ADOLESCENT ACTION-TAKING: ASSOCIATIONS WITH IDENTITY STYLE, POSSIBLE SELVES, AND PARENTAL SUPPORT Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. _________________________________________________ Sarah A. Swart Certificate of Approval: _________________________________ _________________________________ Leanne K. Lamke Jennifer L. Kerpelman, Chair Professor Associate Professor and Human Development and Family Studies Extension Specialist Human Development and Family Studies _________________________________ __________________________________ Gregory S. Pettit Stephen L. McFarland Alumni Professor Acting Dean Human Development and Family Studies Graduate School ADOLESCENT ACTION-TAKING: ASSOCIATIONS WITH IDENTITY STYLE, POSSIBLE SELVES, AND PARENTAL SUPPORT Sarah Swart A Thesis Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, Alabama May 11, 2006 iii ADOLESCENT ACTION-TAKING: ASSOCIATIONS WITH IDENTITY STYLE, POSSIBLE SELVES, AND PARENTAL SUPPORT Sarah Swart Permission is granted to Auburn University to make copies of this thesis at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. ______________________________ Signature of Author ______________________________ Date of Graduation iv VITA Sarah Anne (McCullough) Swart is the daughter of Larry and Elaine McCullough, Lawrence, Kansas. She was born in Abilene, Kansas on September 8, 1980. Sarah graduated from Lawrence High School in 1999 and spent four years in Abilene, Texas where, in May 2003, she graduated summa cum laude, receiving a Bachelor of Science degree in Psychology, with a minor in Biblical Text, from Abilene Christian University. In August 2003, Sarah entered Graduate School at Auburn University. On July 24, 2005 she married Dustin Wayne Swart. v THESIS ABSTRACT ADOLESCENT ACTION-TAKING: ASSOCIATIONS WITH IDENTITY STYLE, POSSIBLE SELVES, AND PARENTAL SUPPORT Sarah Swart Master of Science, May 16, 2006 (B.S., Abilene Christian University, 2003) 131 Typed Pages Directed by Jennifer L. Kerpelman Action-taking is the intentional behavior in which adolescents engage. Productive action-taking promotes positive youth outcomes and future goal-achievement. The aims of the current study were to examine variation in adolescent action-taking and to explore factors associated with action-taking. Two ways of looking at action-taking were examined. Goal-directed behaviors are actions chosen by adolescents that are in intentional and goal-oriented. Weekly time-use reveals patterns of activity choices of adolescents. Three factors that may be associated with action-taking were explored. Identity style represents the way in which adolescents manage the exploration and commitment of identity issues. Possible selves are hoped for future selves that serve as goals to be attained. Perceived parental support for adolescent? action-taking are the adolescents? experiences of their parents? encouragement and facilitation of action taking. vi Participants were 25 adolescents, ages 15 and 16, and one or both of their parents (parental figures). Participants completed a q-sort about current identity, a q-sort about possible selves, an interview addressing goal-directed behaviors and parental support, and a time-sort assessing weekly time-use. Results indicated that there was variation in adolescent action-taking. Examination of action-taking and the factors expected to be related to it revealed that more complex and diverse goal-directed behaviors were associated with informational identity style (i.e., high exploration and high commitment to identity issues) and maternal support. Particular time-use activities also were associated with identity style and possible selves. Finally, there were differences in the identity style and parental support of high and low action-takers. High action-takers reported a strong use of the informational identity style, specific possible selves goals, and more detailed parental support. Low action-takers, however, showed weak associations with informational identity style, and described vague possible selves and parental support. Findings add to the literature on action-taking, identity formation, and parental support. vii ACKNOWLEDGMENTS The author would like to thank Jennifer Kerpelman for her endless help, support, and encouragement during the writing of this thesis. Thanks also are due to Greg Pettit for the opportunities and assistance he has given me, and to Leanne Lamke for her patience and support throughout my time in Auburn. Special thanks are given to my family. Thanks to my parents, brothers, and Katie for their advice and for giving me shoulders to lean on. Thank you also to the Swarts for all your encouragement. Finally, to my husband Dustin, I owe much appreciation and gratitude. Without you, this process would have been very lonely. viii Style manual or journal used: American Psychological Association, Fifth Edition Computer software used: Microsoft Word 2000 for Windows; SPSS 12.0 for Windows ix TABLE OF CONTENTS LIST OF TABLES????????????????????????....... x INTRODUCTION?????????????????????????? 1 Purpose of Study???????????????????????.. 9 REVIEW OF LITERATURE .???????????????????..?.. 11 Overview of Action and Action Theories?????????????? 11 Goal-directed Behaviors?.??????????????????...... 13 Time-Use??????????????????????????.. 16 Identity Formation??????????????????????.. 20 Parental Support???????????????????????... 33 METHOD?????????????????????????????. 38 RESULTS ????????????????????????????... 50 DISCUSSION???????????????????????????. 77 Conclusions, Limitations, and Future Directions????..??????. 89 REFERENCES??????????????????????????? 93 APPENDICES???????????????????????????. 100 Appendix A?????????????????????????.. 101 Appendix B?????????????????????????.. 112 x LIST OF TABLES TABLE 1. Factor Loadings for Group 1 and Group 2 43 TABLE 2. Means, Standard Deviations, and Ranges for Goal-directed Behavior Variables 52 TABLE 3. Means, Standard Deviations, and Ranges for Time-use Variables 53 TABLE 4. Means, Standard Deviations, and Ranges for Identity Formation and Parental Support Variables 57 TABLE 5. Most and Least Like Possible Selves Items for Q-groups 1 and 2 58-59 TABLE 6. Adolescent and Parent Congruence for GDB and Possible Selves 61 TABLE 7. Associations between GDB and Identity Style Variables 62 TABLE 8. Associations between GDB and Possible Selves 64 TABLE 9. Associations between GDB and Parental Support 65 TABLE 10. Associations between GDB and Q-groups 66 TABLE 11. Associations between Weekly Time-use Totals and Identity Formation Variables 67 TABLE 12. Associations between Weekly Time-use and Identity Formation Variables 68 TABLE 13. Descriptives of High and Low Action-taking from Adolescents? Perspective 70 TABLE 14. Descriptives of High and Low Action-taking from Parents? Perspective 71 1 INTRODUCTION Whether they realize it or not, adolescents? decisions about their daily activities have important implications for their futures. The intentional activity choices that adolescents make can be referred to as actions or action-taking (Brandtstadter, 1998; Marshall, Young, & Domene, in press). The more adolescents participate in productive action-taking (i.e., behaviors that promote positive outcomes), the better positioned they are to achieve their future goals, and the more likely they are to overcome obstacles that may emerge (Gollwitzer, 1996; Gonzales, Burgess, & Mobilio, 2001). However, those who do not engage in productive action-taking may find it more difficult to reach their goals. For this reason, the current exploratory investigation will examine action-taking in adolescence. The overall aims of this study are to examine variation in adolescent action- taking and to explore factors that may be associated with productive action-taking, and therefore, goal achievement. Findings are expected to enhance understanding of why some adolescents are more engaged in action-taking than are others. Action-taking in Adolescence Action, as described in numerous theories (Brandstatdter, 1998) can be considered a process made up of numerous parts. Theories of action define the motivation for, structure of, and contextual circumstances influencing action in different ways. For example, Brandstatdter?s theory of action describes how action-taking behaviors, personal characteristics, and cultural influences work together and influence one another. 2 These theories describe both conceptual aspects of action (e.g., social significance, planning, and personal interpretations of behaviors), as well as structural aspects (e.g., goals, actors, and action steps). Although many differences exist between them, theories of action conclude that people are intentional in their day-to-day behaviors, objectives, and activities (Brandtstadter, 1998; Marshall, et al., in press). For example, von Cranach, Kalbermatten, Indermuhle, and Gugler (1982) define action as ?conscious, directed towards a goal, planned and intentional.? The intentional behaviors performed as part of action are referred to as action-taking (or actions). Action-taking is performed on a daily basis in order to reach short-term goals. In addition, people participate in action-taking while striving to reach their long-term goals (e.g., working for pay to save money for college). Specifically in adolescence, individuals are faced with role transitions, increasing autonomy, and greater demands for responsibility taking. These factors are central in adolescence and are associated with action-taking (Brandtstadter, 1998). During this time, many adolescents are considering their long-term goals and engaging in behaviors and activities that move them forward. Two ways of looking at action-taking are through adolescents? goal-directed behaviors (von Cranach, et al., 1982) and weekly time-use (Larson, 2001). Exploring the purposeful actions in which adolescents engage in order to reach their goals, and the amount of time adolescents spend in certain activities may help in understanding processes that lead to goal achievement. Goal-directed behaviors (GDB). In adolescence, youth have ambitions and goals for education, relationships, careers, and success. However, it is not enough only to have goals; there also must be behaviors that direct actions toward goals (Oyserman, Bybee, 3 Terry, & Hart-Johnson, 2004). Therefore, the purposeful actions in which individuals engage that enable them to reach the goals they set for themselves can be conceptualized as goal-directed behaviors (GDB). In attempting to reach long-term goals, issues such as being easily distracted or foreseeing failure may present problems for goal achievement. GDB serve to make attaining goals possible and may function as a buffer to these volitional problems. GDB serve a function in goal achievement by guiding individuals toward their long-term goals through completion of short-term goals, keeping them aware of opportunities and means of reaching their goals, and by serving as deterrents to volitional problems. In other words, when behaviors are directed toward reaching a goal, they are more likely to be carried out, and the goal is more likely to be realized (Gollwitzer, 1996). The complexity of GDB is an indicator of how well the action will work in reaching goals. Higher GDB complexity promotes goal achievement (Oyserman, et al., 2004). Complexity includes having set goals to which one is striving (e.g., ?I want to be a successful business man? versus ?I?d like to have a good job someday?) as well as having set GDB (e.g., ?I am trying extra hard in economics class? versus ?I will try hard in school?). More complex GDB have been linked to higher participation in school, more time spent in homework, higher GPA and lower attendance in summer school (Oyserman et al., 2004). Weekly time-use. An alternative way of looking at action-taking is by examining the ways in which adolescents typically spend their time. Although it is a more indirect way of assessing action-taking than examining GDB, adolescents? time-use patterns during a typical week offer additional insights about adolescent action-taking that affect 4 subsequent goal achievement. Productive activities such as studying or engaging in extracurricular activities are more beneficial than activities such as playing computer or video games (Gilman, Meyers, & Perez, 2004). Productive discretionary time (i.e., typically non-school time), as well as activities that are supervised by a parent or adult, have been found to be related to a number of positive adolescent outcomes in areas of academic performance, skill building, and psychological functioning (Mahoney, Larson, Eccles, & Lord, 2005; Osgood, Anderson, & Shaffer, 2005) and therefore, promote goal achievement. Conversely, large amounts of unproductive activities, such as excessive TV watching or unsupervised time with friends have been associated with academic and behavior problems (Larson, 2001). Therefore, participating in productive and supervised activities, which provide opportunities for skill building and alternatives to delinquent behaviors, enhance adolescents? preparation for the future by helping to achieve short- term goals while focusing on, and preparing for, long-term goals. Factors Associated with Action-taking Identity Formation During adolescence, identity formation becomes a central task. According to Erikson (1968), identity helps the adolescent interpret a variety of experiences, and provides guidance for action. Berzonsky?s (1990) more recent conceptualization of identity style, where style is defined as the approach an adolescent takes to gathering information about identity, offers a way to associate the process of identity formation with adolescent action-taking. In addition, the way adolescents envision their possible selves, or who they will become in the future (Markus & Nurius, 1986), also may help to guide action-taking in the present. 5 Identity style. To understand adolescent action-taking, a focus on identity development, a crucial task during this developmental period, is needed (Erikson, 1968). Erikson described the process of identity formation as occurring through the life stage of identity versus identity confusion. When the individual reaches adolescence, he or she goes through what Erikson termed a ?crisis.? This crisis marks the time when an adolescent searches for beliefs and direction concerning self-definition (Schwartz, 2001). At some point, the adolescent commits to a certain set of beliefs. Because Erikson never developed his theory into factors that could be tested and measured, others since have created more testable definitions for these concepts. Marcia (1966), for example, developed the idea of four identity statuses that vary between different levels of crisis (i.e. exploration) and commitment. Berzonsky (1990) further defined these identity statuses into three identity processing styles. First, the informational style involves actively seeking out information about one?s surroundings and ambitions in order to form identity. Normative style is exhibited when adolescents conform to what respected authority figures tell them about themselves. Finally, the diffuse style is marked by the lack of active information seeking, with regard to identity, due to an absence of crisis or an active avoidance of seeking identity information. Although adolescents use all three identity styles to a degree, typically one is the primary style and therefore, guides action. Most theorists agree that an information seeking style is most conducive to goal setting and attainment (Schwartz, 2001). Identity style would be expected to have an association with action-taking in adolescence. For example, an adolescent who primarily uses an informational identity style, and whose goal is to further her education past high school, would be more likely to 6 engage in more academically-focused GDB, such as studying, investigating scholarships, seeking out advise from her school counselor or spending time preparing for the SATs, than an adolescent who primarily uses a diffuse identity style. In addition, the adolescent using an informational style is more likely to have weekly patterns of time-use that reflect higher engagement in productive activities than does the diffuse adolescent. At the same time, adolescents who engage in productive time-use have the opportunity to explore and gain self-knowledge (Dworkin, Larson, & Hansen, 2003). Thus, the informational style may help to determine adolescent action-taking in the present that supports her goals and preparation for the future. Likewise, these activities help to shape identity. Because the options the normative oriented adolescent considers are guided by respected authority figures and, therefore, more limited than those considered by the information oriented adolescent, use of the normative style would be expected to show a lower level of GDB complexity than the informational style. The normative style also is expected to be associated with a pattern of weekly time-use that includes engaging in more supervised activities. Identity style, therefore, is expected to be related to action-taking behaviors in terms of GDB complexity and the extent to which weekly time-use is productive and supervised. Possible selves. Another aspect of identity formation is how adolescents envision their possible selves. Although theories about future outlook have focused on self- concept, dreams, or future orientation, most do not specifically explore the interrelation of the current self-concept and the future self-concept. Possible selves are described by Markus and Nurius (1986) as one?s expected, hoped for, or feared future selves. These 7 expectations originate in how people have behaved in, and understood, their pasts, as well as how they see their goals for the future (Markus & Nurius, 1986). Possible selves can be thought of as goals that one is striving to obtain. These goals can be classified by the types of capital needed to attain them (Coleman, 1988). Capital is the collective resources a person attains that further the ability to reach future goals. Three types of capital are human, economic, and social. Human capital can be described as resources that enhance a person?s talents, skills and abilities such as knowledge, athletic ability, or interpersonal skills. Human capital possible selves would include having a high level of education, having a career of high respect, and being a person who has developed a particular talent or skill. Economic capital is comprised of the material or financially related resources a person possesses as well as how they are used. Economic capital possible selves would include being a person who has savings in the bank, being the owner of material possessions such as cars or houses, and being someone who donates money to charity. Social capital includes the interpersonal resources a person has, such as a supportive group of friends, a large extended family, or mentors within the community. Social capital possible selves would include being someone with supportive parents, being a member of community organizations, and being an involved parent. Adolescents typically envision their futures as containing each type of capital, however, the amount of focus on each type varies from person to person. Possible selves can be related to action-taking by analyzing what adolescents focus on in their possible selves and whether their actions, through GDB and time-use, reflect the degree to which they emphasize human, economic and social capital. For example, an adolescent who is most focused on human capital possible selves, would be 8 expected to spend significant amounts of time each week building her skills in areas such as academics, athletics, or creative arts, as they relate to her human capital goals. Parental Support Parents also play an important part in the weekly activities of their adolescents, and the extent to which adolescents engage in GDB (Brandtstadter, 1998). Positive adolescent outcomes that may affect action-taking and goal achievement (e.g., higher self-esteem, higher grades, and lower drug use) are associated with such parental supports as success in balancing individual freedom permitted to their adolescents during individuation (i.e., autonomy-gaining) and continued monitoring and guidance (Aquilino & Supple, 2001). Parental support (e.g., praise, encouragement, affection, warmth) has been positively associated with adolescent academic achievement and autonomy and negatively associated with deviant behaviors (Aquilino & Supple, 2001). Therefore, it is important to examine the role of parental support in adolescent action-taking behaviors. Parents may or may not offer support for the adolescent?s actions, however, and adolescents may differ in their action-taking behaviors according to whether they feel encouraged or not in their efforts. In summary, adolescence is an important developmental stage in which many action-taking behaviors take place. Both GDB and time-use can be studied to understand the actions in which adolescents engage that have implications for their future goal achievement. Because identity styles determine how adolescents decide to deal with information available to them, they are expected to be related to GDB, as well as the ways adolescents typically decide to spend their weekly time. Possible selves are goals an adolescent hopes and expects to achieve in terms of who they are becoming. Adolescents? 9 GDB are expected to be shaped, in part, by the extent to which their possible selves reflect human, economic and social capital. Weekly time-use also may be associated with possible selves in that adolescents will spend greater amounts of time in activities that help to build the kinds of capital they need to realize their desired possible selves. Finally, adolescents? perceptions of parental support for their future goals would be expected to have an association with the adolescents? action-taking. Purpose The purpose of the current study was to investigate adolescent action-taking. Variation in adolescent action-taking and identity formation from the perspectives of adolescents and their parents was examined, as was variation in the adolescents? perceptions of parental support. Associations between adolescent action-taking and the factors assumed to be associated with action-taking also were explored. Finally, parents? perceptions of GDB, identity styles and possible selves were explored to enhance the understanding of adolescent action-taking. It was expected that there would be a positive relationship between informational identity style and GDB. Normative identity style also was expected to have a positive relationship with GDB; however, the association was not expected to be as strong as that found for informational style given that the normative style is associated with considering a more narrow range of options that fit with expectations of respected others. The diffuse identity style was hypothesized to be associated with a lack of GDB. Possible selves, according to their emphasis on human, economic, and social capital, were expected to be most strongly related to GDB aimed at attaining human, economic or social capital, 10 respectively. Greater parental support, as perceived by the adolescent, was expected to be associated with greater adolescent action-taking in terms of greater GDB complexity. Associations between identity style and categories of weekly time-use also were explored. The informational and normative styles were expected to be positively associated with productive time-use, and the normative style was expected to have the strongest association with supervised time-use. Diffuse style was expected to be negatively associated with productive time-use. Different patterns of weekly time-use were expected according to the types of possible selves adolescents reported desiring. 