<%BANNER%>

Possible Selves and Perceived Competence in Relation to Cognition and Health

Permanent Link: http://ufdc.ufl.edu/UFE0025007/00001

Material Information

Title: Possible Selves and Perceived Competence in Relation to Cognition and Health
Physical Description: 1 online resource (112 p.)
Language: english
Creator: Dark-Freudeman, Alissa
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: aging, cognition, efficacy, health, self, wellbeing
Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Possible Selves and Competence in Relation to Cognition and Health Markus and Nurius defined (1986) possible selves as elements of the self-concept that represent what an individual could become, would like to become, or is afraid of becoming. These personally meaningful hopes and fears are of central importance in the self-concept. Believing that we have the ability to attain our most important hopes and avoid our most dreaded fears may be central to maintaining positive psychological functioning or well-being. The central focus of this research was to examine age differences in possible selves related to cognition and health and to examine how beliefs about the competent self influence psychological well-being. We found that younger adults reported a greater number of cognitive selves than both middle-aged and older adults. The cognitive selves younger adults reported tended to be hopes related to education; the cognitive selves middle-aged and older adults reported tended to be hopes related to learning new skills and continuing to engage in cognitively stimulating activities, as well as fears related to cognitive decline. In contrast, middle-aged adults reported a greater number of health fears than both older and younger adults. These fears included statements related to specific diseases such as having a heart attack or stroke, or general statements related to maintaining or losing health. To examine the relationship between age, competence and psychological well-being several structural equation models were tested. The paths between competence and psychological well-being were significant for health but not for cognition. We belief this was due to the fact that our younger, middle-aged, and older adults all reported high levels of competence related to cognition. Our sample was high functioning and reported relatively high levels of domain-specific competence and as well as high levels of positive psychological well-being. Unfortunately we were not able to test the relationship between low levels of competence and psychological well-being. Our results still support that idea that believing we have the ability to attain our most important hopes and avoid our most dreaded fears, especially when faced with declines in health or cognition that may threaten our ability to do so, may be central to maintaining positive psychological well-being and skilled performance as we age.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Alissa Dark-Freudeman.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: West, Robin L.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025007:00001

Permanent Link: http://ufdc.ufl.edu/UFE0025007/00001

Material Information

Title: Possible Selves and Perceived Competence in Relation to Cognition and Health
Physical Description: 1 online resource (112 p.)
Language: english
Creator: Dark-Freudeman, Alissa
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: aging, cognition, efficacy, health, self, wellbeing
Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Possible Selves and Competence in Relation to Cognition and Health Markus and Nurius defined (1986) possible selves as elements of the self-concept that represent what an individual could become, would like to become, or is afraid of becoming. These personally meaningful hopes and fears are of central importance in the self-concept. Believing that we have the ability to attain our most important hopes and avoid our most dreaded fears may be central to maintaining positive psychological functioning or well-being. The central focus of this research was to examine age differences in possible selves related to cognition and health and to examine how beliefs about the competent self influence psychological well-being. We found that younger adults reported a greater number of cognitive selves than both middle-aged and older adults. The cognitive selves younger adults reported tended to be hopes related to education; the cognitive selves middle-aged and older adults reported tended to be hopes related to learning new skills and continuing to engage in cognitively stimulating activities, as well as fears related to cognitive decline. In contrast, middle-aged adults reported a greater number of health fears than both older and younger adults. These fears included statements related to specific diseases such as having a heart attack or stroke, or general statements related to maintaining or losing health. To examine the relationship between age, competence and psychological well-being several structural equation models were tested. The paths between competence and psychological well-being were significant for health but not for cognition. We belief this was due to the fact that our younger, middle-aged, and older adults all reported high levels of competence related to cognition. Our sample was high functioning and reported relatively high levels of domain-specific competence and as well as high levels of positive psychological well-being. Unfortunately we were not able to test the relationship between low levels of competence and psychological well-being. Our results still support that idea that believing we have the ability to attain our most important hopes and avoid our most dreaded fears, especially when faced with declines in health or cognition that may threaten our ability to do so, may be central to maintaining positive psychological well-being and skilled performance as we age.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Alissa Dark-Freudeman.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: West, Robin L.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025007:00001


This item has the following downloads:


Full Text

PAGE 1

1 POSSIBLE SELVES AND PERCEIVED COMPETENCE IN RELATION TO COGNITION AND HEALTH By ALISSA DARK FREUDEMAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENT S FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Alissa Dark Freudeman

PAGE 3

3 To my Mom, Darlene Ann Dark, I miss you.

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my advisor and mentor, Robin Lea West, for he r guidance and encouragement over the years. I cannot say enough about the impact she has had on my life. I can only hope that as I move forward in my career I am able to provide the same guidance and support to my own students. I would also like to tha nk the National Institute on Aging for funding my research. This research project was supported by a National Research Service Award: 5F31AG031832 02 I would like to thank my mother and my g randparents for teaching me the valu e and importance of educatio n. I woul d like to thank my f ather, Robert Freudeman for never giving up on me. I would like to thank my husband, Jason Thomas Fleming for his unwavering support. I could not have done this without him. I would like to thank my son Benjamin for teachin g me about my prio rities. I miss him and will always carry him in my heart Last, but certainly not least, I would like to thank my son, Siler for always giving me a reason to smile. He is the light of my life.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 7 LIST OF FIGURES ................................ ................................ ................................ ......................... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 10 INTRODUCTION ................................ ................................ ................................ ......................... 12 Possible Selves ................................ ................................ ................................ ........................ 12 Self Perce ived Competence ................................ ................................ ................................ .... 14 Psychological Well B eing ................................ ................................ ................................ ...... 15 Possible Selves, Self Perceived Competence, and Psychological Well Being ...................... 16 Hypotheses ................................ ................................ ................................ .............................. 17 METHODS ................................ ................................ ................................ ................................ .... 19 Participants ................................ ................................ ................................ ............................. 19 Procedure ................................ ................................ ................................ ................................ 19 Possible Selves and Goal O riented Activities ................................ ................................ 20 Self Perceived Cognitive Compet ence ................................ ................................ ............ 22 Self Perceived Health Competence ................................ ................................ ................. 23 Psychological Well B eing ................................ ................................ ............................... 24 Demographic and Functional Assessment ................................ ................................ ....... 27 Identification of Domain Specific Possible Selves ................................ ......................... 30 Analyses ................................ ................................ ................................ ................................ .. 30 Power ................................ ................................ ................................ ................................ ...... 32 PRELIMINARY ANALYSES ................................ ................................ ................................ ...... 36 RESULTS: COGNITIVE ................................ ................................ ................................ .............. 42 Cognitive Selves ................................ ................................ ................................ ..................... 42 Identifying Individuals for Whom Cognition is Important ................................ ..................... 44 Measurem ent Model: Cognitive Competence ................................ ................................ ........ 45 Measurement Model: Psychological Well Being ................................ ................................ ... 45 Latent Variable Path Model ................................ ................................ ................................ .... 47 Additional Analyses ................................ ................................ ................................ ................ 49 RESULTS: HEALTH ................................ ................................ ................................ .................... 65 Health Selves ................................ ................................ ................................ .......................... 65 Identifying Individuals for Whom Health is Important ................................ .......................... 67

PAGE 6

6 Measurement Model: Health Competence ................................ ................................ ............. 68 Measurement Model: Psychological Well Being ................................ ................................ ... 68 Latent Variable Path Model ................................ ................................ ................................ .... 70 Additional Analyses ................................ ................................ ................................ ................ 71 DISCUSSION ................................ ................................ ................................ ................................ 89 Possible Selves ................................ ................................ ................................ ........................ 89 Cognition ................................ ................................ ................................ ......................... 90 Health ................................ ................................ ................................ .............................. 92 Psychological Well Being ................................ ................................ ................................ ...... 94 Hopes and Fears ................................ ................................ ................................ ............ 100 Limitations ................................ ................................ ................................ ............................ 105 Conclusion ................................ ................................ ................................ ............................ 106 LIST OF REFERENCES ................................ ................................ ................................ ............. 108 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 112

PAGE 7

7 LIST OF TABLES Table page 2 1 Mean education and health ratings ................................ ................................ .................... 33 2 2 Percentage of agreement between independent raters ................................ ....................... 33 3 1 Number of individuals reporting types of selves by age group: Hopes ............................. 38 3 2 Number of individuals reporting types of selves by age group: Fears .............................. 39 3 3 Number of individuals reporting types of most important hopes by age group ................ 4 0 3 4 Number of individuals reporting types of most important fears by age group .................. 41 4 1 Mean number of cognitive selves reported by age group : Total ................................ ....... 51 4 2 Percentage of individuals reporting cognitive selves by age group ................................ ... 51 4 3 Mean number of cognitive selves by age g roup: Hopes and fears ................................ .... 51 4 4 Type of cognitive selves reported by age group ................................ ................................ 51 4 5 Factor loadings for self perceived cognitive competence (SPCC) ................................ .... 52 4 6 Factor loadings for negative psychological well being ................................ ..................... 52 4 7 Factor loadings for positive psychological well being ................................ ...................... 52 4 8 Mean scores on psychological well being composites by age group ................................ 52 4 9 Mean cognitive efficacy scores by age group ................................ ................................ .... 52 4 10 Intercorrelations between efficacy and psychological well being: Younger adults .......... 53 4 11 Intercorrelations between efficacy and psychological well being: Middle aged adults .... 54 4 12 Intercorrelations between efficacy and psychological well being: Older adults ............... 55 4 13 Latent path model: Cognition ................................ ................................ ............................ 56 4 14 Latent path model: Cognition plus objective cognitive performance ................................ 57 4 15 Latent path model: Cognitive hopes excluding negative psychological well being .......... 57 4 16 Latent path model: Cognitive fears excluding positive psychological well being ............ 58 4 17 Mean word list recall by age group ................................ ................................ .................... 58

PAGE 8

8 4 18 Mean memory efficacy and Metamemory in Adulthood Scale ratings by age group ....... 58 4 19 Mean psychological well being ratings by age group ................................ ....................... 59 5 1 Mean number of health selves reported by age group: Total ................................ ............. 75 5 2 Percentage of individuals reporting health selves by age group ................................ ........ 75 5 3 Mean number of health selves by age group: Hopes and fears ................................ .......... 75 5 4 Type of health selves reported by age group ................................ ................................ ..... 75 5 5 Factor loadings for self perceived health competence (SPHC) ................................ ......... 76 5 6 Factor loadings for negative psychological well being ................................ ..................... 76 5 7 Factor loadings for positive psychological well being ................................ ...................... 76 5 8 Mean scores on psychological well being composites by age group ................................ 76 5 9 Mean health efficacy scores by age group ................................ ................................ ......... 76 5 10 Intercorrelations between efficacy and psychological well being: Younger adults .......... 77 5 11 Intercorrelations between efficacy and psychological well being: Middle a ged adults .... 78 5 12 Intercorrelations between efficacy and psychological well being: Older adults ............... 79 5 13 Latent path model: Heal th ................................ ................................ ................................ .. 80 5 14 Latent path model: Health hopes excluding negative psychological well being ............... 81 5 15 Latent path model: Health fears e xcluding positive psychological well being ................. 81 5 16 Mean SF 36 ratings by age group ................................ ................................ ...................... 82 5 17 Mean health efficacy and locus of con trol ratings by age group ................................ ....... 83 5 18 Mean psychological well being ratings by age group ................................ ....................... 84

PAGE 9

9 LIST OF FIGURES Figure page 2 1 Proposed model for cognition ................................ ................................ ............................ 34 2 2 Proposed model for health ................................ ................................ ................................ 35 4 1 Latent path model for cognition ................................ ................................ ......................... 60 4 2 Modified latent path model for cognition ................................ ................................ .......... 61 4 3 Modified latent path model for cognition with objective perform ance ............................. 62 4 4 Latent path model for cognitive hopes and positive psychological well being ................. 63 4 5 Latent path model for cognitive f ears and negative psychological well being ................. 64 5 1 Latent path model for health ................................ ................................ .............................. 85 5 2 Modified latent path model for health ................................ ................................ ............... 86 5 3 Latent path model for health hopes and positive psychological well being ...................... 87 5 4 Latent path model for health fears and negati ve psychological well being ....................... 88

PAGE 10

10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy POSS IBLE SELVES AND PERCEIVED COMPETENCE IN RELATION TO COGNITION AND HEALTH By Alissa Dark Freudeman August 2009 Chair: Robin Lea West Major: Psychology Markus and Nurius defined (1986) possible selves as elements of the self concept that represent what an individual could become, would like to become, or is afraid of becoming. These personally meaningful hopes and fears are of central importance in the se lf concept. Believing that we have the ability to attain our most important hopes and avoid our most dreaded fears may be central to maintaining positive psychological functioning or well being. The central focus of this research was to examine age differe nces in possible selves related to cognition and health and to examine how beliefs about the competent self influence psychological well being. We found that younger adults reported a greater number of cognitive selves than both middle aged and older adult s. The cognitive selves younger adults reported tended to be hopes related to education; the cognitive selves middle aged and older adults reported tended to be hopes related to learning new skills and continuing to engage in cognitively stimulating activi ties, as well as fears related to cognitive decline. In contrast, middle aged adults reported a greater number of health fears than both older and younger adults. These fears included statements related to specific diseases such as having a heart attack or stroke, or general statements related t o maintaining or losing health. To examine the relationship between age, competence and psychological well being several structural equation models were tested. The paths between competence and psychological well bei ng

PAGE 11

11 were significant for health but not for cognition. We belief this was due to the fact that our younger, middle aged, and older adults all reported high levels of competence related to cognition. Our sample was high functioning and reported relatively hi gh levels of domain specific competence and as well as high levels of positive psychological well being. Unfortunately we were not able to test the relationship between low levels of competence and psychological well being. Our results still support that i dea that believing we have the ability to attain our most important hopes and avoid our most dreaded fears, especially when faced with declines in health or cognition that may threaten our ability to do so, may be central to maintaining positive psychologi cal well being and skilled performance as we age

PAGE 12

12 CHAPTER 1 INTRODUCTION As we move through life, declines in physical and mental abili ties will occur across multiple domains. Our beliefs and expectations about the aging process may impact how we perceive, respond to and manage experienced decline. Of course some domains are more important than others. When a given domain like health o r cognition is personally meaningful to us, the value we place on that domain is reflected within our future hopes and fears. These personally meaningful hopes and fears are called possible selves (Markus & Nurius, 1986). These selves are of central impo rtance to who we are and who we hope to become. Believing that we have the ability to attain our most important hopes and avoid our most dreaded fears, especially when faced with declines that may threaten our ability to do so, may be central to maintaini ng positive psychological functioning or well being. Possible Selves Markus and Nurius defined (1986) possible selves as elements of the self concept that represent what an individual could become, would like to become, or is afraid of becoming. These per sonally meaningful hopes and fears are of central importance in the self concept. Although each individual has certain selves that are central to his or her identity, the meanings of these selves may change over time (Cross & Markus, 1991). Possible selv es develop with us as we move through life. Which selves will be dominant depends on the individual and the context he or she experiences at any given point in time (Markus & Wurf, 1987). Recent studies have examined age related differences and changes in possible selves. Differences exist not only in the types of selves reported by age, but in the number of selves reported. Older adults consistently report fewer possible selves than younger adults (Cross & Markus, 1991; Hooker, 1992). Older adults also report more specific, more realistic possible

