Cookson, P. (1986). A framework for theory and research on adult education participation. Adult Education Quarterly, 36(3), 130-141. i Covey, H. (1983). Higher education and older people: Some1 theoretical considerations. Educational Gerontology, 9,\ 95-109. Crabtree, A. (1965-67). Education: The key to successful aging. Adult Education, 16-17, 157-165. i ! i I I i Cropley, A. (1985). Participation in adult education. In T.' Husen, & T. Postlethwaite (Eds.), International | Encyclopedia of education (pp. 3787-3790). Oxford: i Pergamon Press. Cross, K. P. (1981). Adults as learners. San Francisco: | Jossey-Bass. j i : Cross, W. , & Floria, C. (1978). You are never too old to I learn. New York: McGraw-Hill Book Company. I Davenport, J., & Davenport, J. (1984). An application of; Verner's principles of adult learning. Lifelong; 1 Learning, *5(1), 11-12. I i Doyle, D., & Forehand, M. (1984). Life satisfaction and old' | age. Research on Aging, 6(3), 432-448. ! Ebell, R. (1973). Some limitations of basic research in. education. In H. Broudy, R. Ennis, & L. Krimerman: (Eds.), Philosophy of educational research (pp.124-130). j New York: John Wiley. Elias, J., & Merriam, S. (1980). Philosophical foundations 1 of adult eeducation. Huntington, New York: Robert E. : Kreiger Publishing Company. j i Even, M. (1987). Why adults learn in different ways, j Lifelong Learning, 10,(8), 22-27. i Falconer, R. (1974). The effect of locus of control on; ; participation in adult learning activities. Dissertation ' Abstracts International, 3_4(3-4), 1233A-2437A. i i i 107 Fisher, J. (1986). Participation in educational activities j by active older adults. Adult Education Quarterly, 36' I (4), 202-210. | Flynn, P. & DeVoss, R. (1986). Systematic educationalj programming at retirement communities. Educationali Gerontology, 12, 139-150. George, L. (1979). The happiness syndrome: Methodological and | substantive issues in the study of social psychological I well-being in adulthood. The Gerontologist, 19(20), 210- 216. 1 I i Graney, M. (1980). Participation in education among older; j people. Alternative Higher Education: The Journal of Non-1 j traditional Studies, 5,71-86. j I i i i Graney, M. , & Hays, W. (1976). Senior students: Higher. i education after age 62. Educational Gerontology, 1, 343-i ! 3 5 9 - j i . i Gross, R., (Ed.). (1982). Invitation to lifelong learning. | Chicago: Follett Publishing Company. i Hale, W. , & Cochran, C. (1986). Locus of control over the I lifespan. Psychological Reports, 59(1), 311-313. ; i Harris, L. (1975). Myth and reality of aging in America. Washington, DC: National Council on Aging. Havighurst, R. (1958). The sociological meaning of aging. | Geriatrics, 13, 43-50. Heisel, M., Darkenwald, C. & Anderson, R. (1981). | Participation in organized educational activities among! i adults age 60 and over. Educational Gerontology, 6, 227- j 240. i Hentges, K. (1980). Education for successful aging. Lifelong | Learning, 3, 20-30. Hiemstra, R. (1976). Older adult learning: Instrumental and expressive categories. Educational Gerontologist, 1, 227-236. Holtzclaw, L. (1979). The adult learner's need for inner i freedom. Lifelong Learning, 2(9), 18-24. 108 Hooper, J. (1981). Older adults earning university credits: a description. Educational Gerontology, 6, 385-394. Hooper, J. & March, G. (1978). A study of older students1 attending university classes. Educational Gerontology,1 3, 321-330. Hooper, J., & Traupman, J. (1983). Older women, the student role and mental health. Educational Gerontology, 9, 233- 242. Johnstone, J., & Rivera, R. (1965). Volunteers for learning. ! Chicago: Aldine. j Kahana, H., Fairchild, S. & Kahana, R. (1982). Adaptation.! In D. Mangen & W. Peterson (Eds.), Research instruments in social gerontology: Vol. 1. Clinical & Social Psychology (pp. 145-194). Minneapolis: University of j Minnesota Press. j Reeves. J. (1985). Multivariate analysis. In T. Husen, & T.' ■ Postlethwaite (Eds.), International Encyclopedia of Education (pp. 3445-3450). Oxford: Pergamon Press. 