Homework compliance, perceptions of control, and outcome of cognitive-behavioral treatment of social phobia

Behaz'. Res. Ther. Vol. 34, No. 5/6, pp. 423~,32, 1996

Pergamon S0005-7967(96)00014-9

Copyright ~-'~'1996 ElsevierScienceLid Printed in Great Britain. All rights reserved 0005-7967/96 $15.00 + 0.00

HOMEWORK COMPLIANCE, PERCEPTIONS OF CONTROL, AND OUTCOME OF COGNITIVE-BEHAVIORAL TREATMENT OF SOCIAL PHOBIA ANNA W. LEUNG and RICHARD G. HEIMBERG* Social Phobia Program, Center for Stress and Anxiety Disorders, University at Albany, State University of New York, Albany, NY 12205, U.S.A.

(Received 4 December 1995) Summary--This study examined the relationship of compliance with homework assignments and posttreatment anxiety in patients who received cognitive-behavioral group therapy (CBGT) for social phobia. Greater homework compliance measured in the first and latter periods of CBGT was associated with lower levels of social interactional anxiety after treatment. Surprisingly, homework compliance during the middle sessions of CBGT was positively related to posttreatment fears of scrutiny and criticism. Perceptions of control in social phobia and their potential effect on homework compliance and the homework compliance/treatment outcome relationship were also examined using the Levenson (Journal of Consulting and Clinical Psychology, 41, 397-404, 1973) Locus of Control Scale. Social phobics were less likely to believe in their own control over events than a comparison sample of community subjects but attributed greater control over events to other powerful persons. Among patients, higher Internality and lower Powerful Others subscale scores were associated with higher levels of pretreatment anxiety. However, neither subscale was significantly related to measures of homework compliance. Furthermore, when included in multiple regression analyses, neither subscale or its interaction with homework compliance added to the prediction of posttreatment anxiety. Limitations of this study and future research to improve assessment of homework compliance and perceptions of control among social phobic patients are discussed. Copyright © 1996 Elsevier Science Ltd.

INTRODUCTION

Social phobia is defined [DSM-IV; American Psychiatric Association (APA), 1994, p. 41 !] as "a marked and persistent fear of one or more social or performance situations in which the person is exposed to unfamiliar people or to possible scrutiny by others". Persons with social phobia fear being observed and critically evaluated by others and perceive a high likelihood of being humiliated, disliked, or thought worthless as a result of social scrutiny (Heimberg, 1990). Social phobia, as defined in DSM-III-R (APA, 1987), is the third most common psychiatric disorder, with a lifetime prevalence of 13.3% (Kessler, McGonagle, Zhao, Nelson, Hughes, Eshelman, Wittchen & Kendler, 1994). In the NIMH ECA study, social phobia was associated with increased odds of alcoholism and substance abuse (Schneier, Johnson, Hornig, Liebowitz & Weissman, 1992). Moreover, it is commonly associated with depression (Heimberg, Klosko, Dodge, Shadick, Becker & Barlow, 1989), suicidal ideation (Amies, Gelder & Shaw, 1983; Schneier et al., 1992), and significant impairment in educational, occupational, and social functioning (Liebowitz, Gorman, Fyer & Klein, 1985). Social phobic patients who received cognitive-behavioral group therapy for social phobia (CBGT; Heimberg, Juster, Hope & Mattia, 1995) have shown significant improvement in a number of studies (Brown, Heimberg & Juster, 1995; Gelernter, Uhde, Cimbolic, Arnkoff, Vittone, Tancer & Bartko, 1991; Heimberg, Becker, Goldfinger & Vermilyea, 1985; Heimberg, Dodge, Hope, Kennedy, Zollo & Becker, 1990; Hope, Herbert & White, 1995; Lucas & Telch, 1993; see review by Heimberg & Juster, 1995). CBGT consists of training in cognitive restructuring, within-session exposures to anxiety-evoking situations, and homework assignments for in vivo exposure to those situations and associated self-administered cognitive restructuring activities. While homework *Author for correspondence. 423

