Journal of Consulting and Clinical Psychology 2011, Vol. 79, No.

1, 123128

2010 American Psychological Association 0022-006X/10/$12.00 DOI: 10.1037/a0021889

BRIEF REPORT

Making Self-Help More Helpful: A Randomized Controlled Trial of the Impact of Augmenting Self-Help Materials With Implementation Intentions on Promoting the Effective Self-Management of Anxiety Symptoms
Rachel Varley, Thomas L. Webb, and Paschal Sheeran
University of Sheffield
Objective: The effectiveness of self-help materials may be constrained by failures to undertake recommended exercises or to deploy the techniques that one has learned at the critical moment. The present randomized controlled trial investigated whether augmenting self-help materials with ifthen plans (or implementation intentions) could overcome these problems and enhance the self-management of anxiety symptoms. Method: At baseline, participants who reported anxiety symptoms completed the Hospital Anxiety and Depression Scale (HADS) and the state version of the StateTrait Anxiety Inventory (STAI). Participants were then randomized via a computer program to standard self-help (n 86), augmented self-help (n 90), or no-intervention (n 86) conditions. Eight weeks later, 95% (n 249) of the participants completed the HADS and STAI again. Results: Findings showed a significant reduction in anxiety in the augmented self-help condition compared with both the standard self-help and nointervention conditions (caseness rates on the HADS at follow-up were 21%, 49%, and 44%, respectively). Mediation analyses indicated that the benefits of augmented self-help materials were explained by improved detection of anxiety-related triggers and greater experienced benefits of the self-help techniques. Conclusions: These findings suggest that implementation intentions offer a valuable supplement to self-help materials that can enhance their impact on outcomes. Keywords: self-help, self-regulation, anxiety, implementation intentions, relaxation

Self-help, self-care, self-management, self-examination therapy, and bibliotherapy are all terms used interchangeably to describe interventions where the client uses literature and techniques presented in a variety of formats to manage his or her mental health difficulties with minimal or no direction from relevant professionals (Richards, 2004). Reviews of the literature indicate that interventions using self-help materials can be effective in alleviating anxiety and other disorders (e.g., Apodaca & Miller, 2003; Scogin, Bynum, & Stephens, 1990). However, there is considerable variability in effect sizes, and several recent studies have failed to obtain significant improvements in outcomes (e.g., Haeffel, 2010). A key factor determining the effectiveness of self-help materials is compliance with relevant instructions or adherence to the selfhelp regimen. Gould and Clum (1993) defined compliance as

This article was published Online First December 20, 2010. Rachel Varley, Thomas L. Webb, and Paschal Sheeran, Department of Psychology, University of Sheffield, United Kingdom. The contributions of Thomas L. Webb and Paschal Sheeran were supported by ESRC grant RES-060-25-0044: Emotion regulation of others and self (EROS). Correspondence concerning this article should be addressed to Thomas L. Webb or Paschal Sheeran, Department of Psychology, University of Sheffield, Western Bank, Sheffield, United Kingdom S10 2TN. E-mail: t.webb@sheffield.ac.uk or p.sheeran@sheffield.ac.uk 123

actually using the self-help material (reading the manual or watching the videotape) or as applying and practicing what is suggested by that material (p. 170). Lack of compliance is a significant problem, with estimated rates of adherence as low as 50% (Glasgow & Rosen, 1978; Phillips, Johnson, & Geyer, 1972; Rakos, Grodek, & Mack, 1985). Clearly, even clients who are motivated to use self-help materials often fail to do so. Our hypothesis is that self-help materials could be more effective if they are augmented by ifthen plans or implementation intentions (Gollwitzer, 1999; Gollwitzer & Sheeran, 2006)that specify when and where the person will use the techniques that he or she has learned. Implementation intentions link suitable opportunities to act (specified in the if part of the plan) with instrumental responses to those opportunities (specified in the then part of the plan). For example, a client could link an anxiety-provoking situation with the use of a relaxation exercise from his or her self-help booklet: If I feel under pressure, then I will immediately use my breathing tactic to relax! Similarly, the client might plan a time and place to regularly practice the specified techniques: If I have finished putting the kids to bed, then I will use my relaxation exercise! Implementation intentions have been shown to promote compliance in a number of domains. For example, Brown, Sheeran, and Reuber (2009) found that epilepsy patients who planned when and where to take their medication each day were more likely to take their medication on schedule compared with control participants who merely completed a questionnaire con-