1 Kleinbaum, K. & Kuypers, T. (1978). Applied regression i analysis and other multivariate methods. N. Scituate, | Mass: Duxbury Press. | i Knowles, M. (1980). The modern practice of adult education: ! From pedagogy to andragogy (rev. ed.). Chicago: Follett.j Krantz, D., & Stone, V. (1978). Locus of control and the; effects of success and failure in young and community-, i residing women. Journal of Personality, 46(3), 536-551.1 F I Kuypers, J. (1972). Internal-External locus of control, ego! 1 functioning and personality characteristics in old age.| | The Gerontologist, 12,(2), 168-173. ' I I Lachman, M. (1986). Locus of control in aging research: A ' case for multidimensional and domain-specific assessment. | Journal of Psychology and Aging, 1,(1), 34-40. Lao, R. (1974). The developmental trend of the locus of ! control. Personality and Social Psychology Bulletin, 1,, 348-350. 109 Larson, R., (1978). Thirty years of research on the! subjective well-being of older Americans. Journal ofj Gerontology, 3_3(1), 109-125. Leeb, J. (1983). Self-directed learning and growth toward i personal responsibility: Implications for a framework for ; health promotion. Dissertation Abstracts International, I 41, 3814A. i Lemon, B., Bengston, V. & Peterson, J. (1972). An exploration' I of the activity theory of aging: Activity types and lifei satisfaction among in-movers to a retirement community.* : Journal of Gerontology, 27, 511-523. ! ! ; ! . I Leonard, W. (1981-82). Successful aging: An elaboration of: ! social and psychological factors. International Journal ! of Aging and Human Development, 14,(3), 223-232. i Levenson, H. (1974). Activism and powerful others: : Distinctions within the concept of internal-external * control. Journal of Personal Assessment, 38, 377-383. ! i I Lindeman, E. (1961). The meaning of adult education. Montreal: Harvest House. (Original work published 1926) Lohmann, N. (1977). Correlations of life satisfaction, morale and adjustment measures. Journal of Gerontology, 32, 73- 75. Lohmann, N. (1980). Factor analysis of life satisfaction, adjustment and morale measures in elderly adults. International Journal of Aging and Human Development, 11, 35-43. J Lohnes, P. R. (1985). Discriminant Analysis. In Husen, T. & T. Postlethwaite (Eds.), International Encyclopedia of I Education (pp. 1418-1423). Oxford: Pergamon Press. ! j Long, H. B. (1983). Adult learning: Research and practice, j New York: Cambridge Press. Lumpkin, J. (1986). The relationship between locus of control j and age: New evidence. Social behavior and personality, 1(2), 245-252. Mahrer, A., & Gervaize, P. (1985). Humanistic theory of development. In T. Husen & T. Postlethwaite (Eds.), ; International Encyclopedia of Education (pp. 2351-2352). ; Oxford: Pergamon Press. ■ 110 .Karascuilo, L., & Levin, J. (1983). Multivariate statistics in the social sciences; A researcher’s guide. Monterey, CA: Brooks/Cole. Merriam, S. (1977). Philosophical perspectives on adult education: A critical review of the literature. Adult Education, _27(4), 195-208. Merriam, S., & Dimmock, K. (1985). Measuring older adults' attitudes. Educational Gerontologist, 11, 134-143. Mezirow, J. (1982). A critical theory of adult learning and j education. Adult Education, 32(3), 3-24. j Mezirow, J. (1985). Concept and action in adult education.! | Adult Education Quarterly, 35(3), 142-151. I i : Mizer, M. (1975). Differences between educationally active | and non-educationally active older adults in reading,; ! interpersonal values, and life satisfaction.; j Dissertation Abstracts International, 36, 2507A. | Molinari, V. , & Niederke, G. (1984-5). Locus of control, ! depression, and anxiety in young and old adults: A I comparison study. International Journal of Aging and; | Human Development, 20, 41-52. i Moody, H. (1976). Philosophical presuppositions of education : for old age. Educational Gerontology, 1, 1-16. ! Moss, H., & Susman, E. (1980). Longitudinal study