424

Anna W. Leungand Richard G. Heimberg

assignments are an integral part of this treatment package, few studies have investigated the potential effects of homework compliance on treatment outcome. Recently, Hope, Herbert, and Bellack (1991) reported that social phobic patients who complied with homework assignments throughout CBGT showed greater improvement than those who were less compliant. Laguna, Hope and Herbert (1994) further demonstrated that homework compliance in the first two-thirds of treatment was related to outcome although homework compliance later in treatment was not. These authors suggested that the benefits of homework completed during the final weeks of CBGT might be more apparent if measured sometime after the end of treatment, and, in fact, Edelman and Chambless (1995) recently reported that homework-compliant patients were less anxious 6 months after the completion of treatment. The current study further examined the role of homework compliance in CBGT outcome. The effect of homework compliance on treatment outcome during various stages of treatment was examined. In addition, we examined whether perceptions of control, which have been theoretically and empirically linked to anxiety (Barlow, 1988; Emmelkamp & Cohen-Kittenis, 1975; Rotter, 1966; Watson, 1967), might have an impact on homework compliance and thereby influence treatment outcome. Stated otherwise, failure to improve in cognitive-behavioral treatment might result from poor homework compliance which, in turn, might be related to the patient's perceived lack of control over outcomes. Perceptions of control (or uncontrollability) may vary among the different anxiety disorders. Social phobia, for instance, is marked by preoccupation with thoughts of uncontrollable events such as negative evaluation or criticism by others while panic disorder is marked by preoccupation with thoughts of uncontrollable events such as recurrent unexpected panic attacks (APA, 1994). Using Levenson's (1973) Locus of Control Scale (LOCS), Cloitre, Heimberg, Liebowitz and Gitow (1992) demonstrated that patients with social phobia differ from those with panic disorder in the nature of their attributions of control. Both patient groups reported less internal control (i.e. lower scores on the Internality subscale) than normal Ss, but they differed in their explanations of this lack of control. Social phobic patients achieved the highest scores on the Powerful Others subscale of the LOCS, suggesting that they believe events are controllable but only by people other than themselves. Patients with panic disorder, in contrast, achieved their highest score on the Chance subscale, attributing events to luck, chance, or fate. The present study sought further evidence of the relation between social phobia and perceptions of control and examined the clinical correlates of LOCS scores among social phobic patients. As mentioned above, it also investigated whether these perceptions might influence homework compliance and whether homework compliance facilitates treatment effectiveness. METHOD

Subjects Social phobics. The clinical sample consisted of 104 patients who sought treatment for interpersonal or performance anxiety at the Center for Stress and Anxiety Disorders of the University at Albany, State University of New York (CSAD). All were assessed with the Anxiety Disorder Interview Schedule-Revised (ADIS-R; DiNardo & Barlow, 1988) and met DSM-III-R criteria for social phobia. Patients who received a comorbid diagnosis of current major depression, bipolar disorder, psychotic disorder, or active alcohol or drug dependence within the past 3 months were excluded. Patients must also have received a rating of 4 or greater on the ADIS-R Clinician's Severity Rating Scale, a 0-8 global rating that incorporates both level of anxiety and degree of functional interference. The ADIS-R has been characterized by high rates of interrater agreement for the principal diagnosis of social phobia (DiNardo, Moras, Barlow, Rapee & Brown, 1993). Comparison group. A group of nonanxious persons was included in this study for comparison to the social phobics on measures of perception of control. Fifty individuals were solicited from the Albany community by bulletin board advertisements for nonanxious adult volunteers to participate in studies conducted by the CSAD. Each respondent was first screened by telephone to rule out anxiety disorder, affective disorder, substance abuse or psychological treatment (except for marital problems) within the past 2 years.

Homework compliance and social phobia

425

Demographic differences. Descriptive information about the social phobic and comparison Ss is reported in Table I. The two samples were similar in terms of age, gender, and marital status but differed on education and employment. Social phobic patients were more likely to have had college and/or graduate education than the community Ss (Z2(3, n = 154)= 14.90, P < 0.002). However, they were less likely to be employed on a full-time basis than the community Ss (X2(I, n = 154) = 7.36,P < 0.007). Questionnaire measures Locus of control assessment. The Levenson (1973) Locus of Control Scale (LOCS) consists of 24 items grouped into three subscales. Each subscale ranges in value from 8 to 48, with higher scores indicating greater endorsement. The Internality Subscale measures belief in one's ability to influence events (e.g. "When I get what I want, it's usually because I worked hard for it"). The Powerful Others Subscale examines the tendency to believe that success or failure is determined by persons with power (e.g. " I f important people were to decide they didn't like me, I probably wouldn't make many friends"). The Chance Subscale reflects attribution of causality to chance or luck (e.g. "It's not wise for me to plan too far ahead because many things turn out to be a matter of good or bad fortune"). All subscales have been reported to have acceptable reliability and validity (Cloitre et al., 1992; Levenson, 1981). Social anxiety. The Social Interaction Anxiety Scale (SIAS; Mattick & Clarke, 1989) contains 20 items which are rated from 0 (not at all characteristic or true of me) to 4 (extremely characteristic or true of me). Items describe one's typical cognitive, affective, or behavioral reaction to a variety of situations requiring social interaction in dyads or groups, such as going to a party, talking to an attractive member of the opposite gender, or expressing one's feelings. Total scores range from 0 to 80, with higher scores representing higher levels of social interactional anxiety. The reliability and validity of the SIAS as a measure of social interactional anxiety have been repeatedly demonstrated (Brown, Turovsky, Heimberg, Juster, Brown & Barlow, 1996; Heimberg, Mueller, Holt, Hope & Liebowitz, 1992; Mattick & Clarke, 1989). Three other questionnaire measures commonly used in research on social phobia were also included in this study: (1) the Fear of Negative Evaluation Scale (FNE; Watson & Friend, 1969), a measure of concern about the disapproval of others; (2) the Social Phobia Scale (SPS; Mattick & Clarke, 1989), which examines anxiety about being observed by others; and (3) the Social Avoidance and Distress Scale (SADS; Watson & Friend, 1969), which evaluates distress caused by interpersonal interactions and the deliberate avoidance of those situations. Assessment of homework compliance Homework was assigned to CBGT patients at the end of each treatment session (except the final session) and reviewed at the beginning of the following session. The nature of homework varied according to the phase of treatment and individual patients' progress. The initial stage of CBGT Table 1. Demographic characteristics of social phobics and community subjects Social phobics