PAGE 13

13 selves than younger adults (Cross & Markus, 1991; Hooker, 1992). In general, the possible selves older adults report are more in line with their current identity; they typically involve the continuation or ma intenance of current activities and relationships, but may also include continued growth (Cross & Markus, 1991; Frazier, Hooker, Johnson, & Kaus, 2000; Hooker, 1992; Smith & Freund, 2002). Age differences in specific domains have also been found. Hooker and Kaus (1994) found that health related selves are common in later life and begin to increase in frequency in middle age (Hooker, 1992; Hooker & Kaus, 1992; Hooker & Kaus, 1994). Research has also shown that older adults report a greater number of fears related to memory and cognition than younger adults (Dark Freudeman, West, & Viverito, 2006). Thus, the frequency of health and cognitive selves increases with age as health and cognition become more salient during later life when many individuals are co nfronted with personal declines in physical and mental abilities. Changes in possible selves result in changes in levels of motivation and feelings of self efficacy (Cameron, 1999; Cross & Markus, 1991; Hooker, 1992). Possible selves have also been relate d to psychological well being and life satisfaction (Cross & Markus, 1991; Smith & Freund, 2002). Cross and Markus (1991) found that individuals lower in life satisfaction were more likely to report hoped for selves related to personal characteristics and feared selves related to being alone. Individuals scoring lower in life satisfaction also tended to report more extreme hopes and fears (Cross & Markus, 1991). Further, feared selves in the health domain have been associated with declines in life satisf action, satisfaction with aging, and subjective health (Smith & Freund, 2002). The presence of feared health selves have also been associated with greater reports of engaging in health behaviors, suggesting that fears can actually have a positive motivati onal effect (Hooker & Kaus, 1994). The proposed study will help clarify the

PAGE 14

14 relationship between possible selves and well being by focusing on two specific domains that are likely to be highly valued (cognition and health) and by including multiple indica tors of competence, and both positive and negative well being. Self Perceived Competence The need for competence is a universal need that exists across different cultures, individuals, and domains (Skinner, 1995). Although what it means to be competent at a given activity may change over time, perceiving oneself as competent is important throughout the life span (Brandtstadter, 1989; Brandtstadter, Rothermund, & Schmitz, 1998). Self perceived competence includes beliefs about self efficacy and personal co ntrol, and expectations regarding the possible outcomes surrounding a given domain or situation (Skinner, 1995). These beliefs and expectations work together to motivate action and interpret outcomes (Bandura, 1997). There is considerable evidence for ag e declines in feelings of competence, particularly in the memory domain. Given a well developed literature on memory self evaluation and relatively little on self perceptions regarding general cognition (e.g., speed of processing, attention, reasoning), t he cognitive focus of this research will be on memory self perceptions. Declines in cognitive abilities are seen as an inevitable and normative developmental process by adults of all ages (Heckhausen, Dixon, and Baltes, 1989; Lineweaver & Hertzog, 1998; Ry an, 1992). Further, older adults report lower levels of memory self efficacy than younger adults (Berry, West, & Denneh y, 1989; Hultsch, Hertzog, & Dixon, 1987), and less control over memory than younger adults (Hultsch et al., 1987; Lachman, Bandura, Wea ver, & Elloitt, 1995). These average declines in perceived memory are well documented, but the impact these beliefs have on well being has not been investigated. That is a key goal of this research.

PAGE 15

15 Psychological Well B eing Well being is multidimensional (Ryff, 1989) and involves both positive and negative emotions and evaluations (Diehl, Hastings, & Stanton, 2001; Diener, Emmons, Larsen, Griffin, 1985; Updegraff, Gable, & Taylor, 2004). The current proposal defines well being as positive psychological f unctioning. Positive psychological functioning refers to a general satisfaction negative well being may be related to biological influences (genetic or neurolo gical factors), environmental influences (negative life events or stress), or beliefs about the self such as perceived discrepancies between who one is and who one wishes to be (Berenbaum, Raghavan, Le, Vernon, & Gomez, 1999). In general, older and middle aged adults consistently report equivalent or greater levels of well being compared to younger adults (Carstensen, Pasupathi, Mayr, & Nesselroade, 2000) despite age related declines in multiple abilities. Ryff (1989) found that although older and younger adults reported similar levels of happiness and self esteem, older adults also reported higher levels of depression, and lower levels of purpose in life and personal growth than younger individuals. This suggests that measures of general life satisfaction may be missing part of the picture. For this reason, the current proposal will use multiple measures of positive and negative psychological well being. When an individual perceives a large gap between their current self and their most meaningful hoped f or possible self, or when the gap between the current self and a feared possible self narrows, well being may be compromised. If the individual no longer feels capable of attaining the hope or preventing the fear, depression or anxiety may result. Percep tions of competence in a valued domain could control variations in positive and negative aspects of well

PAGE 16

16 being. We predict that positive psychological functioning (higher levels of positive well being and lower levels of negative well being) will be great er for individuals with higher levels of perceived competence within their highly valued, self defining domains. Possible Selves, Self Perceived Competence, and Psychological Well Being The central focus of this research is examining how beliefs about the competent self influence psychological well being. This study will allow us to address this key question: what are the consequences of perceiving that you have low capability in a domain of function that is central to your self concept? This research emph asizes the importance of domain specific beliefs about the self. In particular, we believe that positive well being depends on feelings of success in domains that are concept, and that a negative sense of well being may be most com mon in individuals who are sensing declining capabilities in domains of personal importance to them. As part of the self concept, possible selves motivate and guide actions and emotions in personally meaningful ways (Markus & Nurius, 1986). For instance, we know that middle aged and older individuals who reported a possible self in the health domain also reported engaging in a greater number of health behaviors like exercising, dieting, and seeking regular medical care. In fact, reporting a health self w as more strongly related to health behaviors than was placing a high value on health alone (Hooker & Kaus, 1992; Hooker & Kaus, 1994). Like health ( Hooker, 1992; Hooker & Kaus, 1992; Hooker & Kaus, 1994 ), cognition is believed to become more salient with age (Berry & West, 1993; Cavanaugh, Feldman, & Hertzog 1998; Dark Freudeman, West, Viverito, 2006; Hultsch, et al., 1987). Our previous research has shown that cognitive ssible selves, as older adults anticipate and experience declines. However, the practical importance of having a

PAGE 17

17 possible self within the cognitive domain has yet to be fully understood. For aging individuals reporting a cognitive or health self, higher levels of SPC in that domain may be essential for maintaining higher levels of PWB. Hypotheses This project examines how hopes and fears related to health and cognition, self perceived competence (SPC) in the domains of health and cognition, and psycholog ical well being (PWB) are related. The specific aims of the current study are as follows: 1. Identify age differences in possible selves related to cognition and health. Earlier research on health possible selves has established age differences in health sel ves, but generally has not separated health selves (e.g., fears about losing physical strength) from cognitive selves (e.g., fears about losing memory). This research will consider both domains separately. Earlier work on health selves showed that older a dults report a greater number of selves related to health than middle aged and younger adults (Hooker, 1992; Hooker & Kaus, 1992; Hooker and Kaus, 1994). We expect to replicate these results. Our pilot work on cognitive possible selves has shown that cog nition is important to both older and younger adults. Older adults are more concerned with maintaining or losing cognitive functions, whereas younger adults are more concerned with cognitive activities like learning and performing well in educational and career settings. This investigation will replicate those past findings, distinguish between health selves and cognitive selves, and extend the cognitive data to middle aged adults. It is critical to include middle aged adults, to understand when feared s elves begin to emerge, and to understand whether cognitive selves and health selves emerge at different times. 2. Examine the interrelationship between age, self perceived competence (SPC), and psychological well being (PWB). PWB will be evaluated in relatio n to age differences in SPC for the health domain and for the cognitive domain. Consistent with past research, age differences are expected in SPC in both domains, whereas age differences in overall PWB are not expected. For older adults in general, anti cipating or experiencing declines in cognition and health is more salient than it is for younger adults, and those older adults with particular concerns about their declining competence should have lower PWB. Therefore, we expect the relationship between SPC and PWB to interact with age: to be stronger for older adults than for younger adults, with middle aged adults falling in between. 3. Predict psychological well being for individuals for whom cognition or health is of central importance. Those individua ls who spontaneously report cognitive or health selves see that particular domain as central to who they are. Therefore, their PWB should be dependent on their SPC in that domain. When cognition or health is not a central

PAGE 18

18 feature of the self, SPC should not be highly related to PWB. Thus, for individuals with health selves, SPC for health should predict PWB; for individuals with a cognitive self, SPC for cognition should predict PWB. For the domain of cognition alone, we will also evaluate the potential moderating effects of objective performance on this relationship

PAGE 19

19 CHAPTER 2 METHODS Participants The sample included 347 participants. Younger adults, ages 18 to 39 (M = 21.16, SD = ticipant pool. Middle aged adults, ages 40 to 64 (M = 54.84, SD = 7.24) and older adults, ages 65 to 90 (M = 73.47, SD = 5.66) were the parents and grandparents of younger participants. Six participants were excluded from the study. Four participants we re excluded for scoring below a 30 on the Telephone Interview of Cognitive Status. Two additional participants were excluded for failing to answer the questionnaire in a consistent manner. Health and education were examined as a function of age. Signifi cant differences were found for health, F (2, 338) = 8.30, p < .001. Younger adults reported significantly better health than both middle aged and older adults. Significant differences were also found for education, F (2, 338) = 11.48, p < .001. Middle aged adults reported significantly more years of education than both younger and older adults (see Table 2.1). Procedure Participants completed a take home survey on possible selves, self perceived competence, and well being, which took approximately 2 hou rs. Each packet began with the open ended portion of the possible selves questionnaire. Participants then completed the remaining items concerning memory and health beliefs. Last, participants completed the SF 36 health survey and a Participant Informat ion form, to provide basic demographic information. Younger adults were asked to return their surveys to the psychology building. Middle aged and older adults returned their surveys in prepaid envelopes. After the surveys were returned, participants wer e contacted

PAGE 20

20 by phone to clarify any unclear responses (such as illegible or blank items) and to complete a cognitive test battery by telephone. Pos sible Selves and Goal O riented A ctivities Participants completed the open ended questionnaire developed by Cr oss and Markus (1991), and used extensively in aging research (Hooker, 1999). This questionnaire was designed to spontaneously elicit both hoped for and feared possible selves. These spontaneously generated selves are believed to be central to the indivi asked to list all of their hoped for selves, then to select their three most important, hoped for selves. Next, participants listed the goal oriented activities they have recently undertaken to accomplish each of their three most important hoped for selves, and answered 6 questions to assess self efficacy, and perceived control over attaining each of these three most important selves. Participants were asked to rate the following statements on a seven point Likert scale: efficacy and outcome expectancy respectively. Next, participants listed all of their feared selves, a nd selected their three most dreaded, feared selves. For each most dreaded self, a matched series of questions about recent goal oriented activities, self efficacy and perceived control was answered. Participants also rated personal importance for all sp ontaneously mentioned most important and most dreaded selves. To examine the content of the spontaneously generated possible selves, thirteen categories of selves were created: personal characteristics, health, social, career, financial, life events, cogn itive statements, cognitive activities, cognitive educational, cognitive miscellaneous, memory, dependency, and other. Personal characteristics included statements about personal Health related

PAGE 21

21 related selves also included more specific statements related selves included statements related to working or changing related included statements related to getting married, having children, and losing a spouse. Cognitive Cognitive activities included statements abo ut engaging in cognitively engaging activities such as reading and chess. Cognitive education included statements about learning new skills and doing well in school. Cognitive miscellaneous category included statements related to volunteering, traveling, and hobbies such as woodworking or knitting. Memory related selves and dance steps. Dependency other selves included statements related to hopes and fears about other people such as I hope that my categories. Two independent raters categorized the possib le selves according to these categories. To examine the agreement between these two independent raters, interrater reliability, also known as

PAGE 22

22 kappa, was calculated. Kappas can range from 1.0 to 1.0, with larger numbers indicating better reliability. Ka ppas were .68 for hopes and .75 for fears. For most important hopes, kappas were between .76 and .80. For most important fears, kappas were between .84 and .93. After assessing reliability of the coding system with kappas, d isagreements were resolved by discussion to select on e category for each participant response See Table 2.2 for the percentage of agreement between raters for health and cognitive selves. Self Perceived Cognitive Competence For many older adults, the key aspect of cognition that tri ggers functional concerns about decline is memory, therefore, additional assessments of self perceived cognitive competence (SPCC), which are not derived from the possible selves questionnaire, focus on memory. Measures of SPCC included the possible selve s items on efficacy and perceived control related to achieving a most important hoped for or avoiding a most dreaded feared cognitive self as well as the measures described below. We created a single index from these possible selves, called PSQ. To create the PSQ measure several steps were required. First we calculated the average of the outcome expectancy responses for both cognitive hopes and fears. Next we calculated the average of the self efficacy responses for cognitive hopes and fears. These two averages were then summed to create one measure of PSQ ranging from 2 to 14. The Metamemory in Adulthood scale (MIA) developed by Dixon, Hultsch, and Hertzog (1988) is a standard measure of memory beliefs that has shown consistent validity and reliability The MIA subscales known to assess SPCC are capacity, change, and locus (Hultsch, Hertzog, item will be rated on a 5 point Likert scale, (1 = agree strongly, 5 = disagree strongly). The

PAGE 23

23 subscales had high internal consistency with alpha coefficients between .63 and .91. To create each scal e, items were summed. Sco res ranged from 1 to 5 on each scale. General Memory Self efficacy (GME) represents a brief 3 item measure of SPCC with strong internal consistency; scores on the GME are significantly correlated with recall scores and longer scales of memory self efficacy (West & Yassuda, 2004). Participants are first primed to all, 7 = ve ry important). This is followed by the three critical scale items on a 7 point Likert e your with an alpha coefficient of .90. Responses across the thre e critical scale items were summed to get the final score. evaluation of their memory skill. Self Perceived Health Competence Just as with cognition, m easures of self perceived he alth competence (SPHC) include d the possible selves items on efficacy and perceived control related to achieving a most important hoped for or avoiding a most dreaded feared health self To create this PSQ scale for health, several steps were required. F irst we calculated the average of the outcome expectancy responses for both hopes and fears for health Next we calculated the average of the self efficacy responses for both hopes and fears related to health These two averages were then summed to creat e one measure of PSQ for health, ranging from 2 to 14.

PAGE 24

24 The Multidimensional Health Locus of Control Form B (MHLC) is an 18 item questionnaire that assesses the degree to which an individual feels that maintaining his or her general health is primarily due to internal control, external control, or chance (Wallston, Wallston, & DeVellis, 1978). Participants responded using a 6 point Likert scale (1 = Strongly make m consistency with alpha coefficients between .60 and .75. Scale s cores were calculated by summing the items for each subscale and ranged from 6 to 36. The Self Rated Abilities for Health Practices scale (SRAH P) is a 28 item scale that measures health related self efficacy on four subscales: exercise, stress management, nutrition, and health practices (Becker, Stuifbergen, Oh, & Hall, 1993). Participants responded on a 4 consistency both overall (alpha = .94) and for the individual exercise (alpha = .92), nutrition (alpha = .81), well being (alpha = .90), and responsible health practices subscales (alpha = .86). Items were summed to create each subscale and ranged from 0 to 28 for these individual subscales. The individual subscales were also summed to create an overall measure of Health S elf E fficacy (HSE) and ranged from 0 to 112. Psychological Well B eing Positive psychological well being was assessed using multiple measures including positive affect, self esteem, and a multidimensional measure including six different dimensions of posit ive psychological functioning (Ryff, 1989). These three measures were combined to create

PAGE 25

25 a composite measure of positive psychological well being ( PWB ) Negative psychological well being was assessed using a measure of negative affect, anxiety, and depres sion. These three measures were combined to create a composite measure of negative PWB. Scales of Psychological Well Being (SPWB) w ere used to assess six dimensions: self acceptance, positive relations with others, autonomy, environmental mastery, purpose in life, and personal growth (Ryff, 1989). Each subscale included 9 items on a 6 point scale (1 = decisions are not usually influenced by what everyone else when trying to plan person over ups and dow autonomy (alpha = .83), environmental mastery (alpha = .86), personal growth (alpha = .85 ), positive relations (alpha = .88), purpose in life (alpha = .88), and self acceptance (alpha = .91). Items were summed to create each subscale. Each subscale consisted of 7 items, with the

