REMINISCING, TIME ORIENTATION, AND SELF -CONCEPT IN AGED MEN Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY LEONARD J. POSTEMA 1970 LIBRAR y Bdkifiganfflgne 4% University This is to certify that the thesis entitled HEMINISCING, TIME ORIENTATIOI‘I, AND SELF-CONCEPT IN AGED MEN presented by Leonard J. Postema has been accepted towards fulfillment of the requirements for Ph. D. degree inmmgy ///w§%/W_. Major professor Date 7km 2 .5: /f7/ 0-169 “’0‘ II 3' BINDING BY U HDAG & SUNS' 800K BINDERY INC. LIBRARY BINDERS SPRINBPIIT. IlCIlfll . l .ait III I 1: II4\ 34'... «I ‘NIIU ‘ I “9 "1T '1.“ Ci‘lz' \‘ ‘ n . z to!“ fizz): . ' '. " «.5 '7. ' ' ion. ."A'l-LI .-.- Ln- major .‘-‘.'U-> "‘wure inn: instead t 1 -t v '5 . 1’ vcrmafinlea .2 ‘ lacs -v 50., I.” m cutotstsj 0! r1:? «as T ._ J ’J ’ ABSTRACT REMINISCING, TIME ORIENTATION, AND SELF-CONCEPT IN AGED MEN BY Leonard J. Postema Investigators of personality in the aged have emphasized the need for more research on normal adaptation in old age. The few extant studies of reminiscing suggest that this common behavior serves an adaptive function in contrast to prevailing opinion that it is irrelevant and/or maladaptive. ~However, this suggestion is inconsistent with the popular View in time orientation literature that orientation to the past, assumed to characterize the aged, is maladaptive. I The present study explores the interrelationships be— tween reminiscing quantity, type of reminiscing, time . orientation, and self-concept in aged males. Age and type f def residence (community vs. nursing home) were also studied 3 L§Léin relation to the major variables of the study. ‘No specific ”.Lnyypotheses were formulated due to conflicting findings on lme of the variables and lack of research on the others. The sample consisted of 60 aged male volunteers of at a average intelligence. One-half were self-sufficient Leonard J. Postema community volunteers (CV), and one-half were nursing home (NH) residents. Both subsamples were divided equally into three age groups, i.e., sixties (60's), seventies (70's), and age eighty or over (80+). The data were obtained in individual interviews, usually at each §'s residence. Self—concept was assessed by the Total P score on the Tennessee Self-Concept Scale (TSCS). Reminiscing was elicited by use of a standard set of general questions focusing each g's attention on his past. The measure of reminiscing quantity was the number of statements referring to the past. Two judges classified the gs into four reminiscence types, viz., Conflict, Avoidant, Defensive, and Well Adjusted reminiscers based on clinical criteria and descriptions in past studies. Time orientation data were obtained (a) by direct questions about current relative time direction usage, and (b) by the percentage of Thematic Apperception Test (TAT) story responses referring to past, present, and future for three TAT cards. Neither time orientation measure proved adequate. Social desirability, response inconsistency, and confusion affected direct question responses while stimulus pull re- sulted in an artifactual preponderance of "Present" responses to the TAT. Results indicated that (a) reminiscence quantity is not significantly related to any other variable in the study, (2) self-concept levels varied significantly with reminiscence Leonard J. Postema types from lowest to highest as follows: Conflict, Avoidant, Defensive, Well Adjusted, and (g) the present sample had a significantly higher mean self-concept level than the TSCS norm group. Community gs especially in the 80+ and 60's age ranges (3) had a higher self—concept level, (b) tended to reminisce more, (9) produced more TAT responses, and (g) tended to have a more "balanced" (evenly distributed) time direction usage than did nursing home residents. Increasing age was sig- nificantly associated with decreased TAT productivity though not with reminiscence quantity or type. Reminiscence Avoiders tended to be bound to the present. It was concluded that optimal adaptation to the prob- lems of old age (Erikson's Ego Integrity?) is associated with greater openness (e.g., TAT productivity) and the capacity to flexibly use all three time directions (balance), while the poorest prognosis for adjustment (Despair?) is associated with the primitive defense of avoiding past conflicts and a rigid adherence to present concerns. Integrating the present results with past research and theory, it was suggested that (a) reminiscing can serve as a valuable entre to effective therapeutic intervention where appropriate, (b) reminiscence type appears to be an important ~\ diagnostic and prognostic indicator for geriatric therapy, I (g) reminiscing content can serve to pinpoint unresolved ) Leonard J. Postema “filly overlooked, is an important naturally-occurring is: in the aged that may possess creative potential for ding personality integration. REMINISCING, TIME ORIENTATION. AND SELF-CONCEPT IN AGED MEN By” Leonard J. Postema A THESIS .'I Submitted to ' Michigan State University = in partial fulfillment of the requirements , . _ ‘ 5 for the degree of DOCTOR OP PHILOSGPHY 'Bepartment'offiPSychology .1970 7 _' ' ' v}- Iu" ‘ -'. a“ ngfir‘l' " , h) a 1.1-: - tth_-a . V.- ( To my wife Dorothy A db»? v‘ - and Our children ' "’ - 5 .C~ .Jim. Debbie, and Becky : ._‘i Lg} g". . ~A.. ' um; and , I! . _l I ' . :.-.."‘ - ': V “L "5‘ 1‘39. Chic! 1-; ‘ Scrvluz—u n'1‘t r. '..-; 4‘ .’ 1"" 58:2X05i :3 ’ fitn- Admin. s- r at: :c . ' a; .‘ ‘ ' ‘ 215m? lhflflt; t't " ‘ Ac.“ J.- N‘RCherd nova-re I": «w v . ~- sat-15?"? 3, . a - h: alluwxoa as ch» ~itr -ws: ates-a- rw' ACKNOWLEDGMENTS The author wishes to express his great appreciation and warm regards to Dr. Albert I. Rabin, Chairman of the doctoral committee, not only for his invaluable assistance throughout the planning and execution of this study but also as a teacher and one whose continuing trust helped sustain me through some rough periods of my graduate student career. A debt of gratitude is also due Drs. Bertrum P. Karon, Dozier Thornton, and James Uleman for their helpful sugges— tions and worth-while criticisms as members of my dissertation committee. Special thanks are due my colleagues, Dr. JoAnn Sinclair and Ross Whitman for their time consuming work as independent judges of some of the variables of this study, and also to Dr. Sinclair and Lawrence Schwartz in the proofreading and editorial assistance on prior drafts. In addition I wish to thank Dr. Stewart Armitage, Chief of Psychology Service, and all my colleagues at the Battle Creek Veterans Administration Hospital for their continuing support and encouragement. Thanks are also due to Dr. Martin Buell and Mr. Richard Bowers for their encouragement and consideration in allowing me the time away from my regular duties to complete this study. 9". a. . I also wish to thankers. Berna Harter, director of the D-ygattle Creek Senior Recreation Center, and its staff plus - aiithe personnel of the many Battle Creek area nursing homes ‘ fifiho so willingly assisted in this research. A special debt of gratitude is owed the many fine senior citizens who gave of their time and of themselves in partici- pating as subjects. Finally, I wish to thank my wife Dorothy for the in- numerable hours of tedious work she contributed in her typing and retyping for this project, for her scoring of tests, and above all for her love and consideration that have sustained me now and throughout our married life. iv "m“-lr r1" ’ TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . . . . . . REVIEW OF THE LITERATURE . . . . . . . . . . . . . . Senescence and Ego Development . . The Concept of Coping and Developmental Crises of Later Life . . . . . Theoretical Analyses and Approaches to Aged Personality Change. . . . . . Some Studies Relevant to Individual Coping in Old Age . . . . . . . . . . . . . . . . . . Summary. . . . . . . . . . . . . . . . . . . . Reminiscing . . . . . . . . . . . . . Some Functions of Reminiscing in the Aged. . . Butler' 3 Life-Review Hypothesis. . . . . . Adaptive and maladaptive aspects of the life- review . . . . . . . . . . . . Interpersonal aspects of the life-review. . Theoretical problems with the life-review . Reminiscing and Stress . . . . . . . . . . . . Patterns of Experiencing, Reminiscing, and Age Summary. . . . . . . . . . . . . . . . . . . . Time Orientation. . . . . . . . . . . . . . . . . Terminology . . . . . . . . . . . . . . . Significance of temporal experience . . . . Time Orientation and Perspective in the Aged . General empirical findings. . . Some theoretical factors affecting .futurity Planning, desire for change, satisfaction, and age. . . . . . . . . . . . . . . . . "Temporality," experiencing, and patterns of reminiscence with age. . . . . . . . . . Summary. . . . . . . . . . . . . . . . . . . . Page viii TABLE OF CONTENTS--continued Page Self-Concept. . . . . . . . 42 Problems with the Concept of Adjustment and Rationale for Use of Self-Concept . . . . . . 42 Aging and Self-Concept . . . . . . . . . . . . . 45 Summary. . . . . . . . . . . . . . . . . . . . . 44 PROBLEM. . . . . . . . . . . . . . . . . . . . . . . . 45 Formulation. . . . . . . . . . . . . . . . . . 45 Summary of the Major Relationships to be Investi- gated . . . . . . . . . . . . . . . . . . . . 48 METHOD........................ 50 Subjects . . . . . . . . . . . . . . . . . . 50 Institutional subsample . . . . . . . . . . . 50 Community Volunteer subsample . . . . . . . . 50 Tests and Measures . . . . . . . . . . . . . . 51 Intellectual functioning. . . . . . . . . . . 51 Reminiscing . . . . . . . . . . . . . . . . . 55 Self-Concept. . . . . . . . . . . . . . . . . 55 Time Orientation. . . . . . . . . . . . . . . 56 Procedures . . . . . . . . . . . . . . . . . . 58 Procedures with each S. . . . . . . . . . . . 58 Recruitment of gs . . . . . . . . . . . . . . 58 Scoring of measures and development of remin— iscence classification system. . . . . . . 59 RESULTS........................ 64 Reminiscing and Self-Concept . . . . . . . . . . 64 Reminiscing quantity and self-concept . . . . 64 Reminiscence type and self- -concept. . . . . . 64 Reminiscing and Time Orientation . . . . 66 Reminiscing quantity and time orientation . . 66 Reminiscence type and time orientation. . . . 66 Time Orientation and Self-Concept. . . . . . . . 69 Institutionalization . . . . . . . . . . . . . . 70 Self-concept. . . . . . . . . . . . . . . . . 7O Reminiscing quantity. . . . . . . . . . . . . 72 Reminiscence type . . . . . . . . . . . . . . 74 Time orientation. . . . . . . . . . . . . . . 75 Additional Results . . . . . . . . . . . . . . . 78 TAT productivity. . . . . . . . . . . . . . . 78 Number of deviant signs (NDS) as an alterna- tive measure of adjustment . . . . . . . . 86 TABLE OF CONTENTS--Continued DISCUSSION. . . . . . . . . . . . . . . . . . . . . . Limitations of Measures . . . . . . . . . . . . Self-concept . . . . . . . . . . . . . . . . Reminiscing quantity . . . . . . . . . . . . Reminiscence type. . . . . . . . . . . . . . Time orientation . . . . . . . . . . . . . . Sampling Considerations . . . . . . . . . Characteristics Associated with Levels of Self- Concept. . . . . . . . . Integration of Present Results with Past Research . . . . . . . . . . . . . . . . . . Reminiscing. . . . . . . . . . . . . . . . . Reminiscing and self-concept . . . . . . Relative temporal direction usage and self- concept . . . . . . . Present results and adaptation to aging. . . Therapeutic implications . . . . . . . . . . Implications for Future Research. . . . . . . . I SWRY O O O O O O O O I O I O O O O O O C I O O C 0 REFERENCES. . . . . . . . . . . . . . . . . . . . . . APPENDICES I I I O O I O I O O O I I O O I C I O I O O A. Life Experience Questionnaire . . . . . . . . . B. Personal Data Questionnaire . . . . . . . . . . C. Criteria Used to Define and Distinguish Remin- iscence Types. . . . . . . . . . . . . . . . D. Classification of Statements and Their Temporal Referents. . . . . . . . . . . . . . . . . Page 102 102 104 110 115 117 121 125 129 156 156 157 158 140 1. 10. 13.. 12. I I i TABLE I LIST OF TABLES Comparative Demographic Data for the NH and CV Subsamples. . . . . . . . . . . . . . . . . . . Time Orientation Classification Symbols and their Respective Temporal Referent(s) . . . . . . . . . Self—Concept Score Means and Standard Deviations for Four Reminiscence Types, the Total Sample and the Tennessee Self-Concept Scale (TSCS) Norm Group . . . . . . . . . . . . . . . . . . . . . . Frequencies of §s Classified by Reminiscence Type and Time Orientation Usage. . . . . . . . . . . . Frequencies of Ss Classified by Reminiscence Type and Age Group . . . . . . . . . . . . . . . . . Self-Concept Analysis of Variance: Age Group by Location of Residence . . . . . . . . . . . . . . Self-Concept Score Means and Standard Deviations for Locatibn of Residence by Age Group. . . . . . Reminiscing Quantity Analysis of Variance: Age Group by Location of Residence. . . . . . . . . . Reminiscing Quantity Means and Standard Devia- tions for Location of Residence by Age Group. . . Frequencies of Se Classified by Reminiscence Type and Location. . . . . . . . . . . . . . . . . . Time Orientation Analyses of Variances: Age by Location, Using Percent TAT Responses to Past, Present, and Future as Dependent Variables in the Separate Analyses . . . . . . . . . . . . . . . . Time Orientation Means and Standard Deviations of Percent TAT Responses Referring to Past, Present, and Future for Location of Residence by Age Group viii Page 52 6O 65 68 69 71 71 75 75 74 75 76 LIST 15. 14 15. 16. 17. 18. 19. 20. 21. 22. OF TABLES—~continued Page Analysis of Variance for Overall Productivity on the TAT: Age by Location . . . . . . . . . . . . 79 Overall TAT Productivity Means and Standard Devi- ations for Age Group by Location of Residence . . 79 Past, Present, and Future TAT Productivity Analy- ses of Variance: Age by Location, Using the Number of TAT Statements Referring to Past, Present, and Future as Dependent Variables in the Separate Analyses . . . . . . . . . . . . . . . 80 Past, Present, and Future TAT Productivity Means and Standard Deviations of the Number of TAT Statements Referring to the Respective Time Direc- tions for Location of Residence by Age Group. . . 81 Means and Standard-Deviations of Total Sample for Each TAT Card (Repeated Measures Analysis of Variance) . . . . . . . . . . . . . . . . . . . . 82 Time Direction Used Most by gs Responding to the TAT . . . . . . . . . . . . . . . . . . . . . . . 85 Time Direction Used Least by Ss ReSponding to the TAT I O I O I O I O O O O I O O O O O l I I 85 Frequencies of Se Using the Respective Time Directions Most as Measured by the TAT and Ss' Own Estimates . . . . . . . . . . . . . . . . . . 84 Frequencies of Se Using the Respective Tine Directions Least as Measured by the TAT and Ss' Own Estimates . . . . . . . . . . . . . . . . . . 85 NDS Score Means and Standard Deviations for Four Reminiscence Types and the Total Sample . . . . . 