xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200385Ka 4500
controlfield tag 001 002029491
007 cr mnu|||uuuuu
008 090917s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002911
Does goal generation vary as a function of depressive status?
h [electronic resource] /
by Lindsay Brauer.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 59 pages.
Thesis (M.A.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: Although many theories attempt to explain the manner in which deficits in goal generation may lead to an episode of Major Depressive Disorder (MDD), few studies have actually described how persons with depression vulnerability generate goals for themselves. The current project used a multi-dimensional approach to examine the types, numbers, and domains in which goals are generated, comparing individuals with MDD with individuals with remitted depression, and never-depressed controls. In addition, we examined possible mechanisms that may mediate the relationship between depressive status and goal generation. Surprisingly, results suggested little group difference in the types of goals or subjective ratings of goals that were generated. Instead, we suggest that cognitive and behavioral aspects of motivation may underlie an individual's efficacy of achieving goals. Limitations and future directions are discussed.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Jonathan Rottenberg, Ph.D.
t USF Electronic Theses and Dissertations.
D oes Goal Generation Vary as a Function of Depressive Status? by Lindsay Brauer A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Psychology College of Arts and Sciences University of South F lorida Major Professor: Jonathan Rottenberg, Ph.D. Jennifer Bosson, Ph.D. Mark Goldman, Ph.D. Date of Approval: March 9, 2009 Keywords: depression, goals, motivation, female, undergraduate Copyright 2009, Lindsay Brauer
Dedication To Mom, Dad, and Grandpa.
i Table of Contents List of Tables iii Abstract iv Background 1 Diagnostic Criteria for Depression 1 D efinition and Assessment of Goal Generation 3 Goal T ypology 3 Goal Organization 5 Potential Mechanisms for Goal Related Deficits in Goal Generation 7 Broader Motivational Systems 8 Goal Orientation 11 Do Differ ences in Goal Generation Reflect States or Traits? 12 Limitations of Previous Research 13 The Current Study 14 Specific Aims 16 Exploratory Aims 19 Method 20 Participants 20 Materials 21 Group Determination 21 Anxiety 23 Dispositional Behavioral Inhibition and Activation Sensitivities 24 Goa l Orientation Inventory 24 Goal Generation 25 Procedure 26 Goal Coding 27 Typology (Approach v. Avoidance) 27 Complexity (Simple v Complex) 27 Abstracti on (Abstract v. Concrete) 28 Domain 28 Results 30
ii Descriptives 30 Goal Generation ( Specific Aim 1: Comparison of approach and avoi danc e goals across groups) 32 Behavioral System Sensitivity (Specific Aim 2: Examination of the mediational role of BAS sensitivity in goal generation as a function of depressive status ) 32 Goal Or ientation (Specific Aim 3: Examination of the mediational role of goal orientation in the generation of approach and avoidance goals as a function of depressive status ) 32 Number of Goals and Do mains of Generation (Specific Aim 4: Examination of the number of goals generated and domains of generation as function of depression status ) 33 Exploratory Analyses 34 Level of Abstraction (Specific Aim 5: Ex amination of the average level of abstractness of goals as a function of depression status) 34 Leve l of Complexity 34 Sub jective Ratings 35 Depressive Symptoms and Goal Generation 35 Discussion 37 Limitations and Future Directions 45 S ample Demographics 45 Sample Severity 45 Context 46 Conclusion 42 References 48 Appe ndices 54 Appendix A : Revised Goal Task 55
iii List of Tables Table 1 Demographic i nformation 21 Table 2 Goal coding results by group and goal domain 31 Table 3 Scale scores 33 Table 4 Subjective r atings 35 Table 5 Mean number of depressive symptoms by gro up 36
iv Does Goal Generation Vary as a Function of Depressive Status? Lindsay Brauer ABSTRACT Although many theories attempt to explain the manner in which deficits in goal generation may lead to an episode of Major Depressive Disorder (MDD), few studies have actually described how persons with depression vulnerability generate goals for themselves. The current project used a multi dimensional approach to examine the types, numbers, and domains in which goals are generated, comparing individuals with MDD wit h individuals with remitted depression, and never depressed controls. In addition, we examined possible mechanisms that may mediate the relationship between depressive status and goal generation. Surprisingly, results suggested little group difference in the types of goals or subjective ratings of goals that were generated. Instead, we suggest that cognitive and behavioral aspects of motivation may underlie an individuals efficacy of achieving goals. Limitations and future directions are discussed.
1 Background Major Depressive Disorder (MDD) is a chronic illness that afflicts nearly 20% of the population (Kessler, 2002). MDD is also highly recurrent: by some estimates, at least 60% of those who suffer from a single episode of depression will likel y experience another; 70% who experience two episodes will experience yet another episode, and almost 90% of those who experience three episodes will experience yet another episode (Monroe & Harkness, 2005). In addition, the average depress ive episode lasts four to nine months (Durand & Barlow, 2006). The recurrent nature of MDD has motivated extensive research into factors that may promote and maintain depressive episodes Research on the assessment and treatment of MDD has begun to focus on goal genera tion and the manner in which it is dysregulated by the cognitive, affective, and motivational deficits associated with the disorder ( Strauman et al., 2006; Dimidijian et al., 2006) As such, the current project seeks to investigate differences in goal gen eration (i.e., typology, level of abstraction, level of complexity, and domain) as a function of depressive status Diagnostic C riteria for D epression MDD is characterized by the cardinal symptoms of a persistently depressed mood and/or the presence of an hedonia for two weeks or longer. In addition, to meet criteria for MDD individuals must endorse several of the following associated symptoms: significant change in appetite, weight loss or weight gain, difficulty sleeping,
2 psychomotor agitation or retarda tion, difficulty concentrating or making decisions, excessive or inappropriate guilt, feelings of worthlessness, or suicidal thoughts or attempts. Diagnostic criteria for MDD require that an individual endorses at least one cardinal symptom, with five sym ptoms in total (American Psychiatric Association [APA], 2000). The DSM IV TR symptoms of depression do not, of course, capture the complete array of deficits associated with MDD. For example, individuals with MDD often have difficulty managing everyday t asks such as personal hygiene ( APA, 2000), experience a reduced responsiveness to reward ( Kasch, Rottenberg, Arnow, & Gotlib, 2002), have difficulty recalling detailed id i ographic memories (Williams & Broadbent, 1986) and may have difficulty pursuing and generating specific goals (Street, 2002). Although there is evidence that qualities such as responsiveness to reward and appetitive motivation may be trait like q ualities of the disorder ( Kasch et al., 2002), it is unclear to what extent other characterist ics, such as goal generation, may be state dependent qualities produced by MDD. DSM diagnostic criteria, cl inicians, and research have often asserted that MDD involves deficits in goal related activity. However, upon closer inspection, most descriptions of these deficits are not well specified. The generation of goals potentially refers to a complex set of processes, and the field has often been imprecise and reli ant on lay terms or informal conceptualizations to explore goal related activities in MDD. For tunately, recent advances in social and personality psychology provide additional
3 theoretical and methodological sophistication in this domain. Developments regarding the typology and generation of goals as they pertain to MDD will be discussed. Definitio n and Assessment of Goal Generation Although goal generation is defined slightly differently by researchers working in different fields (APA, 2000; Dickson & MacLeod, 2004, 2006; Higgins 2006; Dykman, 1998), we define goal generation as t he ability to generate personal accomplishments for the future. This definition w as chosen as research suggests that the ability to plan for the future (generate goals) is a process which is disrupted in individuals with MDD (Higgins, 1996; Strauman et al., 2006). To date the paucity of research regarding goal generation has relied upon the individuals capacity and willingness to list desired future events in response to specific prompts. For example, Dickson and MacLeod (2004, 2006) utilized the Goal Task, which requi red individuals to respond to prompts regarding goals to be achieved or avoided in the future. Similarly, Lecci, Karoly, Briggs, and Kuhn (1994) asked participants to list personal goals which they were planning or pursuing in predetermined domains (e.g., academic, health, leisure, family relationship, social/nonfamily). Thus, research has yet to utilize tasks in which individuals respond to less directive prompts, or when goals were examined in a multidimensional manner. The following section explicate s the various theoretical dimensions upon which goals can be examined. Goal Typology One important typology of goals from basic research on personality and motivation has been articulated by Higgins (1996), who proposes that goals fall within