Community subjects

%Male

55.8

50.0

Mean age (SD)

35.8 (9.55)

33.2 (9.46)

Marital s t a t u s Single Married Divorced/separated/widowed

56 (53.8%) 34 (32.7%) 14 ( 13.5%)

19 (38.0%) 23 (46.0%) 8 ( 16.0% )

Education High school or less Some college College graduate Graduate school

13 (12.5%) 15 (14.4%) 50 (48.1%) 26 (25.0%)

16 (32.0%) 13 (26.0%) 12 (24.0%) 9 (18.0%)

Employment Full-time Others

65 (62.5%) 39 (37.5%)

42 (84.0%) 8 (16.0%)

426

Anna W. Leungand RichardG. Heimberg

(weeks 1 and 2) focused on education about the cognitive-behavioral model of social phobia, and assigned homework consisted of self-monitoring of negative thoughts and feelings in anxietyprovoking situations. The middle portion of CBGT (weeks 3 to 7) involved the first several in-session exposures, and assigned homework consisted of self-monitoring, cognitive preparation for individualized in vivo exposures, and the conduct of these self-directed exposures. During the latter phase of CBGT (weeks 8 to 12), in-session exercises continued, and homework was primarily devoted to in vivo exposures and associated cognitive preparation. Patient's compliance with homework assignments was assessed weekly by the therapists using a 0-6 homework compliance scale (HCS; Primakoff, Epstein & Covi, 1986). At the end of each session, the therapists evaluated the degree of compliance of each patient using the HCS, with higher ratings indicating greater compliance ('0' = homework was not assigned; '1' = the patient did not attempt assigned homework; '2' = the patient attempted the assigned homework but was unable to execute it due to extenuating circumstances; ' 3 ' = the patient did homework that was different from that assigned, but could be considered "relevant" to the patient's particular target problems; '4' = the patient did some portion of the assigned homework; '5' = the patient completed the assigned homework; and ' 6 ' = the patient did more of the assigned homework than was requested). An overall average was obtained for each patient by summing HCS scores for sessions 2-12 and dividing by the number of sessions attended. Furthermore, average scores were calculated for the three phases of CBGT described above (HCS1 for weeks 1-2; HCS2 for weeks 3-7; and HCS3, weeks 8-12). Procedure for social phobics

Each potential patient was offered participation in a multicenter study of cognitive-behavioral and pharmacological treatments for social phobia. Treatments included CBGT, education-supportive group therapy, the monoamine oxidase inhibitor phenelzine, and pill placebo. Patients who either refused randomization or were excluded for medical reasons received CBGT as part of another study. The LOCS, FNE, SPS, SIAS and SADS, were administered as part of pretreatment assessment. Patients who received CBGT (n = 91) met in groups of six for twelve 2.5-hr weekly sessions. CBGT was administered by a licensed psychologist and an advanced doctoral student in clinical psychology. CBGT consists of several components: (1) presentation of a cognitive-behavioral explanation of social phobia; (2) training of patients in the skills of identification and disputation of problematic cognitions, and developing alternative rational responses through the use of structured exercises; (3) exposure of patients to simulations of anxiety-provoking situations during therapy sessions; (4) use of cognitive restructuring procedures to teach patients to control their maladaptive thinking before, during, and after simulated exposures; (5) homework assignments for in vivo exposure to situations previously addressed in sessions; and (6) teaching patients a self-administered cognitive restructuring routine for use before and after completion of exposure assignments. More thorough description of CBGT has been provided by Heimberg et al. (1995) and Hope and Heimberg (1993). Upon completion of CBGT, patients were reassessed. This assessment was identical to the pretreatment assessment except that the LOCS was not readministered. Procedure for the comparison sample