124

VARLEY, WEBB, AND SHEERAN

cerning their medication (rates were 79% and 55%, respectively). In a meta-analysis of 94 studies involving a wide variety of behavioral domains, implementation intentions had a medium-to-large effect (d 0.65) on the achievement of intended goals (Gollwitzer & Sheeran, 2006). The present research investigated the impact of augmenting selfhelp materials with implementation intentions that invite participants to plan when and where they will use specified relaxation techniques. Participants were randomly allocated to receive standard self-help materials, self-help materials augmented by implementation intentions, or no intervention. Levels of anxiety were assessed at baseline and 8 weeks later. We predicted that the augmented self-help intervention would engender greater reductions in self-reported anxiety symptoms compared with both standard self-help and no-intervention conditions. We also tested whether frequency of relaxation technique use, effective detection of triggers for anxiety, user-friendliness of the booklet, or experienced benefits of using the relaxation techniques mediated any impact of the augmented self-help materials on reductions in anxiety.

Method Participants
We conducted a randomized controlled trial with a parallel group design and a balanced allocation ratio (Moher et al., 2010). In April 2008 an e-mail was sent to approximately 25,000 staff and students at a university in the United Kingdom inviting them to participate in research that could be of interest and value to

individuals who were experiencing anxiety symptoms. Individuals who were already receiving professional help for anxiety were asked not to take part in the study. Figure 1 shows the flow of participants through the different stages of the trial. A total of 262 participants completed a baseline questionnaire and were randomly allocated via an html program to one of three conditions (participants were unaware of the randomization procedure): (a) Participants in the standard self-help condition received a self-help booklet incorporating relaxation techniques; (b) participants in the augmented self-help condition received the same booklet but also additional materials that asked them to form an implementation intention specifying when they would use the techniques; and (c) participants in the control condition did not receive any intervention. Participants who were assigned to the two self-help conditions were then taken to a Web page where they could download their respective self-help booklets. Eight weeks later, a link to a follow-up questionnaire was e-mailed to respondents; 249 participants (95%) replied, providing 95% power to detect a small-to-medium-sized effect (f 0.18) on outcomes (Faul, Erdfelder, Lang, & Buchner, 2007). Table 1 shows the sample characteristics by condition. Participants were predominantly White British (68.7%), but other ethnicities were also represented (3.2% Chinese, 2.0% White Irish, 0.4% Indian, 0.4% Pakistani, 0.4% Bangladeshi, 1.2% Black African). Ethnicity was not specified by 17.5% of the participants. Participants defined themselves as students (64.7%), academic staff (11.1%), clerical staff (7.1%), and other (17.1%).

Enrollment

E-mailed participation invite ( n = ~25,000) Participants who were already receiving professional help for anxiety were asked not to participate Randomized (n = 262) Allocation

Allocated to control condition (n = 86)

Allocated to standard selfhelp intervention (n = 86) Follow-Up

Allocated to augmented self-help intervention (n = 90)

Did not reply to follow-up e-mail (n = 8)

Did not reply to follow-up e-mail (n = 5) Analysis

Did not reply to follow-up e-mail (n = 0)

Analysed (n = 78)

Analysed (n = 81)

Analysed (n = 90)

Figure 1.

CONSORT 2010 flow diagram (from Moher et al., 2010).