88 ix INTRODUCTION In recent years, students of the aging personality have begun to probe for the determinants of personality change and adaptation in later life. Empirical and descriptive studies in gerontology have given rise to theory building and to studies of the developmental nature of old age in the whole scheme of life. Some developmental crises and goals of later maturity have been identified. Currently more emphasis is on individual coping patterns in these crises in order to determine how the developmental goals of senescence are achieved. Recent studies of well adjusted normal aged subjects indicate that a reappraisal of some common be- haviors, previously thought to be symptomatic of deteriora- tion or maladaptation is in order. These behaviors may actually aid adaptation. The present research is an explora— tory study into the relationship of one such common behavior, viz., reminiscing, to time orientation and to self-concept in aged males. w- .v: v‘ REVIEW OF THE LITERATURE Senescence and Ego Development Most students of aging no longer consider senescence to be merely a period of deterioration and decline but a distinct phase of human development (e.g., Clark, 1967; Kastenbaum, 1964b; Neugarten, et al., 1964; Peck, 1956; Rabin,.1945; Sarnoff, 1962). This developmental view has arisen despite most personality theorists' almost complete disregard of old age in the formulation of their theories (Riegel, 1959). Since development implies growth and un- folding or maturation of the organism, how can old age with its obvious decline be considered developmental? Erik Erikson (1965), in his brilliant extention and elaboration of psychoanalytic theory, has pointed the way. Outlining in broad strokes his eight epigenetic stages of man, Erikson makes clear that the essence of human develop- ment is ego development. The final goal of man then, accord— ing to Erikson, is an ego state transcending mere narcissistic gratification; one which dynamically encompasses and inte- grates the strivings both of self and mankind in the face of physical decline and death (cf. definitions of personal adjustment in Havighurst, 1956). wW~ While this notion fits well with what is commonly con- sidered the "wiSdom of the ages", there is much research literature to establish the fact that most aged people apparently do not attain such a state. On the other hand many investigators, e.g., Butler (1967b) as well as one's common experience, can easily point to many people who have, to a large extent, attained this ego state and even con- tributed significantly to society in their senescence. The arrested development of some is inadequate evidence to deny that senescence is a developmental stage. The Concept of Coping and Developmental Crises of Later Life What are the factors that promote development toward ego integrity in the aged and what can be done to enhance the probability of attainment of the state in any given aged individual? Ego psychologists such as Hartmann and Erikson, and others such as White and Maslow have emphasized the posi- tive, adaptive functions of behaviors previously considered primarily defensive in nature. Almost all behaviors can be viewed as attempts to deal with potentially threatening stimuli or problem situations arising in one's life. Of course, not all behaviors are equally effective or adaptive. The range of human conditions from extreme maladaptation to eminently creative lives is ample evidence of this. The problem then becomes one of determining what behaviors or what aspects of behaviors are useful in promoting successful coping and which are not. This paper presupposes that the period of life commonly called “old age“ or senescence contains certain developmental crises inherent in it, which while not exclusive to this period, are uniquely and predominantly associated with it. Overshadowing all others is the imminence of death. Death, of course, is a crisis with which no living person has had any experience and which must remain insoluble. However, Erikson sees the problem as, the expectation of death with the two alternative approaches being "Ego Integrity vs. Despair."1 The former produces a creative acceptance of one's existential situation while the latter is basically a fear of death and an ultimate frustration with one's whole I life. Kastenbaum (1964a) refers to this dilemma as “the crisis of explanation (p. 516)." More specifically, however, some problems important in senescence are maintenance of self-esteem in the face of declining physical and intellectual capacities; coping with grief and depression resulting from personal losses; finding means to contribute significantly to a society 1Peck (1956) suggests a further elaboration of these stages for the second half of life. He would divide Erikson's eighth stage into "Middle Age" and “Old Age" with four "sub aspects" of the former and three of the latter. Middlg Age 3 Valuing wisdom vs. valuing physical powers. (2) Socializing vs. sexualizing in human relationships. (9) Cathectic (emotional) flexibility vs. cathectic im- poverishment. (g) Mental flexibility vs. mental rigidity. mess (g) Ego differentiation vs. work-role preoccupation. (2) Body transcendence vs. body preoccupation. (g) Ego transcendence vs. ego preoccupation. rPeck and Berkowitz (1964) report some empirical data based somewhat on this conceptual scheme (see p. 11 below). 5 of which older persons are still members; and retaining some sense of identity in an increasingly estranged environment (McMahon & Rhudick, 1967, p. 66). Others have emphasized the fear of isolation and loneliness and the fear of dependency in old age (e.g., Clark, 1968). How can the aging individual cope with these problems? Theoretical Analyses and Approaches to Aged Personality Change While most theorists and therapists have traditionally 'written off' the aged as incapable of change or not worth the effort, the increase in the proportion of aged in our population has forced society to take a new look at the aged and to make attempts at ameliorating their problems. In an early review of psychoanalytic thinking regarding old age Kaufman (1940) recounts Freud's well-known impression that the aged were unworkable and then, the first tentative attempts by Karl Abraham, Jellife, Helene Deutch, and a few others to use psychoanalysis with older patients. These at- tempts did establish that the aged were not as unworkable as believed, and resulted in initial hypotheses to explain the dynamics operating in the aged. These explanations, based as they were on limited experience and focused on pathology, seemed to be coalesceing into the somewhat naive view that old age was merely development in reverse (Klopfer, 1965, p. 852). They failed to see the continuing ego development in normal and superior aged adults. A ~u-.—..-————x\. —_——_ Carl Jung's practice included a large proportion of older patients, and he made pioneering attempts at analyzing and attempting to deal therapeutically with the psycho- logical problems of later maturity. Jung (1954) believed that the psychology of youth differed basically from that of old age. In youth one was forced to develop the will in order to achieve. This produced a lopsided development of the conscious achieving part of the personality which he termed "ego", and a consequent underdevelopment of the "unconscious." Using dreams diagnostically, not only as indications of past traumas but also as harbingers of the future and the genus of creative thought, Jung encouraged his patients to express their unconscious fantasies by means of painting. This presumably functioned to restore a balance to the personality and by the process of "individuation" to arrive at "self“, a sort of adaptive middle ground between ego and unconscious. Erik Erikson's classic Childhood gpd Society (1965), first published in 1950, among other things served as a sort of theoretical global map of the entire life cycle. As such it served to open up and legitimize studies of ego develop— ment in adulthood and old age (Neugarten, 1964). Step by step each developmental crisis had to be successfully resolved in order to ensure final attainment of ego integrity. HOwever, though penetrating in analysis and comprehensive in scope, Erikson was primarily concerned with childhood and offered no new suggestions to enhance the probability of a person's attaining ego integrity in senescence.2 Others have theorized about the nature of senescence in the life cycle, and an increasing number have attempted to deal therapeutically with psychopathology of the senium. These approaches span the spectrum from individual psycho- analysis, to resocialization of geriatric ward patients, to social psychiatry and anthropology. However, there is still a lack of systematic studies with normal subjects into specific coping mechanisms which may facilitate adaptation and ego integrity in senescence. Henry (1956) states: It is often the same situations of complexity out of which spring both the complex normal and the self defeating neurotic. . . . The dynamics leading . . . to neurotic solutions are more clearly understood. The dynamics of being 3 copplex normal adult are not (p. 55). Henry further emphasizes that "the manner of adaptation to stress is the variable of significance (p. 54)." Brief mention should be made of two prevalent social psychological theories of aging. These are, first, what has become known as the “activity theory" implicit in the 1949 studies of Cavan, Burgess, Havighurst, and Goldhamer, and the later "disengagement theory" of Cumming and Henry (1961). 2Gruen (1964) has made an interesting attempt at opera- tionalizing Erikson's eight stages as personality dimensions. Rating scales were devised and applied to interview data from a cross-sectional sample of pe0ple with ages ranging from 40 to 65. This exploratory study, while not offering clear—cut findings offers some tentative confirmation for the theory and a good first step methodologically. The former assumes that good adjustment in old age is based on the active continuance in, or adequate substitution for, the person's earlier adult roles. The latter theory appeared to hold the opposing view that gradual disengagement of the individual from society is inevitable and presumably con- ducive to good adjustment. However, in a later (1964) modification of her theory, Cumming has attempted to extricate it from this apparent opposition. She asserted that "activity and engagement are not on the same dimension (p. 6)." Her reasoning, though, is murky and unconvincing. In any event both "theories" are group oriented, give no indication as to how a particular aged person should cope with establishing 1113 optimal level of engagement or disengagement and empirical studies of the two positions have so far proved inconclusive (Havighurst, 1961; Poorkaj, 1967; Rupp, Duffy, & Danish, 1967; Tallmer 8c Kutner, 1969; Tobin 8c Neugarten, 1961). Some Studifi Relevant to Individu_a_]_. Coping in Old Age Some community-wide studies such as the Langley-Porter- San Francisco study (Lowenthal, Berkman, 8c associates, 1968) llave established that there is a great degree of overlap in Psychopathology present in aged confined to mental hospitals arui community resident aged. This indicates that one's ability to cope well enough to maintain oneself in the com- munity is not dependent solely on the absence of obvious Psyohopatho logy . On the other end of the continuum an intensive multi- disciplinary National Institutes of Health (NIH) study of optimally healthy, well—functioning aged males established that even supposedly maladaptive defenses such as denial, counterphobic activity, obsessive-compulsive behaviors, and schizoid emotional withdrawal may be very adaptive in coping with the problems of old age (Perlin & Butler, 1965; Butler, 1967). In an earlier (1961) study with these same healthy, conmmnity resident subjects, Werner, Perlin, Butler, and .Pollin reported that while most subjects reported decrements in their physical and cognitive functioning, increments were c3ften mentioned in their social—psychological and personality- aiffective spheres. Even where decrements were reported, tlieir data show many adaptive ways of dealing with the per- ceived decrements (e.g., use of a memo pad to compensate for :Eeailing memory) and relatively few complaints concerning these cieecrements. These results reflect both ego development and alzpropriate self-acceptance. In a five year follow-up study of this NIMH sample (Ihatler, 1967), both improved adaptation (4§s) as well as decreased adaptation (6§s) was noted in the 10 _S_s who had changed in this regard out of 29 who were restudied. Of the orig‘inal 47 fig, 59 survived and of these, 10 could not or WC“11d not participate in the follow—up. "Individual adapta- tixbn and widowerhood appeared to be important to survivor- ship (p. 1256)," indicating the significance both of intrapsychic and interpersonal/environmental events. l A .