4 two distinc t categories that facilitate self regulation: promotion or prevention. Promotion goals are characterized by themes of advancement, achievement, experience of positive affect (i.e. cheerfulness), and focus on positive outcomes (i.e., pursuing a promotion at work). Prevention goals, conversely, are characterized by themes of safety, responsibilities, experience of feelings of calmness, and focus on preventing negative outcomes (i.e., not losing a current job). Strauman et al. (2006) draws upon this distinc tion between promotion and prevention goal pursuit and applies it to individuals with MDD. The authors suggest that individuals with a history of failed attempts to attain promotion goals (particularly in a social domain) have future difficulty attaining these goals, as they likely have not developed adequate strategies to achieve promotion goals. In addition, an inability to attain promotion goals is thought to increase an individuals vulnerability for depression ( Scott & OHara, 1993). For example, individuals who fail to achieve promotion goals may not view social situations as venues in which promotion goals can be pursued (whereas they may for nondepressed individuals) thus limiting opportunities to pursue promotion goals and experience positive a ffect (Strauman et al., 2006). As environmental cues relating to the goal being pursued provide feedback to the individual regarding his/her progress towards the goal (Klinger, 1975), the continued perception of cues of failure are likely to decrease incen tive and motivation to pursue such goals, and decrease an individuals feelings of self worth (Strauman et al., 2006). Dickson and MacLeod (2004, 2006) use similar language about in their theoretical discussion of how depressed and dysphoric individuals ge nerate promotion
5 and prevention goals. Dickson and MacLeod suggest that approach goals, akin to promotion goals, encourage the achievement of positive outcomes and maintenance of positive situations. Avoidance goals, similar to prevention goals, encourage the prevention of negative outcomes and avoidance of negative situations. Results from this body of research suggest that depressed and dysphoric individuals generated fewer approach goals but did not differ in the number of avoidance goals compared to control participants (Dickson & MacLeod 2004, 2006). These findings support the notion that individuals with low mood may have a deficit in the generation of approach or promotion goals, as suggested by Strauman et al. (2006). For the purposes of this r eview, we will refer to promotion/approach goals from this point on simply as approach goals, and prevention/avoidance goals as avoidance goals. Goal Organization It does not appear that goal generation is a simple linear process. Instead, Carver and Scheier (1990) suggest that goals are construed within a hierarchy. More specifically, small concrete goals act as intermediary steps towards achieving a large, abstract goal. Concrete goals refer to those which are clearly defined and have a defini tive point of achievement ( e.g., say hello to my neighbors every morning this week ), while abstract goals refer to those which are not clearly defined, and are broad in content (e .g., become a model citizen ; Emmons, 1992). Further, abstract goals may be m ore difficult to achieve than concrete goals, as abstract goals are usually more complex and intricate than concrete goals. More specifically, without a clear endpoint it is difficult to ascertain when an abstract goal has been achieved (Emmons, 1992) S treet
6 (2002) claims that the generation and pursuit of abstract goals has been linked to increases in negative affect and depression. In addition, Carver and colleagues (1988) suggest that abstract goals are more often connected to an individuals self wo rth (as compared to concrete goals), and as such, are more difficult from which to disengage. In addition to the type of goals a person generates, the number of goals to which an individual strives may differ across individuals with and without a history of MDD. Champion and Power (1995) suggest that the rigid fixation upon a single goal increases an individuals cognitive vulnerability to depression. With a unitary focus, the individual is thought to exert all of her/his resources to ward achieving a si ngle goal. When this goal is not achievable, this may lead to excessive persistence and difficulty in goal disengagement. Moreover, if an individual fails to attain this single goal, the failure may decrease the individuals self worth more intensely tha n if the individual had diversified goals. As such, individuals who generate a single goal may be more vulnerable to MDD than individuals who generate multiple goals. Conversely, according to Champion and Power (1995) and Street (2002) generating multip le goals may enhance a specific type of cognitive flexibility that buffers the individual from experiencing intense feelings of loss after failed goal pursuit, and may also allow an individual to disengage from an unachievable goal before resources are wasted. An alternative though related, hypothesis is that depression vulnerability may reside more in the number of domains in which these goals are generated rather than in the absolute number of goals an individual generates or pursues. For example, depr essed individuals may generate an equivalent amount of goals as nondepressed individuals, but
7 the goals of the depressed individuals may fall within a single domain while non depressed individuals may generate goals that span multiple domains. Within thi s conceptualization, depressed individuals put effort forth within a single domain, which in turn may have the same negative consequences as pursuing a single goal. Meaning, an individuals self worth is dependent upon achievement within a single domain, which increases an individuals difficulty disengaging from a goal or goals in the single domain, which is thought to increase an individuals vulnerability to depression (Champion & Power, 1995; Klinger, 1975; Nesse, 2000). Lam and Power (1991) examined t he correlation between number of goals and domains in which these goals were generated and future onset of depression in an elderly sample. Results suggest that valuing or pursuing goals in a single domain, as opposed to multiple domains was pr edictive of future depression. The absolute number of goals generated, however, was not predictive of future depression. Although this information provides insight as to how the number and domains in which goals are generated correspond to MDD vulnerability, process es that underlie both goal generation deficits and MDD vulnerability remain unclear. Potential Mechanisms for Depression Related Deficits in Goal Generation Research and theory have provided an outline of areas in which deficits in goal generation may e xist. What are the mechanisms that explain these deficits among individuals with MDD? As MDD is conceptualized as a disorder of motivation, affect, and cognition (Strauman et al., 2006), the current project proposes to examine the manner in which these t hree areas interact to promote deficits in goal generation. More
8 specifically, we propose that deficits in the behavioral motivation system and the ideographic interpretation of goal pursuit (i.e., goal orientation) mediate the relationship between depressive status and deficits in goal generation. Broader M otivational S ystems. The concept of motivation, similar to that of goal generation, is a broad construct that can be conceptualized in various ways. Motivation, as related to goal generation and goa l pursuit, can be defined in terms of responsiveness to reward (Carver & White, 1994), an individuals desire to change his/her current state of being (Higgins, 1996; Strauman et al., 2006), or the ability to persist in goal oriented activities (Ne sse, 2000). The variety of conceptualizations of motivation has been complimented by an array of assessment tools, ranging from experimental measures of reward response bias in goal attainment tasks (ex Johnson, Ruggero, & Carver, 2005) to self report measures o f personality, attitudes, or cognition (ex.Carver & White, 1994). For the purposes of this project, the term motivation is used to describe of the process by which cognitive, affective, and behavioral components drive an individual to engage with his /her environment, strive towards achieving a goal, and allow the individual to persist in these strivings until a desired state or goal is achieved. This conceptualization is premised upon the proposed connection between the core MDD symptom of anhedonia, (reduced experience of pleasure), and reduced motivation. Interestingly, anhedonia appears to encompass both the affective quality of finding activities less pleasurable than they once were (APA, 2000), as well as a reduced responsiveness to reward (Henriques & Davidson, 2000; Kasch et al., 2002), which are both integral components of goal generation and pursuit (Strauman, 2002). When
9 pleasure and responsiveness to reward associated with a given activity are reduced, the likelihood of an individual engagi ng in that activity in the future is reduced (Dickson & MacLeod, 2004, 2006). When this model of motivation is applied to goal generation, it can be theorized that reduced experience of pleasure and reward associated with goal generation and pursuit, coupled with a history of failed goal strivings (Strauman et al., 2006), may cause an individual with MDD to be apprehensive to generate goals, unresponsive to the reward of pursuing and achieving a generated goal, and may be less likely to generate and pursue similar goals than an nondepressed individual. Perhaps the most commonly used motivational constructs used in the literature are the Behavioral Activation System (BAS) and the Behavioral Inhibition System (BIS). These systems were conceived from animal models of affect and motivation, which suggested that two separate systems controlled engagement with and avoidance from the environment ( Fowles, 1980). Additionally, it is conceived that these systems are controlled by distinct areas of the brain (see Carver & White, 1994). The BAS is associated with appetitive motivation, concentration, reward responsiveness, as well as producing the emotions of elation, hope, happiness (Carver & White, 1994), and frustration and anger in situations when motivation to wards a goal is thwarted (C arver, 2004). The BIS is theoretically responsible for aversive motivation (withdrawal behaviors), responsiveness to threat or punishment, as well as feelings of anxiety, frustration, and sadness (Carver & White, 1994). While B AS sensitivity is associated with the promotion of goal related behaviors, the BIS is associated with the inhibition of goal related behaviors (Carver & White, 1994).