Comparison Ss completed the LOCS and other questionnaires at home. Questionnaires were returned to the CSAD in sealed envelopes within a l-month period. These individuals received $30 for their participation in this and other studies. RESULTS Group differences on the L O C S subscales

As shown in Table 2, social phobic Ss had a significantly lower mean score on Internality than the community Ss. They also had a significantly higher mean score on Powerful Others than the community Ss, but no differences were noted on the Chance subscale. Thus, social

Homework compliance and social phobia

427

Table 2. Group comparison on Levenson Locus of Control Subscales Group Social phobia (n = 104) Subsca[es Internality Powerful Others Chance

Comparison (n = 50)

M

SD

M

SD

t (152)

P

35.82 24.12 23.16

5.24 7.04 7.29

38.66 21.34 22.62

5.71 7.24 8.35

- 3.06 2.27 0.41

0.003 0.025 ns

phobics were more likely to attribute control to powerful others and less likely to attribute control to internal resources than community Ss.

Association between perceptions of control and social anxiety The Internality and Powerful Others subscales scores were correlated with measures of pre- and posttreatment anxiety within the social phobic sample (see Table 3). All pretreatment anxiety measures were significantly and negatively related to Internality. The strongest relationship was observed between Internality and the SIAS while the weakest relationship was observed with the SPS. A significant positive relationship was found between Powerful Others and each of the pretreatment self-report measures. Pretreatment anxiety was related to attributions of control to external sources, specifically to other powerful persons. Posttreatment social anxiety as indexed by the SIAS was significantly and negatively related to Internality. However, posttreatment FNE, SADS and SPS scores were unrelated to Internality. Powerful Others scores were significantly and positively associated with posttreatment scores on the SADS and FNE, but not the SIAS or SPS. However, the effect was not strong or demonstrated consistently across measures.

Association between homework compliance and social anxiety No significant relationship occurred between the overall homework compliance (HCS) score and any measure of pretreatment social anxiety. Homework compliance during the various phases of treatment, as indexed by HCS1, HCS2, HCS3, also failed to show significant relationships with pretreatment measures of social anxiety. Overall HCS was not significantly related to the measures of posttreatment anxiety with the exception of the SIAS (r = - 0.24, P < 0.028). HCS1 was generally unrelated to posttreatment anxiety, although the inverse correlation between HCSI and the SIAS approached significance (r = - 0 . 2 0 , P < 0.056). HCS2 showed a significant but positive relationship with the FNE (r = 0.21, P < 0.04) and SPS (r = 0.22, P < 0.04), but was not significantly related to the SIAS. These rs suggest that greater homework compliance during the middle portion of treatment was associated with greater posttreatment concern over scrutiny and negative evaluation by others. HCS3 was significantly and negatively associated (r = - 0.29, P < 0.008) with SIAS scores only, suggesting that greater homework compliance in the latter portion of treatment was associated with less posttreatment anxiety.

Relationship between homework compliance, perceived internal control, and posttreatment anxiety To examine the overall relationship between homework compliance and patients' perceptions of control, overall HCS and the component homework compliance scores were correlated with the Table 3. Correlations of Internality and Powerful Others with pretreatment and posttreatment anxiety measures

Posttreatment

Pretreatment

Internality Powerful Others

SIAS

SPS

FNE

SADS

SIAS

SPS

FNE

SADS

-0.44*** 0.36***

- 0.18" 0.29**

- 0.31"** 0.32***

-0.34*** 0.18"

- 0.31"* 0.19

-0.01 0.04

-0.10 0.38***

-0.23 0.28*

Note. SIAS, Social Interaction Anxiety Scale; SPS, Social Phobia Scale; FNE, Fear of Negative Evaluation Scale; SADS, Social Avoidance and Distress Scale. Pretreatment n varies between 102 and 104 as a result of missing data; posttreatment n varies between 67 and 83 as a result of missing data. *P < 0.05; **P <001; ***P < 0.001.

428

Anna W. Leung and Richard G. Heimberg Table 4. Hierarchical multiple regression of pretreatment SIAS, Internality, overall homework compliance scale and the interaction between internality and overall homework compliance on posttreatment SIAS Variable Pretreatment SIAS Internality HCS Internality × HCS

r2

0.64 0.64 0.70 0.71

r2

change

0.64 0.0003 0.06 0.01

F change

fl

t

114.46 0.05 13.21 2.61

0.80 0.02 0.25 0.11

10.70" 0.22 3.64* 1.62

Note: SIAS, Social Interaction Anxiety Scale; HCS, Overall Homework

Compliance Scale. For the SIAS (pretreatment and posttreatment), a higher score denotes a higher level of social interactional anxiety. For internality, a higher scoredenotes a greater tendencyto attribute control to internal resources.For the HCS.higherscoresdenotegreater compliancewith homeworkassignments, n = 67. *P < 0.001.