MAKING SELF-HELP MORE HELPFUL

125

Table 1 Sample Characteristics and Anxiety Scores, by Condition

Condition Measure Age Percentage female HADS (depression)baseline HADS (depression)follow-up HADS (anxiety)baseline HADS (anxiety)follow-up STAIbaseline STAIfollow-up Control 28.81 (10.46) 67.95 5.97 (3.60) 4.90 (3.51) 10.94 (4.01) 9.72 (3.91) 53.03 (10.37) 45.71 (16.18) Standard self-help 29.37 (9.90) 74.07 6.32 (3.70) 4.98 (3.23) 10.95 (3.77) 10.14 (4.88) 53.32 (10.74) 47.69 (17.14) Augmented self-help 30.72 (11.33) 74.44 6.11 (3.78) 3.54 (3.33) 11.41 (4.03) 6.81 (4.83) 52.30 (11.72) 39.00 (19.49)

Note. Standard deviations are in parentheses. HADS Hospital Anxiety and Depression Scale; STAI State Trait Anxiety Inventory.

Baseline Questionnaires
At baseline, participants provided demographic information and completed the Hospital Anxiety and Depression Scale (HADS; Zigmond & Snaith, 1983) and the state version of the StateTrait Anxiety Inventory (STAI; Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983). The HADS includes 14 items, half of which measure anxiety (e.g., I get sudden feelings of panic) and the other half depression (e.g., I feel as if I am slowed down). Responses are made on a 0 3 scale, so scores range from 0 to 21 for each subscale. A substantial body of evidence supports the predictive and concurrent validity of the HADS (for a review, see Bjelland, Dahl, Haug, & Neckelmann, 2002). The state version of the STAI asks participants to rate how they currently feel on 20 items (e.g., calmness, worry). Responses are made on a 1 4 scale ranging from not at all to very much so. Therefore, scores range from 20 to 80. Although research on treatment evaluation tends to use the trait form of the STAI (e.g., Fisher & Durham, 1999), there are also studies attesting to the efficacy of the state form. For example, Kvaal, Ulstein, Nordhus, and Engedal (2005) found a clear relationship between the state form of the STAI and psychiatrist diagnoses according to Diagnostic and Statistical Manual of Mental Disorders (fourth edition, text revised; American Psychiatric Association, 2000) criteria. In the present study, both the anxiety and depression subscales of the HADS proved reliable (anxiety .86, depression .79), as did the STAI ( .93).

short breathing exercise that could be undertaken whenever it was neededfor example,
Put one hand on your chest and one hand on your stomach. Think about your breathing for a moment or two. Dont change your breathing just yet. You just want to see how you are feeling. Now breathe in through your nose and out through your mouth. Count to 5 slowly on your in breath and count to 5 on the out breath. Think about filling your lungs as full as they can be, and make sure that your stomach rises rather than your chest.

There was also a longer, deep muscle relaxation exercise that should be used daily at homefor example,
We are going to relax the muscles in your body so that you can see what it is like to feel tense and then fully relaxed. Clench your hands and arms as tight as you can. Hold it for as long as you can, then relax. Notice how different your muscles feel.1

Self-Help Interventions
The self-help booklet, Feeling Less Worried: A Three-Step Plan to Help You Manage Your Anxiety, was based on principles for using cognitive behavioral therapy without therapist intervention (e.g., Cuijpers, van Straten, & Andersson, 2008). The booklet was designed to be straightforward and easy to use; to provide a clear rationale as to why it would be helpful; and to break the task of anxiety management into small, specific steps that participants could work through at their own pace. The booklet was eight pages long and included psychoeducation (Step 1), diary sheets for monitoring anxiety-related triggers and feelings so that participants could construct their own individual anxiety profile (Step 2), and two relaxation techniques (Step 3). The techniques included a

Both the standard and augmented self-help booklets made it clear that participants should practice the relaxation techniques and try to use them to control their anxiety. However, only participants in the augmented self-help condition were also prompted to form two ifthen plans that specified when they would use each of the two relaxation techniques. The first ifthen plan asked participants to identify a particular anxiety-related symptom and to associate it with using their newly-learned breathing tactic (i.e., If I feel [self-nominated symptom], then I will use my breathing tactic to relax!) The second plan invited participants to identify a suitable opportunity to fit the deep muscle relaxation exercise into their daily routine (e.g., If I am in my bedroom before I go to bed, then I will use my relaxation tactic!).