10 In nondepressed individuals, it is assumed that a balance between BAS sensitivity and BI S sensitivity exists, such that individuals are not euphoric or excessively responsive to reward (high BAS sensitivity) or completely uninhibited or unaware of potential harm (low BIS sensitivity); instead, non depre ssed persons are motivated by potential rewards, and aware of potential consequence of its pursuit. In individuals with MDD, however, it is theorized that individuals are thought to be anhedonic and particularly unresponsive to reward (low BAS sensitivity), and are also speculated to be nervous or anxious, and have a heightened sensitivity to cues of potential harm (high BIS sensitivity; Depue & Iacono, 1989). Furthermore, research by Kasch and colleagues (2002) support this theory, as individuals with MDD reported lower BAS sensitivity and higher BIS sensitivity on self report measures of the behavioral system sensitivity (e.g., BIS/BAS scales; Kasch et al., 2002; Carver & White, 1994), when compared to nondepressed controls. Although the mediating role of BAS in goal generation has been all uded to in previous research (Dickson & MacLeod, 2004), this relationship has yet to be tested empirically. As the BAS is associated with appetitive motivation and reward responsiveness (Carver & White, 1994), and approach goals are typified by engagement with the environment to ensure a positive outcome (Higgins, 1996; Dickson & MacLeod, 2004, 2006), it is plausible that BAS activity mediates approach goal generation. Likewise, as the BIS is associated with avoidance of aversive or potentially harmful si tuations (Carver & White, 1994), and avoidance goals are typified by those which attempt to prevent the occurrence of negative outcomes (Higgins, 1996; Dickson &
11 MacLeod, 2004, 2006), it is also plausible that BIS activity mediates avoidance goal generatio n. As evidence suggests that diminished BAS sensitivity is a trait like marker for MDD (Peeters, Nicolson, & Berkhof, 2003), and BAS sensitivity is theorized to mediate deficits in approach goal generation (Strauman et al., 2006; Dickson & MacLeod, 2004, 2006) it is possible that the deficits in promotion/approach goal generation thought to typify MDD may be sustained in individuals with a history of MDD, even when the depressive episode remits. Conversely, as there is a lack of evidence to substantiat e the importance of BIS (Johnson et al., 2003) and avoidance goals in MDD ( Strauman et al., 2006; Dickson & MacLeod, 2004, 2006) or how individuals with and without MDD differ in these domains (Dickson & MacLeod, 2004, 2006), the examination of BIS sensit ivity and avoidance goals across currently depressed, remitteddepressed, and never depressed individuals was of secondary interest in the current project. Goal Orientation. The meaning assigned to goal pursuit may be equally important mediating fact or in goal generation. Dykman (1998) has distinguished between growthseeking (GS) and validation seeking (VS) orientations. Individuals with a GS orientation have the tendency to view goals as an opportunity to learn and grow. It is theorized that assum ing a GS orientation increases a persons resistance to MDD, as failures in these pursuits are considered an inherent part of the learning process, and not reflective of the individuals competency, likability, or self worth. Conversely, an individual with a VS orientation is presumed to be particularly vulnerable to MDD, as
12 validation seekers base their self worth on the achievement of goals, and experience failure as a product of a lack of competency, likability, and worth (Dykman, 1998). In sum, Dykma n (1998) suggests that adopting a GS orientation decreases an individuals vulnerability to MDD, whereas adopting a VS orientation increases an individuals vulnerability. Based upon Dykmans (1998) depiction of goal pursuit, it may be that individuals wi th a validation seeking orientation (who are thereby vulnerable to MDD), who continually fail to achieve promotion goals (thought to characterize individuals with MDD), may not only feel dejected from not attaining the goal and receiving environmental cues of failure, but may also feel as though their self worth is continually invalidated. Theoretically, adopting a VS orientation has the potential to interfere with goal generation. As Dykman (1998) suggests, individuals with a VS orientation experience anticipatory anxiety during goal pursuit, and tend to view goals as events in which individuals can fail, or their self worth can be threatened. As a result, goals themselves might become aversive. We theorize that this aversion to goals may result in reduc ed goal generation by VS individuals, when compared to GS individuals. Unfortunately, the association between goal orientation and goal generation in MDD has rarely been examined. Do Differences in Goal Generation Reflect States or Traits? Although sever al theories posit differences in goal generation between depressed and nondepressed individuals a dearth of research has been conducted in terms of how these processes may differ in individuals with MDD. Moreover, it is unclear if these deficits remit a s an individuals depressive status changes. If deficits in goal generation
13 are trait like, these deficits may be related to vulnerability to MDD onset, to relapse, or to the maintenance of depressive episodes. If, on the other hand, these deficits are s tate like features of MDD and fade as mood improves, these deficits are transient features of MDD. Thus, one aim of the current study is to address the functional significance of these deficits, by comparing goal generation, motivation, and goal orientati on in individuals who are experiencing a depressive episode at the time of testing with individuals who had a depressive episode that has remitted, and never depressed controls. Limitations of Previous Research Although Dickson and MacLeod (2004, 2006), and Dykman (1998) paved the way for examining deficits in goal generation in depression, the theories set forth by these researchers have not been adequately tested with individuals who are currently depressed, or remitted depressed. These studies, howeve r, have implemented unique strategies to assess goal generation. As such, we have utilized the strategies developed by Dickson and MacLeod and Dykman to assess the manner in which goal generation varies as a function of depressive status. Dickson and MacLeod (2004, 2006) examined approach and avoidance goal generation in dysphoric and depressed individuals. In each study, the groups of interest were determined by cut off severity scores on self report depression symptom scales Although self report me asures add incremental knowledge regarding how intensely a low mood is being experienced by an individual, it does not connote MDD diagnosis. As previously mentioned, MDD diagnosis requires the endorsement of at least one cardinal symptom and multiple ass ociated symptoms of MDD, yielding a total of 5 depressive
14 symptoms that have caused impairment for at least two weeks. The self report measures used in the studies by Dickson and MacLeod ( 2004, 2006), namely the Beck Depression Inventory II (BDI II; Beck, Steer & Brown, 1996) only assess the severity of depressive symptoms. Because persons with a high symptom score on the BDI II may or may not have a diagnosis of MDD it is unclear i f the results from the dysphoric samples tested in Dickson and MacLeod (20 04, 2006) studies generalize to individuals who suffer MDD Extant studies provide insight into how individuals with low mood may differ from individuals without low mood in generating goals (Dickson & MacLeod, 2004, 2006), and how goal orientation may predict depressive symptoms amongst individuals who are experiencing stressful events (Dykman, 1998). Yet, it remains unclear as to how goal generation and goal orientation may vary as a function of depressive status. As such, the current study seeks to m odel the designs utilized by Dickson and MacLeod (2004, 2006) and Dykman (1998) in an attempt to examine how goal generation and goal orientation vary as a function of depressive status. The Current Study As MDD can be construed as a disord er that affects the domains of motivation, cognition, and affect (Strauman et al., 2006), the construct of goal generation, which involves all of these domains, is potentially integrative ( Dykman, 1988; Dickson & MacLeod, 2004, 2006; Strauman et al., 2006) Specifically approach goal and avoidance goal generation was examined in currently depressed, remitteddepressed and never depressed control participants. The format in which these goals were examined was loosely based upon the design of Dickson and MacLeod (2004, 2006). To allow
15 better generalization to MDD, groups were constituted from scores on the Inventory for Diagnosing Depression (IDD; Zimmerman, Coryell, Corenthal, & Wilson, 1986) to determine current depressive status and the Inventory for Diagnosing Depr ession Lifetime (IDD L: Zimmerman & Coryell, 1987b ) to determine depression history (instruments which correlate highly with diagnostic criteria for MDD)
16 Specific Aims The general aim of the study was to compare goals and goal orientations across c urrently depressed (CD) remitted depressed (RD) and never depressed (ND) individuals. The specific aims were as follows: Specific Aim 1: Compare the generation of approach and avoidance goals across groups. Hypothesis 1a: Currently depressed individual s will generate fewer promotion/approach goals than never depressed controls. This prediction is supported by theory (Strauman et al., 2006; Dykman, 1998)) and empirical findings (Dickson & MacLeod, 2004, 2006) that suggest that individuals with low mood generate fewer promotion/approach goals than individuals without low mood. Hypothesis 1b: Remitted depressed individuals will generate fewer approach goals than never depressed controls, but more approach goals than currently depressed individuals. As depression is a highly recurrent disorder, we reasoned that deficits in approach goal generation would be trait like. We also expected that CD individuals would exhibit more severe deficits in approach goal generation than RD individuals, as current mood impairment is likely to exacerbate goal generation deficits. Thus we predicted that RD individuals w ould generate more approach goals than CD individuals, but not be able to generate as many promotion/approach goals as ND controls.
17 Hypothesis 1c: Current ly depressed, remitted depressed, and neverdepressed control individuals will not differ significantly in the generation of avoidance goals. Although research by Dickson and MacLeod ( 2004, 2006) suggests that dysphoric individuals generated more avoidan ce goals than nondysphoric individuals, the difference was not statistically significant. Given the limited theoretical and empirical basis for predicting differences in the generation of avoidance goals, we did not expect group differences in this domai n (Dickson & MacLeod, 2004, 2006). Specific Aim 2: Exam ine the mediational role of BAS sensitivity in goal generation as a function of depressive status Hypothesis 2: BAS sensitivity will mediate the relationship between depressive status and approach goal generation. As the behavioral systems are thought to have a key role in motiva tion and affect within MDD, we expected that sensitivities to the BAS system would mediate the relationship between MD D status and approach goal generation. Specific Aim 3: Examine the mediat i onal role of goal orientation in the generation of promotion/approach and avoidance goals as a function of depressive status Hypothesis 3: Validation seeking orientation will mediate the relationship between depressive status and approach goal generation. Dykman (1998) suggests that adopting a validation seeking goal orientation not only increases an individuals vulnerability to future depression, but is also predictive of current symptoms. For this reason, we expected that individuals with a history of MDD would be more likely to
18 adopt a validation seeking goal orientation than never depressed controls. This orientation, in turn, may cause an individual to avoid generating and pursuing approach goals, as they may be more difficult to achieve for individuals with a history of MDD (Strauman et al., 2006), thus posing a greater threat to a validation seeking individuals self worth. As such, we predicted that goal orientation would mediate the relationship between depressive status and approach goal generation.