-~‘4 10 Weinberg (1956, 1967) asserts that much of what appears to be denial in the aged is actually an adaptive mechanism which he terms "exclusion of stimuli."3 He states that denial implies that a stimulus has been perceived, has been cathected to and invested in and then cathexis is withdrawn. “Exclusion of stimuli," on the other hand, is a mechanism which is used to block out, unconsciously, all but stimuli which are emotionally pertinent. This appears to be the other side of the coin of denial. Weinberg (1956) lists two other types of adaptive mechanisms, viz., conservation of energy and regression but does not elaborate on them. His conclu- sions are based mainly on psychoanalytic case studies. "Intellectual control“ as opposed to "impulsivity" is suggested as a favorable factor in differential mortality rates, and as such may be an adaptive mechanism in aging. This tentative conclusion was made by Slater and Scarr (1964) based on a factorial study of MMPI scales with older and younger gs. However, since the data are cross-sectional, the finding of greater intellectual control in older gs may simply reflect personality change in this direction with experience (and, therefore, with age) and be entirely inde- pendent of mortality causes. Studies by Neugarten and Gutmann (1964) and Gutmann (1964) indicate lessened rather than greater intellectual control with age. 3This mechanism was also stressed by Watson in his 1954 review (p. 511 ff.) though he did not attach a label to it. 11 Peck and Berkowitz (1964) studied seven attributes which they considered necessary to adjust to developmental crises of middle life. These attributes were drawn from their ex- pansion of Erikson's stages of ego development. These were: (2) Cathectic flexibility, (3) mental flexibility, (g) ego differentiation, (g) body transcendence, (g) ego transcendence, (g) body satisfaction, and (g) sexual integration. These were then correlated with a measure of "adjustment." Their results showed consistent differences only with regard to socio-economic status and not by age or sex. They conclude that the lack of age differences tend to show that adjustment to developmental crises of aging is largely due to good ad— justment in one's earlier life. They do not deal with the question of how adaptation can be improved or if some of these attributes can be acquired in middle and late life. Many of these studies were part of a much larger research project of adult human development in Kansas City. In another study of this population (Rosen & Neugarten, 1964), con- sistent age differences were found using the Thematic Apper- ception Test. Older gs introduced fewer non—pictured characters, less conflict, a lower activity-energy level, and a lower affective intensity level into their TAT stories than did younger gs. This was taken as support for their hypothesis that with increased age there is less energy available to the ego for responding to, or maintaining former levels of involvement in, the outside world. The implication is that the older person tends to respond to inner rather than to outer stimuli, to withdraw 12 emotional investments, to give up self-assertiveness and to avoid rather than to embrace challenge (p. 99). A five—year follow-up study (Lubin, 1964) confirmed a drop in ego energy on a longitudinal basis using the same type of analysis. However, a reworking of the Rosen-Neugarten data with a similar type of analysis failed to reveal the same age differences (Neugarten & Miller, 1964). Summing up data from their extensive cross-sectional and longitudinal community-wide study of aging in Kansas City, Neugarten et a1. (1964) found the following trends. There is a change with age from "active mastery" to "passive mastery." There is increased preoccupation with the inner life; decreased emotional cathexis toward persons and objects in the outer world, and generally what they term greater "interiority" with age. This goes along with constriction in ability to integrate wide ranges of stimuli (cf. Weinberg's "exclusion of stimuli") and less willingness to deal with complicated and challenging situations. I On the other hand they found that age was not consistent- ly related to socioadaptive functioning and that even in the face of obvious physical and cognitive deterioration many aged continue to function satisfactorily in the community. Reflecting on this they state: How do those men and women who give evidence of inef- fective thought processes continue to appear integrated? Although these questions did not fall within the direct line of inquiry in these studies, the implication is that there are coping and synthesizing processes which presumably provide continuity (p. 197). 15 They then suggest habitual processes as one explanation.4 McMahon and Rhudick (1967) contend that reminiscing too, is such an adaptive coping mechanism in the aged. They based this contention on (a) finding much reminiscing in a sample of males of above average capability and longevity (25 Spanish American War veterans), and (p) the significant rela- tionship found between ability to reminisce and survival in a follow-up study one to three years later. These authors as well as Butler (1965a) challenge the commonly held opinion that reminiscing is a sign of senile deterioration or mal— adjustment in the aged. Summary In this section an attempt has been made to review rele- vant personality theory, and the theoretical background of senescence as a distinct phase of ego develOpment with its own unique developmental crises. Erikson's theory of ego development has been most fruitful here. The concept of coping was examined as the means by which the aged individual overcomes these crises in the attainment of final ego in- tegrity. Also, various behaviors or coping mechanisms which have been advanced as being adaptive were reviewed. ‘One would also suspect continuity of environment as a significant factor in maintaining integration since there is much evidence to show how discontinuities in environment such as displacement by urban renewal or superhighways often pre- cipitate institutionalization, or death, e.g., see report in Geriatric Focus, April 15, 1969, 8, 2. 14 Studies of the aging personality have revealed that behaviors previously thought to be evidence of deterioration or maladaptation may actually be adaptive means of coping with the problems of aging. Extensive personality studies have also listed many attributes thought to be necessary for adjustment, but conclude that these must be acquired at an early age. Most studies have continued the trend of chronicling the effects of deterioration with age. The con- tinuing dearth of research on coping mechanisms that are adaptive in normal aging and lack of research on methods of enhancing ego develOpment in old age.was pointed out. Neugarten et a1. (1964) have called attention to the increased “interiority” with age and have noted the seemingly parodoxical ability of many aged to continue adequate socio- adaptive functioning despite physical and cognitive deteriora- tion. They suggested that coping mechanisms which provide continuity are the key to this paradox. Reminiscing has been advanced as such a continuity pro- viding coping mechanism. The following section will deal with the relevant theory and research on reminiscing in the aged. Reminiscing For the purposes of this discussion reminiscing is de- fined as “the act or habit of thinking about or relating one's past experiences“ (Webster's International Dictionary, 15 2nd ed.). This is to be distinguished from the technical use of the term in experimental psychology as "recall or recognition, without intervening overt practice, of items previously not recallable; an increment in a practiced act after a period of non—practice" (English & English, 1958). Some Functions of Reminiscing in the Aqu McMahon and Rhudick (1967) classified their gs into four categories on the basis of their reminiscing's content and function. Based on their descriptions, these categories might be called: (3) regressive, (p) obsessive-compulsive? (g) well-adjusted, and (d) depressive. J The “regressive" type is characterized by denial of present inadequacies and the use of reminiscing to enhance self-esteem by idealizing the past. The "obsessive—compulsive“ seem preoccupied with reviewing their life and deeds in a judgemental—evaluative way. The “well-adjusted" seem to be .) able to tell a story for the sheer pleasure of telling it and for the enjoyment and information of their audience. These people had no apparent "ax to grind" nor compulsion to review their lives. The "depressives“ seemed preoccupied with current bodily ills and problems and had difficulty reminiscing at all. When they did reminisce it appeared fitful and was constantly interrupted by present concerns which apparently erupted into consciousness. Unfortunately the authors present no data on these four types of subjects and do not give any objective criteria or method of 16 classification. Their implication is that reminiscing, in and of itself, is adaptive. gptler's Life-Revigw Hypothesis Another investigator who has studied reminiscing in the aged extensively is Butler (1965a, 1965b, & 1968). Based on case study experience with his aged patients Butler postulates the universal occurrence in older people of an inner experience or mental process of reviewing one's ; life. [He] propose[s] that the process helps account for the increased reminiscence in the aged, that it con— tributes to the occurrence of certain late-life dis— ‘ orders, particularly depression, and that it participates in the evolution of such characteristics as candor, serenity, and wisdom among certain of the aged (1965a, p. 65). In other words, Butler theorizes that reminiscing is only an overt (i.e., verbal) manifestation of a larger biologically determined mental process of reviewing one's life in old age and that this can be either adaptive or maladaptive. Butler characterizes the life-review as "the progressive return to consciousness of past experiences, and, particularly, the resurgence of unresolved conflicts (1965a, p. 66)." He hypothesizes that this process is precipitated by an aware- ness of the imminence of death and occurs in younger persons, similarly, in the face of impending demise. However, he further speculates that this may be a general response to crises of various types, of which imminent death is only one instance. rThe life-review is seen as an active process which potentially proceeds toward personality reorganization. The more intense the unresolved conflicts, the more work remains if. 17 to be done to achieve reintegration, with the outcome de- pendent mainly on personality or character structure and to some extent environment. Structurally speaking, Butler does not make clear where the life-review originates. He does imply that there may be either conscious or unconscious initiation. However, once in motion the process of surveying, observing, and reflecting upon the past, and especially reintegration, is said to be an ego function as is memory itself upon which reminiscence depends. McMahon and Rhudick imply that their "life-review" subjects seemed to be exercising largely super—ego type activities, i.e., self-critical moral evaluations. Adaptive and maladaptive appects of the life-revigp. ‘ Theoretical considerations aside, Butler comments on some of the adaptive manifestations of the life—review. Through re-evaluation, presumably drawing on subsequent experience, the aged person is often capable of expanded understanding of previous conflict situations. The resulting greater h understanding and acceptance of life is able to mitigate one's fear of death. Where this is obviously creative he states I the effects can be very constructive. He feels that in the majority of the elderly a substantial reorganization of personality does occur. However, Butler believes that the life-review can con— tribute to psychopathology as well, i.e., severe depressions, states of panic, intense guilt and constant obsessional 18 rumination. Those especially prone to these bad effects, Butler (1965a) states, are those who have always put a heavy investment on the future, the future which their old age cannot deliver; those who have "consciously exercised the human capacity to injure others (p. 70)" i.e., those with real guilt, and the characteristically arrogant and prideful whose narcisscism is disturbed by the prospect of death. In these instances the life-review may contribute to the known higher suicide rate in the aged.”‘ An interesting theoretical problem is implicit in the differing positions of Butler and some psychoanalytic writers. Other writers, e.g., Kaufman (1940) emphasize the “return of the repressed“ in old age as an indication of lessened ego energy, while Butler posits the life-review as an active process of attempted personality reorganization in which the ego is actively engaged in resolving long- standing conflicts in preparation for death. Such an attempt at reintegration would seem, if anything, to call for an expenditure of ego energy far in excess of anything the individual would be required to muster at any previous time I in his life. The latter position has interesting economic implications for psychoanalytic gerontology. Perhaps, in- stead of a decline of ego energy in the 60's and 70's, there is a redistribution of ego energy to more internal affairs (Neugarten et al., 1964). The decline, if any, may come much later than usually thought (see discussion of Gorney, 1968, p. 24 ff, below). 19 Interpersonal aspects of the life-revigp. Besides the intrapsychic functions of reminiscing Butler also views reminiscing as serving an interpersonal function. "Oppressive garrulity" may reflect a person's loneliness and his wish to be heard according to Butler. Implied here is that reminisc— ing combined with common courtesy on the part of the other person may serve the function of relieving loneliness by keep— ing the other person there. Furthermore, revelations from the past may drastically alter an aged person's object relationships, either strengthening, weakening, or even destroying long time interpersonal bonds. Butler also notes that the other person's reaction may have a profound effect on the aged reminiscer. 'Bursting the balloon' of a regres- sive idealization of the past may have what Butler calls the Shangri-La effect, i.e., changing an alert youthful looking face into an aged careworn one. Based on his life-review hypothesis Butler states that "probably at no other time in life is there as potent a force toward self—awareness operating as in old age. Yet the capacity to change according to prevailing stereotype, de- creases with age (1965a, p. 75)." He notes that the major portion of gerontological research is concerned "almost/ enthusiastically" with measuring decline in old agey/ His call for comparable attention to studies of the individual, focus— ing on growing wisdom, and the meaning of experience is, I believe, very appropriate. 20 Eggpretiqgl problgpp with the life-reyigy. While Butler's study of reminiscence and adaptation in old age appears fruit— ful, the usefulness of postulating the "life-review“ (essentially a past-review) as a universal biologically determined response to approaching death seems questionable. Butler, himself, notes that the name "life-review" unfortunate- ly implies an orderliness which is usually lacking. Often‘ only certain critical periods of one's life are dwelt upon, not one's entire past (1965a, p. 67). Evaluating the evidence cited for the universality of the "life-review," one could just as easily cite evidence to posit a universal response to “view-the-present" or "preview-the future." Perhaps a stronger case could be made for conceptualizing the life-review as a special instance of one phase of problem- solving. Whenever, a situation is encountered in which grati- fication is not immediate or not quickly assured by the use} I of habitual patterns of response, an apparently automatic t; memory scan for previous similar situations and successful patterns of response takes place. The more difficult the problem or accumulated complex of problems, the more one must draw on a broader and broader focus of past experience in order to accomplish the ego's task of integrating id impulses for gratification with reality as perceived and tested by the individual. In old age one has an almost complete life- time of experiences to draw upon, but for most, if not all, one is forced to deal with a problem so massive and unknown 21 that it defies solution, viz., one's imminent personal demise. In addition to this are all the other developmental crises listed earlier. Is it any wonder that attempts at problem- solving trigger off a memory scan for solutions of "life- review" proportions? Even for those who have confidently worked out their own phiIOSOphical answers to death, e.g., belief in personal immortality, long standing personal con- flicts or other problems of aging may require extensive memory scan. The tentative solutions may often involve exercising one of the few remaining abilities left, i.e., verbal expression, and hence reminiscing. Reminiscing in - turn possesses the creative potential of any medium of selfd\ expression. As such it may well serve the same adaptive / function of individuation that painting served for Jung's 5 patients. Reminiscing and Stress Tobin and Etigson (1968) studied the effect of stress on earliest memory. Their's is one of the few empirical at- tempts to study the reconstruction of reminiscence as reflect- ing ''the synthesis of personally meaningful events from the past within the context of current environmental transactions (p. 455)." These investigators found a significantly greater, amount of personal loss themes in the earliest memories of /) aged gs after being subjected to the stress of admission to 3 old age homes than before. 22 Experimental §s were asked for their earliest memories four months prior to admission (Time 1) and again two months after admission (Time 2). 51.1% of the experimental gs showed increased loss vs. 20.0% of control gs. There were four con- trol samples, and all samples were similar as to sex, marital status, socioeconomic class, level of organic impairment, education, ability to care for their own physical needs, and ethnicity. The age range was 68 to 92 with a mean age of 78. The four control samples included one in which earliest memories were elicited before and after a similar six month period but while they were community residents the entire time. The second control group were §s also on the waiting list at Time 1 (as were the experimental gs at Time 1) but these continued on the waiting list at Time 2 as well. A third control group were short term institutional gs from which earliest memories were elicited two months post- and again one year postinstitutionalization. The last control sample consisted of long term institutionalized gs with earliest memories elicited one to three years then again two to four years postadmission. Earliest memories were rated as to level of introduced loss on a 5-point ordinal scale from nonloss themes to extreme loss involving death themes. Their data clearly show "that an adverse environmental change is reflected in the earliest memory, as manifested by an-increase in the introduction of loss in the repeat recon- struction (p. 459)." 25 Interestingly (and unexpectedly) 50% of all their aged gs gave different incidents as their earliest memories at Time 2, i.e., the manifest content differed. This was sig- nificantly associated with shifts in latent content as well for all groups. This appears to differ from other studies of younger gs who more typically give the same incident as their earliest memory after similar periods of time (Hedvig, 1960 & Winthrop, 1958, as cited in Tobin & Etigson, 1968). However, despite the shifts both in manifest and latent content, it was the strong impression of the investigators that the underlying dynamics were the same at both times. They tentatively conclude that the "surprisingly high per- centage of 50% intraindividual variability in manifest content may be obscuring the dynamic consistency that is portrayed in the reconstruction of reminiscence (p. 442).“ Further, they feel that at least three factors must be considered for determining the consistency of the eafliest memory: (3) the \ availability of a storehouse of early incidents from which one is selected, (p) its expression in the synthesis of present adaptive concerns, and (g) the portrayal of characte - logical dispositions in the reconstruction. In attempting to understand the increased variability of aged respondents they consider increased stress with age, cognitive deficits with age, and the possibility that Butler's postulated "life-review" with its increased reminiscent activity may be operative. 24 What does seem clear from this study is that reminiscence, though it may reflect a basic consistency of characterological dispositions presumably established in early life, is also subject to the influences of current environmental transac- tions. Though this study dealt only with stress and its reflected themes of loss in reminiscence, the possibility exists that current gratificatiopp might also be reflected by themes of past personal gains in the reconstruction of reminiscence. Pattegns of Experiencing, Remingpcing, and Age In another study, drawn from the same research project that the Tobin and Etigson (1968) study was a part of, Gorney (1968) dealt extensively with reminiscence, Butler's "life- review, and its relationship to Eugene Gendlin's concept of "experiencing." Expefiencing is conceptualized as the will- ingneiswto intpqspect upon feelings. In an unpublished pilot study Gorney found a significant negative correlation between "experiencing" and chrono- logical age from the 60's through the 90's. While this willingness to introspect upon feelings consistently lessened with age beyond age 80, he found a bimodal distribution in the 60's and 70's, i.e., he found both high and low experi- encers in the younger age range. In an attempt to comprehend these results within a develOpmental framework, some implications of Butler's 25 life-review theory were invoked. Gorney reasoned that the life-review, as a developmental process leading to personality reorganization, should be reflected in reminiscence by active attemppgwat conflict resolution followed by_a more quiescent Per 10d Pf SSFSEi..P¥..yh_en,..resolutionwhadmeen-ashievsd- Tying this to chronological age he hypothesized that persons in their 60's and 70's should more typically reflect active at- tempts at resolution of long-term conflicts while those in their 80's and 90's would largely have reached the resolution stage. He hypothesized his "Old-High" experiencers (60's and 70's) were active life-reviewers, and his "Very Old-Low" experiencers (80's and 90's) were those who had achieved a serene or resigned life-review resolution. To explain his third group, the "Old-Low" eXperiencers (60's and 70's), he hypothesized that these were peOple who refused to engage in life-review activity by avoiding past conflicts and deal- ing only with present concerns. His sample consisted of the entire study population of the parent project (E?172) on which a mass of data had al- ready been collected. There were three subsamples: 100 gs on the waiting list for admission to an old age home, 55 community residents who were still self-sufficient, and 57 Se who had lived at least one year in an old age home. For various reasons, not all §s were included in many of his analyses. 26 Gorney's primary measure of experiencing was developed by Gendlin. This was an Affect Questionnaire of open-ended questions focusing on gs' experiencing of eight major affects (loneliness, depression, anger, shame, guilt, happiness, satisfaction and pride).‘ Responses were rated on a 7-point experiencing scale ranging from complete lack of intrOSpec- tion to active, insightful understanding of one's inner processes. Gorney validated this by use of a modified version of Gendlin's Focusing Manug; as a criterion measure adminis- tered to some members of the Community sub-sample. A categorical measure of reminiscence type was con- structed called the Evaluation of Life Questionnaire. This 17 question instrument was included in the interview schedule immediately after an extensive series of detailed life history questions. The Questionnaire was designed to encourage an evaluative set in each §_to make judgments about his past life rather than to elicit associative reminiscing pp; pg. Initially a qualitative analysis of responses produced six categories: (3) refusal to engage in reminiscence, (2) dis-\ tortion of the past involving total flight, (g) distortion f; of the past involving partial flight, (Q) manifestation of {‘ conflict indicating active life review, (g) partial resolu-:> tion with one or a few conflict(s) outstanding, and 5 (g) complete resolution with serenity. Problems of inter- rater reliability were resolved by eventually collapsing these into three categories. The first three categories 27 became (I) Flight from the past; category (g) remained as itf was, (II) Manifestation of Conflict (Life-Review); while categories (g) and (g) became (III) Resolution. A remarkably consistent distribution of types across subsamples was demonstrated with one-half diSplaying Flight, one-sixth displaying Manifestation of Conflict, and one-third Resolution. It was also established that type I (Flight from Past) §s evaluated their lives rather neutrally, with some- what flattened affect, type II (Life—Review) gs evaluated their lives quite negatively, with intense affect and type III (Resolution) gs evaluated their lives more positively with moderate affect. The main hypothesis of the study assumed that willing- ,_...n .,y ness to introspect upon feelings (experiencing) was a neces- W— ~_.—.-. process of resolving old conflicts through the life-review in old age would lead to a state of serenity in very old age : which would render further intense introspection develop- mentally irrelevant. The results generally suPported the three predicted patterns, however, two important non-predicted patterns emerged as well. As many of the Old-High experiencers (59%) fell into the Resolution category as into the predicted active life-review category. Also, 40% of the Very Old-Low experi- encers displayed a non-predicted Flight from the Past as well as the 55%‘who were in the predicted Resolution category. 28 These results were found to be independent of the type of subsample, i.e., the five trends, three predicted and two non-predicted, were also replicated across all subsamples. In an attempt to determine whether the two non-predicted trend §s differed in some systematic way from those who followed the predicted patterns, additional analyses were employed using data from the large data pool of other psycho- logical characteristics. However, these analyses failed to disclose any significant differences. The variables were grouped as follows: I. Affect Variables, i.e., Cattell Global Anxiety, Emotional Reactivity, and Life Satisfaction Rating. II. Temporality Variables, i.e., Expectations re Institutional Life, Self Compared to Earlier, and Denial of Institutionalization. III. Reminiscence Variables, i.e., Reminiscence Affect, and Reminiscence Evaluation. Gorney concludes that his investigation provides general support for the hypothesis of a decline in willingness to foo-u. .....-a- from the 60's through the 9013, and that this decline is related to developmental changes in patterns of reminiscence. However, he further concludes that reminiscence type alone most clearly differentiated the aged gs he studied. Stated another way, the most important finding was that "Butler's modal developmental pattern of an active life review cul- minating in the achievement of resolution tends to occur at particular ages (p. 147)." 29 Summary All of the studies reviewed in this section appear to agree that reminiscence is an important psychological process in old age. In addition its importance as an adaptive or at least potentially adaptive coping behavior was stressed by McMahon and Rhudick, Butler and Gorney. Butler has made the major theoretical contribution here in conceptualizing reminiscence as a manifestation of a a. _ _ , _ 1...---.rr._.-...—.t. «< ~- - - l? 2‘ .7 «mm—.,- —..-. -m._~ ,./ largeifldevelopmental process 9f reviewing gnels life-in order / ' to attain final ego integpipy preparatorymto death, Both the Gorney and, to a lesser extent, the McMahon and Rhudick studies support this View and all of the studies lend support to Erikson's contention that old age is a truly developmental stage. Some interesting differences are also apparent in these studies. While Butler includes all thoughts about the past as part of the life review, McMahon and Rhudick classify only those gs with obsessive-compulsive needs to judgementally evaluate their lives as "life-review" reminiscers. They identify at least two other types of reminiscing, i.e., regressive and well-adjusted informative story-telling. Gorney, apparently also interpreted Butler's "life-review" as primarily judgemental evaluative reminiscing since his measure was expressly designed to elicit this. What the latter investigators seem to have overlooked is Butler's emphasis on the interpersonal/environmental aspects of 5O reminiscing as well as intrapsychic attempts at reintegration. Perhaps it is precisely here that the concept of the life- review needs to be viewed in the broader theoretical context of problem-solving, whether of the intrapsychic conflict resolution variety or of the ongoing interpersonal/environ- mental-transaction variety. The Tobin and Etigson study is especially pertinent to the latter point, since taken together with the other studies, it suggests both the reciprocal effect of stress on remin- iscence and the creative selectivity involved in drawing on one's past experiences to assist in current problem-solving. Another difference in empirical findings should be noted. Gorney's results indicate a negative correlation between chronological age and reminiscing (of at least the intra- psychic conflict resolution type) while McMahon and Rhudick's very-old gs were apparently active reminiscers. Their veterans ranged in age from 78 to 90 with a mean age of 84, and according to Gorney's hypothesis should have largely ceased reminiscing. Perhaps these discrepent findings about reminiscing activity in very old age can be accounted for by the select nature of McMahon and Rhudick's sample. Gorney's sample was undoubtedly more typical of our aged population and suffering more from the problems of old age. This very difference suggests another possibility. The decline of experiencing and reminiscing may reflect declining ego energy in very old age. Gorney's "resolution“ gs consisted not only 51 of the serene but also of the resigned. The continuing of active reminiscence in McMahon and Rhudick's gs might then reflect either their superior constitutions with larger reserves of ego energy, or that their active reminiscing was actually enhancing their survival potential. Another factor may well be the evaluative reminiscence set given to Gorney's gs as compared to the nondirective approach used in the Boston area study. However, these differences need clarifi- cation. The next section will be concerned with temporal orien- tation in senescence and its relationship to reminiscence and adjustment. Time Orientation Research on temporal experience in the aged somewhat parallels the study of reminiscence. Time orientation is a molar time concept referring to the direction of one's thoughts, fantasies, and expressions, i.e., to the past, present, or future. Obviously orientation to the past is I intimately bound up with reminiscing. While the literature on temporal experience has three major aspects, viz., estimation or perception of time intervals, time perSpective, and time orientation, this discussion will primarily be limited to the latter two. For more extensive reviews of the literature, the reader is referred to Wallace and Rabin (1960), Fraisee (1965), Laffey (1965), and Smeltzer (1968). 52 Terminology. Differences in terminology have plagued research in this area, therefore, a clarification of terms at this point is desirable. Most of the above discussions employ Wallace's (Wallace & Rabin, 1960) distinction between time orientation and time perspective. Time perspective refers to the Span of time covered by the person's thoughts, fantasies, feelings, etc. into the past and into the future, while time orientation, as mentioned, deals with their direction, i.e., to past, present, or future. Various aspects of time perSpective have been differen- tiated and studied. Extension refers to the distance into the future or past conceptualized. Density refers to the number of events with which one's perSpective is p0pulated. Coherence refers to the logical ordering of events in sequence. Kastenbaum (1965) in a study of future time perspective which he terms "futurity" also distinguishes directionality, i.e., the sense of moving from the present moment into the future, and separates futurity into personal and cognitive futurity. Personal futurity refers to the ability to project oneself into one's own future while cognitive futurity refers to one's ability to use time as an abstract cognitive category for organizing and interpreting experience in general. While the concept, time orientation, has the disadvantage of being confused with one of the cardinal psychiatric diagnostic signs, i.e., whether the patient knows what day it is, for the sake of uniformity in temporal experience research it will con- tinue to be used in this study. 55 Significance of temporal experience. Only a little re- flection is necessary to realize the psychological importance of time in our lives. Fraisse (1965) has written: Man must somehow free himself from the state of change which carries him through life, by keeping the past available through memory and conquering the future through anticipation. This control over time is essen- tially an individual achievement conditioned by every- thing which determines personality, age, environment, temperament, eXperience (p. 177). Indeed, research on temporal experience has almost matched the breadth of personality research itself. Much has been written about the develOpment of the "time sense" in child- hood and of the changing time orientation in the adult years (from future to past). Social class differences have been explored (e.g., LeShan, 1952), cultural differences noted (e.g., Hughes, 1961, and Smith, 1961), and disturbed temporal eXperience in various psychopathologic states (e.g., Dilling & Rabin, 1967), character disorders (e.g., Laffey, 1965) and drug addictions (e.g., Einstein, 1964) have been studied. Many other aspects of temporal experience and their psycho- logical relationships have been studied, however, only studies relevant to aging and adaptation will be discussed here. Time Orientation and Perspective in the Aged General empiricalifindings. Fink (1955) investigated time orientation and its relationship to age, institutionali- zation, and activities in institutionalized and non- institutionalized older males. Each group of gs was 54 subdivided into younger (aged 50-61) and older (aged 61-76) gs. He found institutionalized gs were more oriented to the past, and also that older gs were more oriented to the past. He found work-related activities related to future time orientation though not organizational activities. Fink implies that orientation to the past is maladaptive: Many people use such reSponses as nostalgia for past social arrangements, general non-adjustive reactions of the type characterized by cynical attitudes, a "paranoic" approach to newly emerging arrangements, "nervousness", suicide or other means of withdrawing from social situations (p. 17). This attitude is identical to the prevailing attitude toward reminiscing and is implied by other investigators and theorists such as Albrecht, Vissher, and Arieti (as cited in Fink, 1955). Time and again, the literature cited links past time orientation or lack of future orientation to instances of maladaptation. In addition to the aged, one other category of gs is consistently linked to a past time orientation, viz., depressives (Strauss, 1947 as cited in Wallace & Rabin 1960; Dilling & Rabin, 1967). This relationship is inter- esting because of the prevalence of depressive states in the aged. Cath (1965) has explored depression in the aged and characterized it with an even more morbid title, "depletion." Both old age and depression have been described as periods when the future seems blocked. Despite the many references to the aged and depressives being oriented to the past, Laffey, (1965),notes the paucity of studies dealing with 55 past time orientation or perspective. Seven years later, this research gap still exists. Eson (as cited in Laffey, 1965) studied the time orien- tation of different age groups (10, 16, 25, and 65 year olds) and interestingly all showed a tendency toward greater future orientation, including the 65 year olds. Kastenbaum (1955) studied the future time perspective of 24 relatively intact older people (median age 77, range 66-89) as compared with 24 younger peOple (median age 19, range 18-24) differentiating between "cognitive and personal futurity." Cognitive futurity was assessed by completions of three story roots aimed at future, past, and unSpecified temporal orientations. Personal futurity was assessed by the Important Events technique. gs were asked to give the nearest personally important event in their past, the next nearest and the temporal distance between them. Future events were requested in a similar way. With reSpect to cognitive futurity there were no signifi- cant differences in extension or density while older gs stories were significantly more complex and coherent. However, with respect to personal futurity olders §s showed signifi- cantly less extension and density. Thus, though these older gs had difficulty projecting themselves into their personal futures, their ability to use the future in an abstract cognitive way to order experience in general equalled that of young adults. 56 Some theoretical factors affecting futurity. An observa— tion by Fraisse (1965) seems appropriate here. Generally Speaking, the future only unfolds in so far as we imagine a future which seems to us to be realizablp. This is clearly shown by our attitude to death. Obviously we all know that this is the end that awaits us, but it causes anxiety in us or a form of religious behavior which . . . is a defense against the unknown; but, as Merleau-Ponty remarks, it never becomes part of our temperal perSpective, whatever our age. . . . When seen as an absolute end, death is not an objective to be reached (pp. 172—175). Fraisse also emphasizes the importance of the temporal distance between the present and future situation, implying that the trauma of approaching death is greater for the aged than for younger people. Butler (1965a) implies that this trauma is greater in those aged who have always put a heavy investment on the future; the future which their old age cannot deliver. Relevant to this is Einstein's (1964) finding that following eXperimentally induced loss of "self-integrity," his adoles- cent gs showed a significantly extended (implying unreasonably so) future time perspective. Perhaps the aged Butler speaks of have always had a low level of self-esteem compensated for by the vague hope that one day their "ship would come in," only to realize in old age that their "ship" has foundered on the rock of too little time left. Fraisse (1965) asserts that on the plane of our experi- ence the future is practically nothing but the projection of desire or fear. In addition, a future perSpective depends on the possibility of escaping from a present determined by 57 the situation or from the domination of the past. There is no future without at the same time a desire for something else and an awareness of the possibility of realizing it. When in old age the potential for ordinary need gratification is increasingly restricted, the abstract goal of ego integrity would appear to be the only logical 'something else' which could motivate a continuing personal future perspective. As Erikson has observed, this demands an existential accept- ance of one's past life, and this in turn may require considerable working through of one's past conflicts. Thus, a creative reconstruction of one's past in old age may be necessary to maintain future perSpective. The results of a study by Costa and Kastenbaum (1967) lend some support to this notion. They found that in a sample of centenarians, those who were most able to draw on their storehouse of experi— ence by responding to questions about their earliest memory, the most salient historical event recalled, and their most personally exciting event, were the most likely to respond to a question about their future ambitions. Nawas and Platt (1965) and Platt and Taylor (1967). based on an Adlerian approach, take somewhat the opposite tack. Though not dealing with aged gs, they focus on the lack of future time persPective in nostalgia (cf. reminiscence) and homesickness cases and suggest that treatment should proceed by reassessment of future goals. While this may work with younger adults, one wonders about the applicability of 58 this approach to the aged where restriction of future goals is reality based. In connection with another problem of the aged Halpern (1967) found that placing a high valuation on the importance of work is negatively related to future time perspective and satisfaction with retirement. He found no relationship between future time perspective and chronological age in his sample of 65 retirees with adequate retirement income. Halpern concludes that the extent of future orientation is not only a develOpmental phenomenon but is mediated by cul- tural and personality factors as well. Bienningl desire for change, satisfaction, and age. Recently Spence (1968) published a study of "futurity" and adaptation in a sample of 226 community residents who took part in the Langley-Porter San Francisco study (Lowenthal et al., 1968). In this study, however, "futurity" was assessed mainly on the basis of whether or not the gs planned their activities a week or two in advance and whether or not they desired a change in their present situation. Active planning as opposed to the more passive thinking about the future implied in other future time perspective studies would appear to alter the entire concept and makes comparisons difficult. The measure of adjustment also raises questions of the comparability of findings. Spence used a morale score based on §fs mood-state which, through cluster analysis, then perudtted separating gs on the basis of high vs. low 59 satisfaction with their current state. On the basis of these measures his gs were divided into four grOUps: (g) unsettled planners--those who planned and desired change (52.7%), (p) composed planners--those who planned but desired no changes (29.2%), (g) the disgruntled--those who did not plan but desired change (19.5%), and (Q) the complacent-- those who neither planned nor desired change (18.6%). Spence's results showed that planning is positively related to satisfaction, desiring change is negatively re- lated to satisfaction, and that there is no interaction effect between the three variables. He states that the two principal relationships are statistically significant and independent. Spence separately analyzed the effects of socio- economic status, sex, and age on these relationships. With respect to age he found (g) an inverse relationship between making plans and age holds primarily for those who do not desire change, and (2) an inverse relationship between desir- ing change and age holds primarily for those who do not make plans. This implies that those who desire change continue to plan; however, with increased age, i.e., age 75 and over, this relationship tends to disappear. The latter finding appears to fit the picture of Gorney's (1968) Very-Old Life- Review resolution gs. "Temporality," experiencing, andgpatterns of reminiscence with age. Gorney (1968) also furnishes data on "temporality" 40 for some of his §5.5 "Temporality" appears to refer to the complexity of one's past or future time perspective. Gorney's "waiting list" gs were assessed for future "temporality" on the basis of their expectations or denial of forthcoming institutional life. In addition all of his §s were assessed on past "temporality“ based on a.comparison of themselves at the time of the study with what they felt themselves to have been at age 40. The two measures of complexity of past and future temporal organization correlated positively and sig- nificantly with experiencing ability for all gs, while denial of impending institutionalization was negatively related to experiencing; all as predicted. With respect to reminiscence types a more complex pat- tern emerged. "Flight from the Past" gs had poorly articu- lated expectations, "Active Life Review" gs had moderately articulated expectations, and "resolved" §s had highly articu- lated expectations. Effective articulation of the past was most characteristic of active life-review gs while denial of institutionalization characterized flight from the past gs. Gorney interprets these findings as essentially congruent with Butler's (1965a) description of the life review process. Summary The bulk of temporal experience literature reviewed generally characterizes the aged as primarily oriented to the 5The reader is referred to the earlier section on "Reminiscing" (pp. 24-28) for a more complete discussion of this study. 41 past with constricted and sparse future time perspectives, though there are some discrepant findings. These temporal characteristics are also consistently linked to situations of maladaptation and depression, and have been associated with institutionalization, as well. .However, there is a distinct lack of adequate research into past time perspective as well as a dearth of research into the temporal experience of the aged. Some of the explanations for a past orientation were discussed. Fraisse (1965) has pointed out that people seem to concentrate on the longer portion of their life, i.e., in youth--the future, and in old age--the past. Perhaps more significantly, Fraisse writes that when the future seems blocked, the result appears to be concentration on the past. The future must have a goal or goals, and be realizable to maintain a reasonable future perSpective. The possibility of unreasonably long and vague future time persPectives resulting from poor self-esteem and leading to maladaptation was also raised. Kastenbaum (1965) has demonstrated that the commonly noted lack of futurity in the aged is limited primarily to their personal future, while, in intellectually intact aged, the capacity to use the future in an abstract cognitive way equals that of younger adults. Contrary to the implications of most writers in this area, the present author holds that concentration of the aged 42 on the past is not necessarily a symptom of maladaptation but perhaps a necessary step in the resolution of life con- flicts in order to attain ego integrity. Perhaps only when an adequate resolution has been achieved can a reasonable future time perSpective be maintained. Some of the continuing problems with terminology were noted as in the Spence study, as well as problems of differ- ing assessment methodology. Some indirect SUpport for Butler's life review process was furnished by Spence's study and more directly by Gorney's research. Many of the studies reviewed could also be used as support for the interpretation of a decline in ego energy among the very old. The following section will deal with some aSpects of adjustment and self concept in the aged. Self-Concept Problems with the Concept of Adjustment and Regionale for Use of Self-Concept As initially conceived, the purpose of this research was to study the relationship of reminiscence and time orienta- tion to adjustment or adaptation in the aged. It soon became clear, however, that the concept of 'adjustment' is so broadly and nebulously defined that the plethora of research defini- tions are almost idiosyncratic to particular investigators. One need only briefly review the various meanings attached to the concept and the variety of assessment methods in the 45 studies reviewed to verify this. Some have limited the concept to experimenter judgments of narrow categories of strictly observable behavior while others use subjective judgments of happiness, morale, life-satisfaction, etc. Some criteria are sociologically defined while others con- sider only absence of psychOpathology. Then again should adjustment be defined as a statistical norm or an ideal norm? Do we consider only passive adjustments, i.e., adaptations within the individual to conform to the environment, or the active impinging of the individual on the environment, or both? These questions raise fundamental issues of semantics and methodology in science which are beyond the scope of this research. Because of these problems it was decided to substitute a concept which is generally considered to be a significant determinant and reflection of "adjustment", however defined, viz., self-concept. Though this concept too has been variously defined and assessed, it was felt that use of a reliable standardized measure with adequate validity data would aid in interpreting the generality of the findings without adding to the confusion of terminology and assessment. Aging and Self-Concep; Since self-concept is widely known and used, no extensive review of self-concept literature will be undertaken here. In general, the literature on aging and self-concept indicates an overall pattern of positive self-concept, happiness, morale, 44 self-acceptance, etc. rising to its highest point in the late 20's and 50's and gradually declining after that (Bloom, 1961: Cameron, 1967: Dodge, 1961; and Kuhlen, 1948). However, Grant (1967) using the Tennessee Self Concept Scale, found a tendency to report a more positive self-concept with increas- ing age. He suggests though, that this may be a function of increased denial in the aged rather than an actual increase in self-esteem. He also found that older people tend to view themselves as more passive, less mobile, and less active than younger people. One could argue that the differences Grant found are obviously reality based and that the higher self-concept scores in his aged gs do represent a greater degree of self- acceptance rather than denial. Investigator values of these traits may well bias the interpretation of such findings and obscure the values that the aged themselves place on these traits. Grant's findings also point to the interaction of sex and socioeconomic status with age as determinants of self-concept. Such factors, therefore, must be considered in any study of aging and self-concept. Summary In this section the rationale for the use of self- concept rather than adjustment in the present study was dis- cussed and a brief review of the trends of covariance between chronological age and self-concept was given. The possibility of investigator bias in interpreting self-concept data with 441. the aged was noted as was the need to consider sex and socio- economic factors in any study of age and self-concept. The next section will attempt a formulation of the present research problem and the relationships to be investi- gated. PROBLEM Formglation The research literature on adaptation to aging suggests that the present need is for more intensive study of normal aging in order to determine what coping mechanisms can en- hance the attainment of ego integrity. Several studies have shown that many so-called "maladaptive" defenses actually do have adaptive value. McMahon and Rhudick (1967) have sug- gested that reminiscing is an adaptive c0ping response. Therefore, one purpose of the present research is to assess the adaptive potential of reminiscing in a less select sample of aged males than those used in their Boston area study. The finding of a positive correlation between quantity of reminiscing and positive self-concept would support their contention. However, Butler's (1965a) life-review hypothesis and Gorney's (1968) research (with what can be considered a more typical aged population) suggest that the reminiscing-self- concept relationships is not so clear-cut. In a cross- sectional study, such as is within the scope of the present research, these latter studies would suggest that a high Quantity of reminiscing, cepecially of the intrapsychic 45 46 conflict variety, would actually be associated with a low self-concept. Lack of reminiscing then could be interpreted two ways: (3) as maladaptive avoidance of life-review activity, or (p) as adaptive life-review "resolution." Therefore, one could only expect a low non-significant over- all correlation between reminiscing quantity and self-concept. Whatever adaptive potential is inherent in reminiscing or the life—review process might require an ideally controlled longitudinal study to determine ultimate outcomes of resolu- tion or presumably irreversible maladaptation. Even these results would probably be confounded by the intervention of death. Perhaps questions about the adaptive potential of remin- iscing can only be dealt with by a qualitative analysis of the reminiscing content and by a rough separation on the basis of age. Following Gorney (1968), those with little reminiscing should then fall into the categories of life-review avoiders in their 60's and 70's, and life-review resolved gs in their 80's and 90's. However, other types of qualitative differ— ences in the content of reminiscing have been mentioned by McMahon and Rhudick (1967), and by Butler (1965a) though the nature of these differences has not been adequately explored. This study will eXplore qualitative differences between high and low self-concept gs at various age levels in order to determine if life-review conflict resolution activity is the Host significant type of reminiscing, whether other types 47 such as regressive idealization of the past and well adjusted "story telling" are also important, or whether some other set of variables can better eXplain the reminiscing-self—concept relationship. The next relationships to be explored are those dealing with time orientation of aged male gs and their self-concept. The aged have typically been assumed to be oriented to the past, and a past orientation has typically been linked with maladaptation. However, there is some conflicting evidence as to the past orientation of the aged, and also evidence suggesting that concentration on one's past in old age may be a necessary developmental step leading to the goal of final ego integrity. Another relationship that has not as yet been investi- gated is that between reminiscing and time orientation. Though one would assume that, for example, much concentration on the past would automatically preclude any significant con- centration on the future, implications from various studies would suggest that a reasonable future orientation may only be possible when one can and does freely use one's past ex— periences. Perhaps most maladaptive is the inability to effectively use more than one temporal direction, whether past, present, or future. The final set of relationships to be investigated, while not the primary concern of this study, are of interest and are relevant to much past research in gerontology. These 48 have to do with the relationship of institutionalization (or the complex of factors associated with institutionaliza- tion) to the three major variables of the study, viz., reminiscing, time orientation, and self concept. Previous research (e.g., Fink, 1955) has shown that institutionaliza- tion is associated with a past time orientation and presum- ably with increased reminiscing and a poorer self concept. Within the limits of the design this study will test these relationships. Summary of the Major Relationships to be Investigated Because of conflicting or non-existant research findings concerning the relationships of the major variables of this study no eXplicit hypotheses will be tested. Rather, since the purpose of this study is to generate hypotheses, the major relationships to be explored will simply be stated. I. Reminiscing by self-Concept. Two aspects of reminiscing are to be investigated; overall quantity and qualitative differences between gs that consti- tute distinct reminiscence types. II. Reminiscing by Time Orientation. III. Self-Concept by Time Orientation. IV. Institutionalizatione-by: 6The present study is not designed or intended to investi- gate effects of institutionalization as such. Rather, it is designed to investigate differences between aged males living in nursing homes as opposed to those living independently in 49 (g) Self-Concept (2) Reminiscing (9) Time Orientation the community. It is assumed that the obtained differences will result not only from the effects of location of residence but also from the complex of factors that determined that location, e.g., general level of health, availability of others to care for the g, general socioadaptive functioning, etc. METHOD Subjects The total sample consisted of 60 men, born in 1910 or before who have achieved a raw score of at least 100 on the Peabody Picture Vocabulary Test (PPVT) (Dunn, 1965). _pptitutional subsagple. -One-half of the total sample was drawn from nursing home (NH) residents. In addition to having achieved an adequate PPVT score these residents were selected by age groups with 10 Se in their sixties (60's). 10 in their seventies (70's), and 10 eighty years old or over (80+). Only gs who were physically capable of the tasks required for this study, and who agreed to participate were selected. The initial recruitment was made at a V. A. Nursing Home Care Unit. The remainder of the subsample was recruited from other area nursing homes until the total sub- sample was complete. Community Volunteer subgamplg. The other half of the total sample was drawn from community volunteers (CV) resid- ing independently in the same metr0politan area (Battle Creek, Michigan). These gs also were drawn by age groupings, i.e., 60's, 70's, and 80+, and the first 50 volunteers to meet 50 51 the PPVT and age requirements were chosen. The initial re— cruitment of CV gs was from people active in the city's Senior Recreation program. The two subsamples were selected in order to insure a wide range of socioadaptive functioning (see Table 1) based on the assumption that community volunteers would be better adjusted and have a better self-concept than aged gs who for various reasons require institutional care to survive. This assumption was tested and confirmed. Only male gs were used since the V. A. Nursing Home Care Unit is limited to male residents and since inclusion of female volunteers would introduce possible sex differences which might confound the results. Testgpand Measures Intellectualgfunctionipg. In order to assure at least average intellectual functioning, only persons with raw scores of 100 or above on the geabody,PictupegVocapulapy .Eggg (PPVT), Form.A (Dunn, 1965) were selected as gs. The PPVT was used as a selection device for the follow- ing reasons: (3) it is easily and rapidly administered (approx. 15 minutes), (p) the ability presumably tapped, i.e., vocabulary, is known to "hold up" well and may even increase with age (Latimer, 1965), (g) vocabulary is the best single index of overall intelligence, and (g) the present investi- gator's personal experience has shown the test to be better tolerated and accepted by geriatric patients than other instruments. 52 Table 1 Comparative Demographic Data for the NH and CV Subsamples . Subsample Class of Informatlon NH CV ‘N 50 50 Mean age by age grOUP (in years--months) 60's 65-6 64-9 70's 75-5 75-0 80+ 85-10 85-2 Marital Status Single (never married) 25.5% 0% Married 20.0% 66.7% Widower 56.7% 50.5% Divorced 20.0% 5.0% Occupational level Professional/Managerial 6.7% 25.5% Entrepreneurial 16.7% 20.0% White collar 20.0% 25.5% Skilled labor 16.7% 5.5% Unskilled labor 56.7% 20.0% Farmer 5.5% 10.0% Education Grades 1 through 4 5.5% 6.7% Grades 5 through 8 26.7% 56.7% High school 50.0% 15.5% College 10.0% 26.7% Other Specialized training 50.0% 16.7% State of health (§]s own estimate) Excellent 10.0% 26.7% Good 55.5% 60.0% Poor 56.7% 15.5% 55 While the investigator is well aware of the deficiencies of the PPVT for this age group (e.g., lack of adequate norms and spuriously high I.Q.s reported for other age grOUps) it is felt that use of raw scores rather than I.Q.s and the above mentioned advantages make its use justifiable for the purposes of this study. The PPVT is an individually administered 150 item vocabu— lary test. Typically only about 1/5 of the stimulus words are presented because a "basal" and "ceiling" are obtained precluding use of words above and below this range in diffi- culty. .g is required to respond to each stimulus word by selecting the most appropriate of four pictures presented simultaneously on the plate for that item. The pictures are line drawings of objects or activities, one of which is re- lated to each stimulus word. For example, to stimulus word I'150 gauntlet", §_mmst choose between pictures of 1) a drill chuck, 2) a slingshot, 5) a radar antenna, and 4) a heavy glove complete with a gauntlet. The choice of a "cut score" of 100 was made because norms for the highest age group (17-6 to 18-5) show this to correspond to an 1.0. of 91 for that group. Though adult gs of above average ability bften go 'over the top' on this test, the PPVT at least establishes their vocabulary ability to be in the range required for inclusion in this study. Reminiscing. Methods used in prior studies of reminiscing include a "non-directive" interview (McMahon & Rhudick, 1967) 54 earliest memory (Tobin & Etigson, 1968), and an "Evaluation of Life Questionnaire" (Gorney, 1968) in addition to the reminiscing which occurred spontaneously in the process of the psychoanalytic case studies of Butler (1965a). A major problem in the elicitation of reminiscence is to provide a sufficiently unstructured situation to allow for Spontaneous reminiscing to occur while still maintaining a sufficiently standardized set of stimulus conditions to make valid comparisons between gs possible. Various inter- view formats were tried with some preliminary gs, ruling out the non—directive approach and a "Public Opinion Sampling" Questionnaire ruse as unproductive of reminiscing. Also a more direct interview failed to yield much reminiscing unless the §fs responses were extensively followed up, thus destroy- ing any comparability between gs. In all of these preliminary interview formats the present investigator was attempting to obtain responses yielding not only reminiscing data but also data that could be used to assess the §fs major temporal orientation. However, finding neither dimension adequately assessed, separate techniques were devised to assess them independently. A shortened and modified version of Gorney's (1968) Evaluation of Life Questionnaire combined with requesting the gfs two earliest memories was used to more directly elicit reminiscing. This questionnaire, dubbed the Life Experience Questionnaire for this study, is contained in Appendix A. 55 It assesses the §fs overall feeling about his life, his earliest memories, his main satisfactions and disappoint— ments and the most difficult and happiest periods of his life. A frequency count of the statements dealing with the past was the measure of reminiscing quantity. A system of reminiscence classification was developed partly on a post hoc basis. Several aSpects of reminiscing were considered including intrapsychic conflict resolution attempts, defensive use of reminiscing, non-conflictual "story telling", success-failure themes, interpersonal aSpects, self vs. other preoccupation and other qualities that emerged from the data. The relationship of these reminiscence types to self-concept in particular was examined with a view to deter- mining which, if any, aSpects of reminiscing appear to aid in the attainment of ego integrity. The reminiscence cate- gories finally chosen were (2) conflictual, (p) avoidant, (g) defensive, and.(g) well adjusted. The description of these categories is contained in Appendix C. Self-Concept. The measure of self-concept was the "Total P Score" on the Tennessee Department of Menpgerealth Self Concept Scale (TSCS). (Fitts, 1965). The TSCS is a self-administered scale consisting of 100 declarative self-referential sentences. For example, "19. I am a decent sort of person.“ §_re8ponds by circling the apprOpriate number on a 5-point scale giving his degree of agreement-disagreement. The answer sheet marked by g has all 56 the item scales arranged the same so that 1 stands for "completely false" and 5 stands for "completely true". However, since the TSCS includes both negative as well as positive valence statements, the scoring sheet (in carbon "snap-out" form), below the answer sheet, has the numerical sequence reversed for the negative valence statements. Thus, if a person circled "2" meaning "mostly false" as his reSponse to "25. I am a bad person.", his scoring sheet would Show the number "4“ circled. The numerical sequence for positive valence statements is the same as on the answer sheet. Total P score is the sum of the numbers circled on the scoring sheet. In this way Total P score reflects §fs positive self—concept. The TSCS standardization sample (NF626) was a broadly based representation of the United States population. The age range was 12 to 68. Further research has shown the norms to be adequate and "the effects of such demographic variables as sex, age, race, education, and intelligence on the scores of the Scale are quite negligible" (Fitts, 1965, p. 15). _Time Orientation. Two methods were used in an attempt to assess each g's temporal orientation. The first method consisted of questions requiring §_to state which temporal direction he currently thought about most, and which he cur- rently though about least, i.e., past, present, or future. He then was required to estimate the percentage of his waking hours he currently thought about each temporal direction. 57 This direct assessment method had not previously been tried and its validity was unknown, therefore, a second less direct assessment instrument, the Thematic Apperception Test (TAT). which has been tried (e.g., Epley & Ricks, 1965) was also used. Three TAT pictures (cards 1, 4, & 7BM) were used to elicit stories. Instructions were fairly standard requesting “g to make up a story about each picture using as much crea- tivity as possible. However, each g was asked to assume that each picture was a scene that was occurring somewhere "right now." He then was requested to tell a story stating what happened in the past, what was happening in the present, and what would happen in the future. The percent of the §fs TAT responses dealing with each temporal direction was the measure of his relative time orientation. The TAT cards were chosen to represent males in three major phases of life. Card 1 (a boy contemplating a violin) represents childhood, card 4 (a young man who appears to be pulling himself away from a young woman) represents young adulthood, and card 7BM (an older man with a younger man who appears sullenly staring into Space) represents later maturity. Use of the TAT for this specific research purpose involves the g_priori assumption that the measures, as outlined above, actually do tap the underlying personality variable desired, viz., time orientation. Rabin (1968, p. 617) rightly insists that such p_priori assumptions with projective techniques 58 must be pretested and experimentally validated. That the TAT can be used to assess temporal eXperience data is clear (Wallace & Rabin, 1960). However, the method and measures in the present study differ somewhat from those used previously. Therefore, while use of these procedures provided a valuable preliminary test of the p_priori assumption, confounding effects were noted and the results can only be considered as tentative evidence of the underlying personality construct, time orientation. Procedures Procedures with each_§v The purpose of the study was briefly explained to each potential §_as "research to learn how people adjust to the problems of later life so that we will know how to help other people adjust as they grow older." When their consent to participate was obtained each potential S was administered the Personal Data Questionnaire (Appendix B) and the PPVT. Any ineligible volunteers were then thanked and excused. Each eligible §_was then given the TSCS followed by a tape recorded interview with the Life Experience Questionnaire, time orientation questions and TAT cards. All of these assessment instruments were individually and, insofar as possible, uniformly administered to all gs. Recruitment of is. The data collection process took approximately five months to complete. When the initial 59 sources of SS were exhausted (i.e., the V. A. Nursing Home Care Unit and the city's Senior Recreation Center), intensive and extensive efforts were necessary to recruit the remaining 60% of the sample. Virtually every nursing home within a 20 mile radius and a few as far away as 50 miles were contacted until the NH subsample was complete. Similarly, newspaper advertisements, a small newspaper article, an article in a V. A. employees newsletter that goes to all retirees, and personal contacts through friends, fellow employees, and the county welfare department were necessary to complete the CV subsample. The investigator had no intimate knowledge of any of the [gs prior to the study and was acquainted with only a few. A few are community leaders and were known somewhat by reputa- tion. Though the sampling method was obviously not random, every practical effort was made to prevent systematic bias. Negroes are under-represented with only two of the total N of 60. For the NH subsample several Negroes were screened but failed to attain the criterion raw score on the PPVT (undoubtedly due in part to discriminatory educational poli- cies extant in their youth). For reasons unknown to the investigator few Negroes responded to the channels used to recruit CV gs. Scoring of measures and development of reminiscence classificationfipystem. Following the data collection, the TSCS protocols were independently scored by a person other 60 than the investigator and recorded on data sheets. In order to obtain the measures of reminiscing quantity and relative time orientation, it was first necessary to get a tally of statements referring to past, present and future from the tape recorded responses to the Life Experience Questionnaire, Time Orientation questions, and TAT stories for each g, Therefore, all taped interviews were transcribed into typescript form. Rules of classification were then develOped to classify (p) which utterances constituted statements, and.(p) to identify the temporal referent(s) of each scoreable statement. These rules are contained in Appendix D. Since some state- ments contained more than one temporal referent, the classifi- cation system in Table 2 was develOped. Table 2 Time Orientation Classification Symbols and their Respective Temporal Referent(s) Symbol P PN N NF F PNF Time Past Past & Present Present & Future Past, Referent(s) Present Future Present & Future Note.--"N" for "Now" was used to denote Present in order to avoid confusion with "P" for "Past." 61 All of the utterances in all of the transcribed inter- views were separated by the investigator into scoreable statements and classified as to their time referents. A sample of five interviews was then drawn at random from the total sample, another judge was trained in the classification system, and he independently judged all the utterances for that subsample as to scoreable statements and their time referents. A Pearson correlation coefficient calculated for the total of 1594 paired decisions on the five interviews yielded an inter-judge reliability coefficient of .77. Analysis of the disagreements revealed that most were due to a trivial difficulty which had no effect on the accuracy of the principal judgments.7 After establishing the reliability of the classification, the number of each §fs statements referring to the past on the Life Experience Questionnaire and past, present, and future statements in the TAT stories were recorded on data 7Many4§s indulge in Speech mannerisms with repeated use of phrases such as "you know," or "yes, sir" which, while clearly not scoreable statements, can be included either at the end of one or the beginning of the next scoreable state- ment. In many instances both judges classified adjacent com- plete statements identically but differed in their choice of which statement the intervening nonscoreable phrase should be tacked onto. Since disagreements as to what constituted a scoreable statement resulted in no score for one judge, these zero ratings reduced the computed correlation. The Pearson p.was calculated by assigning the following values to the time classifications: (no time classification PNF made by one judge) P PN N NF F O 1 2 5 4 5 There were only three PNF statements out of the total of 1594. 62 sheets. The former constitutes the measure of reminiscence quantity. To arrive at the measure of relative time orien- tation, each g's total number of TAT statements were summed and the percent past, percent present, and percent future recorded. The use of percents rather than raw numbers of statements, serves to equalize individual differences in TAT productivity. Following qualitative examination of the interview material these reminiscence categories were identified: (g) conflictual reminiscing, (p) avoidance of reminiscence, (g) defensive reminiscing, and (g) well adjusted reminiscing. More detailed descriptions are found in Appendix C. Other aspects such as success-failure, self vs. other preoccupa- tions, etc., were considered as well. However, these latter aspects have been taken into account in the classification criteria used, while use of the four categories listed has the advantage of providing more direct comparisons with, and extensions of prior research (Butler, 1965a; Gorney, 1968; McMahon & Rhudick, 1967). Based on the criteria in Appendix C, the investigator classified each interview into one of the four categories. Since the SS were drawn by age group and type of location (NH vs. CV), forming Six groups of 10 gs each, three inter- views from each of the six groups (18 interviews) were randomly selected for a reliability check. Another Clinical Psycholo— gist was given the criteria and independently classified this 65 18 interview subsample. Only two disagreements resulted yielding an inter-judge reliability contingency coefficient of 0.76.8 Discussion of the disagreements indicated that in each case each judge had considered the only alternative classification to be the one chosen by the other judge. Considering the type of data it was agreed that the accuracy of classification was adequate and the criteria should permit replicability of the findings in further research. 8Edwards (1954, p. 581) points out the limitations of this statistic and notes that it can approach 1.00 only when the number of categories for both criteria is very large. For the number of categories in this case Q cannot exceed 0.816. However, Q_is the only useable statistic for this type of data. RESULTS Reminiscing and Self-Concepp Reminiscing quantity and self-concept. The Pearson product moment correlation coefficient calculated between reminiscing quantity and self-concept for the entire sample is .15 (pp). McMahon and Rhudick's (1967) implication that reminiscing in and of itself is adaptive is not supported by these data. Likewise, inspection of a scattergram re— vealed no indication of any significant correlation between reminiscing quantity and self-concept for the separate age groups. Separate correlations coefficients were calculated to determine whether age itself correlated significantly with either of these two variables. This was done in order to focus on any possible deve10pmental trend. Again both E? were nonsignificant (-.05 for reminiscing quantity by age, and .15 for self-concept Total F score by age). No develop- mental trends were found. Reminiscence type and self-concept. Table 5 represents the results of the analysis of self-concept scores between reminiscence types. 64 65 Table 5 Self~Concept Score Means and Standard Deviations for Four Reminiscence Types, the Total Sample and the Tennessee‘Self—Concept Scale (TSCS) Norm Group Reminiscence Type TSCS Well Total Norm a Conflict Avoidant Defensive Adjusted Sample Group ' §_ 6 21 11 20 60 626 Mean 554 544 566 577 556b 545.57 S.D. 26 54 41 26 56 50.70 Note.