Follow-Up Questionnaires
At follow-up, participants completed the anxiety and depression subscales of the HADS and the STAI again. Participants also completed measures designed to index four potential mediators of the effect of the self-help interventions on outcomes. Perceptions of the user-friendliness of the booklet were indexed by six items
Copies of the standard and augmented self-help booklets can be obtained from Thomas L. Webb or Paschal Sheeran.
1

126

VARLEY, WEBB, AND SHEERAN

Table 2 Effect of Standard Self-Help Versus Augmented Self-Help on Potential Mediators

Measure User-friendliness of the booklet Frequency of relaxation technique use Effective detection of triggers Experienced benefits of the relaxation techniques

Standard self-help Augmented self-help F(1, 172) 4.30 (0.64) 3.02 (0.81) 3.82 (0.86) 3.18 (0.67) 4.63 (0.40) 3.65 (0.80) 4.20 (0.81) 3.53 (0.85) 16.15 26.18 9.00 8.98

2 .09 .13 .05 .05

Note. Standard deviations are in parentheses. p .01. p .001.

measured on 5-point scales (ranging from disagree to agree): The instructions in the booklet were easy to follow; The writing in the booklet was easy to read; The language in the booklet was easy to understand; The layout of the booklet was easy to follow; The information about anxiety and its symptoms was useful; and The information about using relaxation was useful ( .93). Frequency of strategy use was measured by three items (5-point scales ranging from never to once or more a day): How often did you use any part of the self-help booklet?; How often did you use the on the spot relaxation technique?; and How often did you use the longer relaxation exercise? ( .76). Single items measured on 5-point scales (disagreeagree) were used to index effective detection of triggers to anxiety (I have been able to identify my own anxious symptoms) and experienced benefits of using the relaxation techniques (I found using the relaxation exercises effective).

scores: Augmented self-help participants had lower STAI scores compared with both the standard self-help participants ( p .01) and the control participants ( p .05); there was no difference between the standard self-help and control conditions ( p .57).3

Mechanisms of Intervention Effects

The second set of analyses examined whether user-friendliness of the booklet, frequency of relaxation technique use, effective detection of triggers for anxiety, and experienced benefits of using the relaxation techniques could explain why participants who received the augmented self-help booklet had lower anxiety scores than participants who received the standard self-help booklet. To observe mediation, there should be a significant effect of (a) condition (augmented self-help vs. standard self-help) on anxiety scores, (b) condition on the putative mediator variables, and (c) the mediators on anxiety scores. Finally (d), the test of the indirect effect of condition through the mediators should be significant (Preacher & Hayes, 2008). The analyses presented in the previous section indicate that the first criterion was met. Multivariate analysis of variance indicated that all four potential mediators met the second criteria, F(4, 169) 8.72, p .001, 2 .17 (see Table 2). Participants who received the augmented self-help materials perceived the booklet as more user-friendly, used the relaxation techniques more frequently, were better at detecting the triggers of their anxiety, and experienced greater
The full analyses undertaken to check representativeness and randomization can be obtained from Thomas L. Webb or Paschal Sheeran. 3 One might wonder whether the observed reductions in anxiety levels were of clinical significance. To assess this issue, we examined caseness scores on the HADS by condition over time. Caseness refers to participants whose anxiety levels imply a demand of resources in terms of professionals and money (Sandanger et al., 1999, p. 53), and cases are defined by HADS scores 11 (Demyttenaere et al., 2009). There was no significant difference between the augmented self-help, standard self-help, and control conditions in caseness rates at baseline (rates were 64%, 56%, and 53%, respectively), 2(2, N 249) 2.69, ns. Caseness rates did, however, differ by condition at follow-up, 2(2, N 249) 17.17, p .001. There were fewer cases in the augmented self-help condition (21%) compared with both the control (44%) and standard self-help conditions (49%). McNemar tests indicated that there was no significant reduction in caseness rates from baseline to follow-up among control participants ( p .35) or standard self-help participants ( p .35). However, there were significantly fewer cases in the augmented self-help condition at follow-up compared with baseline ( p .001). These findings suggest that the augmented self-help condition had a clinically significant impact on anxiety levels.
2

Results Effects of Condition on Anxiety Scores

Representativeness and randomization checks indicated no threats to the validity of the analyses.2 Two 3-way betweensubjects (condition: control vs. standard self-help vs. augmented self-help) by two-way within-subjects (time: baseline vs. followup) analyses of variance (ANOVAs) were therefore conducted with anxiety scores on the HADS and STAI scales as dependent variables. The main effect of time was significant for the anxiety subscale of the HADS, F(1, 246) 34.66, p .001, 2 .12, and the STAI, F(1, 246) 43.23, p .001, 2 .15, as were the main effects of condition, Fs(2, 246) 5.42 and 5.15, ps .01, 2 .04 and .04, respectively. These main effects were, however, qualified by the expected interaction between condition and time for both the HADS, F(2, 246) 10.66, p .001, 2 .08, and the STAI, F(2, 246) 3.19, p .05, 2 .03. The significant interactions were decomposed by examination of the effect of condition at each level of time. ANOVAs showed that the conditions did not differ on the HADS or the STAI at baseline, F(2, 246) 0.41 and 0.20, respectively, ns, 2s .00. However, there was a significant effect of condition on both variables at follow-up, F(2, 246) 13.49 and 5.68, ps .01, 2 .11 and .05, respectively. Scheffe post hoc tests indicated that participants who received the augmented self-help intervention had significantly lower scores on the HADS compared with both the standard self-help and the control conditions (both ps .001). The standard self-help and control conditions did not differ significantly ( p .85). Equivalent findings were obtained for STAI

MAKING SELF-HELP MORE HELPFUL

127

Table 3 Correlations Between Potential Mediators and Anxiety Scores at Follow-Up

Mediator 1. 2. 3. 4. 5. 6.

1 1.00

2 .43 1.00

3 .23 .24 1.00

4 .46 .34 .41 1.00

5 .06 .09 .21 .23 1.00

6 .03 .05 .18 .19 .86 1.00

User-friendliness of the booklet Frequency of relaxation technique use Effective detection of triggers Experienced benefits of the relaxation techniques HADS (anxiety)follow-up STAIfollow-up

Note. HADS Hospital Anxiety and Depression Scale; STAI StateTrait Anxiety Inventory. p .05. p .01. p .001.

benefits of using the relaxation techniques, compared with participants who received standard self-help materials. Next, correlations were computed between the potential mediators and anxiety scores (see Table 3). Findings indicated that effective detection of anxiety triggers and greater experienced benefits but not the other variables had significant correlations with HADS and STAI scores (rs .18, ps .05), and thus met the third criterion. The bootstrapping procedure recommended by Preacher and Hayes (2008) was used to test the direct effect of condition on anxiety at follow-up and the indirect effect of condition through the mediators (effective detection of anxiety triggers and greater experienced benefits). For the HADS, the direct effect of condition on anxiety remained significant even after the mediators were taken into account (B 4.36, SE 0.72, p .001). However, the total mediation effect was significant (B 1.04, SE 0.36, p .05), and effective detection of triggers and experienced benefits both emerged as significant individual mediating variables (B 0.48 and .56, SE 0.23 and 0.24, respectively, p .05). Equivalent findings were obtained for the STAI. The direct effect of condition on anxiety was significant (B 4.36, SE 0.72, p .001), and the total mediation effect was significant (B 3.08, SE 1.23, p .05). Again, effective detection of triggers and experienced benefits were both significant mediators (B 1.49 and 1.60, SE 0.83 and 0.79, respectively, p .05). These findings indicate that effective detection of anxiety triggers and greater experienced benefits of the relaxation techniques partially mediated the relationship between condition and anxiety scores. The direct effect of condition on anxiety remained significant, which suggests that additional variables may act as mediators alongside effective trigger detection and experienced benefits.