Internality and Powerful Others subscales of the LOCS. There were no significant relationships between the measures of homework compliance and either subscale. To further understand the relationship between posttreatment anxiety, homework compliance and perceptions of control, a series of hierarchical multiple regression analyses were conducted.* Only social phobic patients who completed treatment and provided complete data on all study measures (n = 67) were included in these analyses. The SIAS served as the measure of posttreatment social anxiety. Table 4 presents the hierarchical multiple regression of overall homework compliance, Internality, and their interaction on posttreatment social interactional anxiety. The predictor variables were first transformed by "centering", that is, subtracting the mean value of the variable from each score. This transformation is recommended to minimize inflation of standard errors for regression coefficients due to the use of multiplicative terms and to reduce multicollinearity (Jaccard, Turrisi & Wan, 1990). Pretreatment SIAS was entered in the first step as a covariate and accounted for 63.8% of the variance in posttreatment SIAS, F(1,65)= 114.46, P < 0.0001. Internality was entered in the second step since it was conceptualized as a personality characteristic of patients that was present prior to their participation in treatment, followed by the overall homework compliance score in the third step. The addition of Internality did not significantly increase the explained variance, F-change = 0.04, N S . However, overall homework compliance was a significant predictor, accounting for 6.3% of the variance in posttreatment SIAS, F-change = 13.21, P < 0.0006. The interaction of Internality and overall homework compliance was entered last but did not account for significant additional variance, F-change = 2.61, NS. Greater compliance with homework assignments in C B G T significantly predicted reduced posttreatment social interactional anxiety, but neither Internality nor the interaction of Internality and homework compliance contributed to that prediction. Additional hierarchical multiple regression analyses were conducted to identify the specific effect of homework compliance on posttreatment SIAS scores. Rather than overall homework compliance, HCSI, HCS2, and HCS3 were used in this series of analyses. In the first of these analyses, pretreatment SIAS was entered first as a covariate, followed by Internality, HCS1, and finally, the interaction term. This procedure was repeated for HCS2 and HCS3. HCSI made a unique contribution, accounting for 2.47% of the variance in lower posttreatment SIAS beyond that predicted by pretreatment SIAS scores and Internality, F - c h a n g e = 4.41, P < 0.04. Likewise, HCS3 accounted for 2.7% of the variance in lower posttreatment SIAS, F-change = 4.93, P < 0.03. No significant unique contribution was noted for HCS2. Internality and the interaction of Internality and homework compliance failed to contribute to the prediction of posttreatment SIAS in any equation. Overall, these data suggest that homework compliance, specifically during the earlier and later portions of CBGT, has a meaningful relationship to treatment outcome. *These analyses were repeated substituting the LOCS Powerful Others subscale score for the Internality subscale score. Results were basically equivalent to those presented here.

Homework complianceand social phobia

429

DISCUSSION Previous research suggests that compliance with homework assignments in cognitive-behavioral therapy may contribute to treatment success (Neimeyer, Twentyman & Prezant, 1985; Person & Burns, 1985) and that social phobics who show greater improvement after CBGT (Hope et al., 1991; Laguna et al., 1994) or at follow-up (Edelman & Chambless, 1995) have been more compliant with assigned homework. In the present study, we also examined homework compliance among social phobic patients who received CBGT. It was hypothesized that patients who complied with homework assignments would be more likely to respond to treatment and show lower levels of posttreatment social anxiety. In addition, perceived lack of control over events has been associated with anxiety (Barlow, 1988), and social phobia has been characterized by a lack of belief in one's own control of events and belief in the control of these events by powerful others (Cloitre et al., 1992). Thus another purpose of this study was to examine how this pattern of attributions among social phobics differed from that of normal controls and whether it was related to symptom severity. We also hypothesized that perceptions of control would influence homework compliance and moderate the relationship of homework compliance to treatment outcome. Thus, we tested the ability of these variables (and their interaction) to predict posttreatment social anxiety in a series of multiple regression analyses. Homework compliance

Correlational analyses revealed that pretreatment anxiety was not associated with overall homework compliance or homework compliance as measured at each of the three periods of CBGT. Thus, patients' likelihood of completing homework assignments had little to do with how anxious they felt before treatment. Similar findings were reported by Edelman and Chambless (1995). Other investigations have failed to find an association between pretreatment emotional states (e.g. depression) and homework compliance (Bryant, Simons & Thase, 1993; Laguna et al., 1994). Homework compliance in the early stages of CBGT tended to be inversely associated with posttreatment social international anxiety. Laguna et al. (1994) also reported that homework compliance early in treatment was associated with posttreatment anxiety reduction. These investigators also reported an effect on treatment outcome of homework compliance during the middle portion of CBGT, but little impact of homework compliance in the latter stage. In contrast, we found that homework compliance in the third segment of CBGT (but not the second segment) was significantly related to posttreatment anxiety reduction. Reasons for the discrepancy between these two studies are unclear. Interestingly, in our study, compliance in the second segment appeared to be associated with greater anxiety and fear of negative evaluation. The relationship of the three homework compliance scores to posttreatment social interactional anxiety is discussed further in the context of our multiple regression analyses. Perceptions of control

Social phobic patients, compared to community Ss, exhibited lower scores on Internality and higher scores on perceptions of control by Powerful Others. This result is consistent with the findings of Cloitre et al. (1992) as well as current formulations of the nature of anxiety (Barlow, 1988). Social phobic patients, like individuals with other anxiety disorders, perceive a lack of internal control. Events are seen as unpredictable and their outcomes as uncontrollable. However, social phobics do not believe events are totally random. Instead, events are orderly and outcomes are determined by other participants in social situations. Individuals with social phobia may view these participants as potential sources of negative evaluation or criticism, i.e. feared stimuli. The more severe the person's social phobia, the less s/he believes that s/he has control over events and the more s/he believes others control those events. Given that these perceptions are associated with pretreatment level of social anxiety, they might be expected to influence the process of therapeutic change. Patients who have little faith in their ability to control events might be unwilling to take risks in treatment or may be hesitant to expose themselves to feared social situations. Their belief in the control of outcomes by powerful others paired with their generalized expectations of BRT 34/5--B

430

Anna W. Leungand Richard G. Heimberg

negative reactions from others (Leary, Kowalski & Campbell, 1988) may reinforce their belief that social situations are dangerous. Social phobics who, prior to treatment, attributed control over events less to their own abilities reported greater anxiety after treatment. However, this relationship was not consistently demonstrated across posttreatment measures of social anxiety. Internality was associated only with posttreatment scores on the SIAS, while scores on Powerful Others were related only to scores on the FNE and SADS. These data do not provide strong support for the relationship of pretreatment attributions of control to treatment outcome, a point that is underscored by the findings of our multiple regression analyses. Homework compliance, Internality, and posttreatment social interactional anxiety Multiple regression analyses examined the effect of homework compliance, Internality and their interaction on posttreatment social interactional anxiety (SIAS). Initial hierarchical regression analyses revealed that overall homework compliance was a significant predictor of posttreatment social interactional anxiety even when controlling for level of social anxiety prior to treatment. However, neither Internality nor the interaction of overall homework compliance and Internality significantly increased the amount of explained variance in posttreatment anxiety. In other words, patients who complied with homework assignments over the course of CBGT showed less anxiety in a wide variety of situations requiring social interactions. Our findings, in combination with those of Edelman and Chambless (1995) and Laguna et al. (1994), suggest that homework compliance throughout treatment increases the chance of positive treatment outcome. However, the completion of weekly homework which includes in vivo exposures is an act of courage and perseverance, and the benefits of doing so may not be clear to patients at all times. This may be especially so for many patients during the middle phase of CBGT since HCS2 scores were significantly associated with fears of scrutiny and negative evaluation by others. Initial exposures to feared situations may engender increased anxiety and fear of negative evaluation. Even though our multiple regression results showed that homework compliance during this period of treatment was not related to decreased levels of social interactional anxiety after treatment, homework compliance during this time significantly predicted homework compliance in the latter stage of treatment (r = 0.49, P < 0.0001), which, in turn, did predict positive outcome of treatment. In the long run, patients complying with homework appear to complete treatment with lower levels of social interactional anxiety. Contrary to our hypotheses, posttreatment anxiety was not predicted by pretreatment levels of Internality or its interaction with homework compliance. Correlations between pretreatment Internality and measures of posttreatment social anxiety were modest, and Internality contributed no variance to the prediction of posttreatment social anxiety beyond that contributed by pretreatment social anxiety. In other words, social phobics who attributed control over events to internal resources did not necessarily become less anxious by the end of treatment than those who did not. Moreover, homework compliance, especially in the early and latter stages of treatment, was predictive of posttreatment social interactional anxiety, regardless of patient level of internality. The hypothesis that internality is a potential moderating factor in CBGT outcome was not supported. Limitations and future directions Despite a number of significant findings, the present study has several limitations. First, the Homework Compliance Scale measured only the degree of completion of weekly assignments. It did not assess the quality of homework completed or the degree to which the tasks were done correctly. Also, we did not examine patients' subjective evaluation of their homework performance. A more comprehensive evaluation of the execution of homework tasks should enable us to more thoroughly examine the significance of homework in cognitive-behavioral treatment. Additionally, the LOCS was administered only prior to treatment. Whether perceptions of control assessed at posttreatment or change in these perceptions over the course of treatment would show a stronger relationship to posttreatment anxiety is an area for future research attention. Future research should also examine other variables that potentially affect homework compliance, such as problem-solving skills or perfectionistic ideation about homework. One hypothesis