--One way analysis of variance between reminiscence types yields E?5.27, g£_(5 & 57), p_<:.01. Nonsignificant Emax‘ indicates variances not grossly unequal. Individual comparisons using Newman-Keuls method show all adjacent cate- gory comparisons nonsignificant, all comparisons three steps apart significant (p <(.05), and comparison of extreme cate- gories highly significant (p_<:.01). aTSCS data from Fitts (1965, p. 14). bA two-tailed pf test between the means of the total study sample and the TSCS norm group yields p'=2.56. Using g; of study sample for a conservative test, p <:.05. In general these data suggest a hypothesis that the type of reminiscence a person engages in is, at the least, a re— flection of his level of adjustment as measured by his self- concept. In addition to the highly significant results of the reminiscence type analysis of variance, note that the mean self-concept level of the total sample is significantly higher than of the TSCS norm group. 66 . Reminiscing and Time Orientation9 Reminiscing quantity and time orientation. Separate Pearson product moment correlations were calculated for all §S between their reminiscing quantity and‘percent past state- ments on the TAT, percent present statements on the TAT, and percent future statements on the TAT. All three coefficients proved to be very small and statistically nonsignificant (past = -.05; present = .02; and future = -.04). These find- ings suggest that an §fs overall quantity of reminiscing and his tendencies to talk about past, present, or future events in the telling of stories not directly related to his own life are independent. Reminiscence type and time orientation. In the study as initially conceived, the investigator intended to classify gs on the basis of their relative time orientation and then to relate these categories to the various reminiscence type categories. This was to be done in two ways simultaneously. First, gs were to be classified into (g) those using one temporal orientation predominantly, (p) those using two time directions on the TAT roughly equally while neglecting use of the third, and (g) those using all three temporal 9All of the time orientation analyses in the Results section used the TAT measure as described earlier. The direct question method of assessing time orientation, in the opinion of the investigator, was found to lack any validity due to §fs wideSpread inconsistency in responding and lack of comprehension by many gs. .This topic will be discussed in more detail under an Additional Results section. 67 directions in roughly equal prOportions. The standard for predominant use of one time direction was arbitrarily set at 55% or more in the research proposal. Use of any time direction 10% or less was set as the arbitrary standard indicating neglect of one time direction. Secondly, in classifying gs on the above basis the various time directions and combinations of time directions were to be recorded and related to their reminiscence types. The resulting contingency table would be a 4x7 table with four reminiscence types by (3) past, present, and future for the one predominant orientation, (2) past-present, present- future, and past-future, for a two-way orientation, and (g) past-present-future for the "balanced" orientation. However, problems relating to an artifact of the TAT procedure as a measure of time orientation rendered the proposed analysis inapplicable. The mean percentages of the total sample for past, present and future reSponses reSpec- tively, is as follows: 15%, 59%, and 28%. Obviously, most gs tended to respond more to the immediate stimuli (each TAT card) than they did to whatever tendencies they might have had to talk about the less immediate past or future. This "card pull" effect will be discussed in more detail later. However, because of these findings, the arbitrary standard for one predominant orientation was modified from 55% to 65% and the intended analysis was carried out. Only three of the seven time orientation categories had any entries, therefore, Table 4 has been collapsed to reflect this. 68 Table 4 Frequencies of gs Classified by Reminiscence Type and Time Orientation Usage Time Orientation Usage One Two Reminiscence Orientation Orientations Balanced Row Type Dominant Used Use Totals Past- Present- Present- Present Future Future Conflict 5 1 4 8 Avoidant 12 0 9 21 Defensive 4 1 5 10 Well-Adjusted 4 5 11 18 Column totals 25 5 29 57a Note.-- X? (6gfi) = 4.08 (pg). The use of chi-square in the present table follows Karon's (1968) rationale. aThreeg's of the total N,of 60 told no TAT stories. No meaningful inferences can be drawn from these data about a relationship between reminiscence type and time orientation. An analysis of reminiscence type by age group was done and proved to be nonsignificant (see Table 5). To the extent that Gorney's (1968) "Conflict" and "Resolution" cate- gories are comparable to the "Conflict" and "Well Adjusted" categories of the present study, these data do not support 69 Table 5 Frequencies of gs Classified by Reminiscence Type and Age Group Reminiscence Type Age Well Row GrouP Conflict Avoidant Defensive Adjusted Totals 60's 1 6 4 9 20 70's 5 8 5 4 20 80+ 2 7 4 7 2O Column Totals 8 21 11 20 6O Note.-- 1? (6g:) = 5.60 (pg). The use of chi-square in the present table follows Karon's (1968) rationale. Gorney's finding of an age trend from conflictual life review activity in the 60's and 70's to life review resolu- tion in the 80's and 90's. Time Orientation and Self-Concept Self-concept scores for all §S were correlated sepa- rately with the percent measures of (3) past, (2) present, and (2) future reSponses on the TAT. All coefficients were very low and nonsignificant (s-c x past %%.06; s-c x present %%—.14; s-c x future %% .02). Because the three correlation coefficients were computed separately, the artifact of "card pull" producing more "present" responses overall, could not have any effect on 70 these analyses. The correlations for each time direction are independent. The results suggest there is no relation- ship between these aged males' measured self-concept and their tendency to talk about past, present, or future events in telling stories not directly associated with their own lives. Institutionalization Differences in these aged gs associated with their location of residence, i.e., nursing homes vs. community, were analyzed for all of the major variables of this study. Age group differences were also investigated by two-way analyses of variance. ‘gplf-concept. Tables 6 and 7 contain the results of the analyses by location of residence and age group using Total P scores as the dependent variable measure of self- concept. The results confirm the assumption made prior to the study that community residents have a higher level of adjust- ment than nursing home residents as measured by self-concept scores. This differential adjustment level agrees with past research which has demonstrated this relationship repeatedly in many different samples using a variety of assessment methods (e.g., Davidson & Kruglov, 1952; Fink, 1955; Pan Ju-Shu, 1948: Tobin & Etigson, 1968). Note, however, that for this sample the higher self-concept level is contributed 71 Table 6 Self-Concept Analysis of Variance: Age Group by Location of Residence Source d_f_ pp 3 Age (A) 2 904 .69 Location (L) 1 5577 4.12* A x L 2 1854 1.42 Error 54 1504 *3 < .05 . Table 7 Self-Concept Score Means and Standard Deviations for Location of Residence by Age Group . . . Age Groop Row Cla531f1catlon 60's 70's 80+ Totals N 10a 10 10b 50 Community Mean 568 551 584 568 S.D. 41 24 28 52 N 10a 10 10b 50 Nursing Mean 549 552 546 549 Home S.D. 46 44 28 9 N, 20 20 20 60 Column Mean 558 552 565 558 Totals S.D. 44 1.4 85 56 aComparison of these subsamples yields pF.98 (18gf), (pg). bComparison of these subsamples yields t= 5.02 (18d;). p_ .05. Table 9 Reminiscing Quantity Means and Standard Deviations for Location of Residence by Age GrouP . . . Age Gropp Row C13351f1cat1°n 60's 70's 60+ Totals ‘N 10a 10 10b 50 Community Mean 285 140 206 210 S.D. 204 157 199 250 g 10a 10 10b 50 Nursing Mean 155 84 142 127 Home S.D. 278 88 129 120 N_ 20 20 20 60 Column Mean 220 112 .174 169 Totals S.D. 290 125 142 187 = 1.19 (1651:), (as) . .85 (189;) , (as) . aComparison of these subsamples yields kflfl u Comparison of these subsamples yields 74 hypothesis is that reminiscing quantity is one aSpect of a §fs overall willingness to communicate with a relative stranger, the investigator, and therefore, more a measure of "Openness" than of orientation to the past. This tOpic will be discussed in more detail later. Reminiscence type. What effect does an aged male's residence in the community or in a nursing home have on the type of reminiscing or lack of it that he engages in? Table 10 presents the frequencies of §S classified by remin- iscence type and location. Table 10 Frequencies of SS Classified by Reminiscence Type and Location A Reminiscence Type Classification Well Row Conflict Avoidant Defensive Adjusted Totals Community 4 ' 6 6 14 50 Nursing Home 4 15 5 6 50 Column Totals 8 21 11 20 60 Note.--The use of chi-square in the present table follows Karon's (1968) rationale. Using correction for continuity, 7L2 (59.1) = 6.46, 2<-10 but > .05. 75 The results of this analysis also proved to be inconclu- sive. Certainly the distribution in the Conflict and Defensive reminiscence categories could not be more evenly divided. The almost inverse relationship between the frequen- cies of §S in the Avoidant and Well Adjusted categories is, however, noteworthy. Time orientation. In order to independently assess the effects associated with institutionalization and age on the g's time orientation, separate two-way analyses of variance were calculated for each temporal direction. The analysis of variance summary data and presentation of means and standard deviations are contained in Tables 11 and 12. Table 11 Time Orientation Analyses of Variances: Age by Location, Using Percent TAT Responses to Past, Present, and Future as Dependent Variables in the Separate Analyses _A Time Direction as Dependent Variable Class of Data Past Present Future Source _d_f_ Mg pg MS F M; #1; Age (A) 2 19 .51 88 .51 54 .50 Location (L) 1 470 7.85** 1075 6.21* 117 1.04 A x L 2 14 .25 185 1.06 208 1.84 Error 54 60 175 115 Note.--Three gs in the NH 80+ group told no TAT stories. The means of the cells were used in place of the missing data in order to complete the analysis. *2 < .05. ** p_<(.01. 76 .aamcHouooom omumSmom ouo3 mcoHumw>mo oumocmum pom mcmofi omuoommm Has one sumo mchmHE ecu mo oomHm SH poms mnm3 maamo on» we momma one .moeuoum are on each mooum +om ms was on hm.omunem HH NH OH «H mH OH m #.N ¢.H. .Q.m mHmuOB mm hm mm mm mm ¢m mH HH pH comm GESHOU Op On on Om on On Op On On 2 pH O OH mm HH NH m h m .Q.m em new mm Hp mom on NH mHH «H comm. +Om ON OH OH ON OH OH ON OH OH 2 OH OH m H ON h mH m OH .Q.m mm Hm mm mm mm mm #H HH PH snow m.O> ON OH OH ON OH OH Om OH OH 2 OH m m Hm NH .TH mH m h .Q.m mm mm mm mm mm Hm mH OH PH cmwm m.Om ON OH OH ON OH OH Om OH OH 2 mamuoa mEom wuHcsa mHmuoa 080m wuHcsE mamuoa mEom humane coHumo 30m mchusz I800 30m mchusz I800 30m mchHsz IEOU IHMHmmMHO enough pcommum ummm wanmesm> ucoocoaoa mm pomp coHumucofluO OEHB msouo 004 >9 mocoonmm mo coHumooq How oususm one .ucmmoum .ummm Op mcHHHomom noncommom Ems vacuuom mo mcoHuMH>0Q osmocmum ocm memo: coHumucoHuO mEHB NH magma 77 The results of these analyses are biased somewhat due to the fact that three §s in the NH 80+ subsample would not or could not produce any stories in response to the TAT instruc- tions and stimulus cards. Since the number of missing gs was small it was considered preferable to use the mean of the seven remaining gs in that cell who did produce TAT stories and substitute it for the scores of the missing So. .The analysis then was carried out as if the data for the sample were complete. It is doubtful that this procedure had any significant effect on the overall results. -The results of the analysis for past orientation bear directly on the finding of Fink (1955). Here again, one must use caution in interpretation due to the limitations of the measure. However, contrary to Fink's finding of greater orientation to the past in institutionalized gs, these data suggest significantly greater orientation to the past in com- munity gs. Interestingly, a significant location main effect is also found for present orientation, though in this case the institutionalized gs appear more present oriented. There were no differences not attributable to chance on the analysis for future orientation. In summary, the results show that under the TAT stimulus conditions for this study community SS talk proportionately more about the past and proportionately less about the present than their institutionalized peers, while talking about the future in roughly equal proportions. 78 The lack of any significant age effect or interactional effect between age and institutionalization again fails to support Fink's finding though the results certainly cannot be considered conclusive. Additionaeresplpg In view of the questions about the validity of the main measure of time orientation used in the present study, several additional analyses were performed to shed light on factors affecting it. In addition, some analyses were per— formed to directly relate the use of past, present, and future responses on the TAT to each other. These in turn were re- lated to the answers gs gave to the direct questions about their temporal orientation. Finally, another score on the TSCS supposedly related to psych0pathology was examined as a possible cross validation measure of the self—concept analyses. TAT productivity. Because of the artifactual preponder- ance of TAT "Present" reSponses (card pull) found for all sub- samples, productivity on the instrument was analyzed. Tables 15 and 14 contain the results of this analysis for overall productivity. These results Show a significantly higher number of TAT responses for community gs and for the younger age groups. However, the CV 70's group again shows a marked deviation from the overall pattern of means rendering the interaction effect statistically nonsignificant. 79 Table 15 Analysis of Variance for Overall Productivity on the TAT: Age by Location Source _d_f N_S_ N Age (A) 2 1501 4.52* Location (L) 1 4421 15.55** A x L 2 680 2.56 Error 54 288 *- p_<3.05. *9!- p,<:.01. Table 14 Overall TAT Productivity Means and Standard Deviations for Age Group by Location of Residence . . . Age Group Row C13391f1cat1°n 60's 70's 60+ Totals N 10 10 10 50 Community Mean 56 40 49 48 S.D. 19 11 17 26 N 10 10 10 50 Nursing Mean 41 55 19 51 Home S.D. 22 16 16 56 N, 20 20 20 60 Column Mean 49 56 54 59 Totals S.D. 55 15 67 20 80 In order to examine the effect of differential TAT pro- ductivity associated with each temporal direction, separate analyses were computed. Summary data and the diSplay of means and standard deviations appear in Tables 15 and 16. Table 15 Past, Present, and Future TAT Productivity Analyses of Variance: Age by Location, Using the Number of TAT Statements Referring to Past, Present, and Future as Dependent Variables in the Separate Analyses Time Direction as Dependent Variable Class of Data Past Present Future Source if. 118. 1: be 4.; g 1 Age (A) 2 56 2.12 526 2.17 178 4.94* Location (L) 1 508 18.12** 856 5.57** 401 11.14** A X'L 2 4 .24 262 1.75 126 5.50* Error 54 17 150 56 * p, ucoococon on own: coHpmucmHHO 08HB muonw 00¢ an oucoonom mo coHumuoq How mcoHuomHHn QEHB 0>Huuommom ecu on OCHunommm muco8oumum B