lower levels of anxiety at follow-up than participants in both the standard self-help material and control conditions. This was true whether mean scores on the HADS or STAI (or caseness rates on the HADS; see Footnote 3) were used to measure anxiety. Mediation analyses revealed that the effects of the augmented self-help materials were partly explained by improved detection of anxietyrelated triggers and greater experienced benefits of the relaxation techniques. The present findings point to the importance of not merely knowing self-help exercises but also identifying opportunities to use these exercises and acting upon such opportunities. Implementation intentions provide a simple and cost-effective technique for ensuring that self-help exercises are initiated at the critical juncture and could easily be incorporated into future selfhelp interventions in a variety of domains. Two limitations of the present research warrant discussion. First, implementation intention effects were evaluated with respect to self-reports of anxiety symptoms. Although studies have shown that implementation intentions influence emotional responses that are relatively uncontaminated by demand characteristics and social desirability biases (e.g., Webb, Ononaiye, Sheeran, Reidy, & Lavda, 2010), future research might usefully examine the impact of augmented self-help materials on less reactive indices of anxiety such as the Implicit Association TestAnxiety (Egloff & Schmukle, 2002). Second, participants were a self-selected sample of people who were experiencing anxiety symptoms. The efficacy of augmenting self-help materials with implementation intentions should be tested among clinical samples in future studies.

References
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. Apodaca, T. R., & Miller, W. R. (2003). A meta-analysis of the effectiveness of bibliotherapy for alcohol problems. Journal of Clinical Psychology, 59, 289 304. doi:10.1002/jclp.10130 Bjelland, I., Dahl, A. A., Haug, T. T., & Neckelmann, D. (2002). The validity of the Hospital Anxiety and Depression Scale: An updated literature review. Journal of Psychosomatic Research, 52, 69 77. doi: 10.1016/S0022-3999(01)00296-3 Brown, I., Sheeran, P., & Reuber, M. (2009). Enhancing antiepileptic drug adherence: A randomized controlled trial. Epilepsy & Behavior, 16, 634 639. doi:10.1016/j.yebeh.2009.09.014 Cuijpers, P., van Straten, A., & Andersson, G. (2008). Internetadministered cognitive behavior therapy for health problems: A systematic review. Journal of Behavioral Medicine, 31, 169 177. doi:10.1007/ s10865-007-9144-1

Discussion
The present research investigated whether augmenting self-help materials with implementation intention instructions that invite participants to plan when and where they would use the recommended techniques could boost the efficacy of the materials. Participants experiencing anxious symptoms were randomly allocated to receive the augmented self-help materials, standard selfhelp materials, or no intervention. Standard self-help materials did not reduce anxiety levels over the 8-week follow-up relative to the no-treatment control condition. However, there was a substantial beneficial effect of forming implementation intentions; participants who received the augmented self-help materials reported