PAGE 26

26 exception of Purpose in Life, which included 8 items. Scores r anged from 7 to 42 for all subscales, except Purpose in Life which ranged from 8 to 48. The Positive and Negative Affect Scale (PANAS) assesses two primary dimensions of mood. Positive affect refers to the extent to which an individual feels excited, stro ng, and alert. Negative affect refers to the extent to which an individual feels irritable, jittery, and ashamed (Watson, Clark, & Tellegen, 1988). The scale included a total of 20 descriptors rated on a 5 point Likert scale (1 = Very slightly or not at all, 5 = Extremely). Half of these items were negative and half positive. Internal consistency was high for positive affect (alpha between .86 and .90) and for negative affect (alpha between .84 and .87) Items were summed to create a positive and negat ive scale of mood each ranging from 10 to 50. The Self Esteem Scale (SES) assesses the degree to which a person feels satisfied with him or herself (Rosenberg, 1965). The scale included 10 items rated on a 4 point scale (1 = Strongly agree, 4 = Strongly d alpha coefficients between .85 and .88. Items were summed to create the SES scale with sco res ranging from 10 to 40. The Center for Epidemiological Studies Depression Scale (CES D) assesses the degree to which an individual has experienced depressive symptoms. Respondents rated the frequency with which they have experienced particular depressi ve symptoms during the past week such as: Possible responses can range from 0 (less than 1 day) to 3 (5 7 days). The 20 item CES D scale is designed to measure depression in the general population (Radloff, 1977) and has also been shown to be appropriate for use in older adult populations (Hertzog, Van Alstine, Usala, Hultsch,

PAGE 27

27 et al., 1990). Internal consistency is usually high (alpha = .85) in community sample s. Items were summed and scores ranged from 0 to 60. The State Anxiety scale of the State Trait Anxiety Inventory (STAI; Spielberger, 1983) point scale (1 = Not scale of the STAI has well established psychometric validity and reliability as a measure of current individual arousal levels (Spielberger, 1983). The STAI has high interna l consistency with alpha coefficients between .89 and .92 (Spielberger & Sydeman, 1994). Responses to the STAI are summed, with a total score varying from 20 to 80. Demographic and Functional Assessment The Brandeis Test of Adult Cognition by Telephone ( B TACT) was used to assess memory and cognition: verbal memory (immediate and delayed), working memory span, verbal fluency, reasoning, and speed of processing. This test battery is a valid assessment tool, strongly related to traditional laboratory measure s conducted in person (Tun & Lachman, 2004). For the domain of cognition alone, administration of this test battery by telephone will make it possible to test the relationship between cognitive ability, SPCC, and PWB. The BTACT was administered during th e follow up phone call. The verbal memory test was used as an objective indicator of cognitive performance; the other BTACT activities that were completed during this phone call are not relevant to the purpose of the current study. For the verbal memory test, participants were read a list of 15 words, with a one second interval between each word. Participants were Participants were given 90 seconds to recall. Correc t words, intrusions and repetitions were recorded

PAGE 28

28 The SF 36 short form health survey developed by Ware and Sherbourne (1992) assesses This survey was used because it provides information on eight comprehensive subscales in a relatively short and simple format. The SF 36 is composed of the following subscales: General Health, Physical Limitations, Physical Functioning, Bodily Pain, Vitality, Emotional Limit ations, Social Functioning, and Mental Health. General Health included four questions rated on a 5 (1 = definitely true, 5 = definitely false). Physical limitations were examined by four yes have you had any of the following problems with your work or other regular daily activities as a result of your p functioning was measured using ten questions rated on a 3 t, 3 = no, not limited at all). Bodily Pain was assessed by two questions rated on a 6 point Likert scale, severe). Vitality was measured by four questions rated on a 6 the past 4 weeks, have you had any of the following problems with your work or other regular daily activities as a result of any emotional problems (such as feeling depressed or anxious): Cut tioning was measured by two questions rated on a 5 to what extent has your physical health or emotional problems interfered with your normal social

PAGE 29

29 activities with family, friends, neighbors, or groups? mental health was assessed with five questions rated on a 6 5 = none of the time). T he SF 36 will be used to verify that the participants report comparable levels of physical and mental health across age groups. Internal consistency was high for the SF 36 subscales generally exceeding .80. The exception was the social function subscale (alpha = .76). The SF 36 manual was used to calculate scores for each subscale. Items for each subscale were summed. Next a standardized score was created for each subscale ranging from 0 to 100 (Ware, Snow, Kosinski, & Gandek, 1993). The Participant I nformation Form is a simple form that requested basic demographic information from participants such as gender, race, education, marital status current medications, conditions for which they are currently being treated, recent hospitalizations, and any fa mily history of cognitive impairment. The Telephone Interview of Cognitive Status (TICS) is an 11 item dementia screen that can be administered in person or by telephone and takes between 5 and 10 minutes to complete (Brandt, Spencer & Folstein, 1988). It ems include activities such as immediate recall of a 10 item word list, counting backwards from 20 to 1, and several orientation questions such as: maximum score of 41 points. The TICS was administered as needed during the follow up phone call that is, it was only administered as a check for dementia with participants who seemed to be confused during the call or who performed very poorly on the cognitive measures from the BTACT A cutoff score of 30 out of a total possible score of 41 was utilized for dementia

PAGE 30

30 exclusion. The TICS has exhibited high predictive value for the detection of impairment in previous research (Brandt, Spencer & Folstein, 1988) and has a test retest coefficient of .97. Identification of Domain Specific Possible Selves Possible selves were categorized as noted above, to identify those individuals who had a cognitive self and/or a health self as one of their top 3 feared or top 3 hoped for selve s. To verify the centrality of these selves among individuals spontaneously reporting a cognitive or health self, an additional indicator was utilized. Participants had to rate the personal importance of that self with a 4 or better, on a seven point Lik ert scale, at the time of the initial survey. Analyses Aim 1: Possible selves were coded for all individuals. The frequency and types of cognitive and health related selves were compared among the different age groups using ANOVA and Chi Square. We expec ted older adults to report a greater number of cognitive and health related fears than younger adults, younger adults to report more cognitive and health related hopes, and middle aged adults to fall between the younger and older group. Aim 2: Factor ana lyses were conducted to identify latent variables for self perceived competence for cognition and health (SPCC and SPHC). The latent variable for SPCC consisted of the MIA, GME, and PSQ measures (See Figure 2.1). In Figure 2.1, MIA refers to the capacit y, change and locus of control subscales of the MIA, GME refers to memory self efficacy measured by the GME, and PSQ refers to the efficacy and outcome expectancy items on the Possible Selves Questionnaire. The latent variable for SPHC included the MHLC, HSE, and PSQ (See Figure 2.2). In Figure 2.2, HLC refers to health locus of control measured by the MHLC, HSE refers to health self efficacy measured by the SRAHP, and PSQ refers to the efficacy and outcome expectancy items on the Possible Selves Question naire. Confirmatory factor analyses based on previous work by Diehl, Hastings, and Stanton (2001) were conducted

PAGE 31

31 to identify positive and negative dimensions of PWB from the multiple indicators of each construct described above (See Figure 2.1 and Figure 2.2). In addition to factor analyses, ANOVAs were conducted to identify age differences in composite measures of SPC for health and cognition. Lower levels of SPC were expected among older adults. Based on age increases in the salience of cognition and health, we also expected SPC to be correlated more strongly with PWB (negative and positive) for the middle aged and older adults than for the younger adults. Aim 3: Structural equation modeling was used to test the proposed model in which SPC interacts with the presence of a domain specific self to predict PWB. Positive PWB included the SES, + PANAS, and SPWB measures. SES refers to the self esteem scale, + PANAS refers to the positive affect questions of the PANAS scale, and the SPWB refers to the Sc ales of Psychological Well Being (Ryff, 1989). Negative PWB includes the CES D, STAI, and PANAS. The CES D refers to the depression scale, the STAI is the state anxiety measure, and the PANAS refers to the negative affect questions of the PANAS scale Tentatively, we expected the relationship between a cognitive self, SPCC, and PWB to be stronger for older adults than for younger adults, because older adults generally report fewer possible selves and because the selves they do report are more often g rounded within their current identities. We also examined the possibility that individuals with a feared cognitive self would show a relationship between SPCC and the negatively valenced items for PWB, whereas individuals with a hoped for cognitive self w ould show a relationship between SPCC and the positively valenced items for PWB (See Figure 2.1). For the domain of cognition, we also tested the possible moderating effects of objective performance on this relationship, to determine whether actual cognit ive performance contributed to the prediction of psychological well being independently, or in interaction with SPCC. For the domain of health, the relationship between

PAGE 32

32 self perceived health competence SPHC and PWB was also tested for individuals reportin g a health related self. See Figure 2.2. Power Estimations of sample size and statistical power were considered for analyses of variance (ANOVAs) and structural equation analyses. Based on previous research, expected effect sizes are estimated to fall ne ar .20 for self perceived competence and .25 for cognitive selves. According to Cohen (1988), adequate power at the .05 significance level to detect a medium effect size of .25 requires a total of 195 participants, with approximately 65 participants per a ge group. For structural equation analyses, Bentler (1985) suggested that the ratio between sample size and the number of estimated parameters should range from 5 to 10. The total sample size of the current study (n = 339) was more than adequate. Howeve r several analyses involved a subset of participants: analyses examining individuals identifying cognition as a most important hope or fear ( n = 92) and analyses examining individuals identifying health as a most important hope or fear (n =213). Although the number of participants included in the health analyses was adequate, the number of participants included in the cognitive analyses was low.

PAGE 33

33 Table 2 1. Mean education and h ealth r atings Standard Standard Age group Mean education deviation Mea n health deviation Young 13.87 2.30 2.62 1.48 Middle 15.53 2.91 3.31 2.19 Old 14.13 2.86 3.72 2.09 N = 339 Table 2 2. Percentage of agreement between independent r aters Type of selves Percentage agreement Hoped for Healt h 98% Cognitive 99% Feared Health 94% Cognitive 95% Note: Percentage agreement based on comparison of a randomly selected subset of possible selves data.

PAGE 34

34 Figure 2 1. Proposed model for c ognition SPCC PWB +PWB PANAS CES D STAI +PANAS SES Aut onomy Cognitive Self PSQ GME MIA AGE Environ Mastery Purpose in Life Personal Growth Self Acceptance Positive Relations

PAGE 35

35 Figure 2 2. Proposed m odel for h ealth SPHC PWB +PWB PANAS CES D STAI +PANAS SES Autonomy Health Self PSQ HSE HLC CCC AGE Environ Mastery Purpose in Life Personal Growth Self Acceptance Positive Relations

PAGE 36

36 CHAPTER 3 PRELIMINARY ANALYSES As mentioned above, our younger participants were recruited through the University of family members to participa te in our study. This method of recruitment led to groups of related individuals who had participated in our study, for example a child, parent, and grandparent. A family group code was created so that each family group had a distinct code which could be used A univariate analysis of variance was then conducted to compare family groups across two basic demographic variables: years of education and general health, with famil y group entered as a random factor with and without age as a covariate. Family groups did not significantly differ on general health ratings, F (1, 292) = 1.21, p < .05 ; however, a significant difference was found for years of education with F (1, 292) = 1.34, p < 05 or without F (1, 291) = 1.64 p < 05 age in the analyses. Further examination of these differences revealed that some grandparents did not have a high school diploma; however many did. In addition to these differences, many parents had a Next, to ensure that younger adults who participated in our study and recruited family members were not significantly different from younger adults who participated and did n ot recruit family members, we also compared these two groups of students on years of education and general health. No significant differences were found for education F (1, 68) = 1.94, p > .05 or for health, F (1, 68) = .007, p > .05 Analyses of varianc e were conducted to examine the number of spontaneous possible selves (hopes and fears) reported as a function of age group. Overall a significant difference was found for the total number of hopes, F (2, 336) = 3.33 p < .05, eta 2 = .020, reported by eac h age

PAGE 37

37 group. Younger adults reported a significantly greater number of hopes than older adults. There was no significant difference between the number of hopes reported by middle aged and older adults. A significant difference was also found for the tot al number of fears, F (2, 336) = 21.51 p < .001, eta 2 = .114, reported by each age group. Younger adults reported a significantly greater number of fears than both middle aged and older adults. Again, there was no significant difference between the numb er of fears reported by middle aged and older adults. W e also examined the possible selves (hopes and fears) most frequently reported by each age group (Table 3.1 and Table 3.2). We also examined the types of most important hopes and most dreaded fears reported by each age group (see Table 3.3 and Table 3.4). For younger adults, the most frequently nominated hopes were in the following domains: social, career, and life events; similarly, the most frequently nominated fears were social, other (e.g. hopes related to world peace), life events and finance. For middle aged adults, the most frequently nominated hopes were in the following domains: health, social, other (e.g. hopes related to success of their children), and finance. The most frequently nomina ted fears were health, social, and life events. For older adults, the most frequently nominated hopes were: health, other (e.g. hopes related to the success of their children or grandchildren), social and cognitive miscellaneous; the most frequently nomin ated fears: health, dependency, and social.

PAGE 38

38 Table 3 1. Number of i ndividuals reporting types of selves by age g roup: Hopes Self Younger Middle aged Older category adults adults a dults Personal t raits 24 17 23 Health 34 60 80 Social 49 70 65 Career 77 54 33 Financial 37 62 43 Life ev ents 62 29 27 Cognitive s tatements 1 2 7 Cognitive a ctivities 5 13 22 Cognitive e ducation 39 28 24 Cognitive m iscellaneous 47 67 77 Memory 0 0 2 Dependency 3 7 18 Other 43 60 80 N = 339

PAGE 39

39 Table 3 2. Number of individuals reporting types of selves by age g roup: Fears Self Younger Middle aged Older category adults adults a dults Perso nal t raits 32 20 8 Health 42 78 91 Social 60 58 50 Career 40 20 1 Financial 42 52 46 Life e vents 41 44 36 Cognitive s tatements 1 3 10 Cognitive a ctivities 0 1 0 Cognitive e ducation 23 0 0 Cognitive m iscellaneous 3 1 3 Memory 0 13 12 Dependency 3 34 70 Other 56 38 48 N = 339

PAGE 40

40 Table 3 3. Number of i ndividuals reporting types of most important hopes by age g roup Self Younger Middle aged Older category adults a dults a dults Personal t raits 22 13 19 Health 17 49 78 Social 39 65 64 Career 60 37 20 Financial 15 41 32 Life e vents 51 13 15 Cognitive s tatements 1 0 2 Cognitiv e a ctivities 0 7 10 Cognitive e ducation 22 12 15 Cognitive m iscellaneous 12 39 58 Memory 0 0 1 Dependency 2 4 12 Other 29 43 68 N = 339

PAGE 41

41 Table 3 4. Number of individuals reporting types of most importan t f ears by age g roup Self Younger Middle aged Older category adults adults a dults Personal t raits 28 15 7 Health 25 84 95 Social 66 51 49 Career 18 11 1 Financial 29 37 37 Life e vents 29 3 8 31 Cognitive s tatements 0 3 7 Cognitive a ctivities 0 0 0 Cognitive e ducation 23 0 0 Cognitive m iscellaneous 0 1 2 Memory 0 11 8 Dependency 1 23 65 Other 40 27 47 N = 339

PAGE 42

42 CHAPTER 4 RESULTS: COGNITIVE Initially, a nalyses of variance were conducted to examine the number of spontaneous cognitive selves reported, as a function of age group. In addition to the analyses of variance, chi square tests were also conducted, to determine whet her the number of individuals listing cognitive selves differed significantly by age group. Significant group differences were Confirmatory factor analyses were conducted and compo site variables were created. Next, examination of the proposed structural equation models were carried out. Last, follow up analyses of variance were conducted to examine differences in cognitive performance, memory beliefs, and psychological well being by age group. Cognitive Selves An analysis of variance was conducted to examine whether the three age groups (young, middle, and old) differed by the total number of cognitive selves spontaneously listed. Cognitive significant difference in the total number of cognitive selves reported by age group was found, F (2, 336) = 4.10, p < .05, eta 2 = .024. Younger adults reported significantly more cognitive selves th an middle aged and older adults. There were no significant differences between middle aged and older adults (see Table 4.1). To further explore the spontaneous reports of cognitive selves, participants were divided into those who spontaneously mentioned a cognitive self and those who did not. Given that the total number of cognitive selves reported by each individual was typically small (between 0 and 2), we felt it was prudent to confirm these age differences by examining the frequency with which individ uals reported a cognitive self as a function of age. To do this, a chi square analysis

PAGE 43

43 was conducted comparing those who reported no cognitive selves with those who reported at least one cognitive self, in relation to age group (young, middle, old). The results were significant, Pearson X 2 (N = 337, df = 2) = 9.21, p < .05. Follow up comparisons revealed a significant difference between the number of younger and middle aged adults reporting a cognitive self: Pearson X 2 (N = 200, df = 1) = 6.17, p < .05; as well as a significant difference between the number of younger and older adults reporting a cognitive self: Pearson X 2 (N = 226, df = 1) = 8.17, p < .005. In general a greater number of younger adults reported a cognitive self compared with middle aged and older adults. No significant difference was found between number of middle aged and older adults reporting a cognitive self: Pearson X 2 (N = 248, df = 1) = .08, p > .05 (see Table 4.2). Next, a closer look at the types of cognitive selves (hoped for or feared) reported by the three age groups was examined. An analysis of variance was conducted to examine whether the three age groups (young, middle, and old) differed by the total number of cognitive hopes spontaneously listed. No significant differen ces were found for cogniti ve hopes, F (2, 336) = 1.88, p > .05 An analysis of variance was also conducted to examine whether the three age groups (young, middle, and old) differed by the total number of cognitive fears spontaneously listed. Significant differences were found for cognitive fears, F (2, 336) = 3.95, p < .05, eta 2 = .023. Younger adults reported a significantly greater number of cognitive fears than middle aged and older adults. No differences were found between middle aged and older adul ts (see Table 4.3). To further examine the types of cognitive selves reported, participants were categorized into: those who reported cognitive hopes and those who did not, and those who reported cognitive fears and those who did not. A cross tabulation o f the three age groups (young,

PAGE 44

44 middle aged, old) by cognitive hope was significant: Pearson X 2 (N = 337, df = 2) = 7.51, p < .05. A cross tabulation of the three age groups (young, middle aged, old) by cognitive fear was also significant: Pearson X 2 (N = 337, df = 2) = 6.45, p < .05. Follow up comparisons revealed a significant difference between the number of younger and middle aged adults reporting a cognitive hope: Pearson X 2 (N = 200, df = 1) = 5.32, p < .05; and between the number of younger and midd le aged adults reporting a cognitive fear: Pearson X 2 (N = 200, df = 1) = 4.86, p < .05. A greater number of younger adults reported cognitive hopes and cognitive fears compared with middle aged adults. A significant difference was also observed between the number of younger and older adults reporting a cognitive hope: Pearson X 2 (N = 226, df = 1) = 6.22, p < .05, and between the number of younger and older adults reporting a cognitive fear: Pearson X 2 (N = 226, df = 1) = 4.58, p < .05. Again, a greater number of younger adults reported cognitive hopes and cognitive fears compared to older adults. No significant difference was found between the number of middle aged and older adults reporting a cognitive hope: Pearson X 2 (N = 248, df = 1) = .007, p > .05 or cognitive fear: Pearson X 2 (N = 248, df = 1) = .040, p > .05 (see Table 4.4). Identifying Individuals for Whom Cognition is Important Before continuing with the proposed analyses based on the selves that were spontaneously identified by our respondents it was necessary to identify individuals who not only reported a cognitive possible self, but who also rated that self as highly important. Two requirements had to be met for a participant to be included. First, individuals who listed a cognitive hope or cognitive fear as one of their three most important hopes or most important fears were identified. From our study sample, 97 participants identified a cognitive self as one of their three most important hopes or three most important fears. Second, we examined how important these participants rated these cognitive selves on the possible selves questionnaire. For cognitive

PAGE 45

45 hopes, participants were asked to rate the following statement on a 7 point Likert scale (1 = Not important at all, 7 = Very importa cognitive fears, participants were asked to rate the following statement on a 7 point Likert scale occur most important cognitive hope or fear lower than a 4, they were excluded from further analysis. These criteria resulted in the exclusion of 5 participants. Measurement Model: Co gnitive Competence A factor analysis was conducted to identify a latent self perceived cognitive competence factor proposed in Aim 2. General Memory Efficacy (GME), the capacity, change, and locus subscales of the Metamemory in Adulthood Scale, and PSQ (m easured by the efficacy and outcome expectancy items answered for most important cognitive hopes and fears on the Possible Selves Questionnaire) were all expected to load on the Self Perceived Cognitive Competence (SPCC) factor. These measures did not loa d on the hypothesized SPCC factor strongly (see Table 4.5 for factor loadings). Additional analyses were conducted to identify a SPCC factor using additional subscales of the Metamemory in Adulthood Scale (for example including anxiety and achievement sub scales) and combinations of PSQ (for example including efficacy items only or outcome expectancy items only) without success. Subsequent analyses were therefore carried out with individual variables (PSQ and GME) representing SPCC. Measurement Model: Psyc hological Well Being Factor analyses were also conducted to identify two latent psychological well being factors (positive and negative). Negative Affect, Depression, and Anxiety were expected to load on the negative psychological well being factor. The proposed negative psychological well being variables loaded strongly and significantly on their proposed factor. See Table 4.6 for

PAGE 46

46 standardized and unstandardized factor loadings. Positive Affect, Self Esteem, and the six Scales of Psychological Well Bei ng (positive relations, personal growth, purpose in life, self acceptance, environmental mastery, and autonomy) were expected to load on the positive psychological well being factor. The proposed positive psychological well being variables loaded strongly and significantly on their proposed factor. See Table 4.7 for standardized and unstandardized factor loadings. Next, composite measures were created for both positive and negative well being. Composite measures were created by first creating z scores fo r each individual psychological well being variable. Next these z scores were regression weighted by multiplying each z score by the corresponding factor loading. Last the weighted scores for the individual scales were summed to create a composite measur e. An analysis of variance was conducted to examine whether the three age groups (young, middle, and old) differed on the composite measures of positive and negative psychological well being. No significant differences were found between the three age gro ups for positive psychological well being, F (2, 90) = 1.75, p > .05 or for negative well being, F (2, 90) = 2.42, p > .05 (see Table 4.8) as expected. Next we examined age differences in self efficacy. An analysis of variance was conducted to examine wh ether the three age groups (young, middle, and old) differed on composite measures of General Memory Efficacy and PSQ. General Memory Efficacy included a sum across three memory efficacy items of the GME. The PSQ included the sum of the efficacy and outc Possible Selves Questionnaire. No significant differences were found between the three age groups for General Memory Efficacy, F (2, 90) = .95, p > .05 In contrast, a significant

PAGE 47

47 difference was found for PSQ, F (2, 90) = 5.17, p < .05, eta 2 = .105. As expected, y ounger adults reported significantly greater levels of efficacy and outcome expectancy for their most important cognitive selves than both middle aged and o lder adults (see Table 4.9). Last correlations were examined for General Memory Efficacy, PSQ, positive psychological well being and negative psychological well being for each age group. As expected, efficacy was more strongly correlated with psychologica l well being for the middle aged and older adults than for younger adults (see Tables 4.10 through 4.12). Latent Variable Path Model Next a latent variable path model was examined for individuals who identified a cognitive self as a most important hope or most dreaded fear. Based on the results of the factor analyses, the SPCC factor was removed from the original path model and instead General Memory Efficacy and PSQ were added as individual predictors (s ee Figure 4.1 ) Thus, the revised model proposed th at age directly impacted both General Memory Efficacy and PSQ and that both General Memory Efficacy and PSQ directly impacted negative and positive psychological well being. The fit of this original path model was poor. Modification indices suggested add ing paths between positive and negative well being and between GME and PSQ as well as allowing several errors to covary. See Figure 4.2 for the modified model which includes the new paths (covariances are not pictured in the model) The Chi square for t he modified model was not significant, X 2 (66) = 76.70, and goodness of fit indices indicated good fit: CFI = .981, RMSEA = .042. The effects of age on PSQ and the effects of General Memory Efficacy on PSQ were both significant. The effect of negative ps ychological well being on positive psychological well being was also significant. The remaining effects were not significant. Standardized and unstandardized effects are reported in Table 4.13.

PAGE 48

48 Next we tested the model with objective cognitive performanc e as a predictor in the model to examine whether actual cognitive performance contributes to the prediction of psychological well being ( See Figure 4.3 ) Paths were added from age to cognitive performance and from cognitive performance to GME, positive ps ychological well being and negative psychological well being. The path from age to GME was removed because it was not significant in the previous model. Objective cognitive performance was measured by the immediate word list recall activity from the BTAC T. The Chi square for this model was not significant, X 2 (7 7 ) = 8 2.27 and goodness of fit indices indicated good fit: CFI = .9 91 RMSEA = .0 27 Again, the effects of age on PSQ and the effects of General Memory Efficacy on PSQ were both significant. Th e effect of negative psychological well being on positive psychological well being was significant. The effects of cognitive performance on positive psychological well being were also significant. The remaining effects, including the effect of age on cog nitive performance and the effect of cognitive performance on General Memory Efficacy, were not significant. Standardized and unstandard i zed effects are reported in Table 4.14. Next we examined a latent path model for individuals reporting cognitive hopes (n = 56) with the negative well being factor excluded from the model (Figure 4.4). The Chi square was not significant, X 2 ( 40 ) = 51. 99 ; however, the goodness of fit statistics indicated only moderately good fit: CFI = .93 8 RMSEA = .07 4 The effects of age on PSQ and the effects of General Memory Efficacy on PSQ were both significant. Standardized and unstandardized effects are reported in Table 4.15. Last we examined a latent path model for individuals reporting cognitive fears (n=42) with the positive well being factor excluded from the model (Figure 4.5). The Chi Square was not significant, X 2 ( 8 ) = 6. 31 and goodness of fit indices indicated good fit: CFI = 1.00, RMSEA =

PAGE 49

49 .000. The effects of age on PSQ and the effects of General Memory Efficacy on PSQ were both significant. Standardized and unstandardized effects are reported in Table 4.16. Additional Analyses In addition to the proposed analyses, several follow up analyses were conducted. The purpose of these analyses was to examine age differenc es in cognitive ability, psychological well being, and cognitive competence among individuals reporting important cognitive selves. A multivariate analysis of variance was conducted to compare the three age groups (young, middle aged, old) across the fo llowing dependent variables from the BTACT: word list recall, word list intrusions, and word list repetitions. Age differences were not significant, F (6, 162) = .72, p > .05 (see Table 4.17). A multivariate analysis of variance was also conducted to comp are the three age groups (young, middle aged, old) across the following dependent variables: General Memory Efficacy and the Metamemory in Adulthood Scales (anxiety, achievement, capacity, change, and locus). Age differences were significant, F (12, 168) = 3.65, p < .001, eta 2 = .207, and this significant result was further explored in univariate tests. Significant age differences were found for capacity, F (2, 90) = 6.85, p < .005, eta 2 = .135, as younger adults reported higher levels of memory capacity than both middle aged and older adults. Significant age differences were also found for change, F (2, 90) = 14.12, p < .001, eta 2 = .243, as younger adults expected greater changes in their memory ability over time than both middle aged and older adults. Last, significant age differences were found for anxiety, F (2, 90) = 3.57, p < .05, eta 2 = .075; as older adults reported feeling greater anxiety related to memory tasks than both younger and middle aged adults. No significant age differences were found for locus, achievement, or General Memory Efficacy (see Table 4.18)

PAGE 50

50 Last, a multivariate analysis of variance was conducted to compare the three age groups (young, middle aged, old) across the following dependent variables: depression, anxiety, negative affect, positive affect, self esteem, autonomy, environmental mastery, purpose in life, personal growth, self acceptance, and positive relations. A s expected, a ge differences were not significant, F (22, 158) = 1.39, p =.126 (see Table 4.19)

PAGE 51

51 Table 4 1. Mean number of cognitive s elves reported by age g roup: Total Age g roup Mean number of cognitive selves Standard d eviation Young 1.12 1.35 Middle .68 1.01 Old .74 1.18 N = 339 Table 4 2. Percentage of indivi duals r eporting c ognitive selves by age g roup Age group Yes cognitive self No cognitive self Young 61.8% 38.2% Middle 44.1% 55.9% Old 42.3% 57.7% N = 339 Table 4 3. Mean number of cognitive selves by age g roup: Hopes and f ears Age Mean Standard Mean Standard group h opes deviation f ears deviation Young .78 1.06 .35 .64 Middle .51 .90 .16 .42 Old .56 1.04 .18 .47 N = 339 Table 4 4. Type of c ognitive selves reported by a ge g roup Age g roup Feared selves Fe ared s elves Hoped for selves Hoped for selves Yes No Yes No Young 27% 73% 49.4% 50.6% Middle 14.4% 85.6% 33.3% 66.7% Old 15.3% 84.7% 32.8% 67.2% N = 339

PAGE 52

52 Table 4 5. Fac tor loadings for self p erce ived cognitive c ompetence (SPCC) Standardized Unstandardized Variable loadings l oadings PSQ .323 1.00 Efficacy .524 2.19 Locus .408 2.47 Capacity .907 12.45 Change .804 14.07 N = 92 T able 4 6. Factor loadings for negative psychological w ell b eing Standardized Unstandardized Variable l oadings l oadings Depression .914 1.00 Anxiety .852 1.21 Negative a ffect .822 .645 N = 92 Table 4 7. Factor l oadings fo r positive psychological well b eing Standardized Unstandardized Variable loadings l oadings Self e steem .823 1.00 Positive a ffect .461 .881 Self a cceptance .799 1.21 Environmental m astery .802 1.23 Personal g rowth .568 .689 Purpose in l ife .693 .964 Positive r elations .577 .781 Autonomy .631 .974 N = 92 Table 4 8. Mean scores on psychological well being composites by age g roup Age Mean Standard Mean Standard g roup PWB+ devi ation PWB deviation Young .67 4.33 .60 2.86 Middle 1.19 3.85 .74 1.73 Old .26 3.61 .05 2.18 N = 92 Table 4 9. Mean cognitive efficacy scores by age g roup Age Mean Standard Mean Standard g roup GME deviation PSQ deviation Young 15.44 3.06 12.27 1.69 Middle 16.44 3.15 11.30 2.45 Old 15.59 2.73 10.59 2.15 N = 92

PAGE 53

53 Table 4 10. Intercorrelations between efficacy and psychological w ell being: Younger a dults GME PSQ PWBneg PWBpos GME --PSQ .003 --(32) P= .986 PWBneg .021 .187 --(32) (32) P= .909 P= 305 PWBpos .059 .041 .630** --(32) (32) (32) P= .748 P= .822 P= .000 N = 92 **Correlation is significant at the .01 level (2 tailed).

PAGE 54

54 Table 4 11 Intercorrelations between e fficacy and psychological well being: Middle aged a dults GME PSQ PWBneg PWBpos GME --PSQ .345 --(27) P= .068 PWBneg .228 .072 --(27) (27) P= .257 P= .722 PWBpos .382* .336 .585** --(27) (27) (27) P= .049 P= .086 P= .001 N = 92 **Correlation is significant at the .01 level (2 tailed). *Correlation is significant at the .05 level (2 tailed).

PAGE 55

55 Table 4 12. Intercorrelations between efficacy and p sychologica l well b eing: O lder a dults GME PSQ PWBneg PWBpos GME --PSQ .570** --(32) P= .001 PWBneg .170 .279 --(32) (32) P= .352 P = .122 PWBpos .383** .299 .774** --(32) (32) (32) P= .031 P= .096 P= .000 N = 92 **Correlation is significant at the .01 level (2 tailed). *Corre lation is significant at the .05 level (2 tailed).

PAGE 56

56 Table 4 13. Latent path m odel: Cognition Standardi zed Unstandardized Path loadings l oadings Age GME .025 .003 Age PS Q .323 .030 GME PSQ .298 .220 GME +PWB .157 .183 GME PWB .131 .327 PSQ +PWB .087 .137 PSQ PWB .013 .043 PWB +PWB .734 .342 N = 92 *Effect is significant at the .05 level.

PAGE 57

57 Table 4 14. Latent path m odel: Cognition plus objective cognitive p erformance Standardized Unstandardized Path loadings l oadings Age Word l ist .068 .007 Age PSQ .323 .030 Word l ist GME .00 7 .009 Word l ist +PWB .167 .248* Word l ist PWB .020 .064 GME PSQ .298 .220 GME +PWB .1 61 .18 7 GME PWB .13 0 .32 6 PSQ +PWB .08 5 .13 3 PSQ PWB 01 4 .04 7 PWB +PWB .7 29 .34 0* N = 92 *Effect is significant at the .05 level. Table 4 15. Latent path model: Cognitive hopes excluding negative p sychologi cal well b eing Standardized Unstandardized Path loadings l oadings Age PSQ .25 5 .023 GME PSQ .24 5 .178 GME +PWB .208 .259 PSQ +PWB .071 .122 N = 92 *Effect is significant at the .05 level

PAGE 58

58 Table 4 16. Latent p ath m odel: Cognitive fears excluding positive psychological well b eing Standardized Unstandardized Path loadings l oadings Age PSQ .53 1 .049 GME PSQ .3 16 .229 GME PWB .27 3 .463 PSQ PWB .10 5 .246 N = 92 *Effect is significant at the .05 level. Table 4 17. Mean word list r ecall by a ge g roup Recall Standard Intrusions Standard Repetitions Standard Age g roup deviation deviation deviation Young 8.31 1.97 .31 .71 .38 .7 3 Middle 7.38 2.39 .27 .53 .58 1.47 Old 7.77 2.76 .43 .77 .37 1.03 N = 92 Table 4 18. Mean memory efficacy and Metamemory in Adulthood Scale ratings by age g roup Capacity Standard Achieve. Standard Anxiety Standard Age g roup deviation deviation deviation Young 61.34 8.23 60.00 5.56 39.69 8.14 Middle 55.32 8.57 60.12 8.49 39.30 9.51 Old 53.14 10.41 61.16 7.59 44.47 7.86 Change Standard Locus Standard Efficacy Standard Age g roup deviation deviation deviation Young 61.16 10.65 31.78 3.52 15.43 3.06 Middle 51.19 11.59 31.83 3.98 16.44 3.15 Old 46.94 10.69 30.94 5.21 15.59 2.73 N = 92

PAGE 59

59 Table 4 19. Mean psychological well b eing ratings by age g ro up Depression Standard Anxiety Standard Negative Standard Age g roup deviation deviation affect deviation Young 11.47 10.01 34.56 12.54 18.50 6.76 Middle 7.75 6.50 29.28 9.40 14.81 4.57 Old 9.66 7.57 33.85 9.57 16.39 5.82 Positive Standard Self Standard Autonomy Standard Age g roup affect deviation e steem deviation deviation Young 37.31 6.46 33.81 4.60 31.63 5.67 Middle 36.81 5.88 36.44 3.68 34.37 5. 11 Old 34.96 7.36 34.52 3.87 32.71 5.06 Environ Standard Personal Standard Positive Standard Age g roup mastery deviation growth deviation relations deviation Young 30.41 4.94 35.59 4.49 35.00 4.41 Middle 35.00 5.20 35.41 3.69 34.96 5.31 Old 33.12 5.02 34.51 4.39 35.25 4.60 Purpose Standard Self Standard Age g roup in life dev iation accept deviation Young 35.53 5.00 33.31 5.52 Middle 36.93 5.46 34.52 5.15 Old 36.23 4.12 32.73 5.13 N = 92

PAGE 60

60 Figure 4 1. Latent p ath model for c ognition PWB +PWB PANAS CES D STAI +PANAS SES AGE PSQ GME Autonomy Positive Relations Environ Mastery Personal Growth Purpose in Life Self Acceptance

PAGE 61

61 Figure 4 2. Modified latent path model for c ognition +PWB PWB PANAS CES D STAI +PANAS SES AGE PSQ GME Autonomy Positive Relations Environ Mastery Personal Growth Purpose in Life Self Acceptance

PAGE 62

62 Figure 4 3. Modified latent path mod el for cognition with objective p erformance +PWB PWB PANAS CES D STAI +PANAS SES AGE PSQ GME Autonomy Positive Relations Environ Mastery Personal Growth Purpos e in Life Self Acceptance Word Recall

PAGE 63

63 Figure 4 4. Latent path model for c ognitive h opes and positive p sychological w el l b eing +PWB +PANAS SES Autonomy GME PSQ AGE Environ Mastery Purpose in Life Personal Growth Self Acceptance Positive Relations

PAGE 64

64 64 Figure 4 5. Latent path model for c ognitive fears and negative p sychological well b eing PWB PANAS CES D STAI AGE PSQ GME

PAGE 65

65 CHAPTER 5 RESULTS: HEALTH Analyses of variance were conducted to examine the number of spontaneous health selves reported, as a function of age group. In addition to the analyses of variance, chi square tests were also conducted, to determine whether the number of individuals listing these health selves differed significantly by age group. Significant group differences were identified in each case by Confirmatory factor analyses were conducted and compos ite variables were created. Next, examination of the proposed structural equation models was carried out. Last, follow up analyses of variance were conducted to examine differences in health, health beliefs, and psychological well being by age group. Hea lth Selves An analysis of variance was conducted to examine whether the three age groups (young, middle, and old) differed by the total number of health selves spontaneously listed. Health difference in the total number of health selves reported by age group was found, F (2, 338) = 3.04, p < .05, eta 2 = .018. Younger adults reported significantly fewer health selves than middle aged adults. There were no signi ficant differences between middle aged and older adults, or between younger and older adults (see Table 5.1) which was not expected To further explore the spontaneous reports of health selves, participants were divided into those who spontaneously mention ed a health self and those who did not. Given that the total number of health selves reported by each individual was typically small (between 0 and 3), we felt it was prudent to confirm these age differences by examining the frequency with which individua ls reported a health self as a function of age. To do this, a chi square analysis was conducted comparing those who reported no health selves with those who reported at least one

PAGE 66

66 health self, in relation to age group (young, middle, and old). The results were significant, Pearson X 2 (N = 339, df = 2) = 17.26, p < .001, further supporting the notion that these age groups did differ in their spontaneous report of health selves (see Table 5.2). Follow up comparisons revealed a significant difference between the number of younger and middle aged adults reporting a health self: Pearson X 2 (N = 201, df = 1) = 8.91, p < .005; as well as a significant difference between the number of younger and older adults reporting a health self: Pearson X 2 (N = 228, df = 1) = 15.46, p < .001. No significant difference was found between number of middle aged and older adults reporting a health self: Pearson X 2 (N = 249, df = 1) = .67, p >.05 (see Table 5.2). Overall, a greater number of older adults reported health selves, fo llowed by middle aged and younger adults respectively. Next, a closer look at the types of health selves (hoped for or feared) reported by the three age groups was examined. An analysis of variance was conducted to examine whether the three age groups (yo ung, middle, and old) differed by the total number of health hopes spontaneously listed. No significant differences were found for health hopes, F (2, 338) = 1.34, p > .05 An analysis of variance was also conducted to examine whether the three age group s (young, middle, and old) differed by the total number of health fears spontaneously listed. Significant differences were found for health fears, F (2, 338) = 3.07, p < 05, eta 2 = .018. Middle aged adults reported a significantly greater number of healt h fears than younger adults. No differences were found between middle aged and older adults or between younger or older adults (see Table 5.3) which was unexpected. To further examine the types of health selves reported, participants were categorized into : those who reported health hopes and those who did not, and those who reported health fears and those who did not. A cross tabulation of the three age groups (young, middle aged, old) by

PAGE 67

67 health hope was significant: Pearson X 2 (N = 339, df = 2) = 9.38, p < .05. A cross tabulation of the three age groups (young, middle aged, old) by health fear was also significant: Pearson X 2 (N = 339, df = 2) = 13.14, p < .001. Follow up comparisons revealed a significant difference between the number of younger and mi ddle aged adults reporting a health hope: Pearson X 2 (N = 201, df = 1) = 5.29, p < .05; and between the number of younger and middle aged adults reporting a health fear: Pearson X 2 (N = 201, df = 1) = 11.51, p < .005. A significant difference was also obs erved between the number of younger and older adults reporting a health hope: Pearson X 2 (N = 228, df = 1) = 8.89, p < .005, and between the number of younger and older adults reporting a health fear: Pearson X 2 (N = 228, df = 1) = 8.33, p < .005. No sign ificant difference was found between the number of middle aged and older adults reporting a health hope: Pearson X 2 (N = 249, df = 1) = .383, p > .05 or a health fear: Pearson X 2 (N = 249, df = 1) = .529, p > .05 (see Table 5.4). Again, a greater number of older adults reported health hopes and health fears as expected. Identifying Individuals for Whom Health is Important Before continuing with the analyses of the proposed model, it was necessary to identify individuals who not only reported a health rela ted possible self, but who also rated that self as highly important. Two requirements had to be met for a participant to be included. First, individuals who listed a health hope or health fear as one of their three most important hopes or most important fears were identified. From our study sample, 220 participants identified a health self as one of their three most important hopes or three most important fears. Second, we examined how important these participants rated these health selves on the possib le selves questionnaire. For health hopes, participants were asked to rate the following statement on a 7 were asked to rate the following statement on a

PAGE 68

68 7 most importa nt health hope or fear lower than a 4, they were excluded from further analysis. These criteria resulted in the exclusion of 5 participants. Measurement Model: Health Competence A factor analysis was conducted to identify a latent self perceived health co mpetence factor proposed in Aim 2. Health Self Efficacy (measured by the Self Rated Abilities for Health Practices Scale: SRAHP), Health Locus of Control (measured by the Internal Control subscale of the Multidimensional Health Locus of Control Form B: MH LCB), and PSQ (measured by the efficacy and outcome expectancy items answered for most important hopes and fears on the Possible Selves Questionnaire) were all expected to load on the Self Perceived Health Competence (SPHC) factor. These measures did not load on the hypothesized SPHC factor strongly (see Table 5.5 for factor loadings). Additional analyses were conducted to identify a SPHC factor using individual subscales of the SRAHP (nutrition, exercise, responsible health practices, stress reduction) a nd combinations of PSQ (for example including efficacy items only or outcome expectancy items only) without success. Subsequent analyses were therefore carried out with individual variables (Health Self Efficacy and PSQ) representing SPHC. Measurement Mod el: Psychological Well Being Factor analyses were also conducted to identify two latent psychological well being factors (positive and negative). Although these factor analyses had confirmed a negative and positive well being factor for the earlier analys es, those earlier analyses were conducted with a different sample this was a sample of individuals with health selves in contrast to the earlier sample representing those with cognitive selves. As such, these analyses represented a further confirmation of these well being factors. Negative Affect, Depression, and Anxiety were

PAGE 69

69 expected to load on the negative psychological well being factor. The proposed negative psychological well being variables loaded strongly and significantly on their proposed factor. See Table 5.6 for standardized and unstandardized factor loadings. Positive Affect, Self Esteem, and the six Scales of Psychological Well Being (positive relations, personal growth, purpose in life, self acceptance, environmental mastery, and autonomy) w ere expected to load on the positive psychological well being factor. The proposed positive psychological well being variables loaded strongly and significantly on their proposed factor. See Table 5.7 for standardized and unstandardized factor loadings. Next, composite measures were created for both positive and negative well being. Composite measures were created by first creating z scores for each individual psychological well being variable. Next these z scores were regression weighted by multiplying each z score by the corresponding factor loading. Last the weighted scores for the individual scales were summed to create a composite measure. An analysis of variance was conducted to examine whether the three age groups (young, middle, and old) differe d on the composite measures of positive and negative psychological well being. As expected, n o significant differences were found between the three age groups for positive psychological well being, F (2, 217) = 1.88, p > .05 In contrast, a significant d ifference was found for negative psychological well being, F (2, 217) = 5.33, p < .005, eta 2 = .047. Younger adults reported significantly greater levels of negative well being than middle aged and older adults (see Table 5.8). Next we examined age differ ences in self efficacy. An analysis of variance was conducted to examine whether the three age groups (young, middle, and old) differed on composite measures of Health Self Efficacy and PSQ. Health Self Efficacy included a sum across four

PAGE 70

70 health efficacy scales including nutrition, exercise, stress management, and responsible health practices. The PSQ included the sum of the efficacy and outcome expectancy questions for each e. No significant differences were found between the three age groups for Health Self Efficacy, F (2, 217) = 2.77, p > .05 In contrast, a significant difference was found for PSQ, F (2, 217) = 6.69, p < .005, eta 2 = .059. Younger adults reported signif icantly greater levels of efficacy and outcome expectancy for their most important health selves than both middle aged and older adults (see Table 5.9). Last correlations were examined for Health Self Efficacy, PSQ, positive psychological well being and ne gative psychological well being for each age group. As expected, efficacy was more strongly correlated with psychological well being for the middle aged and older adults than for younger adults (see Tables 5.10 through 5.12). Latent Variable Path Model Ne xt a latent variable path model was examined for individuals who identified a health self as a most important hope or most dreaded fear. Based on the results of the factor analyses, the SPHC factor was removed from the original path model and instead Heal th Self Efficacy and PSQ were added as individual predictors ( s ee Figure 5.1 ) Thus, the revised model proposed that age directly impacted both Health Self Efficacy and PSQ and that both Health Self Efficacy and PSQ directly impacted negative and positive psychological well being. The fit of this original path model was poor. Modification indices suggested adding paths from age to both positive and negative psychological well being, from Health Self Efficacy to PSQ, and from negative well being to positi ve well being as well as allowing several errors to covary. See Figure 5.2 for the modified model (covariances are not pictured in the model)

PAGE 71

71 The effects of the modified path model are mixed. The Chi square was significant, X 2 (58) = 122.02; however, goodness of fit indices indicated only moderately acceptable fit: CFI = .957, RMSEA = .072. The effects of Health Self Efficacy on PSQ, positive psychological well being and negative psychological well being were all significant. The effects of age on bo th Health Self Efficacy and negative psychological well being were significant, and the effects of negative psychological well being on positive psychological well being were also significant. The remaining effects of age on Health Self Efficacy and posit ive psychological well being, as well as the effects of PSQ on both positive and negative psychological well being were not significant. Standardized and unstandardized effects are reported in Table 5.13. Next we examined a latent path model for individua ls reporting health hopes (n=126) with the negative well being factor excluded from the model (Figure 5.3). The Chi square was significant, X 2 (35) = 80.23, and goodness of fit statistics indicated poor fit: CFI = .927, RMSEA = .102. Standardized and uns tandardized effects are reported in Table 5.14. Last we examined a latent path model for individuals reporting health fears (n=169) with the positive well being factor excluded from the model (Figure 5.4). The Chi Square, X 2 (6) = 8.47, was not significan t, and Goodness of fit indices indicated good fit: CFI = .993, RMSEA = .049. The effects of the path model are largely significant. The exceptions are the non significant paths of age on Health Self Efficacy and PSQ on negative psychological well being. Standardized and unstandardized effects are reported in Table 5.15 Additional Analyses In addition to the proposed analyses, several follow up analyses were conducted. The purpose of these analyses was to examine age differences in health, psychological well being, and health competence among individuals reporting important health selves.

PAGE 72

72 A multivariate analysis of variance was conducted to compare the three age groups (young, middle aged, old) across the following dependent variables from the SF 36: general health, physical functioning, role physical, pain, vitality, social, role emotional, and mental health. Age differences were significant, F (16, 416) = 5.75, p < .001, eta 2 = .181, and this significant result was further explored in univariate tes ts. Significant age differences were found for physical functioning, F (2, 216) = 14.06, p < .001, eta 2 = .116, and role physical subscales, F (2, 216) = 3.95, p < .05, eta 2 = .036, due to the fact that older adults reported experiencing poorer physical fu nctioning than both the younger and middle aged adults. Significant age differences were also found for pain, F (2, 216) = 3.26, p < .05, eta 2 = .030; as older adults reported experiencing more pain than younger adults. Next, significant age differences were found for role emotional, F (2, 216) = 7.76, p < .005, eta 2 = .068 and mental health subscales, F (2, 216) = 6.24, p < .005, eta 2 = .055, as younger adults reported poorer emotional and mental health than both middle aged and older adults. No signifi cant age differences were found for the general health, vitality, or social subscales of the SF 36 (see Table 5.16). A multivariate analysis of variance was also conducted to compare the three age groups (young, middle aged, old) across the following depen dent vari ables: Health Self Efficacy for nutrition, exercise, stress management, and responsible health practices, and Heath Locus of Control which included internal, chance and powerful others subscales. Age differences were significant, F (14, 376) = 3. 81, p < .001, eta 2 = .124, and this significant result was further explored in univariate tests. A univariate analysis of variance examining overall Health Self Efficacy as a function of age group was not significant, F (2, 206) = 2.81, p > .05. When exam ining the individual

PAGE 73

73 subscales, s ignificant age differences were found for nutrition self efficacy, F (2, 195) = 3.78, p < .05, eta 2 = .038, due to the fact that younger adults reported feeling significantly less capable of maintaining a healthy diet than the middle aged adults. Significant age differences were also found for exercise self efficacy, F (2, 195) = 5.23, p < .05, eta 2 = .051; as older adults reported feeling less capable of engaging in regular exercise than both younger and middle aged adults Next, significant age differences were found for responsible health practices, F (2, 195) = 3.25, p < .05, eta 2 = .033, as younger adults reported feeling less capable of taking care of their own health than middle aged adults. Last, significant age di fferences were found for the powerful others subscale, F (2, 195) = 9.66, p < .001, eta 2 = .091, as older adults reported that powerful others exert ed greater control over their personal health than did both younger and middle aged adults. No significant age differences were found for self efficacy for stress management or for the internal control or chance subscales of the Health Locus of Control measure (see Table 5.17). Last, a multivariate analysis of variance was conducted to compare the three age gro ups (young, middle aged, old) across the following dependent variables: depression, anxiety, negative affect, positive affect, self esteem, autonomy, environmental mastery, purpose in life, personal growth, self acceptance, and positive relations. Age dif ferences were significant, F (122, 360) = 2.77, p < .001, eta 2 = .145, and this significant result was further explored in univariate tests. Significant age differences were found for depression, F (2, 191) = 5.08, p < .05, eta 2 = .051, due to the fact tha t younger adults reported experiencing a significantly greater degree of depressive symptoms than both middle aged and older adults. Middle aged and older adults did not significantly differ from one another. Significant age differences were also found f or negative affect, F (2, 191) = 5.33, p < .05, eta 2 = .053; again younger adults reported

PAGE 74

74 experiencing greater levels of negative affect than both middle aged and older adults. Next, significant age differences were found for autonomy, F (2, 191) = 5.15, p < .05, eta 2 = .052, and environmental mastery, F (2, 191) = 6.57, p < .05, eta 2 = .065 as younger adults reported less autonomy and less environmental mastery than both middle aged and older adults. Significant age differences were also found for perso nal growth, F (2, 191) = 5.01, p < .05, eta 2 = .050, as older adults reported lower levels of personal growth than both younger and middle aged adults. Last, significant differences were found for positive relations, F (2, 191) = 3.37, p < .05, eta 2 = .03 4, as older adults reported greater engagement in positive relations with others than younger adults. No significant age differences were found for positive affect, self esteem, purpose in life, self acceptance, or anxiety (see Table 5.18).

PAGE 75

75 Table 5 1. Mean n umber of health selves r eported by age g roup: Total Age g roup Mean num ber of health selves Standard d eviation Young 1.57 1.88 Middle 2.18 1.98 Old 1.91 1.45 N = 339 Table 5 2. Percentage of i ndividua ls r epo rting health selves by a ge g roup Age group Yes health self No health self Young 60% 40% Middle 79.3% 20.7% Old 83.3% 16.7% N = 339 Table 5 3. Mean number of health selves by age g roup : Hopes and f ears Age Mean Standa rd Mean Standard group h opes deviation f ears deviation Young .72 1.33 .84 1.12 Middle .97 1.19 1.21 1.15 Old .89 1.03 .97 .95 N = 339 Table 5 4. Type of h ealth selves reported by age g roup Age g rou p Feared selves Feared s elves Hoped for selves Hoped for selves Yes No Yes No Young 46.7% 53.3% 37.8% 62.2% Middle 70.3% 29.7% 54.1% 45.9% Old 65.9% 34.1% 58% 42% N = 339

PAGE 76

76 Table 5 5. Factor l oadings for self perceived health c ompetence (SPHC) Standardized Unstandardized Variable loadings l oadings Self e fficacy .430 1.00 Locus of c ontrol .447 6.09 PSQ .531 1.43 N = 215 Tabl e 5 6. Factor l oadings for negative psychological well b eing Standardized Unstandardized Variable loadings l oadings Depression .913 1.00 Anxiety .810 1.10 Negative a ffect .842 .671 N = 215 Table 5 7. F actor loadings for positive psychological well b eing Standardized Unstand ardized Variable loadings l oadings Self e steem .843 1.00 Positive a ffect .560 1.06 Self a cceptance .848 1.04 Environmental m astery .819 1.27 Personal g rowth .579 .84 Purpose in l ife .662 .75 Positive r elati ons .547 .98 Autonomy .650 1.40 N = 215 Table 5 8. Mean s cores on psychological well being composites by age g roup Age Mean Standard Mean Standard g roup PWB+ deviation PWB deviation Young 1.34 4.24 1.17 2.74 M iddle .28 4.46 .0005 2.41 Old .22 4.10 .36 2.05 N = 215 Table 5 9. Mean health efficacy scores by age g roup Age Mean Standard Mean Standard g roup HSE deviation PSQ deviation Young 89.16 15.56 12.08 2.1 7 Middle 93.24 14.67 10.60 2.25 Old 87.85 16.60 10.42 2.33 N = 215

PAGE 77

77 Table 5 10. Intercorrelations b etween e fficacy and p sychological w ell being: Younger a dults PWBneg PWBpos PSQ HSE PWBneg --PWBpos .555* --(32) P= .001 PSQ .080 .112 --(32) (32) P= .664 P= .543 HSE .191 .286 .207 --(32) (32) (32) P= .295 P= .113 P= .255 N = 215 **Correlation is significant at the .01 level (2 tailed). *Correlation is significant at the .05 leve l (2 tailed).

PAGE 78

78 Table 5 11. Intercorrelations between efficacy and psychological well being: Middle aged a dults PWBneg PWBpos PSQ HSE PWBneg --PWBpos .645* --(84) P= .000 PSQ .255* .316** --(84) (84) P= .019 P= .003 HSE .450 .445** .116 --(84) (84) (84) P= .000 P= .000 P= .295 N = 215 **Correlation is significant at the .01 level (2 tailed). *Correlation is significant at the .05 level (2 tailed).

PAGE 79

79 Table 5 12. Intercorrelations between effic acy and psychological well being: Older a dults PWBneg PWBpos PSQ HSE PWBneg --PWBpos .718** --(102) P= .000 PSQ .214* .336** --(102) (102) P= .031 P= .001 HSE .373** .580** .334** --(102) (102) (102) P= .000 P= .000 P= .001 N = 215 **Correlation is significant at the .01 level (2 tailed). *Correlation is significant at the .05 level (2 tailed).

PAGE 80

80 Table 5 13. Latent p ath m odel: Health Standardized Unstandardized Path loadings l oadings Age Health e fficacy .055 .046 Age PSQ .204 .024 Age +PWB .055 .008 Age PWB .264 .111 Health e fficacy PSQ .235 .033 Health e fficacy +PW B .233 .053 Health e fficacy PWB .349 .164 PSQ +PWB .093 .150 PSQ PWB .097 .319 PWB +PWB .657 .324 N = 215 *Effect is significant at the .05 level.

PAGE 81

81 Table 5 14. Latent path m odel: Heal th hopes excluding negative psychological well b eing Standardized Unstandardized Path l oadings l oadings Age Health e fficacy .091 .080 Age PSQ .174 .021 Age +PWB .276 .061 Health e fficacy PSQ .350 .047 Health e fficacy +PWB .416 .104 PSQ +PWB .300 .560 N = 215 *Effect is significant at the .05 level Table 5 15. Latent path m odel: Health f ears excluding positive p sycholo gical well b eing Standardized Unstandardized Path loadings l oadings Age Health e fficacy .051 .046 Age PSQ .206 .025 Age PWB .247 .105 Health e fficacy PSQ .209 .028 Health e fficacy PWB .356 .167 PSQ PWB .105 .368 N = 215 *Effect is significant at the .05 level.

PAGE 82

82 Table 5 16. Mean SF 36 r ating s b y age g roup Physical Standard Role Standard Pain Standard Age group function deviation p hysical deviation deviation Young 92.27 15.62 89.39 25.79 79.52 21.47 Middle 82.78 25.64 84.88 33.93 70.77 24.24 Old 68.59 26.89 73.30 37.42 67.48 23.72 General Standard Vitality Standard Social Standard Age group h ealth deviation deviation deviation Young 73.91 24.94 55.61 17.89 78.41 25.05 Middle 75.17 23.12 62.22 21.97 84.72 22.71 Old 7 3.47 21.67 63.11 18.87 86.04 22.02 Role Standard Mental Standard Age group emotion deviation h ealth deviation Young 57.58 42.71 70.18 14.56 Middle 85.19 31.18 75.46 17.66 Old 80.58 34. 15 80.54 14.03 N = 215

PAGE 83

83 Table 5 17. Mean health e fficacy and locus of control ratings by age g roup Nutrition Standard Exercise Standard Stress Standard Age g r oup deviation deviation m gmt deviation Young 21.97 5.24 22.36 4 .34 21.52 4.66 Middle 24.40 3.91 21.87 5.91 22.15 4.91 Old 23.26 4.34 19.13 7.12 21.70 4.88 Health Standard Internal Standard Chance Standard Age g roup p ractices d eviation c ontrol deviation deviation Young 22.8 9 4.97 26.40 3.59 16.73 4.24 Middle 24.95 3.49 26.36 4.28 16.33 5.44 Old 24.5 9 3.91 25.89 4.63 15.24 5.09 Powerful Standard Age group o thers deviation Young 17.61 3.81 Middle 18.92 5.26 Old 2 1.80 5.96 N = 215

PAGE 84

84 Table 5 18. Mean psychological w ell being r atings by age g roup Depression Standard Anxiety Standard Negative Standard Age g roup deviation deviation affect deviation Young 13.00 10.19 34.27 10.06 19.24 7.24 Middle 8.82 8.90 31.70 10.18 15.98 6.31 Old 7.58 6.92 29.97 10.09 15.24 5.24 Positive Standard Self Standard Autonomy Standard Age group affect deviation e steem deviation deviation Young 34.64 7.65 3 3.33 4.96 29.12 6.80 Middle 36.07 6.57 34.56 4.65 32.55 5.81 Old 34.91 7.19 34.43 4.16 32.93 5.76 Environ Standard Personal Standard Positive Standard Age g roup mastery deviation growth deviation relati ons deviation Young 30.21 5.35 35.67 4.31 33.36 4.31 Middle 33.15 6.26 34.90 5.32 35.25 5.91 Old 34.38 5.13 32.72 5.81 36.03 4.43 Purpose Standard Self Standard Age g roup in life deviation accept d eviation Young 36.30 5.31 32.33 5.97 Middle 36.71 5.66 33.40 6.84 Old 36.78 5.43 34.00 5.64 N = 215

PAGE 85

85 Figure 5 1. Latent path model for h ealth PWB +PWB PANAS CES D STAI +PANAS SES AGE PSQ HSE Autonomy Positive Relations Environ Mastery Personal Growth Purpose in Life Self Acceptance

PAGE 86

86 Figure 5 2. Modified latent path model for h ealth PWB +PWB PANAS CES D STAI +PANAS SES AGE PSQ HSE Autonomy Positive Relations Environ Mastery Personal Growth Purpose in L ife Self Acceptance

PAGE 87

87 Figure 5 3. Latent path model for health hopes and positive p sychological w ell b eing +PWB +PANAS SES Autonomy HSE PSQ AGE Environ Mastery Purpose in Life Personal Growth Self Acceptance Positive Relations

PAGE 88

88 Figure 5 4. Latent path m odel for health fears and negative p sychological well b eing PWB PANAS CES D STAI AGE PSQ HSE

PAGE 89

89 CHAPTER 6 DISCUSSION Possible selves are hopes and fears we each have about what we will be like in the future The current study examined possible selves related to cognition and health. Specifically we selected individuals who not only reported a self related to cognition or health but also identified this self as highly important. The central premise was that positive psychological well being should be higher (and negative psychological well being lower) for individuals reporting high levels of perceived competence related to high ly important, self defining domains. Although possible selves have been examined extensively, this is the first study to examine the relationship between highly valued selves, feelings of competence, and psychological well being. Believing that we have t he ability to attain our most important hopes and avoid our most dreaded fears, especially when faced with declines that may threaten our ability to do so, may be central to maintaining positive psychological well being as we age. Possible Selves Possible selves are elements of the self concept that represent what a person hopes to become would like to become, or is afraid of becoming (Markus & Nurius, 1986) Possible se lves are not static goals. The se selves c hange over time with changes in con text, abi lity, and life stage. These future selves represent the developmental trajectory an individual envisions for him or herself. As such, possible selves can be considered a road map; although certain detours may be taken along the way, an individual finds a way to reach the important destinations. T he types of selves reported by individuals in different contexts or life stages especially age differences in possible selves have b een examined extensively. Previous research on po ssible selves has documented differences in the number and types of possible selves reported by different age groups. Older adults report fewer possible selves than younger adults (Cross &

PAGE 90

90 Markus, 1991; Hooker, 1992) and the types of selves they report are considered more in line wi th their current identity. For example older adults often report hopes and fears related to physical functioning, maintaining relationships, and continuing to engage in activities that they already enjoy participating in (Smith & Freund, 2002). In contr ast, younger adults report a more diverse array of hopes and fears. For example, younger adults report hopes and fears related to their education, choosing a career path, finding a mate and starting a family (Hooker, 1992). Our results confirmed these fi ndings. Younger adults reported a significantly greater number of hopes than older adults. Younger adults also reported a greater number of fears than middle aged and older adults. Further, looking at the types of most important hopes and fears most fre quently reported by each age group we again see that our results are consistent with previous research. For younger adults, the most frequently endorsed hope was career related and the most frequently endorsed fear was social. For middle aged adults, the most frequently endorsed hope was social; the most frequently endorsed fear was health related. For older adults, the most frequently endorsed hope and fear was health related. These differences are consistent with the experiences and expectations that accompany different stages of adulthood. Further, because middle aged and older adults report fewer selves and the selves they report are more in line with their current identities, the ability to achieve most important hopes and avoid most dreaded fears may be key in maintaining psychological well being throughout the later half of the lifespan as one navigates through life Cognition The broad body of cognitive aging r esearch has documented expectations of cognitive decline (Lineweaver & Hertzog, 1998; R yan, 1992 ) differences in control and efficacy beliefs related to memory performance (Berry, West, & Dennehy, 1989; Lachman et al., 1995 ) and age related declines in cognitive ability (Lachman et al., 1995; Ryan, 1992 ) Despite these findings,

PAGE 91

91 very litt le research has been done to specifically examine whether these beliefs, expectations and experiences are represented within the possible selves of older adults. Further, what little research has been done on cognitive selves has been mixed and often depe nds on the definition of what a cognitive self is. For example most studies do not include a category for memory or cognition ( Cross & Markus, 1991; Hooker, 1992; Hooker & Kaus, 1994) and the few that do often fail to find many participants reporting selv es that fall into this category ( Westerhof, Katzko, Dittmann Kohli & Hayslip, 2001; Smith & Freund, 2002). For the purpose of the current study, cognitive selves included hopes and fears about memory ability, as well as hopes and fears related to learnin g, education, and the ability or desire to engage in cognitively stimulating activities. Prior research has shown that unlike older adults, younger adults are not concerned with their memory ability; however, when examining a broader definition of cogniti on that included learning and cognitively challenging activities both older and younger adults have similar numbers of cognitive hopes and fears (Dark Freudeman, West, & Viverito, 2006). Co nsistent with this previous research, co gnition was important to young, middle aged and older adults in our sample In fact younger adults reported a significantly greater number of cognitive selves than both middle aged and older adults. Looking at the types of cognitive selves reported by the different age groups, younger adults tended to report both hopes and fears related to learning and education. Middle aged and older adults reported hopes related to engaging in cognitively challenging activities and learning, and fears related to mental decline. The results i ndicate that cognition is important among each of the three age groups albeit in different ways. A longitudinal study of possible selves would allow further exploration of the se age differences It would also allow a better understanding of the differenc es between individuals with and without cognitive selves and how those cognitive selves change over time.

PAGE 92

92 Health Like cognition, age differences in health status have been well documented (Ware, Snow, Kosinski, & Gandek, 1993 ) For example, y ounger adults generally report few physical limitations and high levels of physical functioning. Older adults report gr eater physical limitations and pain and ma y report being treated for chronic health concerns such as high b lood pressure or arthritis Unlike cognit ion, research on possible selves has focused heavily on health, specifically examining age differences in the presence or absence of health selves, as well as differences in the content of these health selves between individuals with and without chronic di seases such as Parkinso ) Health selves include hopes and fears about physical functioning, general statements about maintaining health, as well as concerns about avoiding or managing specific di seases and statements about physical activity or exercise Previous research has shown that middle aged and older adults report a greater number of health selves than younger adults (Hooker, 1992; Hooker & Kaus, 1992; Hooker & Kaus, 1994 ) Further the ty pes of health selves reported by younger adults are often very different than the types of health selves reported by middle aged and older adults. For example younger adults may fear getting fat or may hope to lose weight (Hooker, 1992 ) Middle aged and older adults tend to report health selves related to physical functioning and specific diseases (Hooker, 1992; Hooker & Kaus, 1992; Hooker & Kaus, 1994 ) In the current study, m iddle aged and older adults reported a significantly greater number of health selves than younger adults replicating findings by Hooker and colleagues (Hooker, 1992; Hooker & Kaus, 1992; Hooker & Kaus, 1994). When examining the types of health selves (hopes and fears) report ed by each age group, there was no significant differenc e between the number of health hopes reported by each age group. How ever, when examining health fears a significant difference emerged as middle aged

PAGE 93

93 adults reported a significantly greater number of health fears than younger adults. This may indicate t hat middle aged adults are at a crossroads, currently enjoying good health, but expecting declines in the near future. Although we did expect middle aged adults to report a greater number of health selves compared to younger adults, we also expected older adults to report the greatest number of health selves overall. This was not the case; the number of selves the older adults reported was not significantly differen t from the middle aged adults. Research on possible selves has largely neglected middle ag ed adults. The current results indicate that middle aged and older adults may report similar numbers of selves within certain domains like health. Although middle aged and older adults did not significantly differ in the number of health selves as hypot hesized clearly health is important for both groups This leads us to believe that middle aged adults are thinking about their health and their future differently than younger and older adults. Middle aged adults may be starting to experience age relate d declines in health which lead them to focus on how further declines in health may impact their future and their quality of life Middle aged adults may also have aging parents who are f acing health challenges T hese challenges may serve as a personal w arning about what could be looming on the horizon In contrast, o lder adults are already experiencing notable declines in physical health and functioning. Age comparisons of the SF 36 confirm this to be true. Older adults reported experiencing lower lev els of physical functioning and greater physical limitations than both middle aged and younger adults. Thus middle aged and older adults think about health very differently than healthy younger adults do. For example younger adults may wish to live a hea lthy lifestyle or stay in shape, in contrast, a middle aged adult may fear being diagnosed with diab etes and an older adult may hope to maintain the ability to walk without assistance These

PAGE 94

94 health selves are likely to be more salient to middle aged and o lder adults because they have fewer hopes and fears overall and the selves they have are more closely t ied to their current identity. Psychological Well Being Psychological well being is believed to be an indictor of successful development and successful a ging (Markus & Herzog, 1992). In fact successful aging is often defined as maintaining psychological well being despite losses in multiple domains such as health and social roles (Brandtstadter, 1989 ). Research on psychological well being employs many di fferent definitions and many different measures. Common measures of psychological well being include satisfaction with life, self esteem, and a lack of depression. Regardless of the definitions or measures used, the focus has been on what it means to fun ction well throughout life despite the trials and tribulations we all come to face. Recently psychological well being has been viewed as a more complex, m ultidimensional construct that includes distinct positive and negative dimensions (Diehl, H astings, & Stanton, 2001; Ryff et al., 2006) Diehl, Hastings, and Stanton (2001) examined the factor structure of several common measures of psychological well being and found a positive factor and a negative factor. The positive factor included self esteem, auto nomy, environmental mastery, purpose in life, self acceptance, personal growth, positive relations, and positive affect. The negative factor included depression, anxiety, and negative affect. The current study also examined the same measures of psycholog ical well being used by Diehl, Hastings, and Stanton (200 1 ) Confirmatory factor analyses were conducted and the se latent positive and negative psychological well being factors were supported in both cognitive and health domains. Thus positive and negati ve psychological well being are not on opposite ends of the same continuum, but are distinct, though related constructs (Ryff et al., 2006).

PAGE 95

95 Psychological Well Being and Possible Selves Possible selves represent the developmental trajectory a person envi sions for themselves. As such, possible selves are used as a measuring stick to evaluate progress from current states to desired outcomes. In this respect, possible selves are believed to be related to life satisfaction and psychological well being. F or instance, individuals who feel capable of narrowing the gap between current and desired selves may have higher levels of life satisfaction and lower levels of depress ion, whereas individuals who do not feel capable of making progress toward desired selves may report lower levels of life satisfaction and higher lev els of depression and anxiety. The current study examined the relationship between possible selves, feelings of competence related to achieving or avoiding those selves, and psychological well be ing in two domains: cognition and health. Each domain will be discussed separately below. Cognition Competence, and Psychological Well Being Based on the confirmatory factor analyses for psychological well being described above, composites were created f or both positive and negative psychological well being for our cognitive sample No significant age differences were found for positive or negative psychological well being composites in the cognitive sample, as expected. O lder and middle aged adults con sistently report equivalent or greater levels of well being compared to younger adults (Carstensen, et al., 2000). We also tested age differences in General Memory Efficacy and PSQ. No age differences were found for General Memory Efficacy ; however, a ge differences were found for the PSQ. Younger adults reported feeling significantly more capable of attaining their cognitive hopes and avoiding their cognitive fears than both middle aged and older adults, consistent with previous research on p ossible selv es (Hooker, 1992). Although th ese results seem contrary, General Memory Efficacy and PSQ assess different types of

PAGE 96

96 efficacy. The General Memory Efficacy questionnaire asks a participant to rate how well they have performed on recent memory activities, ho w their memory compares with peers, and how satisfied they are with their recent memory performance. It may be that individuals who report a most important cognitive self value cognition because they have always performed well on cognitively challenging a ctivities. These individuals may identify themselves as intelligent and may regularly seek out cognitive stimulation. In contrast, the PSQ assesses a more specific belief -how capable an individual feels of achieving their most important cognitive hope or avoiding their most dreaded cognitive fear. Middle aged and older adults may be more realistic when assessing capability of achieving or avoiding a self compared to younger adults. It may also be that middle aged and older adults have experienced som e decline in cognitive ability that tempers their responses to these PSQ questions but has not yet impacted their overall efficacy beliefs on memory tasks In addition to examining age differences in these composite variables, w e tested a model in which co gnitive competence (General Memory Efficacy and PSQ) predicted psychological well being for individuals who reported a most important cognitive self. Unfortunately the predicted relationship between competence (General Memory Efficacy and PSQ) and psychol ogical well being was not supported in the cognitive model This may be due in large part to the lack of variability in psychological well being, cognitive performance, and efficacy between younger, middle aged, and older adults in our cognitive sample. Previous research has shown that older adults report lower levels of memory self efficacy than younger adults (Berry, West, & Denneh y, 1989; Hultsch, Hertzog, & Dixon, 1987) and perform more poorly on memory tasks than younger adults ( West, Thorn, & Bagwel l, 2003; West & Yassuda, 2004 ) This was not the case in the current study when examining only those with cognitive selves

PAGE 97

97 Young, middle aged, and older adults all performed very well on the word list recall activity and reported high levels of memory e fficacy as mentioned above In light of these unexpected results, we also examined age differences in BTACT word list recall performance across our entire sample. There were also no significant age differences in word list recall. These results indicate that the BTACT word list was not difficult enough as it did not discriminate well. W e also tested the cognitive model described above with cognitive performance (word list recall from the BTACT) added as a predictor of General Memory Efficacy, positive ps ychological well being, and negative psychological well being. The effect of cognitive performance on positive psychological well being was significant; but the effect of cognitive performance on negative psychological well being was not significant. Alt hough General Memory Efficacy and PSQ were not significant predictor s of well being the fact that cognitive performance was, leads us to believe that the overall conceptualization behind the model may still be valid. Individuals who identify cognition as a most important possible self and who perform better on a cognitiv e task report higher levels of positive psycholog ical well being. Middle aged and older adults who identify a most important cognitive self, may value cognition because they have always en joyed and excelled at cognitively challenging activities. These individuals may be considered schematic for cognition. Self schemas are identities that wh o we are (Cross & Markus, 1994). When individuals have a self schema in a particular domain they feel more control over their performance in that domain, are able to cope with negative feedback more easily, and overcome failures (Cross & Markus, 1994). F uture investigations of cognitive selves should also examine attitudes about education, intelligence,

PAGE 98

98 and ability to determine if there is a difference between individuals who are schematic for cognition and those who are not. Health Competence, and Psych ological Well Being Based on the confirmatory factor analyses for psychological well being described above, composites were also created for both positive and negative psychological well being for our health sample. No significant age differences were fou nd for the positive psychological well being composite in the health sample; however a significant difference emerged for negative well being. This difference was a result of younger adults reporting significantly greater levels of negative psychological well being than both middle aged and older adults. These differences were confirmed in follow up analyses. Younger adults in our health sample reported experiencing greater levels of depression and negative affect, as well as less autonomy and less envir onmental mastery than both middle aged and older adults. Although not entirely inconsistent with prior research, we did not expect age differences in either positive or negative well being. It may be that younger adults who report mo s t important health s elves have health concerns, whether about specific illnesses or about physical appearance, and these concerns may impact t heir mental health. We also know that higher levels of depression are associated with higher repor ts of sickness and fatigue (America n Psychiatric Association, 1994 ). Perhaps younger adults who are experiencing greater levels of negative well being are also experiencing more sickness and fatigue, which in turn, leads to concerns about physical health that are reflected in their possibl e selves This is an interesting idea that should be explored in future research. W e also tested age differences in Health Self Efficacy and PSQ. No significant differences were found for the Health Self Efficacy composite measure. Adults of all ages

PAGE 99

99 re ported high levels of Health Self Efficacy overall. In contrast, s ignificant differences were found for the PSQ, as y ounger adults reported feeling significantly more capa ble of attaining their health ho pes and avoiding their health fears than both middle aged and older adults, consistent with previous research on possible selves (Hooker, 1992). Although these measures both assessed efficacy, like the cognitive measures, they assessed different types of efficacy. The Health Self Efficacy measure assessed how capable individuals felt of engaging in a variety of positive health behaviors such as eating healthy food, exercising regularly, and seeking regu lar preventative medical care. The PSQ assessed the degree to which an individual felt capable of achiev ing a most important health hope or avoiding a most dreaded health fear. It is possible that a person could report high levels of Health Self Efficacy and at the same time not feel capable of avoiding a most important health fear such as losing mobility. To examine the relationship between health competence (Health Self Efficacy and PSQ) and psychological well being for individuals reporting most important health selves we tested a model in which health competence predicted both positive and negative psyc hological well being The effects of Health Self Efficacy on both positive and negative psychological well being were significant. Higher levels of Health Self Efficacy were associated with higher levels of positive psychological well being and lower lev els of negative well being as predicted. Although the effects of Health Self Efficacy were significant, the effects of PSQ on positive and negative psychological well being were not signi ficant. Again these measures assess different types of efficacy. F urther, t he PSQ was also quite brief and this may have limited its predictive power. Future studies should develop a more comprehensive measure of efficacy that is directly related to the accomplishment of most important selves to determine if this type o f self specific efficacy is related to psychological well being Overall, these results support the central

PAGE 100

100 hypothesis of the current study and suggest that competence is a predictor of psychological well being in the health domain. A longitudinal study that examines fluctuations in competence over time within highly valued domains would help us to further understand the relationship between competence and psychological well being. Competence, Psychological Well Being and Age Although competence was high for all three age groups, we also examined the relationship between competence and psychological well being by age group. This relationship was stronger for middle aged and older adults than for younger adults as hypothesized. This was true for both cogn ition and for health. We believe this relationship is stronger for middle aged and older adults because they generally report fewer possible selves than younger adults and the selves they report are generally grounded in their current identity. Because o f this, their psychological well being is more dependent on feelings of competence and success in fewer domains. Further, domains like cognition and health may be more salient for middle aged and older adults because they are threatened by the anticipatio n of declines in cognitive and physical ability. Younger adults have a broader repertoire of selves to choose from and are not concerned with declines in health or cognitive function. Thus, the psychological well being of younger adults is still related to feelings of competence but not to the same degree that it is for older adults. A longitudinal study could examine when and if this changes for younger adults as they move through life and begin to experience changes in health and cognition. Hopes and F ears Possible selves include the hopes an individuals is striving to achieve as well as the fears an individual is trying to avoid. There is some controversy in the literature about whether hopes, fears, or a combination of both types of possible selves a re more motivating. Smith and Freund (2002) suggested that having a balance (a hope and fear) in a given domain may be optimally

PAGE 101

101 motivating, however they found that people who deleted a balanced possible self in the domain of health showed less change in subjective health and life satisfaction over a four year period. Smith and Freund (2002) also found that life satisfaction, satisfaction with aging, and subjective health declined for individuals who added a dreaded health self to their possible self repe rtoire over a four year period. On the other hand, Hooker and Kaus (1994) found, contrary to expectations in health behavior literature, dreaded health selves were associated with reports of engaging in a greater number of health behaviors, indicating tha t some individuals may be more motivated by dread than optimism. Our previous research on cognitive selves indicated that older adults reporting hoped for memory selves actually reported greater levels of anxiety and adults reporting feared memory selves. Currently we do not know the importance of framing a possible self as a hope or a fear. Does it matter if a self is presented as a hope or as a fear? Is there something qualitatively different about a hoped for po ssible self and a feared possible self? Is having both a hope and a fear in the same domain a good thing or a bad thing? Although we did not have specific hypotheses regarding hopes a nd fears, we did test models that examined hopes and fears separately t o shed some light might be shed on these questions. These models were tested for both cognition and for health and will be discussed below. Cognitive Hopes and Fears First we tested a model in which age and cognitive competence (General Memory Efficacy an d PSQ) predicted positive well being for individuals with most important cognitive hopes. T he model examining th is relationship did not fit the data well The paths were largely non significant. These results seem to indicate that having a cognitive hop e and high cognitive competence does not predict higher levels of positive well being. Next we tested a model in

PAGE 102

102 which age and competence predicted negative well being for individuals with most important cognitive fears. In contrast to hopes the model e xamining cognitive fears did fit the data well Although this model fit well, the effect s of both General Memory Efficacy and PSQ on negati ve psychological well being were not signifi cant. Cognitive competence did not predict negative well being for indi viduals reporting a most important cognitive fear. However the paths from age to PSQ and from General Memory Efficacy to PSQ were significant. The results confirmed that lower levels of P SQ are associated with older age. We also saw that higher levels o f General Memory Efficacy predict higher levels of PSQ. Again, we feel that these model s should be tested using a more comprehensive measure of memory self ef ficacy that allows for greater variability in cognitive competence before being discarded. Health Hopes and Fears The model examining the relationship between competence and positive psychological well being for individuals with health hopes did not fit the data well. It appears that having a health hope and high levels of health competence do not ne cessarily predict positive psychological well being. In contrast, the model examining the relationship between health competence and negative psychological well being for individuals with health fears did fit the data well. The effect of Health Self Effi cacy on negative psychological well being was sign ificant for the health sample. Lower Health Self Efficacy was associated with higher levels of negative psychological well being. This result appears to be driven by the younger adults. When examining th e individual subscales of Health Self Efficacy, younger adults reported lower levels of efficacy with regards to both nutrition and responsible health practices. Y ounger adults felt less capable of eating well or taking care of themselves than middle aged and older adults Younger adults in our health sample also report ed greater levels of depression and negative

PAGE 103

103 affect, and lower levels of autonomy and environmental mastery. Thus younger adults with most important health fears are less efficacious and r eport lower levels of positive well being and higher levels of negative well being compared to our middle aged and older adults This is true despite the fact that older adults report the lowest level o f physical health overall. These result s indicate th at there is something diffe rent about individuals who report health fears. It may be that individuals with higher levels of negative well being experience greater health problems and are more likely to be pessimistic about the future which results in fe ars regardless of age Hope s and Fears: Do They Matter? Interestingly, t he results for hopes and fears were similar across domains. Further, t he se results indicate that qualitative differences exist between framing a possible self as a hope or framing as possible self as a fe ar a lthough we have yet to fully understand what those differences are It may be that the psychological impact of moving toward a desired self and moving away from an undesired self are very different. Again, previous research sugg ests that hopes and fears may influence motivation differently; however we did not assess motivation in the current study. Another possibility is that personality traits may impact whether an individual frames something as a hope or a fear. For example i ndividuals scoring higher on neuroticism may be more likely to report fears rather than hopes. This difference may also be related to a their future. Anoth er possibility is that current levels of psychological well being and competence influence whether a self is framed as a hope or a fear. The current study supports this notion Future studies should work to identify specific factors that drive people to select specific hopes or fears, or to frame their personal issues as a hope rather than a fear. This aspect of possible selves research remains relatively unexplored.

PAGE 104

104 Paradox of Well Being Our approach to understanding the relationship between our most im portant selve s competence and well being may help address issues related to the paradox of well being in late life. The paradox of well being acknowledges that even in light of multiple age related declines across multiple domains like health and cognit ion, older and middle aged adults consistently report equivalent or greater levels of well being compared to younger adults (Carstensen, et al., 2000). Consistent with this, older adults in the current study reported fewer possible selves, reported experi encing greater physical problems and limitations, as well as greater pain. Yet in light of this, older adults reported higher levels of positive well being and lower levels of negative well being than younger adults. Because older adults report fewer pos sible selves than younger adults, they are most likely more invested in the selves they do report. As a result older adults may be reporting better psychological well being in part because they also report high levels of efficacy in domains of h igh person al importance. Because these domains are already tied to the c urrent identity of older adults, they serve to reinforce positive self definitions. Brandtstadter (1989) has described coping mechanisms through which older adults maintain well being in the fa ce of age related decline. When striving toward personally meaningful goals, older adults maintain feelings of competence and well being by assimilative processes like adjusting their efforts or by accommodative processes like redefining their goals and w hat it means to be successful (Brandtstadter, 1989; Brandtstadter, Rothermund, & Schmitz, 1998). Research has shown that assimilation and accommodation increase in the later half of life ( Brandtstadter and Baltes Gotz, 1990 ; Rothermund & Brandtstadter, 20 03) and that h igher levels of assimilativ e and accommodative coping are associated with lower levels of depression and social dysfunction (Boerner, 2004) Although the current study examined how perceptions of

PAGE 105

105 competence related to highly valued future ho pes and fears contributed to well being, we did not assess whether an individual had redefined competence or adjusted his or her possible self over time. Examining when and how individuals engage in assimilation and accommodation activities in the cogniti ve and health domains will be an exciting next step in this program of research. Limitations A main goal of this study was to understand the consequences of high and low levels of competence in domains of central importance to the self concept. Unfortunat ely we were not able to address this issue entirely. Our sample had high levels of competence for both cognition and health. This lack of variability is a limitation. As a result we were not able to test what the model would have predicted for individua ls with low levels of competence in the current study. It may be that individuals, who value cognition or health but begin to experience declines in competen ce, revise or drop these selves and as a result were not included in our sample. A longitudinal s tudy would be necessary to examine this possibility. Similarly, our younger, middle aged, and older adults all reported high levels of memory efficacy, contrary to expectations. It may be that our measure of memory efficacy was too brief to capture the va riability within our age groups. Initially this measure was designed to be a part of a composite that included measures from the Metamemory in Adulthood scale and was not intended to be used as an individual predictor. In the future, a more comprehensive measure such as the Memory Self Efficacy Questionnaire (MSEQ) developed by West and Berry ( Berry, West, & Dennehy, 1989 ) may be more appropriate. The cross sectional nature of the current study limits the conclusions we can currently draw. Although we examined younger, middle aged, and older adults, we cannot make any statements about when cognitive or health selves begin to emerge. A longitudinal study of

PAGE 106

106 possible selves, competence, and psychological well being would allow us to understand how possib le selves and feelings of competence change over time and how these changes impact psychological well being and behavior in domains of personal importance. Last, our method of data collection using surveys that participants completed at home, on their ow n, may have limited the amount of data collected on the open ended portion of the possible selves questionnaire. Most of the previous research on possible selves has been conducted using individual or group interview sessions This was done because the q uestionnaire could potentially be confusing to older adults (Hooker, 1992; Hooker & Kaus, 1992; Hooker, 1999). However, our results are consistent with those obtained from individual and group interviews, suggesting that this methodological change was not problematic. For example, on average, 86% of the older participants in these prior studies reported health related selves. Consistent with these previous results, in the current sample, 83% of our older participants reported a health related self. Alth ough it is true that we may have received more detailed descriptions of possible selves in a one on one setting, participants in the current study reported a diverse and rich array of future selves consistent with previous research. Conclusion The relation ship between competence and psychological well being is complex. This has been the first study to examine the relationship between most important possible selves, domain specific competence and psychological well being. Although age differences in the co ntent of possible selves have been examined extensively, the relationship between possible selves, competence and psychological well being has been taken for granted. The current study found that competence does predict psychological well being in the hea lth domain. With a more sensitive measure, we believe that competence will predict psychological well being in the cognitive domain as well. Testing models within different domains of personal importance such

PAGE 107

107 as possible selves and competence related to careers and social relationships would add further support to the overall model. Further, the relationship between competence and psychological well being appears to be different for hopes and fears. The combination of fears within self defined important domains, low competence in those domains, and negative well being may be important in identifying individuals who may need psychological intervention. T he goal of future investigations will be to uncover the importance and utility of possible selves by u nderstanding the relationship between our most important hopes and fears, our feelings of competence related to achieving or avoiding them, our psychological well being and ultimately our behavior.

PAGE 108

108 LIST OF REFERENCES American Psychiatric Association. (1 994). Diagnostic and statistical manual of mental disorders 4th edition. Washington, DC: Americ an Psychiatric Association Bandura, A. (1997). Self efficacy: The exercise of control New York: W. H. Freeman and Company. Becker, H., Stuifbergen, A., Oh, H. S., & Hall, S. (1993). Self Rated Abilities for Health Practices: A health self efficacy measure. The Journal of Health Behavior, Education, and Promotion, 17 42 50. Bentler, P. M. (1985). Theory and implementation of EQS: A structural equations progra m Los Angeles, CA: BMDP Statistical Software. Berenbaum, H., Raghavan, C., Le, H., Vernon, L., &Gomez, J. (1999). Disturbances in Emotion. In D. Kahneman, E. Diener, & N. Schwarz (Eds.) Well being: The Foundations of Hedonic Psychology New York: Russel l Sage Foundation. 267 287. Berry, J.M. & West, R.L. (1993). Cognitive self efficacy in relation to personal mastery and goal setting across the life span. International Journal of Behavioral Development, 16 351 379. Be rry J.M., West, R.L. and Denneh y, D .M. (1989). Reliability and validity of the memory self efficacy questionnaire. Developmental Psychology, 25 701 713. Boerner, K. (2004). Adaptation to disability among middle aged and older adults: The role of assimilative and accommodative coping. Jo urnal of Gerontology 59B, P35 P42. Brandstadter, J. (1989). Personal self regulation of development: Cross sequential analyses of development related control beliefs and emotions. Developmental Psychology, 25 96 108. Brandtstadter, J., & Baltes Gotz, B (1990). Personal control over development and quality of life perspectives in adulthood. In P.B. Baltes and M.M. Baltes (eds.): Successful aging: Perspectives from the behavioral sciences Cambridge, England: Cambridge University Press. Brandstadter, J. Rothermund, K., & Schmitz, U. (1998). Maintaining self integrity and self efficacy through adulthood and later life: The adaptive functions of assimilative persistence and accommodative flexibility. In J. Heckhausen & C. Dweck (Eds.), Motivation and s elf regulation across the life span (pp. 365 388). New York: Cambridge University Press. Brandt, J., Spencer, M., & Folstein, M. (1988). The Telephone Interview for Cognitive Status. Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 1 111 117. C acioppo, J. T., Petty, R. E., & Kao, C. F. (1984). The efficient assessment of need for cognition. Journal of Personality Assessment, 48 306 307.

PAGE 109

109 Cameron, J.E. (1999). Social identity and the pursuit of possible selves: Implications for the psychological well being of university students. Group Dynamics: Theory, Research, and Practice, 3 179 189. Carstensen, L. L., Pasupathi, M., Mayr, U., Nesselroade, J. R. (2000). Emotion experience in the daily lives of older and younger adults. Journal of Personality and Social Psychology 79 ,644 655. Cavanaugh, J. C., Feldman, J. M., and Hertzog, C. (1998). Memory beliefs as social cognition: A reconceptualization of what memory questionnaires assess. Review of General Psychology, 2 48 65. Cohen, J. (1988). Statist ical power analysis for the behavior sciences (2 nd ed.). Hillsdale, NJ: Erlbaum Cross, S., & Markus, H. (1991). Possible selves across the life span. Human Development, 34 230 255. Cross, S.E. & Markus, H.R. (1994). Self schemas, possible selves, and competent performance. Journal of Educational Psychology, 86 423 438. Dark Freudeman, A., West, R. L., & Viverito, K. M. (2006). Future selves and Aging: Older Educational Gerontology, 32 85 109. Diehl, M., Hastings, C. T., & Sta nton, J. M. (2001). Self concept differentiation across the adult life span. Psychology and Aging, 16 643 654. Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction With Life Scale. Journal of Personality Assessment, 49 71 75 Dixon, R. A., Hultsch, D. F., & Hertzog, C. (1988). The Metamemory in Adulthood (MIA) Questionnaire. Psychopharmacology Bulletin, 24 671 688. Frazier, L. D., Cotrell, V., Hooker, K. (2003). Possible selves and illness: A comparison of individuals with adults. International Journal of Behavioral Development, 27 1 11. Frazier, L.D., Hooker, K., Johnson, P.M., & Kaus, C.R. (2000). Continuity and change in possible selves in later life : A 5 year longitudinal study. Basic and Applied Social Psychology, 2 237 243. Heckhausen, J., Dixon, R. A., & Baltes, P. B. (1989). Gains and losses in development throughout adulthood as perceived by different adult age groups. Developmental Psychology, 25, 109 121. Hooker, K. (1992). Possible selves and perceived health in older adults and college students. Journal of Gerontology, 47 P85 95.

PAGE 110

110 Hooker, K. (1999). Possible selves in adulthood: Incorporating teleonomic relevance into studies of the self. In T. M. Hess & F. Blanchard Fields (Eds.) Social cognition and aging (pp. 97 122). San Diego, CA: Academic Press. Hooker, K. & Kaus, C.R. (1992). Possible selves and health behaviors in later life. Journal of Aging and Health, 4 390 411. Hooker, K. & Kaus, C.R. (1994). Health related possible selves in young and middle adulthood. Psychology and Aging, 9, 126 133. Hultsch, D. F., Hertzog, C., & Dixon, R. A. (1987). Age differences in metamemory: Resolving the inconsistencies. Canadian Journal of Psychology, 41, 193 208. Hultsch D. F.,Hertzog, C., Dixon, R. A., & Davidson, H. (1988). Memory self knowledge and self efficacy in the aged. In M. L. Howe & C. J. Brainerd (Eds.), Cognitive development in adulthood: Progress in cognitive development research (pp. 65 92). New York: Springer Verlag. Lachman, M. E., Bandura, M., Weaver, S. L., & Elliott, E. (1995) Assessing memory control beliefs: The Memory Controllability Inventory. Aging & Cognition, 2, 67 84. nd control beliefs regarding memory and again: Separating general from personal beliefs. Aging, Neuropsychology & Cognition, 5, 264 296. Markus, H. & Nurius P. (1986). Possible selves. American Psychologist, 41 954 969. Radloff, L. S. (1977). The CES D sc ale: A self report depression scale for research in the general population. Applied Psychological Measurement, 1 385 401. Rosenberg, M. (1965). Society and the adolescent self image Princeton, NJ: Princeton University Press. Rothermund, K., & Brandtsta dter, J. (2003). Coping with deficits and losses in later life: From compensatory action to accommodation. Psychology and Aging, 18, 896 905. Ryan, E. B. (1992). Beliefs about memory changes across the adult lifespan. Journals of Gerontology, 47, P41 P46 Ryff, C. D. (1989). Happiness is everything, or is it? Explorations on the meaning of psychological well being. Journal of Personality and Social Psychology, 57 1069 1081. Ryff, C. D., Love, G. D., Urry, H. L., Muller, D., Rosenkranz, M. A., Friedman, E M., Davidson, R. J., & Singer, B. (2006). Psychological well being and ill being: Do they have distinct or mirrored biological correlates? Psychotherapy and Psych o somatics, 75 85 95.

PAGE 111

111 Skinner, E. A. (1995). Perceived control, motivation, and coping. Thousand Oaks, CA: Sage publications. Smith, J. & Freund, A. M. (2002). The dynamics of possible selves in old age. Journal of Gerontology, 57B P492 P500. Spielberger, C. D. (1983). State Trait Anxiety Inventory for Adults Palo, Alto, CA: Mind Garden. S pielberger, C. D., & Sydeman, S. J. (1994). State Trait Anxiety Inventory and State Trait Anger Expression Inventory. In M. E. Maruish (Ed.), The use of psychological testing for treatment planning and outcome assessment (pp. 292 321). Hillsdale, NJ: L. Erlbaum Associates. Stevens, J. (1986). Applied multivariate statistics for the social sciences Hillsdale, NJ: Lawrence Erlbaum Associates. Tun, P. A., & Lachman, M. E. (2004). Test of Adult Cognition by Telephone (TACT) Technical Report Brandeis Univ ersity. Updegraff, J. A, Gable, S. L., & Taylor, S. E. (2004). What makes experiences satisfying? The interaction of approach avoidance motivations and emotions in well being. Journal of Personality and Social Psychology, 86 496 504. Wallston, Wallston, & DeVellis, ( 1978 ). Development of the Multidimensional Health Locus of Control (MHLC) s cales. Health Education and Behavior 6, 160 170 Ware, J. E. & Sherbourne, C. D. (1992). The MOS 36 item short form health survey (SF 36). Medical Care, 30(6), 473 483 Ware, J. E., Snow, K. K., Kosinski, M., & Gandek, B. (1993). SF 36 Health Survey: Manual and interpretation guide. Boston: The Health Institute, New England Medical Center. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54 1063 1070. West, R. L., & Yassuda, M. S. (2004). Aging and memory control beliefs: Performance in relation to goal setting and memory self e valuation. Journals of Gerontology, 59B P56 P65. Westerhof, G. J., Katzko, M. W., Dittmann Kohli, F., & Hayslip, B. (2001). Life contexts and health related selves in old age: Perspectives from the United States, India, and Congo/Zaire. Journal of Agin g Studies, 15 105 126.

PAGE 112

112 BIOGRAPHICAL SKETCH Alissa Dark Fre udeman was born in Akron, Ohio She attended Saint Thomas Aquinas High School in Ft. Lauderdale, Florida. She went on to attend the University of Florida and received a Bachelor of Arts in li nguistics with a minor in business administration in 1998. She graduated with highest honors after completing a senior honors thesis on sexism and violence in returned to the University of Florida as a graduate student in developmental psychology. She received her m aster s degree in developmental psychology with a minor in gerontology from the University of Florida in 2004