of personality development. In O. Brim, Jr., & J. Kagan; ; (Eds.), Constancy and change in human development (pp.; I 530-738). Cambridge, MA: Harvard University Press. j National Center for Educational Statistics. (1978)., Statistics of trends in education, 1966-67 to 1986-87.; | Washington, DC: United States Department of Commerce. i Nehrke, M., Hulicka, I., & Morganti, J. (1980). Age1 differences in life satisfaction, locus of control, and self-concept. International Journal of Aging and Human Development, 11_,(1), 25-33. Neugarten, B. (1977). Personality and Aging. In J. Birren,■ & K. Schaie (Eds.), Handbook of the psychology of aging; i (PP« 626-647). New York: Van Nostrand Reinhold. i 111 f Neugarten, G. , Havighurst, R. & Tobin, S. (1961). The measurement of life satisfaction. Journal of Gerontology, 16, 134-143. Nisan, M. (1985). Motivation: Academic. In Husen, T., & T. Postlethwaite (Eds.), International encyclopedia of education (pp. 3430-3434). Oxford: Pergamon Press. O'Connor, D. (1987). Elders and higher education: Instrumental or expressive goals? Educational! j Gerontology, 13, 511-519. ! I 1 Okun, M. , Stock, W., & Covey, R. (1982). Assessing the, j effects of older adult education on subjective well-, being. Educational Gerontology, 8, 523-526. j Palmore, E. (1979). Predictors of successful aging. The1 | Gerontologist, 19,(5), 427-431. Palmore, E., & Luikart, C. (1972). Health and social factors; I related to life satisfaction. Journal of Health and! I Social Behavior, 13, 68-80. i ! Papalia-Finley, D., Dellman, M., Blackburn, J., Davis, E., & i Roberts, P. (1981). Attitudes of older women toward I continuing adult education at the university level: j Implications for program and curriculum development.j Educational Gerontology, ]_, 159-166. j Paulhus, D. (1983). Sphere-specific measures of perceived control. Journal of Personality and Social Psychology,1 44, 1253-1265. Penk, W. (1969). Age changes and correlations of internal-j external locus of control. Psychological Reports, 25,! ! p. 858. j Perkins, D., & Robertson-Tchabo, E. (1981). Retirees return | to college: An evaluation study at one university j campus. Educational Gerontology, 6, 273-287. [peterson, D. (1975). Life span education and gerontology. Thei j Gerontologist, 15, 436-451. i i ■Peterson, D. (1981). Participation in education by older j people. Educational Gerontologist, 1_, 245-256. Peterson, D. (1981b). White House Conference on aging. ! Lifelong Learning, 5, 16-18. Powers, E. (1982). Work and retirement. In D. Mangen & W. j Peterson, (Eds.), Research instruments in social gerontology: Vol. 1. Clinical and social psychology (pp. 231-241). Minneapolis: University of Minnesota Press. Price, W., & Lyon, L. (1982). Educational orientation of the aged: An attitudinal inquiry. Educational Gerontology,' 8, 474-484. ! i Radcliffe, D. (1982). Third-age education: Questions that! j need asking. Educational Gerontology, 8, 311-323. j | I Ralston, P. (1981). Educational needs and activities of older| | adults: Their relationship to senior center programs.I j Educational Gerontologist, 7, 231-244. i , Rotter, J. (1966). Generalized expectancies for inteernal i versus external control of reinforcements. Psychologicalj i monographs, 80, whole number 609. j I I t Rotter, J. (1975). Some problems and misconceptions related to the construct of internal versus external control of | reinforcement. Journal of Consulting and Clinical j Psychology, 43,(1), 56-67. Rubenson, K. (1975). Participation in recurrent education. Paris: Organization for Economic Cooperation and: I Development. ; i Rubenson, K. (1982). Adult education research. Adult; | Education, 32, 57-75. j Russ-Eft, D., & Steel, L. (1980). Contribution of education1 \ to adults' quality of life. Educational Gerontoloqy, 5,; j 189-209. i Ryckman, R., & Malikiosi, M. (1975). Relationship between i locus of control and chronological age. Psychological ; Reports, 36, 655-658. Sauer, W., & Warland, R. (1982). Morale and life satisfaction. In D. Mangen & W. Peterson (Eds.), ! Research Instruments in social gerontology: Vol. l. Clinical and social psychology (pp« 198-236). 