128

VARLEY, WEBB, AND SHEERAN reporting parallel group randomised trial. British Medical Journal, 340, c869. Phillips, R. E., Johnson, G. D., & Geyer, A. (1972). Self-administered systematic desensitization. Behaviour Research and Therapy, 10, 9396. doi:10.1016/0005-7967(72)90014-9 Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879 891. doi:10.3758/ BRM.40.3.879 Rakos, R. F., Grodek, M. V., & Mack, K. K. (1985). The impact of a self-administered behavioral intervention on pediatric asthma. Journal of Psychosomatic Research, 29, 101108. doi:10.1016/00223999(85)90014-5 Richards, D. (2004). Self-help: Empowering service users or aiding cash strapped mental health services? Journal of Mental Health, 13, 117123. doi:10.1080/09638230410001669246 Sandanger, I., Moum, T., Ingebrigtsen, G., Srensen, T., Dalgard, O. S., & Bruusgaard, D. (1999). The meaning and significance of caseness: The Hopkins Symptom Checklist25 and the Composite International Diagnostic InterviewII. Social Psychiatry and Psychiatric Epidemiology, 34, 5359. doi:10.1007/s001270050112 Scogin, F., Bynum, J., & Stephens, G. (1990). Efficacy of selfadministered treatment programs: Meta-analytic review. Professional Psychology: Research and Practice, 21, 42 47. doi:10.1037/07357028.21.1.42 Spielberger, C. D., Gorsuch, R. L., Lushene, R. E., Vagg, P. R., & Jacobs, G. A. (1983). Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press. Webb, T. L., Ononaiye, M. S. P., Sheeran, P., Reidy, J. G., & Lavda, A. (2010). Using implementation intentions to overcome the effects of social anxiety on attention and appraisals of performance. Personality and Social Psychology Bulletin, 36, 612 627. doi:10.1177/ 0146167210367785 Zigmond, A. S., & Snaith, R. P. (1983). The Hospital Anxiety and Depression Scale. Acta Psychiatrica Scandinavica, 67, 361370. doi: 10.1111/j.1600-0447.1983.tb09716.x

Demyttenaere, K., Verhaeghen, A., Dantchev, N., Grassi, L., Montejo, A. L., Perahia, D. G. S., . . . Bauer, M. (2009). Caseness for depression and anxiety in a depressed outpatient population: Symptomatic outcome as a function of baseline diagnostic categories. The Primary Care Companion to the Journal of Clinical Psychiatry, 11, 307315. doi: 10.4088/PCC.08m00748blu Egloff, B., & Schmukle, S. C. (2002). Predictive validity of an Implicit Association Test for assessing anxiety. Journal of Personality and Social Psychology, 83, 14411455. doi:10.1037/0022-3514.83.6.1441 Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175191. Fisher, P. L., & Durham, R. C. (1999). Recovery rates in generalized anxiety disorder following psychological therapy: An analysis of clinically significant change in the STAIT across outcome studies since 1990. Psychological Medicine, 29, 14251434. doi:10.1017/ S0033291799001336 Glasgow, R. E., & Rosen, G. M. (1978). Behavioral bibliotherapy: A review of self-help behavior therapy manuals. Psychological Bulletin, 85, 123. doi:10.1037/0033-2909.85.1.1 Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54, 493503. doi:10.1037/0003066X.54.7.493 Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69 119. doi:10.1016/S00652601(06)38002-1 Gould, R. A., & Clum, G. A. (1993). A meta-analysis of self-help treatment approaches. Clinical Psychological Review, 13, 169 186. doi:10.1016/ 0272-7358(93)90039-O Haeffel, G. J. (2010). When self-help is no help: Traditional cognitive skills training does not prevent depressive symptoms in people who ruminate. Behaviour Research and Therapy, 48, 152157. doi:10.1016/ j.brat.2009.09.016 Kvaal, K., Ulstein, I., Nordhus, I. H., & Engedal, K. (2005). The Spielberger StateTrait Anxiety Inventory (STAI): The state scale in detecting mental disorders in geriatric patients. International Journal of Geriatric Psychiatry, 20, 629 634. doi:10.1002/gps.1330 Moher, D., Hopewell, S., Schulz, K. F., Montori, V., Gtzsche, P. C., Devereaux, P. J., . . . Altman, D. G., for the CONSORT Group. (2010). CONSORT 2010 explanation and elaboration: Updated guidelines for

Received February 15, 2010 Revision received July 6, 2010 Accepted August 6, 2010