11 REVIEW OF LITERATURE The review of literature first briefly addresses theories of action (i.e., a concept or process) and action-taking, or actions, (i.e., the choices or behaviors performed). Second, research on adolescent action-taking, through goal-directed behaviors and weekly time- use, is reviewed. Finally, literature that has examined associations among identity style, possible selves, parental support, and adolescent action-taking is discussed. Adolescent Action-taking Overview of Action and Action Theories It is important to understand action-taking in adolescence not only because this period marks such important developmental transitions as moving from parental dependence to relative independence, but also because action provides a venue for adolescents to be active in their own development (Lerner, Theokas, & Jelicic, 2005). During this period, many goals are being formed, adolescents? choices of goals and actions are greatly increased, and the decisions adolescents make affect their futures. In his review of action and human development, Brandtstadter (1998) discussed how action facilitates development and what forces facilitate action. This theory of action ties together concepts of action-taking, identity formation, and support for action-taking. Brandtst?dter defines actions as: behaviors that (a) can be predicted and explained with reference to intentional states (goals, values, beliefs, volitions); (b) are at least partly under personal 12 control, and have been selected from alternative behavioral options; (c) are constituted and constrained by social rules and conventions or by the subject?s representation of these contextual constraints; and (d) aim to transform situations in accordance with personal representations of desired future states? (p. 815). This definition specifies aspects related to action-taking that are worthy of examination in terms of an adolescent?s preparation for the future. First, actions are intentional behavior chosen over alternative options, and selected in reference to personal goals or values. Second, these options are defined and refined by the context of which they are a part. Therefore, the current study investigated action-taking in terms of goal directed behaviors (GDB) and time-use, as well as factors that facilitate choices of action (i.e., identity style and possible selves) and contextual factors (i.e., perceived parental support). Numerous action theories exist within the fields of biology, economics, philosophy, sociology and psychology, and within the field of psychology, several types of action theories exist. A control-system theory explains action as the use of discrepancy reducing feedback cycles. Through hierarchically organized levels, goals are transformed into behaviors, and activities are guided by these feedback loops. The social- constructivist theory of concepts of action is based on the work of Vygotsky. In this theoretical view, actions and thoughts are organized by mediating the individual?s perception of reality (i.e., cognitive structures) and reality (i.e., objects, tools, cultural symbols). Structural action theories focus on the cognitive operations guiding action through set stages of development, analyses of actions and skills themselves, as well as the persons or elements of action (i.e., actors, goals, instruments). Finally, motivational 13 theories of action focus on expectancy-value models, in which actions are guided and affected by the expectations and evaluation of personal goals and standards. Often, aspects of these theories overlap. In the current study, action is explored in terms of its? structural parts (i.e., characteristics of actors, GDB, time-use), as well as the motivation that guides and promotes it (i.e., possible selves, parental support). Taken together, these theories of action are based on the assumption that actions are intentional and purposeful. Action begins with a recognized goal and the actor uses personal ability or contextual factors to facilitate movement toward the goal. This assumption guides the current study. Adolescents? behaviors are intentional actions used to facilitate their development and affect their futures. Two general types of intentional behaviors that will be examined are the adolescents? choices about their activities (i.e., weekly time-use), and their engagement in GDB. Goal-directed Behaviors Overview of GDB. Theories of action and of planning often use these terms to represent multi-faceted systems. For example, action can be thought of as a set of hierarchical goals across time, where actors set and accomplish short-term goals in pursuit of long-term goals. (Marshall et al., in press). Likewise, planning has been viewed as the action of first deciding to make wishes and desires into goals, second creating plans for action toward these goals, third putting the plans into action in order to complete the goals, and fourth, evaluating the results (Gollwitzer, 1996). However, it is useful to look at the separate steps or components within these concepts. Action steps are one component of action described by von Cranach, et al. (1982). In their book, Goal- Directed Action, von Cranach et al., describe the intentional steps that lead actors to their 14 desired goals as action steps. In subsequent literature by other authors (Friedman, Kofsky Scholnick, & Cocking, 1987; Oyserman, et al., 2004; Young & Valach, 2004), the term goal-directed action, or goal-directed behaviors, has come to describe action steps as they relate to specific goals. In the current study, the term goal-directed behavior (GDB) will be used to describe action steps that are intentional and used to accomplish goals. Empirical examination of GDB. In a short-term longitudinal study Oyserman, et al., (2004) described the importance of specific strategies and GDB created by adolescents in order to reach their academic and other future goals. The researchers hypothesized that adolescents who had set or particular academic goals, including GDB for attaining them (e.g., ?I will study hard in order to pass my test,? ?I will pass each class in order to get into the next grade?), would be more likely to reach those goals and have more academic engagement than adolescents who did not have specific GDB or who simply had numerous academic goals but no GDB (e.g., ?I will be a good student,? ?I will be a high school graduate.?) One hundred sixty-eight eighth graders from low- income households were asked about their academic goals and the GDB they had for reaching their desired goals for the 9 th grade, through a series of open-ended, written prompts. The adolescents were instructed to write down specific goals and what GDB they were using to reach these goals. Responses were coded for the number of set (i.e., particular, not general) academic goals mentioned, as well as for the combination of number of set goals and GDB (i.e., complexity) used for attaining them. Each participant?s goals were separated into categories according to their content (e.g., school achievement, personality traits, health). Goals related to academic achievement were of central interest to the researchers, 15 and notably, they made up 82% of the total goals given by the whole population. Each interview was first coded for the number of set goals and GDB on a 6-point scale. The researchers also wanted to see whether a second coding system comprised only of the number of goals (as opposed to also looking at both) would have the same predictive power as a more complex coding system that included the number of set goals and the GDB related to them. The results showed that adolescents who could articulate GDB for reaching their goals had more academic success than those who simply expressed the desire to reach those goals. The results also suggested that the more complex coding strategy yielded information that was a better predictor of participation in class, time spent doing homework, grade point average, and referral to remedial summer school than was the strategy that simply counted goals. In their assessment of achievement strategies (i.e., processes associated with GDB such as task-avoidance, planning, and active efforts to handle stressful situations) applied by young adults in the transition from school to the work force, M??tt?, Nurmi, and Majava (2002) showed a relationship between GDB, goal achievement, and future success. The participants were 250 (129 male and 121 female) Finnish young adults, who were graduating from college and entering the work force. Each participant was assessed for the GDB and achievement strategies he or she used four months prior to graduation (Time 1). Participants were assessed on the same scales four months after their graduations (Time 2), as well as about their work status at that time and then again 10 months following graduation (Time 3). Results showed that use of passive strategies at Time 1 (e.g., procrastination of GDB), were associated with unemployed at Time 3. Also, participants who showed high 16 levels of planning (i.e., constructing and implementing GDB) at Time 2 were more likely to be employed and less likely to still be a student at Time 3. Therefore, the use of productive GDB was associated with positive outcomes, whereas the lack of GDB was associated with negative outcomes. In summary, GDB are important parts of action that actors use to move toward their goals. Studies have shown that having and using GDB is positively associated with positive outcomes such as academic success and employment. In addition, assessing GDB complexity (i.e., GDB and the goals to which they are intended) aids in fully assessing this part of action-taking. Time-use In adolescence, individuals gain more freedom to make decisions about where and how they spend their time. American youth spend approximately 40-50% of their waking hours in discretionary activities compared to 35-45% in Europe and 25-35% in East Asia (Larson, 2001). The effects of this time-use are determined by how adolescents decide to spend this time. Productive and unproductive time-use in general, and specifically the difference in the activities in which adolescents spend time (e.g., in sports, paid work, homework), are associated with different outcomes. Furthermore, whether adolescents engage in activities alone, with friends, or supervised by parents or adults, gives insight into action-taking, and ultimately, goal achievement. Overview of time-use. Adolescents? time-use has been the focus of much research in recent years. Researchers have studied the effects of unproductive leisure time activities (e.g., watching TV, playing games, surfing the internet), and productive leisure activities (e.g., reading for pleasure, exercising, school-related clubs, scouts, sports). 17 Several links between adolescent well-being and time-use have been suggested. For example, time spent productively by adolescents in school and community activities such as homework, reading, or 4H has been positively associated with knowledge, intelligence, belongingness, self-esteem, intrinsic motivation, academic achievement, and emotional well-being (Gilman, et al., 2004, Jacobs, Vernon, & Eccles, 2004; Larson, 2001). Likewise, most productive leisure activities, such as sports, music, art, or volunteering, have been found to be positively related to high rates of challenge, concentration, motivation, identity development, supportive relationships, initiative, and reduced rates of delinquency (Larson, 2001). Supervised adolescents are at lower risk for negative peer influence that may lead to delinquency and such activities as smoking and sex (Osgood, et al., 2005). Conversely, unproductive leisure time and time unsupervised, either alone or with friends, has been linked to negative adolescent well-being. Excessive TV viewing (i.e., 3- 4 hours a day) has been related to lack of physical exercise and lower school grades (Gilman et al., 2004, Larson, 2001). Furthermore, unsupervised time spent with friends has been shown to be associated with behavior problems and lower academic outcomes, and low or no involvement in productive extracurricular activities, has been linked to school dropout, substance use, and antisocial behavior (Gilman, et al., 2004, Jacobs et al. 2004, Larson, 2001). Factors associated with productive and supervised time-use, such as higher academic performance, initiative, and supportive relationships, promote positive adolescent development (Mahoney, et al., 2005). The skills and resources adolescents attain from participating in productive and supervised time-use activities may better 18 equip adolescents to attain their goals (Tepper, 2001). Conversely, the factors associated with participation in unproductive or unsupervised activities may create obstacles to adolescent goal achievement. Implications of time-use categories. Researchers have investigated specific categories of time-use and the similarities and differences in outcomes associated with them. Bartko and Eccles (2003) assessed the relationships between involvement in specific activities and adolescent outcomes including academic performance (i.e., GPA), problem behavior (e.g., lying, cheating, fighting), and psychological functioning (i.e., depressive symptoms, psychological resilience, self-esteem, and internalizing and externalizing behaviors). One thousand and four ethnically diverse 16 and 17 year olds were assessed. Adolescents were asked to indicate their involvement in activities using a range from less than once a month to usually everyday. Using cluster analysis, six groups of adolescents were identified according to the types of activities in which they spent their time: sports, schoolwork and clubs, volunteer work, paid work, high-involved, or uninvolved. Several differences between the groups were found for each of the outcome variables. Adolescents who focused their time on sports were more likely than the other groups to have problem behaviors, but were also more likely to have lower internalizing problems and higher psychological resilience. Adolescents who spent a majority of their time in schoolwork showed higher academic performance and psychological resilience, and lower problem behaviors and depressive symptoms. Those students who were uninvolved in productive activities were more likely to have behavior problems, depressive symptoms, internalizing, and externalizing problems and lower academic performance and grades than did any other groups of students. Students who spent much 19 of their time in volunteer work had lower depressive symptoms and those who worked extensively in paid employment had higher problem behaviors and externalizing behaviors. Lastly, students who were highly involved in several different kinds of activities were more likely to have higher academic performance and GPAs, higher psychological resilience, and lower problem behaviors and internalizing problems. These patterns of results suggest that different activity choices of adolescent are associated with different positive or negative outcomes. Being highly involved in productive activities, specifically in school-related activities, was positively associated with numerous positive outcomes. Being uninvolved was associated with negative outcomes and interestingly, paid work was associated with negative outcomes as well. Other research has offered implications for specific time-use categories. For instance, time spent in schoolwork has been associated with higher grades and academic scores (Tepper, 2001) and adolescents? focus on homework has been associated with the likelihood of attaining higher education after high school (Jordan & Nettles, 2000). Involvement in religious activities also has been associated with positive outcomes such as engagement in schoolwork and school clubs (Jordan & Nettles, 2000), greater self- esteem and academic persistence, and lower delinquency, problem behavior and substance abuse (Kleiber & Powell, 2005). Findings for other activities have been inconsistent. Household chores may be beneficial for learning specific skills but they do no offer much challenge or developmental content (Larson, 2001). Participation in sports has been related to positive social skill, self-esteem, and academic achievement (Pedersen & Seidman, 2005), but has also been related to delinquency and alcohol use (Eccles & Barber, 1999). Working for pay has also been associated with both positive and negative 20 outcomes. Jordan & Nettles (2000) found that paid work was positively associated with adolescents? perceptions of their life chances but negatively associated with math and science achievement. Furthermore, paid work in excess of 20 hours a week has been related to higher levels of delinquency and school misconduct. However, paid work may be beneficial to adolescents if it is linked to school or building work-related skills (Larson, 2001). Taken together, these studies show the importance of time-use decisions of adolescents. First, how adolescents spend their time affects many outcomes such as well- being, academic achievement, problem behavior, and preparation for the future. Second, productive time-use such as extracurricular activities and time spent in school-related activities are beneficial for adolescents. Finally, understanding the implications of different types of adolescent activities such as paid work, sports, and schoolwork, add to the understanding of action-taking behaviors. Factors Associated with Adolescent Action-taking Identity Formation Identity styles. The adolescent identity formation literature began with Erikson?s observations of normative child development in the 1950s. As Erikson studied and observed children, he developed a theory of the psychosocial stages of human development. One of these stages was identity versus identity diffusion, in which adolescents engage in differing levels of exploration, or crisis, and commitment pertaining to their self-definition. Focusing on the Eriksonian concepts of crisis and commitment, Marcia (1966) developed identity status classifications. Four combinations of self-exploration (e.g., 21 crisis, no crisis) and self-definition (e.g., committed, uncommitted) form identity statuses. Marcia defined Achievers as those who have resolved identity crises and have committed to a set of standards and values that make up the ?self.? Those who are seeking to resolve a crisis yet have made no resolutions or commitments are Moratoriums. Foreclosures have not gone through a state of crisis, yet have a firm commitment to a definition of ?self?, usually through the adoption of norms and standards of significant others. Those who are Diffuse have neither gone though an identity crisis, nor committed to a sense of self. It is often assumed that people go through a sequence of identity exploration starting with Diffuse, moving through Moratorium or Foreclosure and finally becoming Achieved. Marcia?s theory focused on identity statuses as outcomes. However, research has shown that over time, people often shift in identity statuses and can move though the sequence in unexpected progressions (Berzonsky, 1990). In order to understand these findings, Berzonsky offers a process model of identity formation that describes the differences in the way people receive, process, and represent information (Berzonsky, 1990). This allows for movement and shifts in identity statuses, while also defining the typical manner in which one deals with crises and commitment. Therefore, identity style refers to the processes used by individuals when faced with challenges, developmental tasks, or role transitions at different points in their lives. Informational style is used by those who actively seek to gather information, ask questions, and evaluate their decisions before making identity commitments. Normative style is the acceptance of the values and ideas of significant and respected others. Finally, procrastination and avoidance of both identity crises and identity development characterize the diffuse style. 