Homework compliance and social phobia

431

is that social phobic patients may be likely to underestimate their homework accomplishments because they may reach conclusions based on their emotional reactions and find reasons why positive homework outcomes "don't count". These distorted cognitive styles, if not addressed in therapy, may affect patients' compliance with homework or the accuracy of their report of homework completion by negatively biasing their evaluations of their homework performance. The results of this study underscore the need for more adequate measures of homework compliance. In addition to the Homework Compliance Scale, Primakoffet al. (1986) have proposed other measures that may be useful for the meaningful assessment of homework compliance in cognitive therapy in general. They recommended an independent assessment of therapist adherence to the homework protocol to ensure their consistency and compliance in rating homework. They also suggest measuring the quality of completed homework (i.e. the degree to which tasks are done correctly) in addition to the amount of homework completed. Measuring both the amount and the quality of homework compliance may further our understanding of the specific effect of the homework component in cognitive-behavioral treatment and in what way compliance with homework assignments improves treatment success.

REFERENCES American Psychiatric Association (1987). Diagnostic and Statistical Manual of Mental Disorders (3rd edn revised). Washington, DC: Author. American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders (4th edn). Washington, DC: Author. Amies, P. L., Gelder, M. G. & Shaw, P. M. (1983). Social phobia: a comparative clinical study. British Journal of Psychiatry, 142, 174-179. Barlow, D. H. (1988). Anxiety and its Disorders. New York: Guilford Press. Brown, E. J., Heimberg, R. G. & Juster, H. R. (1995). Social phobia subtype and avoidant personality disorder: effect on severity of social phobia, impairment, and outcome of cognitive-behavioral treatment. Behavior Therapy, 26, 467-486. Brown, E. J., Turovsky, J., Heimberg, R. G., Juster, H. R., Brown, T. A. & Barlow, D. H. (1996). Validation of the Social Interaction Anxiety Scale and the Social Phobia Scale across the anxiety disorders. Manuscript submitted for publication. Bryant, M. J., Simons, A. D. & Thase, M. E. (1993). Therapist behavior and patient variables related to homework compliance in cognitive-behavioral therapy of depression. Paper presented at the Annual Convention of the Association for Advancement of Behavior Therapy, Atlanta, GA. Cloitre, M., Heimberg, R. G., Liebowitz, M. R. & Gitow, A. (1992). Perceptions of control in panic disorder and social phobia. Cognitive Therapy and Research, 16, 569-577. DiNardo, P. A. & Barlow, D. H. (1988). The Anxiety Disorders Interview Schedule, Revised (ADIS-R). Albany, NY: Graywind Publications. DiNardo, P. A., Moras, K., Barlow, D. H., Rapee, R. M. & Brown, T. A. (1993). Reliability of DSM-III-R anxiety disorder categories. Archives of General Psychiatry, 50, 251-256. Edelman, R. E. & Chambless, D. L. (1995). Adherence during sessions and homework in cognitive-behavioral group treatment of social phobia. Behaviour Research and Therapy, 33, 573-577. Emmelkamp, P. M. G. & Cohen-Kettinis, P. T. (1975). Relationship of locus of control to phobic anxiety and depression. Psychological Reports, 36, 390. Gelernter, C. S., Uhde, T. W., Cimbolic, P., Arnkoff, D. B., Vittone, B. J., Tancer, M. E. & Bartko, J. J. (1991). Cognitive-behavioral and pharmacological treatments of social phobia: a controlled study. Archives of General Psychiatry, 48, 938 945. Heimberg, R. G. (1990). Cognitive behavior therapy [for social phobia]. In Bellack, A. S. & Hersen, M. (Eds), Handbook of Comparat&e Treatments for Adult Disorders (pp. 203-218). New York: Wiley. Heimberg, R. G. & Juster, H. R. (1995). Cognitive-behavioral treatment: Literature review. In Heimberg, R., Liebowitz, M., Hope, D. & Schneier, F. (Eds) Social Phobia: Diagnosis, Assessment and Treatment (pp. 261-309). New York: Guilford Press. Heimberg, R. G., Becker, R. E., Goldfinger, K. & Vermilyea, J. A. (1985). Treatment of social phobia by exposure, cognitive restructuring, and homework assignments. Journal of Nervous and Mental Disease, 173, 236--245. Heimberg, R. G., Juster, H. R., Hope, D. A. & Mattia, J. I. (1995). Cognitive behavioral group treatment for social phobia: description, case presentation and empirical support. In Stein, M. B. (Ed.) Social Phobia: Clinical and Research Perspectives (pp. 293-321). Washington, DC: American Psychiatric Press. Heimberg, R. G., Mueller, G. P., Holt, C. S., Hope, D. A. & Liebowitz, M. R. (1992). Assessment of anxiety in social interaction and being observed by others: the Social Interaction Anxiety Scale and the Social Phobia Scale. Behavior Therapy, 23, 53-73. Heimberg, R. G., Klosko, J. S., Dodge, C. S., Shadick, R., Becker, R. E. & Barlow, D. H. (1989). Anxiety disorders, depression, and attributional style: a further test of the specificity of depressive attributions. Cognitive Therapy and Research, 13, 21 36. Heimberg, R. G., Dodge, C. S., Hope, D. A., Kennedy, C. R., Zollo, L. & Becker, R. E. (1990). Cognitive behavioral group treatment of social phobia: comparison to a credible placebo control. Cognitive Therapy and Research, 14, 1-23. Hope, D. A. & Heimberg, R. G. (1993). Social phobia and social anxiety. In Barlow, D. H. (Ed.) Clinical Handbook of Psychological Disorders: A Step-By-Step Treatment Manual (2nd ed.) (pp. 99-136). New York: Guilford Press.