19 Exploratory Aims Specific Aim 4: Examine the number of goals generated and domains of generation as function of depression status. Hypothesis 4: Currently depressed and remitted depressed participants will generate fewer goals in fewer domains overall than neverdepressed participants. As previously discussed, individuals with MDD may generate and fixate upon a single goal (Champion & Power, 1995), within a single domain, and may not be able to perceive alternative goals. As such, we explored whether CD and RD individuals would generate fewer goals in fewer domains overall than ND controls. Specific Aim 5: Examine the average level of abstractness of goals as a function of depression status Hypothesis 5: Currently depressed and remitted depressed participants will generate more abstract goals overall than never depressed participants. Abstract goals are thought to be more difficult to achieve (Emmons, 1992), harder to disengage from, as well as being more related to an ind ividuals overriding sense of self worth (Carver & Scheier, 1988) than concrete goals. As the pursuit of abstract goals has been shown to increase negative affect and low mood (Street, 2002), and improperly disengaging from a goal is suggested to promote MDD (Nesse, 2000), we explored whether CD and RD individuals would generate more abstract goals than ND controls.
20 Method Participants Participants were recruited through an online volunteer participant pool coordinated by the University of South Flori da Psychology department. Six hundred twenty seven female undergraduates participated in the online portion of the study. Based on their responses to the online questionnaires, 274 individuals were invited to participate in the laboratory portion of the s tudy. One hundred twelve individuals agreed to participate in the laboratory portion, and 84 participants provided responses that qualified them for the final group assignments. For the purposes of this study, the first 20 participants eligible for each group were included in the analyses, to render the final sample of 60 participants. Final participants were female, between the ages of 18 and 55, and fluent English speakers. The final sample approximated the racial distribution of the University of South Florida undergraduate population : 67.2 % Caucasian, 14.8% African American, 8.2% Asian, 1.6% Latino, 1.6% Native American, and 4.9% other. Table 1 contains demographic information of the sample by group.
21 Table 1. Demographic information Demographics ND CD RD Age [Mean (SD)] 21.90 (5.09) 20.60 (5.29) 21.90 (6.03) Race [Frequency (%)] Caucasian 10 (50) 14 (70) 17 (85) African American 6 (30) 2 (10) 1 (5) Asian 2 (10) 2 (10) 1 (5) Latino 1 (5) 0 0 Native American 0 1 (5) 0 Other 1 (5) 1 (5) 1 (5) Education [Mean (SD)] 14.80 (1.28) 14.10 (1.21) 14.65 (1.14) Marital Status [ Frequency (%)] Single 17 (85) 16 (80) 15 (75) Married 2 (10) 1 (5) 1 (5) Living with Domestic Partner 1 (5) 1 (5) 2 (10) Divorced 0 1 (5) 2 (10) Widowed 0 1 (5) 0 Materials Group Determination. The Inventory to Diagnose Depression (IDD; Zimmerman, et al. 1986) was utilized to determine the presence of a current depressive episode The IDD consists of 22 items that cover the full range of depressive symptoms. Score s range from 1 (no disturbance) to 5 (symptom presence) in ascending order of severity. The IDD also assesses the duration of each sympto m (present for either more or less than 2 weeks). Meeting diagnostic criteria for depression on the IDD requires a score of a 3 for low mood, irritability, or hopelessness, or a 4 or higher on the anhedonia item s Additionally, an individual must score a 3 on four of the eight items that describe the supplementary symptoms of depression, as well as endorse that these symptoms have been present for at least 2 weeks.
22 a high Spearmanaddition, the IDD has high specificity (98.5%) and acceptable sensitivity (54.5%), and high concordance with diagnoses of depression rendered by the Diagnostic Interview Schedule (97.2% agreement; Zimmerman & Coryell, 1987a). The Inventory to Diagnose Depression Lifetime Version (IDD L; Zimmerman & Coryell, 1987b) was utilized to determine history of depressive episodes. The format and scoring is identical to that of the IDD. The directions on the IDD L direct the individual to focus on the period in her/his life during which s/he was feeling the most depressed. The IDDL also includes questions regarding what may have caused or precipitated the depressive episode, and if the individual sought treatment. The IDD L has similar Brown split half reliability (.90). In addition, the IDD L has a high concordance rate with the DIS for lifetime diagnosis of MDD (k=.60) high specificity (93%) and adequate sensitivity (74% ; Zimmerman & Coryell, 1987b). Groups were determined as follows: Currently depressed (CD) met criteria for a current mood episode as based on the score derived on the IDD but was not required to have endorsed a past episode on the IDD L The CD group was not required to have past episodes of depression, as determined by the IDD L. Recovered depressed (RD) required that individual endorsed past experience of a mood episode on the IDD L but no current moo d episode on the IDD. Never depressed (ND) required that an individual did not endorse a current or past mood episode, as determined by scores on the IDD or IDD L.
23 Structured Clinical Interview for the DSM IV TR Axis I Disorders, Research Version, Patient Edition with Psychotic Screen (SCID I/P W/PSY SCREEN; First et al., 2002). In order to increase confidence in the assessment of MDD, 10 individuals from the CD group were randomly selected to have the overview and mood modules of the SCID I completed. Due to researcher oversight, however, only 9 eligible participants were approached. The SCID I is a semi structured interview that assesses for psychopathology as pre scribed by the DSM IV. Upon completion of the laboratory portion of the study, eligible par ticipants were invited to participate in a separate laboratory session, during which the mood module of the SCID I would be completed. Participants were offered additional course credit for participation in this portion of the study. The interviews were audtiotaped and were completed by a clinical psychology graduate student, who w as not blind to participant status. Of the CD participants, nine were approached to participate in the SCID portion of the study, and 8 agreed. Although the IDD and the SCID did not converge completely (4 met MDD diagnosis on both assessments, 1 endorsed the core symptom of hopeless on the IDD not captured by the SCID I, 1 endorsed 4 supplemental symptoms on the IDD but only 3 were captured on the SCID I, and 2 failed to meet criteria for other reasons on the SCID I), the IDD did appear at a minimum to identify persons with significant depressive symptomology. Anxiety. Johnson et al. (2003) suggest that anxiety symptoms that present with depression may confound depressionrela ted group differences To permit a post hoc statistical control for the possible influence of moodstate dependent anxiety, we
24 administer ed the Beck Anxiety Inventory (BAI; Beck, Brown, & Steer, 1988). The BAI is 21item self report questionnaire that assess es the presence of anxious symptom s with scores rated on a scale of 0 (symptom not present) to 3 (symptom is present and severe). Beck, Brown, & Steer, 1988). Dispositional Behavioral Inhibition and Activation Sensitivities. The Behavioral Inhibition System/Behavioral Activation System Scale (BIS/BAS; Carver & White, 1994) is a 20item self report questionnaire that assesses BIS and BAS sensitivity. Responses are coded on a 4 point likert scale, ranging in response of very true for me to very false for me . The measure includes four subscales: BIS or punishment sensitivity, Drive, Fun Seeking, and Reward Responsiveness. The BIS scale loads onto BIS sensitivity, while the other three load on BA S sensitivity. The reliability statistics for slightly lower, and range from .59.69 (Carver & White, 1994). Goal Orientation Inventory. The Goal Orientation Inventory (GOI; Dykman, 1998) is a 36item self report questionnaire that assesses the extent to which an individual possesses a validation seeking (VS) or growthstriving (GS ) goal orientation. Items consist of narratives that are considered to describe facets of validation seeking (proving self worth) and growthstriving (learning) orientations. Below are examples of validation seeking and growthstriving narratives:
25 VS: I nstead of just enjoying activities and social interactions, most situations to me feel like a major test of my basic worth, competence, or likeability. GS: I look upon potential problems in life as opportunities for growth rather than as threats to my se lf esteem. Responses are coded on a 7 point likert scale, ranging in response from strongly disagree to strongly agree. Both scales consisted of 18 items (VS: 1, 4, 6, 7, 9, 12, 15, 16, 18, 21, 22, 24, 26, 29, 30, 32, 34, 36; GS: 2, 3, 5, 8, 10, 11, 13, 14, 17, 19, 20, 23, 25, 27, 28, 31, 33, 35). Subscale scores were calculated by summing ratings of each item on the subscale. Total goal orientation score was calculated by subtracting the GS subscale score from the VS subscale score. Higher scores denote a VS orientation, while lower scores denote a GS orientation. Reliabilities for the VS and GS subscales are quite high retest reliabilities are slightly lower (VS: ykman, 1998). Goal Generation. The Revised Goal Task was created for the current project, and included two timeframes in which individuals were asked to generate goals: current and future. For the current task, participants were given the following promp t: Please list all the goals that you are either actively pursuing, or considering pursuing in the next 6 months. These can be goals that you feel it will be important for you to accomplish, or goals that involve avoiding certain situations or preventing certain outcomes.
26 For the future task, participants were given the following prompt: Please list all the goals that you are NOT currently pursuing, but are considering pursuing in the future. These can be goals that you feel it will be important for you to accomplish, or goals that involve avoiding certain situations or preventing certain outcomes. Future and current subtasks of the Revised Goal Task were administered in counterbalanced order across participants. Goals were be examined by comparing s ums of each type of goal generated, as well as by examining the proportion of abstract goals within the total amount of goals generated. After participants listed their goals, they were asked to make 4 subjective ratings of importance, likelihood of achie vement, impact on ones life if the goal was achieved, and expected pleasure of achieving generated goals. All ratings were based on a scale of 1 (low) to 10 (high). The information generated in this section created subjective ratings of goals (Appendix) Procedure This study was completed in two phases: an online pre screen and a laboratory session. The online portion was conducted via an online psychology volunteer pool coordinated by the University of South Florida Psychology department. During this session, participants were asked to complete the BIS/BAS, IDD, and IDD L. Scores on the IDD and IDDL were used to determine group status. If eligible, participants were provided an invitation code (via email) to signup for the laboratory session. Dur ing the laboratory portion, participants were consented, and asked to complete the following scales: Demographic questionnaire, Revised Goal Task, GOI, IDD, and BAI. The IDD
27 was readministered to confirm group status. A subset of eligible participants c ompleted the SCID interview. Following completion of study participation, all participants were debriefed. Goal Coding The Revised Goal Task asked participants first to list the goals they are currently pursuing (current task) or plan to pursue in the fut ure (future task). The goals generated were coded for orientation, complexity, abstractness, and domain by 4 undergraduate research assistants. All research assistants were blind to all hypotheses, as well as participant status. Each research assistant coded all goals independently. Ratings were then entered into a database, which was then examined for interrater discrepancies. Discrepancies were then discussed between the raters until a consensus rating was reached. The consensus ratings were those us ed in the final analyses. Typology (Approach v. Avoidance). Approach goals were defined as those which describe movement towards achieving a desired state, or promoting engagement with the environment (i.e., apply to graduate school; Higgins, 1996; Dickson & MacLeod, 2004, 2006). Avoidance goals were defined as those which described the prevention of an unwanted event, or withdrawal from the environment (i.e., do not fail a course this semester; Higgins, 1996; Dickson & MacLeod, 2004, 2006). Interrater r eliability for typology was high Complexity (Simple v. Complex). Simple goals were defined as those which involved only a single step, such as (i.e., take a Psychology course next semester, whereas complex goals were larger goals involving multiple steps (i.e., graduate from
28 c ollege with a B.A. in Psychology; Street, 2002; Carver & Scheier, 1990). Interrater reliability for complexity was high ( ). Abstra ction (Abstract v. Concrete). Abstract goals were defined as those which lacked a clear endpoint at which ach ievement could be defined (i.e., become a model citizen) whereas concrete goals, conversely, were those which did have a clearly de fined endpoint (i.e., volunteer in the community next weekend; Emmons, 1992). Interrater reliability for abstraction was high Domain Goals were categorized into the following mutually exclusive domains: social, professional, scholastic, family, financial, appearance, lifestyle, and other. Goals falling within the social domain typically involved themes regarding friendships, social networking, or engaging in social activities. Professional goals typically discussed themes such as the maintenance or promotion of ones job, changing jobs, or developing a career. Scholastic goals usually referred to attaining a specific grade in a class, graduating from college, or applying for graduate school. Goals related to family either involved maintaining or repairing a relationship with ones family of origin, or creating a family of ones own. Financial goals involved those which per tained to money or ones financial status. Appearance goals typically involved wanting to change or maintain ones weight, but could have also involved wanting to change ones hair color or cut, or undergo plastic surgery. Goals which spanned more than 2 domains were classified as lifestyle. Lifestyle goals were classified as those which spanned more than one of the aforementioned domains. Goals such as traveling or relocating were considered lifestyle goals. Goals not falling clearly within any of the aforementioned domains were
29 described as other. Interrater reliability for domain, the consistency with which raters agreed upon the domain which best described each goal, was
30 Results Descriptives. Individuals, irrespective of group, typically generated between 5 and 6 goals on the current and future subtasks of the Revised Goal Task. A significant majority (98%) of these goals were determined to be approach goals. Interestingly, many individuals generated approach goals relating to pursuing an academic degree (scholastic domain), going on vacation or relocating (lifestyle), getting married (family domain), and having children (family domain) (Table 2). Goal Generation (Specific Aim 1: Comparison of approach and avoidance goals across groups). We anticipated that CD individuals would generate the least number of approach goals, followed by RD individuals who were predicted to generate more approach goals than CD individuals, but fewer approach goals than ND individuals. A 3 ( depressive status : CD, ND, RD) by 2 (task: current, future) Multivariate Analysis of Variance (MANOVA) failed to reveal significant group differences in the amount of approach goals gene rated [ F (2, 57)=.92, p=.46] (Table 2) A 3 (group) by 2 (task) MANOVA revealed the anticipated lack of group differences in the generation of avoidance goals [ F (2, 57)=.34, p=.85] (Table 2). As the results suggest that a floor effect for the generation of avoidance goals is present, it is possible that the Revised Goal Task may not be the ideal task for examining group differences in avoidance goal generation.
31 Although we collected BAI data for potential use in covariate analyses, we ultimately decided not to pursue this analysis. First, use of covariates in group analyses when groups differ significantly on the proposed covariate is controversial (Miller & Chapman, 2001). Second, covariates are typically described as nuisance variables. Anxiety is hi ghly comorbid with MDD (APA, 2000),; in this respect, it is unclear whether it represents a substantive or a confounding variable. Table 2. Goal coding results by group and goal domain Group Goal Typology ND CD RD Current Approach 5.20 (2.33) 5. 50 (1.93) 6.10 (3.37) Future Approach 5.40 (1.96) 6.50 (2.42) 5.05 (2.84) Current Avoidance .05 (.22) 0.25 (0.91) 0.25 (0.55) Future Avoidance .05 (.22) 0.10 (0.31) 0.10 (0.45) Total Goals 10.70 (3.83) 11.45 (3.98) 11.50 (6.00) Objective Rating Total Abstract 5.25 (2.34) 4.40 (3.17) 3.55 (3.30) Total Concrete 5.45 (1.85) 7.40 (3.50) 8.30 (4.12) Total Simple 3.20 (2.57) 5.65 (2.18) 5.45 (3.02) Total Complex 6.25 (3.06) 6.15 (2.81) 6.40 (4.11) Domain Total Scholastic 3.20 (1.74) 3.2 0 (2.04) 2.90 (1.55) Total Lifestyle 1.90 (1.55) 2.00 (1.38) 2.60 (1.81) Total Professional 1.70 (0.98) 1.35 (0.81) 1.50 (1.05) Total Family 1.50 (0.89) 1.65 (1.50) 1.55 (1.47) Total Social 0.65 (0.99) 0.95 (1.64) 0.55 (0.83) Total Financial 0.05 (0.2 2) 0.05 (0.22) 0.05 (0.22) Total Appearance 0.65 (0.93) 0.60 (0.88) 1.20 (1.32) Total Other 1.05 (1.82) 2.00 (1.97) 1.50 (2.01) Note: Values represent means and standard deviations of number of goals generated for each objective description Because gro up differences in approachrelated goals failed to emerge, it was not possible to examine the mediational roles of BAS sensitivity and goal orientation. These factors are worthy of future examination in regards to goal generation, as they may
32 influence ho w individuals perceive and value goals. These factors, in turn, may relate to the subjective ratings of goals previously described. As such, we proceeded to examine group differences in behavioral system sensitivity and goal orientation. Behavioral Syste m Sensitivity ( Specific Aim 2: Examination of the meditational role of BAS sensitivity in goal generation as a function of depressive status ). We hypothesized that BAS sensitivity would mediate the relationship between depressive status and approach goal generation. This analysis was precluded by the lack of group differences in approach goal generation. We did, however, examine group differences in BAS and BIS sensitivity. We anticipated that CD individuals would have lower BAS sensitivity than RD indivi duals, who in turn would have lower BAS sensitivity than ND individuals. A one way ANOVA was conducted to examine group differences in BAS sensitivities. Surprisingly, groups did not differ on BAS sensit ivity [ F (2, 57)=.58, p=.56]. We also planned to examine group differences in BIS sensitivity, anticipating that groups would not differ in BIS sensitivity. A 3 (group) by 1 (BIS sensitivity) suggested that groups did in fact differ in BIS sensitivity [ F (2, 57)=4.64, p= .01], such that CD individuals had s ignificantly higher BIS scores than ND individuals (see Table 3) Goal Orientation (Specific Aim 3: Examination of the mediational role of goal orientation in the generation of approach and avoidance goals as a function of depressive status ) As before, we could not test the hypothesis that adopting a validationseeking (VS) goal orientation would mediate the relationship between depressive status and approach goal generation because the groups did not differ in approach goal
33 generation. We did examine group differences in goal orientation. More specifically, we hypothesized that CD individuals would have higher GOI scores (adopt a VS orientation) than RD individuals, who in turn would have higher GOI scores than ND individuals (adopt a GS orientation). A one way ANOVA was performed to examine group differences in GOI scores. Results revealed a main effect of depressive status on GOI orientation [ F (2, 57)=13.44, p= .0001], such that CD individuals adopted a VS orientation, while ND and RD individuals adopted a GS orientation. This suggests that CD individuals tend to view goals as opportunities to prove (or disprove) their self worth, and ND and RD individuals tend to view goals as opportunities to learn and grow. Pairwise comparisons revealed that CD i ndividuals had significantly higher overall GOI scores than ND and RD individuals ( p= .0001), and RD individuals had significantly higher GOI scores than ND individuals ( p= .001 ; see Table 3). Table 3. Scale scores Group ND CD RD IDD Symptom Total 0. 10 (0.31) 6.80 (2.42) b 0.65 (0.88) a b BAI 5.79 (5.07) 19.80 (8.84) b 9.40 (5.87) a BAS b 41.50 (5.53) 39.80 (6.16) 41.35 (4.74) BIS 21.00 (3.42) 24.15 (3.29) b 22.60 (3.08) a GOI ab 46.95 (27.51) 11.75 (42.81) c 22.70 (35.98) a VS b 48.95 (21.61) 84.60 (22.95) b 61.95 (26.26) a GS b 95.90 (11.87) 71.85 (22.26) b 84.65 (17.83) a b a,b,c denote group differences p<.01 Number of Goals and Domains of Generation (Specific Aim 4: Examination of the number of goals generated and domains of generation as functi on of depression status ). We hypothesized that CD individuals would generate fewer goals across fewer domains than RD and ND individuals. A 3 (group) by 2 (task) MANOVA was conducted to
34 analyze group differences in the number of goals generated across gr oups. Results suggest that individuals generated similar numbers of goals across all time points [ F( 2, 57)=1.03, p =.40]. A 3 (group) by 8 (domain: scholastic, professional, family, social, financial, appearance, lifestyle, other) MANOVA revealed that ind ividuals generated similar numbers of goals across domains irrespective of depressive status [ F (2, 57)=.84, p=.64]. In addition, results from a one way ANOVA revealed that groups did not differ significantly in the number of domains in which goals were ge nerated [ F (2, 57)=.29, p=.75] (Table 2). Exploratory Analyses. All exploratory analyses were conducted using overall (current and future combined) ratings of goals, unless otherwise specified. Level of Abstraction (Specific Aim 5: Examination of the average level of abstractness of goals as a function of depression status ). We hypothesized that CD and RD individuals would generate more abstract goals overall than ND individuals. Results from a one way ANOVA suggest that groups generated similar numbers of abstract goals [ F (2, 57)= .83, p=. 44] We also examined the possibility that CD and RD individuals may generate fewer concrete goals than ND individuals; however the results from a oneway ANOVA suggest that groups generated similar numbers of concrete goals [ F (2, 57)= .36, p=.70] (Table 2) Level of Complexity. We hypothesized that CD and RD individuals were generate more complex goals overall than ND individuals. Results from a one way ANOVA
35 suggest that irrespective group status, individuals gene rated similar numbers of complex goals [ F (2, 57)=.03, p=.97] We also examined the possibility that CD and RD individuals would generate fewer simple goals than ND individuals, but results from a one way ANOVA did not support this hypothesis, as it appea rs that groups generated similar numbers of simple goals [ F (2, 57)=1.21, p= .31] (Table 2). Subjective Ratings. Four separate one way ANOVAs were conducted to analyze group differences in achievability, expected pleasure, importance, and impact In all ca ses significance, groups did not differ in their appraisals concerning the likelihood of achieving goals, the amount of enjoyment expected from achieving these goals, the potential impact the goals have on their lives, or the importance of the goals ( ps>.2 T able 4). Table 4. Subjective ratings Group Subjective Rating ND CD RD Achievement 8.34 (1.21) 7.70 (1.10) 8.07 (1.19) Pleasure Expect ed 8.91 (0.76) 8.64 (1.68) 9.03 (1.13) Importance 8.31 (0.92) 8.57 (0.98) 8.50 (0.73) Impact 8.22 (1.01) 8.40 ( 1.07) 8.56 (0.89) Values represent means and standard deviations Depressive symptoms and goal generation. As the results from the group level analyses of goal generation were counterintuitive, contrary to theory, and contradictory to the few empirical st udies of the topic, we also tested our hypotheses utilizing a correlational approach, which should have greater statistical power. As such, we proceeded to conduct several independent regression analyses between the number of depressive symptoms endorsed and ratings of goal generation.
36 Table 5. Mean number of depressive symptoms by group. Status Number of Symptoms ND .10 (.31) CD 6.8 (2.40) RD .65 (.88) Values represent means and standard deviations. First, we conducted several independent regressi on analyses to examine the relationship between depressive symptoms and approach goal generation, avoidance goal generation, and total goal generation. Consistent with grouplevel results, all regressions failed to reach significance ( ps>.1) reinforcing t he initial null findings. In addition, several regression analyses were performed to examine the relationship between depressive symptoms, and the generation of abstract goals, concrete goals, simple goals, and complex goals. Again, all regressions faile d to reach significance ( p s>.1). Lastly, s everal independent regression analyses were performed to examine the relationship between depressive symptoms and subjective ratings of goals. Results revealed a significant relationship between number of depres sive symptoms and ratings of achievability ( R2 =.11, p=.009), such that as the number of symptoms increased, ratings of achievability decreased. Similarly, there was a significant relationship between the number of symptoms and subjective ratings of expec ted enjoyment ( R2= .08, p=.03 ), such that as the number of symptoms increase, ratings of expected enjoyment decreased. Conversely, there were no relationship s between number of symptoms and ratings of importance and impact of goal achievement ( ps >.5).
37 Discussion Although many theories attempt to explain the manner in which deficits in goal generation may lead to an episode of Major Depressive Disorder (MDD), few studies have actually described how persons with depression vulnerability generate goals f or themselves. The current project examined potential differences in multiple aspects of goal generation as a function of depressive status (current, remitted, never depressed). Although we anticipated that individuals with a history of depression would generate fewer approach goals, more abstract goals, more complex goals, and fewer goals across fewer domains than never depressed individuals, our results did not support these hypotheses. Instead we found remarkabl e similarity in multiple aspects of approach goal generation, irrespective of depressive status and regardless of whether continuous or grouplevel analyses were used Although surprising the se findings do not represent a major contradiction of prior empirical work given that much of the pr evious work in this area has been primarily theoretical. Although it is suggested that individuals with MDD should generate fewer goals (Champion & Power, 1995) across fewer domains (Champion & Power, 1995, Strauman et al., 2006), and also should have the tendency to generate more abstract (Emmons, 1992; Carver & Scheier, 1998) and complex goals (Carver & Scheier, 1990;
38 Street, 2002) than never depressed individuals, these theories have yet to be supported empirically. Although limited, previous empiric al work by Dickson and MacLeod (2004, 2006) has loosely supported the theoretical work of Higgins (1997) and Strauman and colleagues (2006), which suggests that depressed individuals should generate fewer approach goals than never depressed individuals. I n samples of adolescents high in self reported ratings of depressive symptoms, Dickson and MacLeod (2004, 2006) found that individuals high in depressive symptoms did in fact generate fewer approach goals than individuals low in depressive symptoms. Because their samples differ in age and severity from the current study, it is unclear whether our null results represent a direct conflict with Dickson and MacLeod or represent differences in study construction. To examine whether individuals with and without a history of depression differ in the types of goals they generate, we followed in the footsteps of Dickson and MacLeod (2004, 2006) and examined the relationship between depressive symptom severity and various aspects of goal generation. Here too, regr ession analyses revealed no significant relationships between symptom severity and approach goal generation, avoidance goal generation, number of abstract, concrete, simple, or complex goals, or the overall number of goals and domains in which goals were generated. As such, it appears that the results of the current project are both counter theory and previous empirical work (see below for discussion) The multi dimensional approach of the Revised Goal Task allowed individuals to generate subjective rating s of the importance, achievability, impact, and expected
39 pleasure of generated goals. Although no specific hypotheses were stated regarding these ratings, grouplevel analyses did not reveal differences in any of the subjective ratings. Following suit of the previous analyses, we conducted several independent regression analyses to examine the relationship between depressive symptom severity and subjective ratings of goals. Results suggest that as the severity of depressive symptoms increases, ratings of achievability and expected pleasure associated with goals decreases. These findings are consistent with both intuition and theory (Carver, 2001). Lastly, it is of worth to note the extremely low number of avoidance goals generated by the Revised Goal Tas k. The meager number of avoidance goals ( (i.e., means < 0.25), made it difficult to examine group differences in avoidance goals because of a floor effect. In previous work, Dickson and MacLeod (2004, 2006) were able to elicit significantly higher numbers of avoidance goals from high anxiety, mixed anxiety and depressed, and depressed, and control adolescents ( M =4.04), and dysphoric and nondysphoric adolescents ( M =3.90). Methodological differences may explain why Dickson and Macleod found higher levels of avoidance goals (as well as the group differences in avoidance goal generation) and we did not Dickson and MacLeod utilized a twopart design: one part focused upon approach goal generation, while the other focused upon avoidance goal generation (as opposed the current projects focus upon time). For the approach goal task, directions were as follows: In the future it would be important for me to [try] and the directions for the avoidance task were as follows: In the future it would be important for me to [try to] avoid As such, it is possible if not likely, that explicit direction to generate goals of a certain type (i.e., avoidance goals) strongly
40 influence on how many goals of that type are listed. It would be useful to examine the types of goals that individuals spontaneously generate for themselves (without directions for orientation or timeframe), to clarify wh ether individuals naturally generate both approach and avoidance goals As there is a dearth of empirical literature in this area the current project was forced to develop hypotheses based on theoretical work. In order to better understand the results of the current project, we examine limitations in our own work and the small amount of previous empirical work in this area. First it is possible that counter to theory and previous research, individuals who are vulnerable to MDD truly do not differ in the generation of approach and avoidance goals, or the number, domains, level of abstraction, and level of complexity of self genera ted goals. Or it is possible that populationlevel group effect s may exist but be smal ler than our study was powered to detect Therefore, we decided to conduct power analyses with the current sample of 60 individuals, in order to determine what type of power we had to determine moderate and large effect sizes. The multivariate analyses conducted to examine differences in approach goal generation, avoidance goal generation, our statistical power to detect a moderate effect size (f2 =.25) at an alpha le analyses conducted to examine group differences in BIS and BAS sensitivity, GOI scores, BAI scores, total number of goals generated, the number of domains in which goals were generated, level o f abstraction, level of complexity, and analyses of subjective ratings of goals our statistical power to detect a large effect size (f2 =.5) at an alpha level
41 examine group differences in the number of domains in which goals were generated, our statistical power to detect a moderate effect size (f2 =.27) at an alpha the current sample was .96 In sum, the current project had adequate power to detec t moderate to large effect sizes with the current sample size. If by chance small effects were present (i.e., f2Although our results show that individuals, irrespective of depressive status, gene rated similar types and numbers of goals for themselves, it is quite possible that differences in behavioral system sensitivity and goal orientation may influence the an individuals ability and motivation to effectively pursue goals. Despite the fact that =.10), our sample size would need to approximate 1548 participants. As small effect sizes are considered rare and difficult to detect (Cohen, 1970), the power analyses presented here increase our confidence that our findings are not a result of an underpowered analyses. I t is also possible that results from the research by Dickson and MacLeod (2004, 2006) do not extend to samples of a different age or a sample with higher severity (i.e., diagnosable MDD). As the authors utilized adolescent samples, classified as being depressed or dysphoric by the number of symptoms endorsed on a single self report measure, it is unclear if similar results w ould be found in older samples with more extreme symptomatology (i.e., actual diagnosis). As we would anticipate that larger group differences would be found in diagnosed samples, more data collection is need on goal generation in diagnosed depressed and nondepressed individuals to determine whether our null findings represent an aberrant result or reveal that diagnosed samples perform differently than symptomatic samples.
42 analyses failed to reveal significant relationships between behavioral system sensitivity and approach or avoidance goal generation, or between goal orientation and approach or avoidance goal generation ( ps<.42), BAS sensitivity was significantly related to ratings of achievement ( R2=.16, p=.002) and expected pleasure ( R2=.11, p=.01). In addition, goal orientation was also significantly related to ratings of achievement ( R2=.11, p=.001) and expected pleasure ( R2Contrar y to expectations, groups did not differ in BAS sensitivity, but did differ in BIS sensitivity, such that CD individuals had significantly higher scores of BIS sensitivity than ND and RD individuals. Heightened BIS sensitivity is associated with =.08, p=.03). More specifically, higher BAS sensitivity and a growthseeking orientation are associated with higher ratings of achievability and expected pleasure of goals. These results suggest that although behavioral system sensitivity and goal orientation are unrelated to the types of goals i ndividuals generate, these factors are related to an individuals confidence in her ability to achieve goals, which in turn may be related to goal pursuit deficits. In line with this hypothesis, work by Strauman and colleagues (2006) suggest that individua ls with a history of MDD have a history of failure experiences. These failure experiences may result from the use of ineffective goal pursuit techniques (i.e., being overly sensitive to signs of threat and withdrawing from goal pursuit), and may influence how individuals view goal generation and striving (i.e., as threats to their sense of self worth). In order to explain this thesis further, we turn to the results regarding group differences in behavioral system sensitivity and goal orientation.
43 anxiety, nervousness, and behaviors which encourage an individual to withdraw from the environment and decrease goal related behaviors (Carver & White, 1994). As anticipated, CD individuals were also higher in self reported levels of anxiety. I n regards to goal or ientation, r esults from the current project suggest (consistent with theory) that CD individuals were characterized as adopting a validation seeking orientation (VS), while RD and ND individuals were characterized as adopting a growth seeking (GS) orientat ion, as identified by the Goal Orientation Inventory (Dykman, 1998). These results suggest that CD individuals tend to view goals as a way in which to prove their self worth, and when not achieved pose a threat to their self worth. Although speculative, we theorize that individuals do generate similar goals for themselves irrespective of depressive history (as results from the current study suggest), but individuals differ in their ability to effectively stride towards these goals. More specifically, we posit that goal orientation and behavioral system sensitivity may influence an individuals goal striving ability. Results from the current project revealed that CD individuals were unique that that they adopted a VS orientation and had heightened BIS sen sitivity when compared to RD and ND individuals. Viewing goals as threats to self worth, while also having a heightened sensitivity to threat may immobilize goal related activities, thereby making goal striving seemingly more difficult. These factors uni que to CD individuals may be related to a history of failed goal strivings Strauman and colleagues (2006) suggest is characteristic of individuals with MDD. Although we are unable to speak directly to this hypothesis with the results of the current study, we feel it is important to speculate the influence of motivational factors such as
44 goal orientation and behavioral system sensitivity in the larger scheme of goal generation and goal pursuit
45 Limitations and Future Directions Sample Demographics Bec ause this was a sample of undergraduate females, findings may not generalize to other populations. Although previous research failed to find gender differences in goal generation (Dickson & MacLeod, 2004, 2006), it is unclear to what extent the specific g oals generated by this particular sample generalize to males (i.e., get married, have a family). In addition, a significant proportion of our participants were in their early 20s and single. Thus, the types of goals and domains in which goals were created (i.e., graduate from college, get married, have a family) may be more prevalent amongst this sample than others, thereby reducing the generalizability of these results. Sample Severity The IDD and IDD L have been shown to converge with the diagnostic interviews, when used with an undergraduate aged sample ( Goldston et al., 1990). Our sample, however, included very few individuals with extreme scores on either measure. The average number of symptoms endorsed by women identified as depressed by Goldst on and colleagues was much higher ( M =13.8) than that of the current sample ( M =6.8). In addition, it appears that only a subset of individuals identified as CD by the IDD also met diagnostic criteria on the SCID I. In sum, it appears that the current proj ects aspirations to recruit a more severe sample than that utilized in previous work (i.e., dysphoric individuals; Dickson & MacLeod, 2004, 2006), did not completely
46 succeed. Thus, it is important to acknowledge that the sample recruited here is likely to be a relatively high functioning subset of individuals with a history of MDD, and as such may not be representative of all individuals with a history of MDD. With the recruitment of individuals with more severe levels of MDD, it is possible that differen ces in the number, level of abstraction and complexity, and domains in which individuals generate goals may emerge. Context As the current project was conducted in an academic setting, as partial fulfillment of course credit (on the part of the participants), it is possible that the recurrent scholastic themes seen in the individuals goals may have been unintentionally cued by the strong academic context surrounding study participation. In addition, it is possible that the groups may differ naturally in the types of cues they generate, but the strong contextual cues may have overridden this tendency, and masked potential group differences in the number and domains of goal generation. Future research should examine goal generation in a more neutral or ambiguous setting in order to better examine the domains in which, and orientation of goals individuals tend to generate.
47 Conclusion Despite the potential limitations of the current project, we feel that the unanticipated findings shed light on the area of goal generation as a function of depressive status. Contrary to theory, individuals with and without a history of MDD did not differ in the numbers of goals within similar domains (Champion & Power, 1995, Strauman et al., 2006), levels of goal abs traction (Emmons, 1992; Carver & Scheier, 1998) or complexity (Carver & Scheier, 1990; Street, 2002). In addition, the findings of the current study challenge those of previous empirical work in this area, in that individuals generated similar types of goals irrespective of depressive history (Dickson & MacLeod, 2004, 2006). As groups did differ in BIS sensitivity and goal orientation, we propose that these aspects of motivation may influence an individuals ability to achieve similar goals.
48 References American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders (4th ed., Text revision, DSM IV TR). Arlington, VA: APA. Baron, R.M., & Kenny, D.A. (1986). A moderator mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51 (6), 11731182. Beck, A. T., Epstein, N., Brown, G., & Steer, R. A. (1988). An inventory for measuring clinical anxiety: psychom etric properties. Journal of Consulting and Clinical Psychology, 56(6), 893897. Beck, A.T., Steer, R.A., & Brown, G.K. (1996). BDI II Manual San Antonio, TX: The Psychological Corporation. Carver, C.S. (2004). Negative affects deriving from the behavior al approach system. Emotion, 4(1), 322. Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality and Social Psychology, 67 (2 ), 319333. Carver, C.S., & Scheier, M.F. (1990). Origins and functions of positive and negative affect: A control process view. Psychological Review, 97( 1), 1935.
49 Carver, C.S., La Voie, L., Kuhl, J., & Ganellen, R.J. (1988). Cognitive concomitants of depression: A further examination of theories of generalization, high standards, and self criticism. Social and Clinical Psychology, 7, 350365. Carver, C.S. (2001). Affect and the functional bases of behavior: On the dimensional structure of affective experience. Personality and Social Psychology Review, 5(4), 345356. Champion, L.A., & Power, M.J. (1995). Social and cognitive approaches to depression Towards a new synthesis. British Journal of Clinical Psychology, 34, 485503. Cohen, J. (1970). A pproximate power and sample size determination for common one sample and two sample hypothesis tests. Educational and Psychological Measurement, 30, 811831. Depue, R.A., & Iacono, W. G. (1989). Neurobehavioral aspects of affective disorders. Annual R eview of Psychology, 40, 457 492. Dickson, J.M., & MacLeod, A.K. (2006). Dysphoric adolescents causal explanations and expectancies for approach and avoidance goals. Journal of Adolescence, 29, 177191. Dickson, J.M., & MacLeod, A.K. (2004). Approach a nd avoidance goals and plans: Their relationship to anxiety and depression. Cognitive Therapy and Research, 28(3), 415432.
50 Dimidijian, S., Hollon, S.D., Dobson, K. S. (1996). Randomized trial of behavioral activation, cognitive therapy, and antidepress ant medication in the acute treatment of adults with major depression. Journal of Consulting and Clinical Psychology, 74(4), 658670. Durand, V.M., & Barlow, D.H. (2006). The essentials of abnormal psychology (4th edition). US: Thomson Wadsworth. Dykman B.M. (1998). Integrating cognitive and motivational factors in depression: Initial tests of a goalorientation approach. Journal of Personality and Social Psychology, 74(1), 139158. Emmons, R.A. (1992). Abstract versus concrete goals: Personal strivi ng level, physical illness, and psychological well being. Journal of Personality and Social Psychology, 62(2), 292300. Fowles, D.C. (1980). The three arousal model: Implications of Grays twofactor learning theory for heart rate, electrodermal activity and psychopathy. Psychophyisology, 17(2), 87104. First, Michael B., Spitzer, Robert L, Gibbon Miriam, and Williams, Janet B.W.: Structured Clinical Interview for DSM IV TR Axis I Disorders, Research Version, Patient Edition With Psychotic Screen (SCID I/P W/ PSY SCREEN) New York: Biometrics Research, New York State Psychiatric Institute, November 2002.
51 Goldston, D. B., O'Hara, M. W., & Schartz, H. A. (1990). Reliability, validity, and preliminary normative data for the inventory to diagnose depression i n a college population. Psychological Assessment, 2 (2), 212 215. Henriques, J.B., Davidson, R.J. (2000). Decreased responsiveness to reward in depression. Cognition & Emotion, 14(5), 711724. Higgins, E.T. (1996). The self digest: Self knowledge serving self regulatory functions. Journal of Personality and Social Psychology, 71(6), 10621083. Johnson, S., Ruggero, C.J., & Carver, C.S. (2005). Cognitive, behavioral, and affective responses to reward: Links with hypomanic symptoms. Journal of Socia l and Clinical Psychology, 24(6), 894906. Johnson, S., Turner, R.J., & Iwata, N. (2003). BIS/BAS level s and psychiatric disorder: An epidemiological study. Journal of Psychopathology and Behavioral Assessment, 25 (1), 2536. Lam, D., & Power, M. (1991). A questionnaire designed to assess roles and goals: A preliminary study. British Journal of Medical Psychology, 64 359373. Kasch, K., Rottenberg, J., Arnow, B., & Gotlib, I. (2002). Behavioral activation and inhibition systems and the severity and course of depression. Journal of Abnormal Psychology, 111(4), 589597. Kessler, R. C. (2002). Epidemiology of depression. In: I. H. Gotlib & C. L. Hammen (Eds.) Handbook of Depression. (pp. 2342). New York: Guilford Press. Klinger, E. (1975). Consequences of commitment to and disengagement from incentives. Psychological Review, 82(1), 125.
52 Lecci, L., Karoly, P., Briggs, C., & Kuhn, K. (1994). Specificity and generality of motivational components in depression: A personal projects analysis. Journal of Abnormal Psychology, 103( 2), 404408. Monroe, S. M. & Harkness, K. L. (2005). Life stress, the kindling hypothesis, and the recurrence of depression: considerations from a life stress perspective. Psychological Review, 112(2), 417445. Nesse, R.M. (2000). Is depression and adaptation? Archives of General Psychiatry, 57(1), 1420. Peeters, F., Nicolson, N.A., & Berkhof, J. (2003). Effects of daily events on mood states in major depressive disorder. Journal of Abnormal Psychology, 112 (2), 203211. Strauman, T.J. (2002). Self regulation and depression. Self and Identity, 1, 151157. Strauman, T.J., Vieth, A.Z., Merril, K.A., Kolden, G.G., Woods, T.E., Klein, M.H., Papadakis, A.A., Schneider, K.L., & Kwapil, L. (2006). Self system therapy as an i ntervention for self regulatory dysfunction in depression: A randomized comparison with cognitive therapy. Journal of Consulting and Clinical, 74(2), 367376. Street, H. (2002). Exploring relationships between goal setting, goal pursuit and depression: A review. Australian Psychologist, 7 (2), 95103. Williams, J.M.G., & Broadbent, K. (1986). Autobiographical memory in suicide attempters. Journal of Abnormal Psychology, 95(2), 144149.
53 Zimmerman, M., & Coryell, W. (1987a). The inventory to diagnose depr ession (IDD): A self report scale to diagnose major depressive disorder. Journal of Consulting and Clinical Psychology, 55(1), 5559. Zimmerman, M., & Coryell, W. (1987b). The inventory to diagnose depression, lifetime version. Acta Psychiatrica Scandinav ica, 75(5), 495499. Zimmerman, M., Coryell, W., Corenthal, C., & Wilson, S. (1986). A self report scale to diagnose major depressive disorder. Archives of General Psychiatry, 43(11), 0761086.
55 Appendix A Goal Generation Task (Current) This study is interested in what types of goals people generate for themselves. We define goals as an ambition or achievement an individual would like to occur for him or herself and is willing to work toward s accomplishing Please list all the goals that you are either actively pursuing, or considering pursuing in the next 6 months. These can be goals that you feel it will be important for you to accomplish, or goals that involve avoiding certain situatio ns or preventing certain outcomes.
56 Goals I am currently pursuing or plan to pursue within the next 6 months Importance (1= not very important, 10= very important) Likelihood of A chievement (1=not at all possible, 10=definitely achievable) Level of Impact (1=no impact, 10=life changing) Experience of Achieving Goal (1= very unpleasant, 5=not unpleasant, but not pleasant, 10=very pleasant)
57 Goa l Generation Task (Future) Please list all the goals that you are NOT currently pursuing, but are considering pursuing in the future. These can be goals that you feel it will be important for you to accomplish, or goals that involve avoiding certain situ ations or preventing certain outcomes.
58 Goals that I plan to pursue at some point in the future (NOT within the next 6 months) Importance (1= not very important, 10= very important) Likelihood of A chievement (1=not at all possible, 10=definitely ach ievable) Level of Impact (1=no impact, 10=life changing) Experience of Achieving Goal (1= very unpleasant, 5=not unpleasant, but not pleasant, 10=very pleasant)
59 Ratings of Goals Please rate on a scale of 1 10, how important these goals are to you (1= not very important, 10=very important). Next to each goal, please rate on a scale of 1 10 how likely y ou are to achieve this goals (1 = not at all pos sible; 10= I will definitely achieve this goal). Next to each goal, please rate on a scale of 1 10 how much impact achieving the goal will have on your life (1= no impact at all; 10= life changing). Next to each goal, please rate on a scale of 1 10 how y ou will feel if you achieve this goal (1=little to no satisfaction, 10= extremely happy).