22 Associations between GDB and identity styles. Limited research has looked at the relationship between identity styles and action-taking. In these studies, action-taking often takes the form of decisional or coping strategies, effective problem-solving, or planning when faced with self-relevant problems or stressors (Berzonsky, 1992). However, these constructs are similar to GDB in that they involve pre-decisional strategies and evaluation of actions. Likewise, striving to reach goals and possible selves can be thought of as stressors or self-relevant problems. Berzonsky has conducted several studies focusing on identity styles and coping or decisional strategies people use when confronted with problems or stressful tasks or events (Berzonsky, 1992; Berzonsky & Ferrari, 1996; Berzonsky, Nurmi, Kinney, & Tammi, 1999). Berzonsky, et al. (1999) evaluated how identity style was related to strategies people used when confronted with difficult tasks or stressful situations (i.e., success expectation, task-irrelevant behavior, reflective thinking or planning, task management, avoidance, and seeking social support.) The participants were 198 undergraduate college students (51 male, 147 female) who were predominately Caucasian. The researchers hypothesized that adolescents using an informational identity style would engage in more effective strategies and behaviors than those who used a diffuse identity style. Specifically, they hypothesized that those using an informational identity style would employ more task-relevant behaviors and adaptive strategies than those using a normative style and that those using a normative style would engage in more adaptive strategies than those using a diffuse style. The participants were asked to complete assessments of identity orientation as well as several subscales on the adaptive or maladaptive strategies they used. The revised Identity Style Inventory (ISI-revised), 23 consisting of three, ten-item scales representing diffuse, informational, and normative identity styles, was used to assign primary identity styles to each participant. Raw scores on each of the identity style scales were transformed into standardized z-scores and identity styles were assigned to participants according to his or her highest z-score. As expected, those using a diffuse identity style were found to engage in maladaptive strategies, and the most task-irrelevant behaviors and avoidance. Informational types scored higher than both diffuse and normative types on measures of reflective thinking/planning. Those using a normative identity style tended to seek social support more than did those using informational or diffuse styles. Another study focused on identity styles and academic problems of college students (Berzonsky, 1992). Participants were 69 males and 102 females in college (ages ranged from 18-25). The ISI-revised was used to identify participants who used informational, normative, or diffuse identity styles. Participants were given checklists and asked to identify aspects of five coping strategies they used when faced with a stressful homework assignment or exam. The coping strategies assessed were: problem-focused coping, wishful thinking, distancing and detachment, tension-reduction, and seeking social support. Correlational analyses revealed that participants using an informational identity style handled stressful tasks with problem-focused coping strategies and the seeking of social support. These participants were not likely to use wishful thinking or distancing when faced with stress. In contrast, participants using a diffuse style used wishful thinking and distancing to reduce stress. Interestingly, normative participants also used wishful thinking, and distancing as well as tension reduction when faced with stressful tasks. Berzonsky concludes that because the normative style was related to 24 identity commitment (r = .36), these participant may be more likely to use these kinds of avoidance strategies when faced with a specific stressor that may threaten their identity commitments. In contrast, the diffuse style was negatively correlated with identity commitment (r = -.34) and therefore those using a diffuse style may use avoidant strategies on a regular basis to avoid emotional distress. In a study of 338 (85 male, 253 female) college students, who were primarily Caucasian, Berzonsky and Ferrari (1996) looked at identity styles and decisional strategies. Participants were asked to complete the ISI-revised, as well as scales assessing vigilance, panic, rationalization, excuse making, and decisional avoidance when making decisions. The researchers found that adolescents with informational identity styles were more likely to use systematic, vigilant, and evaluative criteria when weighing decisions, as opposed to those with a diffuse style who were more likely to panic or use maladaptive decision-making strategies (e.g., excuse making, rationalization, and avoidance). The diffuse style also was strongly related to procrastination in both decisions (.58, p < .01) and behavior (.47, p < .01). The group of normative adolescents were more vigilant than the diffuse adolescents, yet also more maladaptive than the adolescents using the informational style. Therefore, those who used more informational or normative styles of identity processing showed the use of more adaptive action-taking behaviors. Associations between time-use and identity style. Although no research has been conducted to link time-use and Berzonsky?s identity processing styles, one study addresses associations between time spent in productive discretionary activities and identity exploration and identity formation. Time spent in discretionary activities allows 25 adolescents to gain self-knowledge, enhance talents and interests, and explore their identities (Dworkin, et al., 2003) Dworkin, et al. (2003) interviewed 55 adolescents who were active in one of three extracurricular activities: Future Farmers of America, a community based arts program, or a service-learning leadership organization. The researchers used focus groups to better understand salient themes brought up by adolescents when asked about their extracurricular activities. The adolescents were asked to discuss their experiences in terms of developmental areas including identity exploration and development, development of initiative and social skill development. After the data collection, the group discussions were coded for consistent themes of the adolescents? experiences in extracurricular activities. Within each of the areas, several themes were identified, as represented in the following table: Identity Exploration Initiative Development Social Skill Development a) try new things b) gain self-knowledge c) learn individual limits a) set realistic goals b) put forth effort and persevere c) manage time d) take responsibility for behaviors a) learn to work as a group or team b) learn about leadership and responsibility c) learn to take and give feedback d) learn communication skills This study described the ways in which extracurricular activities advanced important identity-related developmental tasks of late adolescence. Involvement in extracurricular activities helped adolescent identity exploration by promoting the gain of important skills, awareness, and connections. These extracurricular activities provided 26 guidance and the setting for adolescents to gain self-knowledge, learn skills such as leadership, goal setting, and communication, and develop initiative. In summary, identity development is clearly an important task in late adolescence. As adolescents anticipate, and are faced with, challenging situations in their lives, their use of identity styles helps to determine how they deal with these issues. Those who use an informational style consider many different choices and circumstances of behavior before making decisions. Adolescents using this style generally have better academic, social, and personal outcomes. This style is also beneficial in spending time in GDB and in setting goals for the future. The normative style involves guidance and cues from important adult influences. Parental support is very important for these individuals and their decisions for GDB and time-use may be influenced by what others are doing or encourage them to do. A diffuse style of dealing with difficult situations and decisions is characterized by procrastination and task-avoidance. Using this style leads to poorer outcomes for adolescents? unproductive time-use and less GDB. Possible selves. When Markus and Nurius (1986) conceptualized possible selves, they expanded the understanding of identity formation, self-concept and self-knowledge. They proposed that how people view their futures affects how they behave and understand their present identity. Possible selves originate from ideas of current identity, reflections of past ideas and behaviors, and social comparisons of respected peers and authorities. Possible selves provide the motivation for action-taking to occur by connecting the concepts of one?s current beliefs about himself to a future self in which he has accomplished important goals (i.e., personalizing the goal). Therefore, adolescents? 27 goals, aims, and ideas about possible selves determine actions they take in the present (Markus & Ruvolo, 1989). Possible selves span many aspects of adolescents? lives. Using Q-methodology to measure possible selves in a sample of high school students, Shoffner and Kerpelman (2005) assessed 51 females, ages 14-16 (22 African American and 29 Caucasian) in order to understand how different adolescents thought about their possible selves and what expected and feared selves were the most salient. The participants were asked to sort 41 possible selves statements twice, resulting in two continuums, one that represented expected possible selves and the second representing feared selves. The sorts included statements about various individual characteristics, life roles, and circumstances. A Q- analysis revealed three groupings of adolescents for each ?expected? and ?feared? possible selves. The groups for expected possible selves focused priorities on either family roles, careers, or marriage and careers over parenthood. Feared possible selves sorts resulted in groups that (1) viewed careers as neutral and motherhood as least feared, (2) saw multiple career possibilities as least feared but viewed motherhood as somewhat feared, and (3) adolescents who strongly feared divorce, as well as math or science careers. This study revealed variability in the salient themes for adolescent females. Specifically, careers, motherhood, and marriage were varied in prominence and positivity when the adolescents thought about their possible selves and future goals. From this study, it can be seen that adolescent have varying ideas about their possible selves. This variance also would be assumed to be associated with varying action-taking in adolescence. 28 In order to assess possible selves and the action-taking variables associated with them, possible selves can be categorized into capital groups. Capital is the concept of exchanges and gains. In order for people to attain goals and possible selves, they use physical, mental, and emotional tools that enhance the possibility of achieving these goals. Human capital is the set of resources someone has that enhance a person?s skills and capabilities, whereas social capital is the sharing across relationships of physical or human capital (Coleman, 1988). Economic capital is physical possessions or monetary resources one has. Adolescents? possible selves can, therefore, be categorized into types of capital they desire to have in the future. For example, possible selves that focus on education, career, or the building of skills would be considered human capital possible selves. Likewise, those focused on earning, saving, or spending money would be economic capital possible selves. Lastly, possible selves that refer to supportive relationships or involvement in social activities or organizations would be referred to as social capital possible selves. Associations between GDB and possible selves. In a study of medical students, Inglehart, Markus, & Brown (1989) looked at two motivational processes: the structuring and focusing of goals and the emotional energizing effect of having goals. The aim of the study was to examine the effects that professional possible selves had on achievement in the medical profession. Participants were measured on academic achievement (i.e. GPA and test scores), as well as how much each student focused on medicine as their only career option and how satisfying they viewed a career in medicine. The results of this study showed that those students who had a strong focus on medicine as a career (i.e. had only one possible professional self), those who saw medicine as the most satisfying 29 possible career choice, and those who viewed their specialization as the most attractive, had higher GPAs at the beginning and end of their programs, and higher test scores throughout their programs than those who viewed their professional possible selves as only somewhat satisfying or attractive. These results suggest that having set possible selves and positive affect toward these possible selves may help focus and motivate individuals? action-taking in reaching their goals. Leondari, Syngollitou, & Kiosseoglou (1998) hypothesized that clear and specific positive possible selves would be related to academic achievement and motivation. Outcome measures of grade point average and task-persistence were evaluated. The participants, 289 high school students (ages 14-15), were asked to write short essays describing their possible selves. Positive possible selves were coded into four categories: (1) clear and specific with success due to hard work, (2) clear and specific with success due to luck, (3) general and vague with success due to hard work, and (4) general and vague with success due to luck. A multivariate analysis of variance showed that in terms of academic achievement, the group with clear and specific positive possible selves with success due to hard work was significantly different than the other three groups. Likewise, in terms of task persistence, this group was significantly different from the group with general and vague possible selves with success due to luck. The group with clear, specific possible selves with success due to hard work was positively associated with both academic achievement and task persistence. Similar to Oyserman, et al.?s (2004) and Inglehart, et al.?s (1998) work, this study suggests that clear and specific possible selves promote positive adolescent action-taking. 30 Hock, Deshler, and Shumaker (in press) also studied the motivational properties of possible selves in the academic domain. They hypothesized that possible selves serve a function in motivating students toward desired end states such as academic achievement goals. For example, individuals would be more motivated to learn when they could identify the usefulness and significance of information as well as future outcomes. In order for students to gain motivation in striving toward their hoped for possible selves they needed to (1) identify valued goals, (2) believe that they could attain these goals, and (3) develop specific GDB to attain them. They proposed that students who could visualize their possible selves would be able to accomplish more GDB in the present. For this reason, the Possible Selves Program was created by Hock et al., (in press) at a large, Midwestern university. Student participants were asked to construct and describe their possible selves. Their next task was to draw a Possible Selves Tree, which was a visual representation of their possible selves. Students were asked to examine their goals and possible selves and to then evaluate what it would take to actualize these possible selves and create GDB for doing so. In analysis of this program, the researchers found that students who went through the Possible Selves Program were able to identify more life goals, earned higher GPAs, and had a higher graduation rate than students in a control group. These results support possible selves? relation to motivation and academic achievement in that those who were able to see and analyze their possible selves scored higher on measures of academic achievement. Associations between time-use and possible selves. In a study of time-use and possible selves, Shanahan and Flaherty (2001) assessed the time-use patterns of 933 high 31 school students over three years. The population was racially and economically diverse and students were enrolled in public high school. Participants answered questionnaires about their daily time-use in work, homework, extracurricular activities, volunteer work, chores, and ?other tasks.? Possible selves also were assessed and adolescents were asked to indicate the level of education they envisioned completing, their marital plans, the importance of becoming a good citizen, and the importance of pay, using one?s skills, and being helpful to others. Results indicated that in 9 th and 10 th grades, students who indicated having higher education possible selves were involved in time spent in homework and extracurricular activities. In 10 th and 11 th grades, students who envisioned becoming good citizens were involved in extracurricular activities and had high levels of school engagement. Also in 10 th grade, students whose possible selves included having a job in which they were able to help others were involved in homework and time spent doing chores. These results suggest that possible selves of adolescents have indications on how they spend their discretionary time. Another study addressed self-perceptions of adolescents in explaining the relationship between productive time-use and positive youth development. Although possible selves and self-perceptions differ in that possible selves are visions of one?s self in the future, whereas self-perceptions are visions of one?s self in the present, they have similar indications because they both measure how one sees one?s self. In a study of 1,259 adolescents (predominately white, middle class), Eccles, Barber, Stone, & Hunt (2003) assessed activity involvement in prosocial activities (church, volunteer, community service), performance activities (band, drama, dance), team sports, school involvement (pep club, cheerleading, student government), and academic clubs (math 32 club, foreign language club, tutoring). Adolescents also were asked to indicate which kind of person they were most like: Princess, Jock, Brain, Basket-Case, or Criminal. These categories were based on the identities of the characters in the movie, The Breakfast Club, which was popular at the time of the data collection. Adolescents were asked to ignore the sex of the character and indicate which character they identified with most. Results indicated that 87% of the Jocks played on a school sports team (although not all athletes identified themselves as Jocks). Other groups were over-represented in certain time-use activities, indicating a link between their self-perception and their choices of activities. Princesses were over-represented in performing arts and school- related activities (e.g., dance, drama, and pep club) and Brains were over-represented in prosocial activities (e.g., band/orchestra, math club, foreign language club). Both Basket- Cases and Criminals were under-represented in all categories except for performing arts and sports, respectively. This study supports the hypothesis that perceptions of one?s self are associated with the time-use choices of adolescents. Also, this study presents two ways of understanding adolescent time-use decisions. Adolescents may choose their activities because of how they see themselves or, in contrast, adolescents may see themselves in certain ways because of the activities they choose. In summary, possible selves help us understand motivation for GDB and time- use. They represent salient themes that can act as motivation for GDB and time-use. As adolescents look at their possible selves, they set long and short-term goals and create strategies for how to attain these goals. Possible selves also influence the types of activities and amounts of time adolescents spend in building their skills and talents in preparation for the future. 33 Parental Support Associations between parental support and GDB. In adolescence, when children are learning about themselves as well as the world around them, parents have a unique opportunity to support their adolescents? goals and goal-directed behaviors. Young (1994) described ways in which parents support the GDB of their adolescents. One way is through intentionality and meaning. An example of this may be a parent who notices a child?s interests and abilities and finds ways to support and encourage those skills. Effective influence of this kind can be accomplished with open communication on the part of the parent and the adolescent, development of responsibility in the youth, active involvement by the parent, encouraging autonomy in the youth, and specific guidance by the parent. A second way in which parents influence GDB in their children is through the use of narrative. This means creating an open communication about past and present roles of the parent and the adolescent in two domains. The first domain focuses on the adolescent?s past experiences and how they have been able to use the parent?s influence on their goals and actions. This helps adolescents understand their own identity development in terms of their goals as well as to see support from their parents and discuss their plans and goals. The second domain focuses on the accomplishments of the parent in helping to foster growth in their child. Finally, an important aspect of parental influence on goals is recognizing sources of conflict. This may be a result of the balance between attachment and autonomy of the parent and the adolescent or from the ability of the parent to influence their child without pushing their values onto them. Therefore, because long-term goals of adolescents can introduce considerable emotion into a discussion, parents need to be aware of sources of conflict. This study provides several 34 examples of how parental support may be important in encouraging adolescents in action- taking behaviors. Bean, Bush, McKenry, and Wilson (2003) assessed parental support and its associations with academic achievement and self-esteem. They were interested in looking at the differences between maternal support and paternal support as well as the differences between African American and European American families. Participants in this study were 75 African American adolescents and 80 European American adolescents. The adolescents all were enrolled in public high schools. Participants were asked to answer questions about support on a 4-point Likert scale ranging from strongly disagree to strongly agree. They also were assessed using a self-esteem scale and self-report of their grades in school. The first step in data analysis was to look at bivariate correlations. Maternal support was positively associated with the academic achievement of both African American and European American adolescents and paternal support was positively associated with the self-esteem of African American adolescents. Secondly, a multiple regression analysis showed that support from African American mothers was positively associated to self-esteem and to academic achievement. As can be seen from this study, parental support, particularly from mothers, was positively related to positive adolescent outcomes. Furthermore, logically, it can be assumed that support also would be associated with the GDB (e.g., studying, doing homework, working on relationships) that may enhance these outcomes. Holahan, Valentiner, and Moos (1995) assessed 241 college freshmen with measures tapping parental support, GDB, and adolescent well-being, self-worth, happiness, and distress. Participants were asked to relate their ?most important problem? 35 of the past year and rate how often they engaged in GDB in the form of coping strategies (i.e., positive reappraisal, problem-solving, cognitive avoidance, and emotional discharge). Participants were separated into either a high parental support or a low parental support group. Results showed that parental support and adolescent adjustment were positively correlated. Those with high parental support showed more adaptive coping strategies (i.e., problem-solving) compared to those in the low parental support group. Participants in the high parental support group also showed higher psychological adjustment and lower distress than those with low parental support. In this study parental support was shown to be an important factor in adolescent GDB in the form of coping strategies, as well as for adjustment and well-being. In studying the influence of parental support on adolescent action-taking, Nurmi (1987) looked at family relationships and the ways in which adolescents view their futures and the use of GDB in preparation for their futures. Finnish adolescents (75 male, 73 female) ages 10-19 were interviewed about planning for their possible selves (i.e., how much they had thought about or participated in GDB), as well as about their home environments. Results showed that adolescents who described their family environments as supportive (i.e., warm, amiable, and secure) had more educational plans than did those whose descriptions of their home environments were negative. However, the relationship between these factors changed according to age-related developmental tasks and expectations. For instance, Nurmi found that among 11 year-olds, future planning was negatively related to negative family relationships. In contrast, a positive relationship between the same variables was found for 18 year olds (i.e., more future planning was associated with negative family relationships). In this study, Nurmi concluded that 36 younger participants were less likely to have parental support (especially in the presence of negative family relationships) to encourage GDB, and older participants would be more likely to be in search of ways of leaving the negative family atmosphere. Taken together, parental support has been shown to be important for adolescents as they and their parents navigate the transition from dependence to independence. Parental support such as warmth, encouragement, and advice has been found to be associated with positive adolescent development and outcomes such as self-esteem, academic achievement, problem-solving and planning. Summary Action-taking in adolescence is an important aspect of positive youth development and goal achievement. Actions are intentional and purposeful behaviors that are defined by, and used to achieve, goals. Two ways of looking at action-taking are goal- directed behaviors and time-use. Goal-directed behaviors are steps that move one toward a goal. Weekly time-use measures the amount of time spent by adolescents in productive or unproductive activities and alone, unsupervised with friends, or supervised by parents or adults. The use of both GDB and productive time-use has been linked to positive adolescent outcomes such as academic performance, psychological functioning, and skill development. Several factors have been shown to be associated with adolescent action- taking. First, the use of different identity styles has been related to the use of GDB. Likewise, productive time-use in extracurricular activities has been related to identity exploration and formation. Second, possible selves have been shown to serve as motivators for action-taking and having specific possible selves has been related to such GDB related factors as task persistence, academic achievement, and goal achievement. 37 Types of possible selves also have been linked to engagement in specific time-use activities. Finally, parental support has been shown to enhance adolescents? engagement in GDB such as planning and problem-solving. 38 METHOD Participants The current exploratory study used data that were collected for a larger project addressing identity formation and capital building efforts of rural youth. The participants in this study were 25 adolescents (ages 15 and 16), and one or both of their parents or parental figures, living in a rural area of the Southeast. Nine of the adolescents were male and 16 were female (12 Caucasian, 12 African American, 1 Hispanic). Two of the female adolescent participants were excluded from parts of the study due to one inaudible interview tape and one missing Time Sort. Parents were 22 mother figures (21 biological mothers and 1 aunt) and 15 father figures (12 biological fathers, and 3 stepfathers). Parents? combined family income ranged from less than $10,000 a year to over $100,000 a year. Fifty-six percent of parents reported making between $25,000 and $55,000 a year (8% reported less than $10,000 - $25,000; 20% reported $55,000 - $85,000; 16% reported $85,000 or more). Parental education ranged from some high school education to earning a graduate degree. Nine percent of mothers had some high school, 27% earned a high school degree, 32% reported some college or a earning a technical school degree, 23% reported earning a college degree and 9% reported earning a graduate degree. Twenty percent of fathers earned a high school degree, 54% reported some college or a technical school degree, 13% reported earning a college degree and 13% reported earning a graduate degree. 39 Procedure Each participant independently completed two q-sorts addressing identity formation and an interview addressing goal-directed behaviors and parental support. After completing these individual tasks, the family was brought together to complete a group activity addressing the adolescent?s weekly time-use. Measures Identity style. Each adolescent initially received three identity style scores, one for informational, one for normative, and one for diffuse. To obtain these scores the adolescent completed a 60-item Current Identity Q-sort (CIQ, see Appendix A.1). The CIQ was used to assess the relative accuracy of different identity descriptors. Each of the 60 cards had sentences relating to identity (e.g., ?I am someone who likes to gather a lot of information about myself,? ?What my parents think I should do is one of the MOST important influences on my life choices,? ?My life feels like a roller coaster changing from day-to-day,?). Parents also sorted the same identity items according to their perceptions of their adolescents. After reading through each of the cards, the participants sorted them into 9 columns ranging from ?most like me/most like my teen,? in column 9 to ?least like me/least like my teen,? in column 1. Pilot data (Kerpelman, Pittman, & Li, 2004) from 31 college students who completed the sort twice, four to six weeks apart, revealed a mean test-retest correlation of .71 (SD = .11, Min r = .51, Max r = 1.00). Prior to data collection, three criterion sorts were created to represent the three identity styles. To create the sorts, the 60 self-assessment items were sent to experts in the identity field. They were asked to sort the items three times; first as an exemplar of 40 the informational style, then as an exemplar of a normative style, and finally as an exemplar of the diffuse style. From the responses of 12 experts, interrater reliability was examined. Interrater reliability was .86 (SD = .31, Min = .29, Max = 1.78), .96 (SD = .38, Min = .00, Max = 1.78), and 1.01 (SD = .45, Min = .00, Max = 2.11) for informational, normative, and diffuse styles, respectively. A smaller mean SD indicated that more items that were similar were placed together by the experts; therefore, the criteria were reliable across experts. The experts? sorts were aggregated to form the three final criterion Q- sorts. Participants? CIQ-sorts were correlated with the criterion Q-sorts in order to create identity style scores, from the perspective of the adolescent, from his or her mother and from his or her father. Larger positive correlations indicate greater use of a particular style. When adolescents? scores for the three styles were correlated with each other, it was determined that the informational style and diffuse style were measuring the opposite ends of the same continuum (r = -.94), whereas the normative style and informational style were orthogonal (r = -.02). Therefore, only the informational style (where closer to +1 = stronger use of informational style; closer to ?1 = stronger use of diffuse style) and normative style scores were used for subsequent analyses. Possible selves. A 99-item Possible Selves Q-sort was used to identify the adolescent?s possible selves, as well as each parent?s perception about his or her adolescent?s possible selves (see Appendix A.2). Each possible self item represented one of three categories: human capital, economic capital, or social capital. Examples of items included ?I (my teen) will have a formal education beyond high school,? ?I (my teen) will have family members who will give me (her) support,? and ?I (my teen) will wear 41 expensive clothes.? The participants sorted these items into 9 columns ranging from column 9 ?most like me (my teen) in the future? to column 1 ?least like me (my teen) in the future.? Pilot data (Kerpelman, et al., 2004) from 44 college students who completed the sort twice, a month apart, revealed a mean test-retest correlation of .82 (SD = .08, Min r = .63, Max r = .94). Possible selves were represented in two ways: first, by the amount of human, economic, and social capital items adolescents and parents placed among their most desired possible selves (i.e., the 21 items placed in their ?most like me/my teen in the future? columns) and second, by the extent to which adolescents were thinking about different types of possible selves relative to other adolescents in the sample. Of the 99 items in the PSQ, 46 represented human capital possible selves, 26 represented economic capital possible selves and 27 represented social capital possible selves. Each participant?s human, economic, and social capital possible selves in the top two ?most like me/my teen in the future? columns were counted and then divided by the total number available for each type. This resulted in a score equaling the percent of the total of each possible selves category that the adolescent or parents placed in his or her top two columns (e.g., 18/46 human capital items = 39% of the human capital items possible were placed in the top 2 columns). Q-analysis was used to assess how the adolescents in the sample were thinking about their possible selves relative to each other. The goal of Q-analysis was to group ?like people? who were similar in how they sorted the Q-sort items. The Q-analytic procedure took advantage of the full set of possible selves and involved several steps that resulted in identifying ?ways of thinking? in a group (Kerpelman, in press). First, the data 42 were transposed so that respondents were the variables to be analyzed. The data were prepared in this way in order to analyze the data according to each person?s full set of responses to the Q-sort. After the data were prepared, a factor analysis was employed, followed by the construction of ?composite Q-sorts? based on those individuals who loaded highest (i.e., .6 or higher) on a factor and did not cross load on other factors. Participants who sorted the Q-sort items similarly loaded significantly on the same factor. The factor analysis revealed two groups with different ?ways of thinking? about possible selves. Table 1 shows the factor loadings for all adolescents. The factor loadings are bolded to indicate those used to create the composite sorts. 43 Table 1 Factor Loadings for Group 1 and Group 2 Participant Group 1 Group 2 100 .749 .097 500 .724 .228 1800 .652 .215 300 .613 .309 2100 .594 .290 1900 .520 .291 700 .493 .458 2400 .428 .390 1200 .101 .790 200 .151 .643 2900 .419 .630 2600 .176 .627 1300 .378 .594 1400 .127 .571 1100 .092 .090 2300 .220 .231 1500 .341 .275 2200 .094 .227 2000 .306 -.175 1700 .252 .515 2800 .410 .128 3000 .089 .192 2500 .445 -.047 1000 .058 .465 2700 .153 .035 44 For each person used to create the composite q-sort a weight was calculated (Brown, 1993; McKeown & Thomas, 1988) and each person?s placement score for each of his or her items in the sort was multiplied by his/her factor weight, and then summed across people for each item, creating the final location score for each item of the composite sort. These scores, along with their respective item numbers, were sorted in descending order to create the composite sort for each factor. Once created, the composite sorts were compared to determine similarities and differences between the groups. Finally, as was used to determine the adolescents? identity style scores, correlations were computed for each adolescent?s PSQ with each of the PSQ composite sorts to indicate the extent to which he or she ?thinks like? each group, regardless of whether he or she was used to create the composite sort. (For a detailed account of the Q- analytic procedure, see Appendix B.1). Parental support. Parental support was addressed in the Possible Selves Q-sort Interview (PSQ Interview, see Appendix A.3). Adolescents were asked to answer the question, ?Who supports you in your efforts to reach your goals and what do they do?? The question was open-ended and adolescents were free to mention and describe support across their goals from any source (i.e., mother, father, sibling(s), peers, kin, and non- kin). Therefore, mothers and fathers could be mentioned as providing similar and/or different types of support. The number of times the adolescent mentioned support given by his or her mother was counted and became the score for ?maternal support.? Likewise, the number of times the father was mentioned became the score for ?paternal support.? Examples of perceived parental support included: my mom backs me in everything; my 45 dad always helps me with my homework; and, my parents help me financially when I need it. Because the interview focused on activities aimed at achieving possible selves, it was likely to elicit reports of perceived ?instrumental? parental support. Goal-directed behaviors. The adolescents? GDB used to strive for their possible selves, as well as the parents? perceptions about the adolescents? possible selves and GDB, were collected through the PSQ Interview (see Appendix A.3). Trained interviewers asked each participant to view a list of his/her ?most like me (like my teen)? possible selves (i.e., columns 8 and 9) obtained from the PSQ. The participant was asked to describe what he or she (or his or her adolescent) was doing currently, or was intending to do, in order to reach the possible selves. Following data collection, each PSQ Interview was transcribed verbatim and prepared for analysis. Three coding schemes were applied to the PSQ Interview transcripts. (See Appendix B.2 for a full description of the coding schemes.) Several preliminary steps were taken to create charts that were used for coding and for further analysis. At least 2 trained coders were assigned to each transcript. The first step in preparing the charts was to read and create a list of every GDB mentioned. A GDB was defined as any action or behavior the adolescent mentioned participating in at the time of the data collection, or would participate in during the high school years, used in obtaining or working toward his or her possible selves. Second, coders determined to which possible self each GDB referred, according to statements made in the PSQ Interview. Examples of GDB and possible selves or goals included: researching universities on the internet to attain higher education, working to save money to be a car owner, and volunteering at a hospital to have a career in the health field. 46 Next, coders grouped the GDB that were similar using letters to designate groups. For example, a list of GDB may include: work hard in school (a), get good grades (a), shadow a doctor for a day (b), save money in a checking account (c), earn money at work (d), and get a job to save for college (d). GDB with the same letter were similar to each other in how the adolescent intended to attain his or her goals. Each coder compiled her lists for each interview into a chart to be used to compare with the charts of the other coders. After coders individually coded assigned transcripts, they came together to discuss what categories were assigned and why (on average, coders agreed 79% of the time before they came to a consensus). This procedure was used in order to reach a consensus (i.e., 100% agreement) between coders and a final revised chart for each interview was made. Three coding schemes were applied to each interview. The first coding scheme was based on Oyserman (2004) and was used to assess the complexity of adolescents? GDB and parents? perceptions of their adolescents? GDB. This coding scheme combined the number of possible selves categories mentioned (e.g., Go to College, Career) and the number of GDB generated in the interview. Complexity was measured, therefore, in terms of how many goals and corresponding GDB were mentioned in the interview. This resulted in a GDB complexity score that ranged from 1 = low complexity to 7 = high complexity. The second coding scheme measured GDB diversity. GDB diversity assessed whether the participant described similar GDB (e.g., ?To get into college I will: study, work on my homework, and try to make good grades?) or diverse GDB (e.g., ?To get into college I will: study, participate in community service activities and apply for 47 scholarships). The GDB diversity score was equal to the number of letters assigned for groupings of similar GDB (see Appendix B.2). Oyserman?s coding scheme (2004) assessed self-generated possible selves and GDB. However, because the participants in the current study had the same possible selves items to choose from when completing the PSQ (i.e., had a restricted range and therefore, did not generate possible selves on their own), a second score (1 = low diversity to 9 = high diversity) was needed in order to assess self-generated answers. This resulted in the ability to assess GDB in a more intensive way than would have occurred only by using Oyserman?s coding scheme. The third and final coding scheme was used to explore the relationship between types of GDB and types of possible selves (i.e., related to human, economic, and social capital). It was hypothesized that the use of GDB to build certain types of capital would be related to the types of items adolescents would choose as their ?most like? possible selves (e.g., an adolescent who placed a majority of economic possible selves in his top two columns would describe more economic GDB). For this coding scheme it was determined which capital category each GDB reflected (e.g., ?took a class on nursing? reflected human capital goals; ?started a checking account? reflected economic capital goals). Although the GDB did not always fit cleanly in one category, to best answer the research question, the GDB were placed in the category that represented the PSQ item it represented. For example, ?balancing a checking account? could be seen as an economic capital GDB or a human capital GDB; however, because the PSQ item ?I will have savings in the bank,? represented economic capital, this GDB was categorized as economic. 48 Next, the number of human, economic, or social capital GDB the individual mentioned was counted. This step yielded three scores for each adolescent: number of human capital GDB, number of economic capital GDB, and number of social capital GDB. The same process was completed for parents? perceptions of GDB. Weekly time-use. Weekly time-use was measured by a Time Sort completed by the adolescent and his or her parents (or parent-figures). Participants were asked to work, as a family, to sort a set of ?hour? cards (total = 168 hours in a week) into 14 mutually exclusive and exhaustive categories according to how many hours were spent in each area by the adolescent during a typical week. After the total hours for each category were sorted, the family sorted, within each category, how much of the time was spent by the adolescent alone, with friends unsupervised, or supervised by parents or adults (see appendix A.4). As part of the larger study from which these data were drawn, adolescents were interviewed by telephone about their time-use over the past 24 hours. The types of categories mentioned and amounts of time devoted to different activities corresponded positively with the weekly time sort, supporting the validity of the Time Sort assessment. Several steps were taken to obtain the best representation of the adolescents? time- use. Fourteen time-use categories were assessed to determine their fit within the research questions. Therefore, categories that did little to aid the understanding action-taking (i.e., meal time) were omitted. Among the remaining categories, those that were mandatory for all of the adolescents, and in which they spent approximately equal time (e.g., School, Sleep), also were eliminated. Due to the small number of adolescent participants in the current study, some categories were removed because too few hours were spent in these categories by any of the adolescents (e.g., Lessons/Tutoring). 49 The remaining categories (n = 9) were assessed and analyzed, when possible, according to the hours adolescents spent alone, with friends but unsupervised, and with parents or adults present (i.e., supervised). When variance in these subcategories was too small (e.g., when only two adolescents mentioned spending time with parents or adults in hobbies), only the total amount of hours for the category was used. The final set of categories used for the current study were: non face-to-face time with friends, chores, paid work, schoolwork, sports, religious activities, extracurricular activities, hobbies, and general leisure. 50 RESULTS The primary aim of the current study was to explore associations between action- taking behaviors (i.e., GDB and weekly time-use) and identity formation, as well as perceived parental support. Data analyses proceeded in several phases: first, descriptive statistics and associations between adolescent and parent responses were examined, second, correlational analyses of associations among action-taking and identity formation and parent support variables were conducted, and third, in depth examination of high and low action-taking adolescents was undertaken. Descriptive Analyses The first phase in exploring the data was to examine the action-taking, identity formation, and parental support variables. Descriptive statistics were used to explore variation in action-taking, as well as variation in the identity factors (i.e., identity style and possible selves) and parental support. Variation in GDB. As can be seen in Table 2, adolescents varied in the complexity of their GDB, as well as in the diversity of their GDB. The widest range of GDB across the group of adolescents was human capital GDB with adolescents mentioning about four human capital GDB on average. There was a moderate range of social capital GDB mentioned and on average adolescents mentioned two GDB related to social capital. The narrowest range of GDB mentioned was in regard to economic capital. 51 It was less common for adolescents to mention this kind of GDB; on average adolescents mentioned one economic capital GDB. Similar to adolescents? reports, parental perception of adolescents varied on GDB complexity, and GDB diversity. Parents reported, on average, three human capital GDB and one economic capital GDB. Mothers and fathers reported between two and three social capital GDB on average. Variation in time-use. Table 3 shows the weekly time-use areas in which adolescents spent their time. Because the Time Sort was completed by the adolescent and his or her parent(s) together, only one set of time-use variables resulted. Adolescents? time was distributed among different activities. A majority of their time was spent in general leisure and non face-to-face time with friends (i.e., talking on the phone or through the computer). Adolescents spent approximately equivalent amounts of time in chores, schoolwork, sports, and religious activities. A modest amount of time was spent working for pay, and in extracurricular activities or hobbies. Adolescents spent the majority of their schoolwork time alone and most of their time doing religious activities with parents or adults present. Sports and chores were the only time-use categories in which adolescents spent relatively similar amounts of time alone, with friends (unsupervised), and supervised. 52 Table 2 Means, Standard Deviations, and Ranges for Goal-directed Behavior Variables M SD Min. Max. Goal-Directed Behaviors Adolescent GDB Complexity 4.00 2.34 1 7 GDB Diversity 4.38 2.41 1 9 Number of Human Capital GDB 3.88 2.80 0 9 Number of Economic Capital GDB 1.08 1.50 0 4 Number of Social Capital GDB 1.88 1.90 0 7 Mother GDB Complexity 4.43 2.54 1 7 GDB Diversity 5.29 2.53 2 10 Number of Human Capital GDB 3.38 2.31 1 11 Number of Economic Capital GDB 1.38 1.63 0 6 Number of Social Capital GDB 2.43 1.89 0 7 Father GDB Complexity 5.07 2.19 1 7 GDB Diversity 5.33 2.61 2 11 Number of Human Capital GDB 3.40 2.29 0 8 Number of Economic Capital GDB 1.33 1.59 0 5 Number of Social Capital GDB 2.73 2.15 0 7 53 Table 3 Means, Standard Deviations, and Ranges for Time-use Variables M SD Min. Max. Time-Use (hours per week) Non Face-to-Face Time with Friends 13.00 8.51 0 30 Paid Work 2.56 5.22 0 16 Chores Unsupervised Alone 1.92 2.33 0 10 Unsupervised With Friends 0.54 1.50 0 5 Supervised 1.85 2.66 9 Total 4.31 3.93 0 15 Schoolwork Unsupervised Alone 4.83 5.42 1 22 Unsupervised With Friends 0.19 0.53 0 2 Supervised 0.48 0.93 3 Total 5.50 5.33 1 22 Sports Unsupervised Alone 0.83 1.95 0 7 Unsupervised With Friends 1.29 3.01 0 13 Supervised 2.42 5.29 19 Total 4.54 5.93 0 19 Religion Unsupervised Alone 0.08 0.24 0 1 Unsupervised With Friends 0.48 1.31 0 6 Supervised 3.56 3.35 12 Total 4.31 3.82 0 15 Extracurricular Activities 2.35 3.81 0 15 Hobbies 3.44 4.75 20 General Leisure 27.40 11.50 4 48 54 Variation in identity formation and parental support variables. The scores for identity style were derived by correlating the criterion sorts with each adolescent?s and parent?s CIQ. Therefore, all scores for identity styles are mean correlations. As can be seen in Table 4, on average, the adolescents had a modest positive association with the informational identity style. The association with the normative style was weak and positive. On average, mothers? and fathers? sorts had a modest positive association with the informational identity style. Mothers? sorts had a weak positive association with normative identity style, whereas fathers? sorts had a modest positive association with normative identity style. Possible selves were indicated by the percent of human, economic or social capital possible selves adolescents or parents placed in the top two ?most like me in the future/most like my teen in the future? columns of the PSQ. As can be seen in Table 4, there was a moderate range of the number of human capital and economic capital items adolescents placed in the top two columns and a wider range of the number of social capital items across adolescents. On average adolescents included items that represented all three areas. Parents? reports of human and social capital GDB were very similar to adolescents, reporting similar means and ranges. Reports on economic capital were similar for mothers, yet appeared slightly lower for fathers. The q-analysis resulted in two types of thinking about possible selves among the adolescents, of which a brief summary follows (for a full description of the Q-analysis groups see Appendix B.3). Table 5 provides a list of the ?most like? 21 items as well as 55 the ?least like? 21 items for each group. The first group was highly focused on human capital possible selves, mainly emphasizing a future career (e.g., having a job that includes saving or protecting lives, having a job that will require a lot of time). This group placed a moderate emphasis on economic and social capital items. Group 1 seemed to be very certain of the career area in which they would like to work. The majority of their ?least like? human capital items were careers they did not want to have (e.g., work in the business field). Group 2 focused mostly on social capital. They wanted to have help and support from others and spend time with family. This group also had possible selves that focused on aspects of economic capital (i.e., financial stability and material goods). Group 2 had possible selves that had a limited focus on human capital. Group 2?s ?least like? possible selves were primarily human capital items (i.e., types of jobs they would not like to have). Although the possible selves of Group 1 and Group 2 were ordered differently, several items were mentioned in both groups? ?most? or ?least like? categories. Both groups focused on social capital possible selves in which family and maintaining close relationships were valued. In addition, being a parent, a spouse, and a best friend were important for each group. Economic capital possible selves such as owning nice things and having savings in the bank were important for each group. Both groups ranked having a formal education and a highly skilled job as important. A few differences separated the groups such as Group 1?s desire to have a job in the medical field that involved saving or protecting lives, have a job that required a lot of time, and work with children or youth, all of which Group 2 placed in their ?least like? columns. Group 2?s 56 ?most like? items included having a job in sports or athletics, volunteering to coach a sport and playing a team sport, which were put in Group 1?s ?least like? columns. Overall, the adolescents in the sample related moderately and positively with both q-analysis groups, yet, for Group 1, individual adolescents ranged from a modest association to a strong association. For Group 2, the associations ranged from weak to strong. Although there is a wide range of how much adolescents in the full sample ?thought like? each group, it is important to note that all adolescents? sorts were positively associated with each Q-group. Finally, there was a wide range of support from both mothers and fathers according to the adolescents as identified in the PSQ Interview. On average, the adolescents mentioned support of future goals from mothers four times and fathers three times during the interview. 57 Table 4 Means, Standard Deviations, and Ranges for Identity Formation and Parental Support Variables M SD Min. Max. Identity Style Informational Adolescent .30 .21 -.15 .67 Mother .24 .21 -.20 .58 Father .27 .23 -.26 .55 Normative Adolescent .14 .13 -.19 .38 Mother .12 .13 -.05 .41 Father .22 .13 .04 .53 Possible Selves Human Capital Adolescent 7.72 (.17) 3.97 (.09) 2 (.04) 18 (.39) Mother 8.82 (.19) 4.86 (.11) 1 (.02) 19 (.41) Father 9.27 (.20) 3.84 (.08) 1 (.04) 18 (.39) Economic Capital Adolescent 6.20 (.24) 2.89 (.11) 0 (.00) 12 (.46) Mother 5.86 (.23) 3.14 (.12) 1 (.04) 13 (.50) Father 4.27 (.16) 1.87 (.07) 1 (.04) 8 (.31) Social Capital Adolescent 7.08 (.26) 3.64 (.13) 1 (.04) 19 (.70) Mother 6.32 (.23) 3.72 (.14) 0 (.00) 16 (.59) Father 7.47 (.28) 3.09 (.11) 0 (.00) 14 (.52) Q-Analysis Correlation Group 1 .52 .18 .21 .86 Group 2 .48 .19 .09 .91 Parental Support Support from Mothers 3.96 2.40 0 10 Support from Fathers 3.00 2.65 0 8 Note: For possible selves, numbers outside the parentheses indicate the number of PSQ cards; numbers inside the parentheses indicate the percentage of items chosen out of total items possible. Table 5 Most and Least Like Possible Selves Items for Q-groups 1 and 2 Most Like Possible Selves Least Like Possible Selves Column 9 Column 8 Column 1 Column 2 Group 1 (A) Have a job of high importance and respect (A) My job will involve taking risks (A) Will have a job in sports and athletics (A) Will work with nature (A) Work in the medical/health field (A) My job will require a lot of time (A) Will be a salesperson (A) Will work in law enforcement (A) My job will allow me to help people (B) Will have savings in the bank (A) Will be in the Military (A) Will work with animals (A) Have a job that involves saving/protecting (B) Will own my own car (A) Will work with food (A) Will have a job that involves religion (A) Have a highly skilled job (B) Will own my own home (A) Will be a stay-at-home parent (A) My job will involve entertaining people (A) Will work with children/teens (B) Will give money to a place of worship (A) Will drive vehicles for work (A) My job will help people improve (A) Formal education beyond HS (C) My parents will help with obstacles (B) Making ends meet will be a monthly challenge (A) My job will allow me to be artistic or creative (A) Will feel happy doing the job I choose (C) My parents will support my goals (C) Will play in a band (A) My job will not take a long time to master (A) Will make a lot of money (C) I will be a spouse (C) Will volunteer to coach a sport (A) Will have a paid job I can do at home (C) Will have family members who will support (C) I will be a churchgoer (C) Will play team sports (A) Will have more than one job at a time (C) I will be a parent (C) Will join a country club 58 Note: A = Human Capital, B = Economic Capital, C = Social Capital. Most Like Possible Selves Least Like Possible Selves Column 9 Column 8 Column 1 Column 2 Group 2 (B) Will have savings in the bank (A) Formal education beyond HS (A) Will have a teaching job (A) Will have a desk job (B) Will earn enough money to cover living (A) Will have a job in sports and athletics (A) My job will help people improve (A) Work in the medical/health field (C) Will have people visiting me in my home (B) Will have the money to buy my spouse a car (A) Will operate machinery for job (A) I will be in the Military (C) Will have family members to support me (B) Will own my own home (A) Will be a researcher/scientist (A) My job will involve entertaining people (C) Will have a close relationship with one or (B) Will take regular vacations (A) Will be a stay-at-home parent (A) Will work in law enforcement (C) My parents will help with obstacles (B) Will own my own car (A) Will have a job that involves religion (A) Will work with children/teens (C) Will have good friends who I can trust and be (C) My parents will support my goals (A) Will work with animals (B) Will have to find way to earn money for (C) Will volunteer to coach a sport (C) Will have a lot of close friends (A) My job will not take a long time to master (B) My future spouse will make plenty of money to (C) Will play team sports (C) Will be a best friend (A) Will have more than one job at a time (C) Will play in a band (C) Will be a spouse (C) Will be a parent (B) Making ends meet will be a monthly challenge (C) Will be a Sunday school teacher (C) Will organize social activities for my (C) Will sing in a choir Note: A = Human Capital, B = Economic Capital, C = Social Capital Table 5 (continued) 59 60 Congruence between Adolescents? and Parents? Perspectives The mean adolescent-mother CIQ correlation was .27 (Min = -.22, Max = .70). The mean adolescent-father CIQ correlation was .36 (Min = .00, Max = .71). The mean adolescent-mother PSQ correlation was .48 (Min = .11, Max = .73) and the mean adolescent-father PSQ correlation was .54 (Min = .33, Max = .71). Thus, associations between parent and adolescent CIQ (identity processing style) were modest, and associations between parent and adolescent PSQ (possible selves) were moderate. Congruence for GDB and possible selves is shown in Table 6. There was a moderate positive association between adolescents? and mothers? GDB complexity and diversity and adolescents? and fathers? GDB diversity. Fathers GDB complexity was weakly associated with adolescents? GDB complexity. Both mothers and fathers had moderate positive associations with adolescents? reported human capital GDB, and weak associations with adolescents? reported social capital GDB. However, there was a strong positive association between mothers? and adolescents? economic capital GDB. Fathers however, had a weak positive association with adolescents? economic capital GDB. Finally, mothers? and adolescents? possible selves positioned in the top 2 columns were weak (and in the negative direction). Conversely, fathers? and adolescents? possible selves positioned in the top 2 columns were positively related and significant in the case of human capital possible selves. It is possible that adolescents and parents may have been more similar in terms of possible selves than theses correlations suggest. These correlations are based on the placement of items in the top 2 columns of the sort. However, placing items in the 3 rd and 4 th top columns still suggest that these items were desired. 61 Table 6 Adolescent and Parent Congruence for GDB and Possible Selves Adolescent-Mother Congruence Adolescent-Father Congruence GDB Complexity .37 .11 GDB Diversity .39 .40 Human Capital GDB .25 .36 Economic Capital GDB .58** .17 Social Capital GDB .00 -.08 Human Capital Possible Selves -.14 .56* Economic Capital Possible Selves -.15 .44 Social Capital Possible Selves -.10 .19 Note: *p<.05 (2-tailed); **p=.01 (2-tailed) Associations between Action-Taking, Identity Formation, and Parental Support After examining the variation that emerged for each of the study variables, the second phase involved a systematic examination of the hypothesized associations among the action-taking variables and the factors proposed to be associated with them. The approach used was examination of correlations. Two-tail significance levels were used except where directional hypotheses were stated. Given the small sample size, there was limited power to detect significant associations. Therefore, patterns of the direction and magnitude of the correlations also was examined. First, correlations were used to assess the linkages of identity styles and possible selves with GDB from the adolescents? and parents? perspectives. Next, correlations between adolescents? GDB and adolescent reports of parental support were examined. Associations between GDB and identity styles. Associations between GDB complexity and GDB diversity and identity styles are reported in Table 7. Overall, there was a difference between the adolescents? and mothers? perspectives and the perspectives 62 of the fathers. It was expected that there would be a moderate to strong positive relationship between GDB and informational identity style and a weaker, but positive, association with normative identity style. Both adolescent reports and mother reports showed a positive relationship (although in most cases, nonsignificant) between both GDB complexity and GDB diversity and informational identity style. However, contrary to the hypothesis, there was a lack of association (and in some cases in the negative direction) between normative style and adolescents? and mothers? GDB complexity and diversity. In contrast, fathers? reported GDB complexity showed no link with informational identity style but was the only perspective to show a positive, though nonsignificant, association between normative identity style and GDB complexity and diversity. Table 7 Associations between GDB and Identity Style Variables Informational Identity Style Normative Identity Style Adolescent GDB complexity .32 -.03 GDB diversity .56** -.16 Mother GDB complexity .25 -.17 GDB diversity .35 .05 Father GDB complexity -.03 .31 GDB diversity .13 .39 ** p<.01 (1-tailed) 63 Associations between GDB and possible selves. Associations between GDB and possible selves were examined by correlating the number of human, economic, and social GDB mentioned in the PSQ Interview with the percent of human, economic and social possible selves participants placed in their top two ?most like me (my teen)? categories. In the PSQ Interview, each participant was focusing on his or her own sort. As can be seen in Table 8, in general, the expected associations were not found, although adolescent reports of human capital GDB were moderately and positively related to human capital possible selves, as expected. There also was a positive, although nonsignificant, association between human capital possible selves and social capital GDB. Associations also were found between human capital GDB and economic capital possible selves. When adolescents and fathers focused on economic capital possible selves, there was a moderate negative association with human capital GDB, although for fathers, this association was nonsignificant. However, when mothers reported more economic capital possible selves, they were more likely to report human capital GDB. Overall, there were no significant results for fathers? reports, but the pattern showed possible positive associations between social capital possible selves and social capital GDB and possible negative associations between economic capital possible selves and human capital GDB, economic capital possible selves and social capital GDB, and social capital possible selves and economic capital GD Table 8 Associations between GDB and Possible Selves Adolescent Mother Father HCPS ECPS SCPS HCPS ECPS SCPS HCPS ECPS SCPS Adolescent her her HGDB .41* -.45* -.09 EGDB .09 .05 -.13 SGDB .21 -.08 -.17 Mot HGDB -.12 .39* -.18 EGDB .04 .02 -.07 SGDB -.07 .19 -.07 Fat HGDB .04 -.28 .12 EGDB .16 .16 -.30 SGDB -.09 -.27 .28 64 Note. HCPS = Human Capital Possible Selves; ECPS = Economic Capital Possible Selves; SCPS = Social Capital Possible Selves; HGDB = Human Capital Goal-directed Behaviors; EGDB = Economic Capital Goal-directed Behaviors; SGDB = Social Capital Goal-directed Behaviors. *p < .05 (1-tailed) 65 Associations between GDB and Q-groups. There were no significant correlations between the Q-groups and adolescent GDB. Table 9 shows patterns of weak negative correlations between GDB complexity and both Group 1 and Group 2. However, for GDB diversity, there was a moderate positive association between Group 1 and a weak negative association with Group 2. Human capital GDB appears to be associated differently between the two groups. There was a positive modest association between human capital GDB and Group 1, but a modest negative association with Group 2. There also was a negative association between social capital GDB and Group 2. Economic capital GDB had no associations with either Q-group. Table 9 Associations between GDB and Q-groups Group 1 Group 2 GDB Complexity -.17 -.18 GDB Diversity .31 -.18 Human Capital GDB .27 -.29 Economic Capital GDB -.01 .06 Social Capital GDB .00 -.25 Associations between GDB and parental support. As expected, there was a positive association between maternal support as perceived by the adolescent and both adolescent GDB complexity and diversity, as shown in Table 10. However, there was no association between GDB and paternal support. 66 Table 10 Associations between GDB and Parental Support Adolescent report of GDB Maternal support Paternal Support Complexity .40* .01 Diversity .31+ .05 *p<.05 (1-tailed); +p=.07 (1-tailed) Associations between time-use variables and identity formation variables. Table 11 shows the relationship between weekly time-use totals and identity formation variables. The purpose of analyzing time-use was to explore patterns between identity formation variables and productive time-use activities. Several significant relationships were found. First, identity style was examined. There were moderate associations between schoolwork and identity styles. Schoolwork was positively related to the use of an informational identity style, but was negatively related to the normative identity style. Also, time spent in religious activities was moderately negatively related to informational identity style. Next, the percent of human, economic, and social capital items placed in the top two PSQ categories were examined. The total time spent in sports was moderately negatively related to human capital possible selves, but positively related to social capital possible selves. Economic capital was moderately positively related to non face-to-face time with friends and paid work. 67 Table 11 Associations between Weekly Time-use Totals and Identity Formation Variables INFO NORM HCPS ECPS SCPS Non Face-to-Face Time with -.06 .13 -.15 .46* -.20 Paid Work .12 .23 -.30 .37* .02 Chores -.15 -.16 .00 -.14 .11 Schoolwork .48** -.45* .29 -.30 -.07 Sports .05 .19 -.40* .07 .37* Religious Activities -.43* -.27 .05 -.21 .11 Extracurricular Activities -.23 -.13 .04 .01 -.05 Hobbies .04 .09 -.14 -.03 .17 General Leisure .24 .08 .07 -.14 .04 Note. INFO = Informational Identity Style; NORM = Normative Identity Style; HCPS = Human Capital Possible Selves; ECPS = Economic Capital Possible Selves; SCPS = Social Capital Possible Selves. *p< .05, **p< .01 A second method of understanding time-use was to examine with whom adolescents? time was spent during a typical week. As can be seen in Table 12, four activities revealed enough variation between alone time, time unsupervised with friends, or time supervised by parents or adults. Adolescents? time alone in schoolwork was positively associated informational identity style, yet negatively related to normative identity style. Analysis also revealed that the informational identity style was negatively related to time spent in religious activities that were supervised by parents or adults. Time spent in sports revealed several associations. Human capital was positively associated with sports when the adolescent was alone but negatively related when the 68 sports were supervised. Conversely, social capital was positively related to time spent in supervised sports, as opposed to time in sports alone. Table 12 Associations between Weekly Time-use and Identity Formation Variables INFO NORM HCPS ECPS SCPS Chores Unsupervised Alone -.21 .08 -.10 .16 -.02 Unsupervised With Friends .30 -.01 -.02 .04 -.01 Supervised -.21 -.30 .10 -.37 .18 Schoolwork Unsupervised Alone .46* -.40* .30 -.26 -.11 Unsupervised With Friends .17 -.28 .14 -.10 -.07 Supervised -.01 -.10 -.14 -.14 .26 Sports Unsupervised Alone .22 -.02 .39* -.15 -.29 Unsupervised With Friends .22 .31 -.19 .11 .11 Supervised -.15 .05 -.49** .07 .46* Religious Activities Unsupervised Alone -.08 .12 -.14 -.07 .20 Unsupervised With Friends .21 -.07 -.04 -.07 .10 Supervised -.44* -.28 .15 -.24 .03 Note. INFO = Informational Identity Style; NORM = Normative Identity Style HCPS = Human Capital Possible Selves; ECPS = Economic Capital Possible Selves; SCPS = Social Capital Possible Selves. *p< .05, **p< .01 69 Associations between high and low action-takers, identity formation and parental support variables. A final phase in the data analysis was to examine the data of high and low action- takers. High and low action was determined by assessing GDB complexity, GDB diversity, and productive time-use (i.e., combined hours in chores, paid work, schoolwork, sports, religious activities, extracurricular activities, and hobbies). To be considered a high or low action-taker, individuals had to score consistently high or consistently low, respectively, on all types of action-taking behaviors. The three adolescents who scored highest were considered high action-takers; the three adolescents who scored lowest were considered low action-takers. High and low action-taking from adolescents? perspective. As can be seen in Table 13, there was contrast in the action-taking of high and low action-takers. High action-takers had GDB complexity and diversity scores that ranged between 5 and 9, whereas low action-takers? complexity and diversity scores ranged between 1 and 4. High action-takers spent between 32 and 42 hours a week in productive time-use, while low action-takers spent between 11 and 23 hours in productive activities. Differences were also found when looking at the identity styles of high and low action-takers. High action- takers had moderate to strong positive informational identity style scores and weak normative style scores. Low action-takers had weak informational identity style scores (in both the positive and negative direction) and weak to moderate positive normative style scores. Finally, perceived maternal support was higher than the sample average for two of the high action-takers. Low action-takers had moderate maternal support scores. Paternal support was low for each high action-taker and two low action-takers. 70 Table 13 Descriptives of High and Low Action-taking from Adolescents? Perspective 1 2 3 4 5 6 7 Adolescents High Action Adolescent A 5 6 42 .39 .10 7 0 Adolescent B 5 9 32 .67 -.01 1 1 Adolescent C 7 8 41 .50 .18 10 0 Low Action Adolescent D 4 3 11 .00 .35 5 1 Adolescent E 1 1 23 -.15 .15 5 6 Adolescent F 1 1 22 .09 .11 3 0 Note: 1 = GDB Complexity Score 2 = GDB Diversity Score 3 = Hours in Productive Time-Use 4 = Informational Identity Style Score 5 = Normative Identity Style Score 6 = Maternal Support Score 7 = Paternal Support Score High and low action-taking from parents? perspective. As can be seen in Table 14, there was contrast in the action-taking of high and low action-takers from the parents? perspectives, although data collection was taken for only one father of the low action- taker group. Parents of high action-takers had GDB complexity and diversity scores that ranged between 2 and 10 (mothers) and 3 and 8 (fathers). Parents of low action-takers had complexity and diversity scores ranged between 1 and 4 (mothers) and 2 and 3 (father). Mothers? scores for informational identity style for of high action-takers were positive and ranged from weak to strong, but fathers? scores were positive and moderate 71 to strong. (Notably, the adolescent-mother congruence on the CIQ for Adolescent A was -.22.) Both mothers and fathers of high action-takers had weak associations with the normative identity style. Mothers and fathers of low action-takers had weak to moderate positive associations with informational identity score. Both mothers and fathers of low action-takers had moderate positive associations with normative identity style scores. Table 14 Descriptives of High and Low Action-taking from Parents? Perspectives GDB Complexity GDB Diversity INFO NORM Mothers High Action Taking Mother A 2 3 .05 -.01 Mother B 7 10 .54 -.03 Mother C 7 10 .35 .17 Low Action Taking Mother D 1 2 .35 .17 Mother E 1 3 .37 .35 Mother F 2 4 .14 .35 Fathers High Action Taking Father A - - - - Father B 3 4 .55 .15 Father C 7 8 .38 .08 Low Action Taking Father D - - - - Father E 2 3 .27 .24 Father F - - - - Illustration of high and low action-taking from the PSQ Interview. The adolescents? and parents? PSQ Interviews illustrated multiple differences of high and low action-takers. As seen in Table 13, high action-takers mentioned more GDB and hours in 72 productive time-use than did low action-takers. PSQ Interviews revealed other aspects about the adolescents? GDB and time-use. High action-takers mentioned specific GDB and areas of time-use in which they were involved. High action-takers and their parents mentioned such GDB as being involved in certain clubs, helping in the community, taking advanced or college coursework, applying for scholarships, researching colleges and careers on the internet, spending specific amounts of hours in schoolwork, and ways of thinking about relationships with friends and dating. For example, one mother described the GDB and time-use of her daughter: ?She already knows over these next two years what courses she needs to be taking in high school? and she?s already looking and planning on scholarships. She has already started taking a lot of the tests that prepare her for college entry level classes or possibilities of scholarships? She is also a member of an organization?that meets once a month at the hospital and they tour different facilities and learn about different occupations throughout the hospital: radiology, anesthesiology, different things like that.? In contrast, low action-takers and their parents talk quite differently about their GDB and time-use activities. When asked to describe what they currently were doing, or were going to do, to attain their goals the adolescents, as well as their parents, talked in very general or tentative terms, and often talked about their GDB and time-use activities in future tense. For example, one adolescent described how she intended to reach her goals: ?Once I get my salary, get my career, whatever, I plan on having some kind of a budget, save some of it, so that I can do the things that I want to do? Well, I 73 guess, to get the money that I want or whatever, I have to go and get some kind of extra education or whatever to get a career that I like. And I guess to get my spouse I have to get myself right first, make sure I am ready for all of it. And the rest of the stuff, I don?t know how. I guess just get myself ready for that too, along the way.? Adolescents and their parents talked about characteristics that related to identity styles. High action-takers, who had positive associations with informational identity style, described themselves as being focused, independent, and determined, and described a commitment to an identity. For example, one adolescent described why she chose certain PSQ cards over the others: ?I think that the cards pretty much go along with my beliefs, my personality, my work ethic and the way I work and strive to achieve my goals? I pretty much know my beliefs and morals and values.? Parents also described identity characteristics of their adolescents. This mother of a high action-taker also talks about characteristics of her daughter?s identity. ?She is very adamant about wanting to know all she can now about the fields she is thinking about going into so she doesn?t get there and realize, ?oh, this is not what I want to do after all.? She doesn?t want to reach that point, she wants to know and be certain this job is absolutely what she wants to do.? Conversely, low action-taking was marked by weak or negative associations with informational identity style (i.e., use of a diffuse identity style). One low action-taking adolescent described an obstacle that could get in the way of her goals: 74 ?Like I said, I was going to be maybe, the nursing career, whatever, I had to go to college or some kind of schooling or whatever. If I find someone [to marry] before that, that might kind of take me off track in my career option, whatever.? Likewise, this adolescent?s mother also stated that getting married or having a family too soon would alter her daughters? plan for her career. This is an interesting example of how an adolescent who uses a diffuse identity style may let life circumstances determine her educational and career goals (i.e., has a lack of commitment) as opposed to one who makes choices about her life circumstances that fit with her goals. The possible selves and the ways in which adolescents and their parents described their possible selves also were very different between high and low action-takers. High action-takers were very focused on specific and set career goals; particularly, on having a career in the medical or health field. They were very clear not only about the area of their desired occupation, but also about certain jobs within the field (e.g., anesthesiologist, forensic scientist, pediatrician). They all mentioned how important it was for them to do well in school and to work especially hard in certain classes that would help them in their future careers. These adolescents also mentioned their involvement in activities such as volunteering at hospitals, shadowing doctors, being involved in special school programs associated with the medical field, or gathering information about their desired careers from friends or family in the field. In contrast, low action-takers and their parents described varied career goals and discussed them in less detail. Low action-takers mentioned careers in sports, the military, and engineering, but did not have specific aims or jobs chosen for themselves. In 75 particular, one adolescent exemplifies how low action-takers described their possible selves when asked to describe why she chose certain possible selves: ?Well, I guess I don?t know why I put those there. Um, I guess those are just things that I think I would like to have in my, in the next ten years of my life?I guess, somewhere in there, I believe those are important, somewhere? [They] are just nice things to do. And a lot of them seem nice to me.? Finally, perceived parental support looked somewhat different for high and low action-takers. Both groups of adolescents mentioned receiving parental support, however, there were differences in how they described that support. High action-takers described specific support characteristics of their parents, whereas low action-takers described support in more general terms. For example, one high action-taker talked about support from her mother: ?My mom, she?s like number one. She?s completely helping me in everything. She?s always there for me and always has been. Like, helping me get a job. She?s helping me out with looking. And going to church, getting to places right now, helping me find a car. You know, testing my boyfriend too, she likes doing that. She likes to make him nervous. But, uh, I really just watch her examples, my whole family, and think ?well, that?s the way I want to be.?? Conversely, low action-takers mentioned very vague and general parental supports. These adolescents used phrases such as: ?They support me,? ?They will help me,? and ?My mother is there for me.? 76 Parents also talked about the support they provided for their adolescents. Similar to their adolescents, parents of high action-takers mentioned different kinds of support they gave their adolescents, while parents of low action-takers mentioned very vague kinds of support. One mother of a high action-taker described what she was doing to help her adolescent with her grades: ?[We are] interested in grades. I don?t want to know just what is your grade at the end of the 9 weeks; I like to see the projects that she is working on throughout.? This same mother goes on to mention that, although she gets ?tired of doing it sometimes,? she and her husband sit down at the computer with their daughter and look at the schools in which she is interested in applying. A father of a high action-taker described how he supports his daughter: ?I help her with her homework, what I can, she comes up with these crazy projects and I will get in there and try it.? Parents of low action-takers also mentioned being very supportive of their adolescents, mostly through open communication. However, very few mentioned instrumental support for their adolescents? goals. When one mother was asked what she was doing to help her adolescent reach her goals, she answered: ?Just whatever I can. I mean, just whatever I can. Whatever it is that I can do to help her out, I?ll just try to do it. I don?t know.? 77 DISCUSSION This study addressed action-taking in adolescence and its associations with identity formation and perceived parental support. Overall, there was variability in the action-taking behaviors of the adolescents. In general, action-taking appeared to be associated with identity formation and parental support. Specifically, the informational identity style was positively associated with action-taking, especially GDB and time spent in schoolwork. This is consistent with the literature in that informational identity style is related to positive youth outcomes (Berzonsky, 1992; Berzonsky & Ferrari, 1996; Berzonsky, et al., 1999). It was hypothesized that the normative style would have a weaker, yet positive relationship with action-taking. Surprisingly, patterns of the results show the normative style was negatively associated with GDB and several productive time-use activities (i.e., schoolwork, religious activities, and extracurricular activities). Consistent with expectations, examination of possible selves revealed that adolescents who reported focusing on human capital possible selves also reported more human capital GDB, however, other associations between possible selves and action-taking were inconsistent with expectation. Although findings for maternal support and adolescent GDB were consistent with the literature in that high levels of support promote positive youth development (Bean, et al, 2003; Holahan, et al., 1995), paternal support was not related to adolescent GDB. 78 GDB and identity styles. As hypothesized, analyses of GDB and identity style showed that the more adolescents used an informational identity style the more complex and diverse their GDB. This may indicate that adolescents using the informational style have an understanding of what they need to be doing to accomplish their future goals (Berzonsky, 1999; Berzonsky & Ferrari, 1996). It also is possible that, through exploration, these adolescents have narrowed their goals and, therefore, better know the activities needed to accomplish these goals. For example, a closer examination of the interviews revealed adolescents who had the highest informational scores mentioned specific careers they wanted to pursue as well as numerous actions they were using that directly related to their stated goals. In contrast, those who scored lowest on the informational style not only mentioned fewer GDB, but also typically mentioned broad categories of career fields (e.g., working in the outdoors). They did not, like their high informational style counterparts, mention specific positions within those fields (e.g., working as a park ranger). In addition, when those with low informational identity style scores were asked about the actions they were taking to reach their goals, they often preceded their actions related to the future in hypothetical or tentative terms (e.g., someday I will make some kind of a budget, I guess I would go to college, I could get a second job for extra income). This supports previous literature indicating that adolescents with low informational style (i.e., high diffusion) have not participated in exploration or commitment of identity issues (Berzonsky, 1990). It was expected that adolescents who used a normative identity style would have a positive, yet weaker, association with GDB than those using an informational style. The weak negative association between these variables could indicate that adolescents using a 79 normative identity style indeed do have a narrower view of their goals and GDB, therefore not putting as much effort into numerous and diverse GDB (Berzonsky & Ferrari, 1996). It also could indicate that these adolescents, who were 15 or 16 years old, have not actively sought to set goals for themselves, but are, instead, influenced by respected adults. For example, Nurmi (1987) found that plans increased with age. The adolescents in the current study still had several years before they would graduate and would have to make decisions for themselves about further education or careers. They may, therefore have had less GDB. The parents? interview data were used, in part to capture action language that the adolescents may not have expressed. It was expected that parent data may capture different action-taking behaviors than the adolescents? data, but would generally show some consistency with the adolescent self-reports (e.g., higher action-takers would be perceived by parents as taking greater action). Mothers? reports of GDB and informational style were consistent with the adolescent?s reports. However, fathers? reports on identity showed an opposite pattern from adolescents? reports. Reasons for this difference are not fully understood. It could be that mothers who viewed their adolescents as having complex and diverse GDB saw this as a form of independence and exploration, whereas fathers may have viewed it as compliance (as seen by positive associations with the normative style from the fathers? perspectives). GDB and possible selves. Analyses of possible selves and GDB revealed inconsistent results. It was hypothesized that adolescents who have certain types of possible selves (i.e., those that are associated with human, economic, or social capital) would have GDB related to those possible selves. Because few significant results were 80 found, there is limited support for this hypothesis. However, the association between human capital GDB and human capital possible selves is consistent with this hypothesis. It is possible that adolescents find it easy to associate actions that need to be done in connection with their human capital goals, such as attaining a higher education or a successful career, as opposed to GDB associated with economic or social possible selves. Because adolescents are at a developmental point in which many of them are thinking about their career goals and how to achieve them, they are presented with, and can recognize, many opportunities to build skills. Human capital possible selves also had a positive, although weak, association with social capital GDB. These adolescents may be using connections with more knowledgeable adults to help them build skills and experiences (Coleman, 1988; Teachman, Paasch, & Carver, 1997). In the PSQ Interviews, adolescents often mentioned working with, shadowing, or talking with adults in the career fields in which they were interested. Interestingly, there were very weak relationships between the possible selves and the GDB of economic and social capital. One possible explanation for this may be that economic possible selves often involved owning material goods in the future. It is possible that since adolescents at this age are still financially supported by their parents and may be less exposed to opportunities to earn or give money, this resulted in less action-taking in this area. Likewise, based on the PSQ items, social goals usually consist of being a spouse or a parent, going to church, or being involved in civic activities or groups. These are goals that are more likely to be viewed as attained through the passage of time as opposed to having to ?make? them happen. 81 The negative association between economic possible selves and human capital GDB was surprising considering that human capital is used to achieve economic capital (Coleman, 1988). It is possible that the adolescents focusing on economic possible selves, given their age, were more idealistic and less realistic than older adolescents or adults would be regarding how economic capital is obtained. Few relationships were found for parents? reports. When mothers reported a higher proportion of economic capital possible selves, they also reported more human capital GDB. Mothers may perceive adolescents? action-taking in building their skills as a sign that they are striving to have quality material goods as well as well-paying careers in their futures. The Q-groups did not show any significant associations with GDB. The Q-factor analysis takes advantage of the entire PSQ sort, but the associations between the Q- groups and GDB may be more complex and not easily understood through simple correlation. For example, analyses revealed that Group 1 was significantly correlated with informational identity style (r = .43, p < .05). Group 2, however, was significantly correlated with normative identity style (r = .55, p < .01). Although these relationships were not examined as part of the current study, they show that there were differences associated with thinking more like one group versus the other. Therefore, future use of more complex designs and analytic strategies may elucidate how possible selves and GDB are related. The patterns revealed for the PSQ groups and GDB were still of some interest. Adolescents? GDB diversity was measured by the different kinds of GDB the adolescents generated on their own, regardless of how many goals to which they were related. In 82 contrast, because each adolescent was working with the same set of PSQ cards, GDB complexity was limited to a certain set of possible selves items. Therefore, it is possible that the relative lack of association between GDB complexity and either group was because each adolescent had the same set of goals from which to choose. This could result in each group talking about approximately the same number of goals, which is the first criterion in coding GDB complexity (Oyserman, 2004). Thinking like Group 1 was positively associated with GDB diversity and human capital possible selves, which could be a result of their focus on one or two specific career goals. It may be that although they were limited in the number of goals to which they related they had more actions associated with these specific goals. This finding is supported by prior research (Inglehart, et al., 1989; Leondari, et al., 1998; Oyserman, et al., 2004) that found that having clear, set, and specific possible selves was positively associated with GDB and goal achievement. In contrast, thinking like Group 2, who were focused mostly on social capital, was not associated with diverse actions for achieving goals. GDB in terms of human, economic, or social capital also revealed patterns that were related to the Q-groups. The positive association between human capital GDB and thinking like Group 1, and the negative association between human capital GDB and thinking like Group 2 may indicate that there is a link between focusing on human capital possible selves and human capital GDB. Group 2, the group more focused on social capital, talked less about human capital GDB for reaching their goals in their PSQ Interviews. However, surprisingly, there was a negative relationship between thinking like Group 2 and social capital. This could, again, be because of the difficulty adolescents may find in relating actions to their social capital goals. Neither Group 1 nor Group 2 83 talked about economic capital GDB. This is probably because those who associated most strongly with these two groups had only a modest focus on economic possible selves. It is interesting to note that the adolescents whose Q-sorts were used to create Group 1 were four females, whereas those used to create Group 2 were three males and one female. Although not part of the current study, these gender differences could contribute to the differences in the two groups, and should be examined in future work. GDB and parental support. In support of the hypothesis, perceived maternal support was found to be positively related to adolescents? GDB. Maternal support may play a role, such as encouraging or advice giving, in the actions an adolescent takes to be able to reach his or her goals (Bean, et al., 2003). Another possible explanation may be that adolescents who are participating in more GDB are receiving, eliciting, or feeling more support from their mothers. Also, past research has shown the importance of maternal support for adolescent future education orientation. Nurmi (1987) asserted that mother?s beliefs about future educational orientation were mirrored by adolescents. Also, Wilson and Wilson (1992) found that mothers? expectations about their adolescents? education were particularly important for adolescents? educational attainment. The lack of findings for paternal support and GDB could be attributed to traditional gender roles. Traditionally, mothers play a more nurturing role than fathers do in raising children. Traditional paternal support (e.g., financial support) may not be as easily recognized as support by adolescents. More research defining the types of support given by parents are needed to better understand the relationships between GDB and parental support. Time-use and identity styles. Weekly time-use was analyzed to understand the patterns of adolescents who used differing identity styles or had differing kinds of 84 possible selves. Findings were not consistent across all productive time-use variables, however, several associations supported the hypotheses. As hypothesized, the more adolescents used the informational style, the more likely they were to be involved in time spent in schoolwork. Informational adolescents may be more focused on doing well in school in order to get scholarships, keep a high GPA, or pass classes that may help them gain entry into college and succeed in the future careers they aimed to attain (Berzonsky, 1990). Informational adolescents spent most of their schoolwork time alone. This could indicate a self-motivation to do the work needed to achieve their goals. Because these adolescent explored issues related to their identities, they may be more able to make a connection between the work they do in school now and their goals and possible selves. Also of interest was that the informational style was negatively related to time spent in religious activities and these adolescents were especially unlikely to spend religious time with adults. Again, this is in contrast to literature on productive time use, which suggests that time spent in religious activities is related to positive adolescent outcomes (Jordan & Nettles, 2000, Kleiber & Powell, 2005). It is possible that these adolescents may be spending time exploring their personal beliefs and making independent decisions about going to church or religious activities with adults. Surprisingly, the informational style was negatively associated with extracurricular activities and positively associated with general leisure (although these findings were not significant). This pattern was contrary to much of the literature on time-use (Gilman, et al., 2004; Jacobs, et al., 2004). Typically, time spent in extracurricular activities is related to identity exploration and self-knowledge (Dworkin, et al., 2003), whereas leisure (often in the form of watching TV or hanging out with 85 friends) is often associated with more negative outcomes (Gilman, et al., 2004). However, the findings in this study could be explained by several factors. First, the Time Sort did not identify specific extracurricular or leisure activities, nor did the participants specify in which activities they were involved. Also, these findings may be a result of the relatively smaller amounts of time spent in extracurricular activities as reported by the sample adolescents, generally. Normative style was negatively associated with schoolwork, especially schoolwork spent alone. This is in contrast to the hypothesis that, in comparison to the informational style, they would have a weaker, yet positive, relationship with productive activities such as schoolwork. Possibly, normatively oriented adolescents depend more on social support (e.g., parental support) for time spent on homework. When support is not available they may be less inclined to complete homework independently, in contrast to informational oriented adolescents (Berzonsky, 1990). Patterns between normative style and time-use showed a positive association between spending time in talking with friends on the computer or phone, doing paid work, and playing sports. The more normative adolescents were, the more they were spending time in sports with friends, presumably for the social aspect of the activity. Time-use and possible selves. Analyzing possible selves also revealed interesting differences in weekly time-use. The more human capital possible selves adolescents had the less likely they were to spend time in sports. Patterns also showed that human capital possible selves were positively associated with time spent in schoolwork and negatively associated with time spent in paid work. It is possible that when adolescents focused on careers and higher education, they spent regular time doing schoolwork to help them 86 reach those goals (Oyserman, et al., 2004). They may have been spending less time in sports and paid work because they are busy focusing on school. Consistent with Eccles, et al.?s research (2003) on social self-perceptions, those who see themselves as highly focused on future education or careers (i.e., have a strong focus on human capital possible selves) would be more likely to spend time in activities that supported this notion (i.e., schoolwork). It also could be that parents who make their adolescents spend a lot of time in schoolwork do not want their adolescents committing too much time to sports or paid work which has been shown to interfere with academic achievement (Jordan & Nettles, 2000). When looking more closely at schoolwork, adolescents who had more human capital possible selves were spending time in schoolwork either alone or with their friends. This may indicate a self-motivation to do well in school. As expected, having economic capital possible selves was significantly and positively associated with paid work. These adolescents also were spending less time in chores at home, likely because they make contributions outside the home by having a job. Beside the fact that these adolescents may work to support their economic capital goals, they may be more focused on economic goals because they are working and earning money (Eccles, et al., 2003). Interestingly, those focused on economic goals reported higher non face-to-face time with friends. This could mean that these adolescents have the ability to buy technology (e.g., cell phones) with which to communicate with friends. It is also interesting that these adolescents were spending less time in schoolwork than those focusing on human capital goals, possibly because they have less time to engage in schoolwork (Jordan & Nettles, 2000). 87 A stronger focus on social possible selves was significantly and positively related to time spent in sports. This could reflect a desire to be involved in team sports and to spend time with others in group activities. It is especially interesting that adolescents who were focused on social possible selves reported spending time in sports with friends or supervised by adults but little time alone doing sports. For those focusing on human capital possible selves the pattern was reversed; they spent the time in sports alone and not with friends or supervised. Often, a strong focus on human capital was associated with wanting to become a professional athlete or working toward a scholarship. Therefore, they may be spending time alone in sports to improve their skills in order to reach these goals. In contrast, those focusing on social capital were more likely to be playing for the social aspect of the sport rather than solely trying to improve their own skills. In a final analysis of time-use, patterns were analyzed to understand how different identity styles and types of possible selves were associated with whom adolescents? time was spent. Two clear patterns were found. First, the use of an informational identity style was consistently associated with time spent with friends but not with parents (i.e., informational identity style was positively associated with time spent with friends in chores, schoolwork, sports, and religious activities; it was negatively associated with time spent with parents in these four activities). This is contrary to what would be expected because time spent with friends is often associated with negative outcomes (Larson, 2001). It could be that these adolescents are more independent or are allowed to do more activities with their friends than are other adolescents because they have shown responsibility to their parents. Also of interest, those with social capital possible selves 88 were spending time with parents or adults as opposed to alone. This could indicate action-taking by these adolescent in building social capital as a way of reaching their goals. Therefore, time-use may be a better indicator than GDB of how adolescents with social capital goals were participating in action-taking. Analysis of high and low action-taking. Examination of the PSQ Interviews of high and low action-takers and their parents revealed striking differences. Not only were there apparent differences between the study variables of the two groups (i.e., identity style, parental support), their language in describing their goals and actions was clearly different as well. High action-takers and their parents used clear and specific language when describing GDB and time-use. The language of low action-takers, however, made it appear that they had not done much thinking about the possible selves they were describing. Similarly, high action-takers had clearly defined career goals they were aiming to reach, whereas low action-takers merely had broad ideas about the types of career they might like. Not having a clearly defined career or job position to work toward may make it harder for these adolescents to begin preparing for that career (Inglehart, et al., 1989; Leondari, et al., 1998; Oyserman, et al., 2004) and possibly would contribute to lower levels of GDB and productive time-use. Differences also appeared in how adolescents and parents described parental support. High action-takers and their parents described high amounts of instrumental support. In contrast, this type of support was underrepresented in the interviews of the low action-takers. Because the interview emphasized instrumental support for action- taking, the lack of instrumental support as described by low action-takers is more meaningful. 89 Finally, there was a distinct difference between the identity-related language of the two groups. High action-takers described themselves as being focused, determined, and independent, as did their parents. These are qualities of someone using an informational identity style (Berzonsky, 1990). Therefore, these qualities likely may be a guiding force in GDB and time-use activity choices. In contrast, low action-takers seemed to avoid decisions about identity and let life circumstances determine their future. These characteristics may contribute to the lack of GDB and productive time-use activities of these adolescents. Conclusions, Limitations, and Future Directions The aims of the current study were to examine variation in adolescent action- taking and to explore factors that may be associated with action-taking in adolescence. Because adolescence is a time of self-exploration and transitions, this developmental period is an ideal time to study action-taking. In the current study, there was a wide range of action-taking expressed by adolescents. High and low action-takers were very different in their identity styles, parental support, and the ways they talked about their possible selves and identity. Identity styles, particularly the use of an informational identity style, appeared to be related to action-taking behaviors in adolescence. Likewise, maternal support seemed to be an important factor in adolescent action-taking. The assessment of possible selves and GDB did not result in the expected associations. A more extensive examination of the motivational aspects of possible selves as goals is needed to better understand this concept. However, consistent with literature (Eccles, et al., 2003) possible selves categories did appear to be related to certain time-use activities. 90 Certain limitations existed in the current study. The study was based on a larger study examining identity formation and capital building efforts of rural youth. Action- taking behaviors, especially GDB were gleaned from an interview with a main focus of addressing possible selves as opposed to action-taking. Although this data collection method resulted in the ability to assess action-taking, data on action-taking was limited, and therefore, the full realm of action-taking may not have been represented. In the future, studies addressing action-taking should probe participants to discuss more detailed aspects of GDB and time-use activities. The understanding of GDB complexity was limited in the current study. In future research, GDB complexity should extend the work of Oyserman by considering not only the number of goals and GDB, but also what possible selves adolescent are aiming to achieve and the diversity of strategies for attaining those possible selves. This study was limited in understanding time-use. First, adolescents and parents completed the Time Sort activity together, and therefore, it is uncertain what influence parents had in directing the time sort activity. Also, the current study did not take into account why adolescents were involved in certain activities (i.e., the intentions, motivations, and requirements behind time-use activities were not addressed). In future studies of time-use, adolescents and parents should do the time sort independent of one another and research should assess the intentions and motivations of time-use. Also, it is unclear from the results the directionality of time-use and identity formation. Longitudinal studies should focus on whether identity influences time-use or time-use activities influence identity (Dworkin et al., 2003). 91 This study was limited in the way possible selves were assessed. Simple correlations did not fully examine the full range of adolescents? goals and the efforts being made in goal-attainment. Future research should focus on how adolescents and parents talked about their goals as well as the characteristics of the goals themselves (e.g., specific versus vague, how long the goals have existed, the reasons for choosing certain goals, etc.). Finally, the current study was designed as a pilot project and included a small sample of adolescents and parents. The small sample size made it difficult to detect significant associations among variables and did not permit analyses that examined multivariate relationships. In spite of these limitations the current study did reveal differences between high and low action-taking in adolescence as well as links between action-taking, identity formation, and parental support. High action-takers, defined by their high scores in GDB and productive time-use, also scored higher than low action-takers on informational identity style and perceived parental support. These adolescents expressed a strong desire and readiness to fulfill the goals they had set for themselves, as evidenced by their own reports as well as by the perspectives of their parents. Low action-takers, however, did not appear to be prepared to take advantage of goal-related opportunities or to address obstacles they may soon encounter. In conclusion, future research is needed to more fully understand the links between action-taking, identity formation and parental support. Samples with adequate representations of male and female adolescents and adolescents of diverse race and SES should be used in future studies of action-taking efforts. Assessing race, SES, and gender 92 differences of participants may capture more complex patterns of action-taking. Furthermore, longitudinal studies addressing goal achievement would add to the literature. Assessment of outcome variables such as academic achievement, educational and career attainment, and other goal achievements are needed to better understand action-taking. The use of simple correlation analysis limits the ability to look at complex associations between the action-taking, identity formation, and parental support variables. Future research should expand the assessment of parental support, specifically by assessing types of support and how support links to particular possible selves. For example, many of the high action-takers in the current study mentioned instrumental support from their parents for specific possible selves while low action-takers talked about parental support in more vague and general terms. Looking at how different types of parental support (e.g., instrumental, encouragement, advise giving) affect action-taking would contribute to the action-taking literature (Young, 1994). Thus, future studies should examine more fully the action-taking language used by adolescents and their parents. The findings of the current study offer some insights regarding why some adolescents are more involved in action-taking than are others, and add to the literature on identity style, possible selves, and parental support as well as action-taking. The results also may have implications for youth program development and implementation. Development of programs or curricula that facilitate adolescent action-taking, help them develop possible selves as goals, and provide them support could promote positive youth outcomes (Ferrer-Wreder, Lorente, & Kurtines, 2002; Hock et al., in press). 93 REFERENCES Aquilino, W. S. & Supple, A. J. (2001). Long-term effects of parenting practices during adolescence on well-being outcomes in young adolescents. Journal of Family Issues, 22, 289-308. Bartko W. T., & Eccles, J. E. (2003). Adolescent participation in structured and unstructured activities: A person-oriented analysis. Journal of Youth and Adolescence, 32, 233-241. Bean, R. A., Bush, K. R., McKenry, P. C., & Wilson, S. M. (2003). The impact of parental support, behavioral control, and psychological control on the academic achievement and self-esteem of African American and European American adolescents. Journal of Adolescent Research, 18, 523-541. Berzonsky, M. D. & Ferrari, J. R. (1996). Identity style and decisional strategies. 20, 597- 606. Berzonsky, M. D. (1990). Self-construction over the life span: A process perspective on identity formation. Advances in Personal Construct Psychology, 1, 155-186. Berzonsky, M. D. (1992). Identity style and coping strategies. Journal of Personality, 60, 771-788. Berzonsky, M. D., Nurmi, J. E., Kinney, A., & Tammi, K. (1999). Identity processing style and cognitive attributional strategies: Similarities and difference across different contexts. European Journal of Personality, 13, 105-120. 94 Brandtstadter, J. (1998). Action perspectives on human development. In W. Damon & R. M. Lerner (Eds.), Handbook of Child Psychology: Volume 1: Theoretical Models of Human Development (5 th ed.) (pp. 807-863). John Wiley & Sons, Inc.: New York. Brown, R. S. (1993). A primer on Q methodology. Operant Subjectivity, 16, 91-138. Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, S95-S120. Dworkin, J. B., Larson, R., & Hansen, D. (2003). Adolescents? accounts of growth experiences in youth activities. Journal of Youth and Adolescence, 32, 17-26. Eccles, J. S & Barber, B.L. (1999). Student council, volunteering, basketball, or marching band: What kind of extracurricular involvement matters? Journal of Adolescent Research, 14, 10-43. Eccles, J. S., Barber, B. L., Stone, M., & Hunt, J. (2003). Journal of Social Issues, 59, 865-90. Erikson, E. H. (1968). Identity, Youth, and Crisis. New York: W.W. Norton. Ferrer-Wreder, L., Lorente, C. C. & Kurtines, W. (2002). Making life choices program: A psychoeducational intervention for promoting critical cognitive and affective competencies in adolescents at-risk for problem behaviors. Journal of Adolescent Research, 17, 168-187. Friedman, S. L., Kofsky Scholnick, E., & Cocking, R. R. (1987). Reflections on reflections: What planning is and how it develops. In S.L. Friedman, E.K., Scholnick, & R.R. Cocking (Eds.) Blueprints for Thinking: The Role of Planning 95 in Cognitive Development (pp. 515-535). New York, NY, US: Cambridge University Press. Gilman, R., Meyers, J., & Perez, L. (2004). Structured extracurricular activities among adolescents: Findings and implications for school psychologists. Psychology in the Schools, 41, 31-41. Gollwitzer, P. M. (1996). The volitional benefits of planning. In P. M. Gollwitzer & J. A. Bargh (Eds.), The Psychology of Action: Linking Cognition and Motivation to Behavior (pp. 287-312). New York, NY, US: Guilford Press. Gonzales, M. G., Burgess, D. J., & Mobilio, L. J. (2001). The allure of bad plans: Implications of plan quality for progress toward possible selves and postplanning energization. Basic and Applied Social Psychology, 23, 87-108. Hock, M. F., Deshler, D. D., & Shumaker, J. B. (in press). Enhancing student motivation through the pursuit of possible selves. In C. Dunkel and J. Kerpelman (Eds.) Possible Selves: Theory, Research and Application, Hauppauge NY: NOVA Science Publishers. Holahan, C. J., Valentiner, D. P., & Moos, R. H. (1995). Parental support, coping strategies, and psychological adjustment: An integrative model with late adolescents. Journal of Youth and Adolescence, 24, 633-648. Ingelhart, M. R., Markus, H., & Brown, D. R. (1989). The effects of possible selves on academic achievement ? A panel study. In J. P. Forgas & J. M. Innes (Eds.) Recent Advances in Social Psychology: An International Perspective (pp. 469- 477). North-Holland: Elsevier Science Publishers 96 Jacobs, J. E., Vernon, M. K., & Eccles, J.S. (2004). Relations between social self- perceptions, time-use, and prosocial or problem behaviors during adolescence. Journal of Adolescent Research, 19, 45-62. Jordan, W. J., & Nettles, S. M. (2000). How students invest their time outside of school: Effects on school-related outcomes. Social Psychology of Education, 3, 217-243. Kerlinger, F. N. (1986). Chapter 32: Q methodology. In Foundations of Behavioral Research (3 rd ed.). New York: Holt, Rinehart and Winston. Kerpelman, J. L. (in press). Using Q methodology to study possible selves. In C. Dunkel and J. Kerpelman (Eds.) Possible Selves: Theory, Research and Application, Hauppauge NY: NOVA Science Publishers. Kerpelman, J. L., Pittman, J., & Li, C. (2004). Using q methodology to examine adolescent identity: The value of a person-centered analysis strategy. A paper presented at Society For Research On Adolescents Biennial Meeting, March 2004, Baltimore. Kleiber, D. A., & Powell, G. (2005). Historical change in leisure activities during after- school hours. In J. L. Mahoney, R. W. Larson, & J. S. Eccles (Eds.), Organized Activities as Contexts of Development (pp. 23-44). Mahway, NJ: Lawrence Erlbaum Associates. Larson, R. W. (2001). How U.S. children and adolescents spend time: What it does (and doesn?t) tell us about their development. Current Directions in Psychological Science, 10, 160-164. Leondari, A., Syngollitou, E., & Kiosseoglou, G. (1998). Academic achievement, motivation, and possible selves. Journal of Adolescence, 21, 219-222. 97 Lerner, R. M., Theokas, C., Jelicic, H. (2005). Youth as Active Agents in Their Own Positive Development: A Developmental Systems Perspective. In W. Greve, K. Rothermund, & D. Wentura (Eds.), The adaptive self: Personal continuity and intentional self-development (pp. 31-47), Ashland, OH, US: Hogrefe & Huber Publishers. M??tt?, S., Nurmi, J. E., & Majava, E. M. (2002). Young adults? achievement and attributional strategies in the transition from school to work: Antecedents and consequences. European Journal of Personality, 16, 295-311. Mahoney, J. L., Larson, R. W., Eccles, J. S., & Lord, H. (2005). Organized activities as development contexts for children and adolescents. In J. L. Mahoney, R. W. Larson, & J. S. Eccles (Eds.), Organized Activities as Contexts of Development (pp. 3-22). Mahway, NJ: Lawrence Erlbaum Associates. Marcia, J. E. (1966). Development and validation of ego identity status. Journal of Personality and Social Psychology, 3, 551-558. Marcia, J. E. (2001). A commentary on Seth Schwartz?s review of identity theory and research. Identity, 1, 59-65. Markus, H. & Ruvolo, A. (1989). Possible selves: Personalized representations of goals. In Goal Concepts in Personality and Social Psychology, Pervin, L.A. (Ed.). Hillsdale, NJ: Lawrence Erlbaum. Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist, 41, 954-969. 98 Marshall, S. K., Young, R. A., & Domene, J. F. (in press). Possible selves as joint projects. In C. Dunkel and J. Kerpelman (Eds.) Possible Selves: Theory, Research and Application, Hauppauge NY: NOVA Science Publishers. McKeown, B., & Thomas, D. (1988). Q Methodology. Series: Quantitative Applications in the Social Sciences. Newbury Park, CA: Sage Publications, Inc. Nurmi, J. E. (1987). Age, sex, social class, and quality of family interaction as determinants of adolescents' future style: A developmental task interpretation. Adolescence, 22, 977-991. Osgood, D. W., Anderson, A. L., & Shaffer, J. N. (2005). Unstructured leisure in the after-school hours. In J. L. Mahoney, R. W. Larson, & J. S. Eccles (Eds.), Organized Activities as Contexts of Development (pp. 45-64). Mahway, NJ: Lawrence Erlbaum Associates. Oyserman, D. (2004). Possible selves citations, measure, and coding instructions. Retrieved April 22, 2005 from: http://sitemaker.umich.edu/daphna.oyserman/files/possible_selves_measure.doc Oyserman, D., Bybee, D., Terry, K., & Hart-Johnson, T. (2004). Possible selves as roadmaps. Journal of Research in Personality, 38, 130-149. Pedersen, S., & Seidman, E. (2005). Contexts and correlates of out-of-school activity participation among low-income urban adolescents. In J. L. Mahoney, R. W. Larson, & J. S. Eccles (Eds.), Organized Activities as Contexts of Development (pp. 85-110). Mahway, NJ: Lawrence Erlbaum Associates. Schwartz, S. J. (2001). The evolution of Eriksonian and Neo-Eriksonian identity theory and research: A review and integration. Identity, 1, 7-58. 99 Shanahan, M. J., & Flaherty, B. P. (2001). Dynamic patterns of time-use in adolescence. Child Development, 72, 385-414. Shoffner, M.F., & Kerpelman, J. L. (2005). Examining the possible selves and career development of rural adolescent girls through the use of Q-methodology. Unpublished Manuscript, University of North Carolina at Greensboro. Teachman, J. D., Paasch, K., & Carver, K. (1997). Social capital and the generation of human capital. Social Forces, 75, 1343-1359. Tepper, R L. (2001). Parental regulation and adolescent discretionary time-use decisions: Findings from the NLSY97. In R.T Michael, (Ed). U Chicago; Irving B. Harris Graduate School of Public Policy Studies. Social Awakening: Adolescent Behavior as Adulthood Approaches (pp. 79-105). New York, NY, US: Russell Sage Foundation. Thompson, B. (1980). Comparison of two strategies for collecting Q-sort data. Psychological Reports, 47, 547-551. von Cranach, M., Kalbermatten, U., Indermuhle, K., & Gugler, B. (1982). Goal-Directed Action, Academic Press: London. Wilson, P. M., & Wilson, J. R. (1992). Environmental influences on adolescent educational aspirations: A logistic transform model. Youth and Society, 24, 52-70. Young, R. A. (1994). Helping adolescents with career development: The active role of parents. Career Development Quarterly, 42, 195-203. Young, R. A., & Valach, L. (2004). The construction of career through goal-directed action. Journal of Vocational Behavior 64, 499-514. 100 APPENDICES Appendix A A.1 Current Identity Q-Sort Items 101 A.1 Continued Current Identity Q-Sort Items 102 A.2 Possible Selves Q-Sort Items Note: A = Human Capital, B = Economic Capital, C = Social Capital 103 A.2 Continued Possible Selves Q-Sort Items Note: A = Human Capital, B = Economic Capital, C = Social Capital 104 A.2 Continued Possible Selves Q-Sort Items Note: A = Human Capital, B = Economic Capital, C = Social Capital 105 A.2 Continued Possible Selves Q-Sort Items Note: A = Human Capital, B = Economic Capital, C = Social Capital 106 107 A.3 PSQ Interview Questions Instructions: We now are going to do a brief interview. I will ask you a few questions that pertain to the importance of your possible selves, plans for obtaining them, and possible obstacles you might encounter. GOALS, PLANS, BARRIERS Ask the participant to look at the items in columns 8 & 9. Have the adolescent respond to the following: 1. Describe why you selected these cards as most like you in the future. probes: Why did you pick these cards over the others? Tell me why these cards best fit who you will be? BEFORE MOVING ON ASK: ?What about any of the other items you placed in these two columns?? CONCLUDE BY ASKING: ?Is there anything else?? 2. Describe what you are planning to do to reach your possible self. probes: What kinds of steps will you take to get there? What kinds of activities will you need to do? BEFORE MOVING ON ASK: ?Are there any other things you are doing or planning to do to reach your possible self?? CONCLUDE BY ASKING: ?Is there anything else?? 3. Is there anyone who will help you to become who you intend to be in the future? What will they do to help? probes: What will this person do to make it easier for you to attain your goals for yourself? In what ways will this person offer you support as you move toward the future? CONCLUDE BY ASKING: ?Is there anyone else who might help you reach your goals? How?? 108 A.3 Continued 4. Is there anything that could get in the way of you becoming who you intend to be in the future? How might you deal with that? probes: What kinds of obstacles might you face? What can you do to overcome with these obstacles? BEFORE MOVING ON ASK: ?Are there any other obstacles you can think of?? CONCLUDE WITH ASKING: ?Is there anything else?? 109 A.4 Time Sort Protocol Set-up: Before the family enters the room, have the following ready: - time sort board - hour cards (168 total) - calculator - note pad - pens/pencils Read the bolded words aloud: Part 1: This last activity, called a Time Sort, involves sorting these hours [point to the hour cards on the table] into these areas of activity [read off a few of the activity areas] according to how much time [name of teen] spends in them during a typical week [show the family where to place the hours by putting one of the hour cards into one of the activity slots]. As a family, discuss how the time should be divided and decide together how [name of teen]?s time during a typical week is best represented. Please do not open any of the flaps (we will do that later). The first thing we would like you to do is to write down your school schedule for us. This list should include all of your classes plus, sports, band, choir, clubs, etc. You have a calculator and notepad to use. For categories with many different tasks (such as leisure) we recommend that you write down these different activities as this will help you with a second part you will be doing for this activity that I will explain later. We suggest that you start with the sleep and school categories, since they will take some of the largest chunks of time. Don?t worry if you do not have hours in some of the categories, any time left over can go in the ?other? category. ** Don?t forget: - Meals eaten at school should count as part of school time and should not be counted again in the meal time category. - Varsity sports practice during school hours could be considered ?sports? or ?school? but not both. - Time spent in self-care can be categorized as ?other.? Do you have any questions before you begin? 110 A.4 Continued Ask the family to report how many hours are in the ?other? category. If there are more than 10 hours in the ?other? category ask the family if they think they did not put quite enough hours into some of the categories. Let them sort some of these additional hours if they choose, or put them back into the ?other? slot if they want to keep them there. Part 2: [Open the ?doors? - except to the ?sleep?, ?school?, and ?non face-to-face time with friends? categories]. Your second task is to take the hours in each of these categories with the open doors and to divide them up according to how much time you spend doing the activities in this category alone, with parents, with friends, or with parents and friends together. [Show where the time card goes for each of these categories. The time spent with parents and friends together should be placed behind the back panel. Also explain that time with siblings should go in the ?friends? slot, and time with other adults should go in the ?parents? slot]. The facilitator should step away from the family, allowing them to sort the time cards, yet remain easily accessible for questions and to distribute ?change? cards if needed. 111 A.4 Continued Time Sort Code Sheet Activity Parent(s)/ Adult(s) Friends Parents & Friends Alone Total Time Sleep School Non Face-to-Face Time with Friends Kitchen/Home/Yard Maintenance Paid Work Homework/Studying Reading Books for School Lessons/Tutoring Sports/Physical Activity Religious Activities Extracurricular/ Volunteer Activities Meal Time Leisure Activities Hobbies Other TOTAL HOURS _______ 112 Appendix B B.1 Q-Analytic Procedure The Q-analytic procedure takes advantage of the full set of possible selves and involves several steps (Kerpelman, in press). First, the Q-sort data placements are recorded on a code sheet and then entered into a data file. The Q-sort items label the variables across the top, and the location scores for each person?s sort are then entered. Next, the dataset is transposed so that the participants label the columns, making the respondents the variables to be analyzed. This makes it possible to analyze the data according to each person?s full set of responses to the Q-sort. Thus, the goal of Q- analysis is to group ?like people? who are similar in how they sorted the Q-sort items. After the data are prepared, a factor analysis is run, which involves the calculation of the correlations among all the possible pairs of Q-sorts and then factor analyzing them (Kerlinger, 1986; Thompson, 1980). The size of each person?s loading on each factor represents the association of that individual?s Q-sort with the ?way of thinking? associated with each factor (Brown, 1993; McKeown & Thomas, 1988). The last step is to create ?composite Q-sorts? from those individuals who load highest on a factor and do not cross load on other factors. Participants who have sorted the Q-sort items similarly will load significantly on the same factor; the higher a person?s factor loading the more that person is consistent with the underlying meaning of that factor (Kerlinger, 1986). For each person to be included in the composite sort, a weight is calculated (w=f/(1-f 2 )) (Brown, 1993; McKeown & Thomas, 1988). After calculating the weights 113 for each person, each person?s placement score for each of his or her items in the sort is multiplied by his/her factor weight, and then summed across people for each item, creating the final location score for each item of the composite sort. These scores along with their respective item numbers can be sorted in descending order to create the composite sort for each factor. Once created, the composite sorts can be compared to determine which items most differentiate the groups. According to Brown (1993), a difference of two or more columns is considered to be significant. These differences help to elucidate the distinct ways of thinking that each group represents. A qualitative description is used to convey the focus of each of the Q-groups that emerge for the adolescents and for the parents. A correlation was computed for each participant with each of the PSQ composite sorts to indicate the extent to which he or she ?thinks like? each group (similar to the procedure used with the identity style criterion sorts). 114 B.2 GDB Coding Procedures The adolescents? GDB used to attain their future goals (i.e., possible selves), as well as the parents? perceptions about the adolescents? GDB, were collected through an open-ended interview called the Possible Selves Q-sort Interview. Trained interviewers asked each participant to view a list of his/her ?most like me (like my teen)? possible selves obtained from the PSQ. The participant was asked to describe what he or she (or his or her adolescent) was doing currently or was intending to do in order to reach the possible selves. Following data collection, each PSQ Interview was transcribed verbatim and prepared for analysis. In order to apply coding schemes to the interviews, several preliminary steps were taken to create charts that were used in further analysis. At least 2 coders were assigned to each transcript. For the first step in preparing the charts, each transcript was read to determine and list every GDB mentioned. GDB were defined as any action or behavior the adolescent mentioned participating in at the time of the interview, or would participate in during the high school years, used in obtaining or working toward his or her possible selves. Second, coders determined to which possible self each GDB referred, according to statements made in the PSQ Interview. For example in the statement: Respondent: I am working on getting a scholarship so I can go to college. ?Working on getting a scholarship? would be the GDB and ?Get a College Education? would be the goal to which it referred. GDB that were mentioned but could not be traced to a specific possible self were assigned to a ?General? category. For example, 115 Interviewer: What are you planning to do to reach your goals or possible selves? Respondent: I will work hard and make good grades in my classes. In this case the GDB, ?work hard? and ?make good grades in classes? would be assigned to a ?General? category because the respondent did not refer to a specific possible self. Next, coders grouped the GDB that were similar together using letters to designate groups. For example, a list of GDB may include: work hard in school (a), get good grades (a), shadow a doctor for a day (b), save money in a checking account (c), earn money at work (d), and get a job to save for college (d). GDB with the same letter are similar to each other in how the adolescent intends to attain his or her goals. Each coder compiled their lists into a chart to be used to compare with each other. The following is an example of what a chart looked like: Goal Categories GDB A. General Trying to get good grades (a) Doing a lot of sports activities (b) B. Play Professional Football Playing football in high school (b) Working hard in weight room (c) C. Own a Business Going to take Economics (d) Is going to work in school store (e) D. Go to College Working hard to do well on Grad Exam (a) Working hard to do well on ACT (a) Studying hard (a) After coders individually coded assigned transcripts, they came together to discuss what categories were assigned and why. This procedure was used in order to 116 reach an agreement between coders and a final revised chart for each interview was made. Three coding schemes were applied to each interview. The first was based on Oyserman (2004) and was used to assess the complexity of adolescents? GDB and parents? perceptions of their adolescents? GDB. This coding scheme combined the number of possible selves categories mentioned (e.g., Go to College, Career) and the number of GDB generated in the interview. Coding scores were as follows: Score Possible Selves Categories Number of GDB 1 1 2 1-4 2 2 2 3-4 3 2 6+ 4 3 3-5 5 3 6+ 6 4 4-5 7 4 5 6+ 5+ Because each adolescent had the same possible selves items to choose from when interviewed (i.e., they did not generate possible selves on their own), a second score was used to assess self-generated answers. GDB diversity is found by counting the letters assigned when finding similar GDB. The higher the number of letters, the more diverse GDB were being used by the adolescent. 117 Therefore, according to the first two coding schemes, the participant represented by the above chart would receive the following scores: NAME OF SCORE HOW TO FIND SCORE RESULTING SCORE GDB Complexity: Use Oyserman?s Coding Scheme 7 GDB Diversity: Count the number of letters given to represent similar GDB 5 The third and final coding scheme was used to explore the relationship between GDB and possible selves. It was hypothesized that the use of GDB to build certain types of capital would be related to the types of capital adolescents choose as their ?most like? possible selves (e.g., an adolescent who placed a majority of economic possible selves in his top two columns would describe economic GDB). For this coding it was determined which capital category (i.e., human, economic, or social capital goals) each GDB reflected (e.g., ?volunteered at the hospital? reflects human capital goals; ?balanced own checking account? reflects economic capital goals). Next, the number of human, economic, or social capital GDB the individual had was counted. This step yielded three scores for each adolescent or parent: number of human capital GDB, number of economic capital GDB, and number of social capital GDB. An example of an adolescent?s capital related GDB would look like: Goal Category GDB A. General Not slipping/letting up in school (HC) B. Have support system Being a good example to friends (SC) C. Religion Go to church (SC) D. Financial stability Started checking account (EC) 118 Therefore, the participant represented by the above chart would receive the following scores: NAME OF SCORE HOW TO FIND SCORE RESULTING SCORE Number of HC GDB Count number of HC GDB 1 Number of EC GDB Count number of EC GDB 1 Number of SC GDB Count number of SC GDB 2 119 B.3 Q-Group Descriptions The q-analysis resulted in two types of thinking about possible selves among the adolescents. The first group was highly focused on human capital possible selves. Many of these possible selves dealt with aspects of working in the medical or health field (e.g., having a job that includes saving or protecting lives, having a job that will require a lot of time). This group placed a moderate emphasis on economic capital. They described wanting to own nice things but also wanting to give money to a charity or church and looking to their parents to help them if they needed financial assistance. They also had a moderate amount of social capital items but most were placed in columns 8 and 7 rather than column 9, where many human capital items were placed. The social capital items were balance between relationship-oriented items (i.e., receiving support and help when they were needed and on maintaining close relationships with family and friends) and on duties they wanted to fulfill (i.e., being a churchgoer, a spouse, a parent, and a best friend). Group 1 seemed to be very certain of the career area they would like to be in. The majority of their ?least like? human capital items were careers they did not want to have (e.g., I want to work in the business field). Their ?least like? social capital items focused on civic groups they did not want to be involved in (e.g., join a county club, play in a band.) Group 2?s ?most like? possible selves focused mostly on social capital. They had a strong desire to maintain close relationships with family and friends. The items focused on having help and support from others as well as the interest in spending time with family and being involved with their future adolescents? activities as well as being 120 involved in several civic groups. This group?s possible selves also focused on aspects of economic capital. Areas of importance were financial stability (e.g., having savings in the bank) and material goods (e.g., owning a home and a car). Group 2?s possible selves had a limited focus on human capital, only listing four items in their top three ?most like? columns. Those items focused on further education and wanting to be happy in the job they chose although having a job in sports or athletics was the only specific type of job mentioned. Group 2?s ?least like? possible selves were primarily human capital items. These included numerous types of jobs they would not like to have as well as having a job that required a good amount of time and that would take a long time to master. Only a few economic possible selves items were mentioned and the only social capital items mentioned referred to civic groups in which they would not be involved. Although the possible selves of Group 1 and Group 2 were ordered differently, several items were mentioned by both groups. Both groups focused on social possible selves in which family and maintaining close relationships were valued. In addition, both groups? sorts had items that focused on being a parent, a spouse, and a best friend. Economic possible selves such as owning nice things and having savings were important in each group. Although Group 2 only had 4 human capital items in their ?like me? columns, both groups ranked having a formal education and a highly skilled job as important. A few differences separated the groups. Items that were placed in Group 1?s ?most like? and in Group 2?s ?least like? were having a job in the medical field that involved saving or protecting lives, having a job that required a lot of time, and working with children or youth. Items placed in Group 2?s ?most like? but in Group 1?s ?least 121 like?? were having a job in sports or athletics, volunteering to coach a sport and playing a team sport. The majority of the similarities in the groups? ?least like? columns referred to the jobs they would not like to have and groups and time-use in which they did not desire to be a part of.