432

Anna W. Leung and Richard G. Heimberg

Hope, D. A., Herbert, J. D. & Bellack, A. S. (1991). Social phobia subtype, avoidant personality disorder and psychotherapy outcome. Paper presented at the Annual Meeting of the Association for Advancement of Behavior Therapy, New York. Hope, D. A., Herbert, J. D. & White, C. (1995). Diagnostic subtype, avoidant personality disorder, and efficacy of cognitive behavioral group therapy for social phobia. Cognitive Therapy and Research, 19, 399~,17. Jaccard, J., Turrisi, R. & Wan, C. K. (1990). Interaction Effects in Multiple Regression. London: Sage Publications. Kessler, R. C., McGonagle, A. M., Zhao, S., Nelson, C. B., Hughes, M., Eshelman, S. E., Wittchen, H. & Kendler, K. S. (1994). Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Archives of General Psychiatry, 51, 8-19. Laguna, L. B., Hope, D. A. & Herbert, J. D. (1994). Homework compliance and improvement in cognitive-behavioral group therapy with social phobics. Paper presented at the Annual Meeting of the Anxiety Disorders Association of America, Santa Monica, CA. Leary, M. R., Kowalski, R. M. & Campbell, C. D. (1988). Self-presentational concerns concepts and social anxiety: the role of generalized impression expectancies. Journal of Research in Personality, 22, 308 321. Levenson, H. (1973). Multidimensional locus of control in psychiatric patients. Journal of Consulting and Clinical Psychology, 41, 397~,04. Levenson, H. (1981). Differentiating among internality, powerful others and chance. In Lefcourt, H. M. (Ed.), Research with the Locus of Control Construct: Vol. I. Assessment Methods. New York: Academic Press. Liebowitz, M. R., Gorman, J. M., Fyer, A. J. & Klein, D. F. (1985). Social phobia: review of a neglected anxiety disorder. Archives of General Psychiatry, 42, 729 736. Lucas, R. A. & Telch, M. J. (1993). Group versus individual treatment of social phobia. Poster presented at the Annual Meeting of the Association .fi)r Advancement of Behavior Therapy, Atlanta, GA. Mattick, R. P. & Clarke, J. C. (1989). Development and validation of measures of social phobia scrutiny fear and social interaction anxiety. Unpublished manuscript. Neimeyer, R. A., Twentyman, C. T. & Prezant, D. (1985). Cognitive and interpersonal group therapies for depression: a progress report. The Cognitive Behaviorist, 7, 21 22. Person, J. B. & Burns, D. D. (1985). Factors associated with dropout and outcome in a naturalistic study of cognitive therapy for depression. Paper presented at the Annual Meeting of the Society for Psychotherapy Research, Evanston, IL. Primakoff, L., Epstein, N. & Covi, L. (1986). Homework compliance: an uncontrolled variable in cognitive therapy outcome research. Behavior Therapy, 17, 433~146. Rotter, J. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 80(1 Whole No. 609). Schneier, F. R., Johnson, J., Hornig, C. D., Liebowitz, M. R. & Weissman, M. M. (1992). Social phobia: comorbidity and morbidity in an epidemiologic sample. Archives of General Psychiatry, 49, 282-288. Schneier, F. R., Martin, L. Y., Liebowitz, M. R., Gorman, J. M. & Fyer, A. J. (1989). Alcohol abuse in social phobia. Journal of Anxiety Disorders, 3, 15-23. Watson, D. (1967). Relationship between locus of control and anxiety. Journal of Personality and Social Psychology, 6, 91-92. Watson, D. & Friend, R. (1969). Measurement of social-evaluative anxiety. Journal of Consulting and Clinical Psychology, 33, 448457.