1 Minneapolis: University of Minnesota Press. 113 i t t s Schick, F., Ed. (1986). Statistical handbook on aginqi Americans. New York: Oryx Press. ! i Sihvola, T. (1985). Education, aging, and quality of life.1 j Convergence, 18, 58-62. | ] i Snyder, T. (1987). Digest of education statistics. Washington! j DC: Center for Education Statistics 1 I Spreitzer, E., & Snyder, E. (1974). Correlates of life! j satisfaction among the aged. Journal of Gerontology, 29, ? 458-464. Staats, S. (1975). Internal versus external locus of control ; for three age groups. International Journal of Aging and i Human Development, 5, 7-10. i l Statistical abstract of the United States, 108th Ed. (1988).; | Washington, DC: Government Printing Office. ! Stevens, G., & Cho, J. (1985). Socioeconomic indexes and the* new 1980 census occupational classification scheme. ; Social Science Research, 14, 142-168. I Strickland, B., & Shaffer, S. (1971). I-E, I-E and F. Journal for the Scientific Study of Religion, 10, 366- 369. Tabachnick, B. & Fidell, L. (1983). Using multivariate statistics. New York: Harper and Row. Thomae, H. 1980). Personality and adjustment of old people., In J. Birren & R. Sloane, (Eds.), Handbook of mental health and aging (pp.285-303). Englewood Cliffs, New Jersey: Prentice-Hall. Thomas, H. (1979). The concept of development and life span, j developmental psychology. In P. Baltes & O. Brim, Jr., | (Eds.), Life span development and behavior: Vol. 2 (pp. i 281-312). New York: Academic Press. i Timmerman, S. (1985). Lifelong learning for self-sufficiency: A theme for the 1980s. Convergence, 18(1-2), 28-37. Toseland, R. , & Sykes, J. (1977). Senior citizens'center participation and other correlates of life satisfaction. I Gerontologist, 17, 235-241. 114 ' Trent, C., & Trent, M. (1977). Education for aging:i Imperatives for the future. Adult Leadership, 25, 231—; 233. U. S. Bureau of the Census. School enrollment: Social and ! economic characteristics of students, October, 1987.| Washington, DC: U. S. Government Printing Office. j I p Ventura, C., & Worthy, E. (1982). Education for older adults:; j A synthesis of significant data. Washington, D. C: The1 j National Council on Aging. I Willis, S. (1982). Concepts from life span developmental' psychology: Implications for programming. In M. Okun ! (Ed.), New directions for continuing education programs i for older adults, 14, (pp.31-96). San Francisco: Jossey- I Bass. i i Willis, S. (1985). Towards an educational psychology of the older adult learner: Intellectual and cognitive bases, i In Birren, J., & K. Schaie (Eds.), Handbook of the ; psychology of aging (2nd ed) (pp. 3-11). New York: Van | Nostrand Reinhold. i Wiltshire, H. (1964). The nature and uses of adult education. Nottingham, England: The University of Nottingham. ■ I i Wolk, S. & Kurtz, J., (1975). Positive adjustment and involvement during aging and expectancy for internal control. Journal of Consulting and Clinical Psychology,! 43, 173-178. i Wood, V., Wylie, M., & B. Sheafor. (1969). An analysis of a! short self-report measure of life satisfaction! 1 correlation with rater judgments. Journal of^ Gerontology, 24, 465-469.

Asset Metadata

Creator MacKenzie, Barbara Joan (author)

Core Title Differences in life satisfaction, future orientation and locus of control between educationally active and non-active older adults

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Education-Counseling Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag education, adult and continuing,OAI-PMH Harvest

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c26-528001

Unique identifier UC11245113

Identifier usctheses-c26-528001 (legacy record id)

Legacy Identifier DP25289.pdf

Dmrecord 528001

Document Type Dissertation

Rights MacKenzie, Barbara Joan

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA