538
JOURNAL OF STUDIES ON ALCOHOL / JULY 2006
Development and Initial Validation of a 12-Step Participation Expectancies Questionnaire CHRISTOPHER W. KAHLER, PH.D.,† JOHN F. KELLY, PH.D., DAVID R. STRONG, PH.D.,† GREGORY L. STUART, PH.D.,† AND RICHARD A. BROWN, PH.D.† Center for Alcohol and Addiction Studies, Brown University, Box G-BH, Providence, Rhode Island 02912
ABSTRACT. Objective: There are no available instruments that assess expectancies for participation in 12-step mutual-help groups despite the impact such expectancies may have on actual participation. The purpose of the present study was to develop a measure of attitudes and expectancies regarding 12-step participation, to conduct preliminary analyses on its psychometric properties, and to explore its concurrent and predictive validity. Method: Alcohol-dependent patients (N = 48) undergoing inpatient detoxification completed a questionnaire that included subscales assessing expected benefits of, concerns about, and barriers to 12-step participation. Participants also completed measures of 12-step group participation and drinking outcomes at 1, 3, and 6 months following discharge. Results: After examining the internal consistency of the items within each subscale and refining the questionnaire accord-
ingly, an exploratory factor analysis showed that the scales could be combined into a higher-order total score. This total score correlated significantly with prior 12-step experience and goals for attending future 12-step meetings. In addition, the Expectancies Total Score at baseline significantly predicted 12-step group participation during follow-up. Conclusions: The measure of attitudes and expectancies regarding 12step group participation demonstrated good internal consistency, concurrent validity, and predictive validity. The measure may have clinical utility in highlighting patients’ expectancies regarding 12-step participation, allowing treatment providers to explore with patients the benefits, concerns, and barriers to involvement that they have endorsed. (J. Stud. Alcohol 67: 538-542, 2006)
T
WELVE-STEP MUTUAL-HELP GROUPS, such as Alcoholics Anonymous (AA) and Narcotics Anonymous (NA), are attractive as adjuncts or alternatives to addictions treatment, because they can be attended free of charge, are easily accessible, and are widely available in most communities. Increasing evidence regarding the utility of AA-NA (e.g., Emrick et al., 1993; Kelly, 2003; Timko et al., 2000; Tonigan et al., 2003) has led to widespread referrals to such groups (Humphreys, 1997). However, many patients do not attend at all, and others discontinue attendance after some initial exposure (Kelly and Moos, 2003; Tonigan et al., 2003). Greater understanding regarding which patients participate in AA-NA, and why, would inform and help target efforts to facilitate 12-step involvement. Decision-making theory suggests that individuals engage in a conscious appraisal of the benefits and drawbacks associated with a given course of behavior before engaging in behavior change (Janis and Mann, 1977). Assessment of beliefs about potential positive and negative outcomes of AA-NA participation may enhance predictive precision regarding who participates in these fellowships and could provide valuable clinical information by identifying specific
barriers to participation. However, we are aware of no validated measures that assess beliefs or attitudes regarding 12-step participation. AA and NA offer a number of potential benefits that may influence decisions to participate. For example, these programs may offer abstinence-specific social support and may act to maintain motivation for recovery through the sharing of personal testimony (e.g., Kelly et al., 2000). Stories of recovery may be uplifting and inspiring for attendees, and participation in 12-step programs may enhance sober living skills and confidence in staying sober (e.g., Morgenstern et al., 1997). AA-NA also may provide a way of structuring sober time, especially during high-risk periods such as evenings and weekends. Assessing the degree to which individual patients perceive these potential benefits of 12-step participation as being likely to occur for themselves may be of predictive value and clinically may provide a means of highlighting and reinforcing the benefits of increased mutual-help involvement. Patients also may perceive 12-step programs in negative ways. For example, some may dislike the group format of AA-NA meetings or may perceive meetings as aversive, causing boredom, embarrassment, or hopelessness. For others, barriers may be more logistical such as difficulty obtaining transportation. Finally, the explicit spiritual emphasis of AA-NA may be a concern for some. Assessing these potential barriers to participation may help clinicians better understand and manage resistance to engaging in 12-step mutual-help groups.
Received: August 19, 2005. Revision: November 3, 2005. †Correspondence may be addressed to Christopher W. Kahler at the above address or via email at:
[email protected]. David R. Strong, Gregory L. Stuart, and Richard A. Brown are with the Brown Medical School and Butler Hospital, Providence, RI.
538
KAHLER ET AL. Study aims The purpose of the present study was to develop a measure of attitudes pertaining to 12-step participation, to conduct preliminary analyses on its psychometric properties, and to examine its concurrent and predictive validity. Method Participants Participants were 37 men and 11 women recruited from a private, nonprofit, inpatient detoxification program to participate in a randomized clinical trial comparing brief advice to attend AA-NA with a motivational enhancement intervention that focused on increasing involvement in 12step mutual-help groups (ME-12; Kahler et al., 2004). Alcohol-dependent patients ages 18-65 were included. Exclusion criteria were current suicidal or homicidal intent, organic impairment, psychotic symptoms or history of psychotic disorder, or use of methadone maintenance. Drug dependence was diagnosed in 22.9% of participants. The participants mean (SD) age was 43 (7.4) years, and 50% had some schooling beyond high school. The sample was 81.2% white, 8.3% black, 6.3% Hispanic/Latino, and 4.2% of other backgrounds. Participants drank on 65.2% (31.3%) of days in the 3 months prior to treatment, an average of 23.5 (17.4) drinks per drinking day. The mean on the Alcohol Dependence Scale (ADS; Skinner and Allen, 1982) was 23.0 (9.1). Procedure Participants were recruited into the study after they had spent at least 24 hours on the detoxification unit. Research assistants completed the baseline assessment after obtaining written informed consent. Participants were randomly assigned to either the brief advice or ME-12 protocol, which was conducted on the unit. Kahler et al. (2004) provide complete details of the recruitment procedures and the treatments received. Participants were re-interviewed at 1, 3, and 6 months; participation rates were 85.4%, 87.5%, and 89.5%, respectively. At baseline, participants reported the total number of 12-step meetings they had ever attended (sample median = 162 meetings). Lifetime AA-NA involvement was assessed with five dichotomous items (Tonigan et al., 1996) regarding whether they had ever considered themselves an AANA member (56.2% of the sample); been to 90 meetings in 90 days (39.6%); celebrated an AA-NA sobriety birthday (31.2%); had a sponsor (50.0%); or had been a sponsor (16.7%). Following our previous work (Kahler et al., 2004), attendance and involvement were standardized and summed to form an AA-NA experience variable. AA-NA attendance
539
goals were assessed with a single item ranging from every day or almost every day (n = 20; 42%), four or five times per week (n = 6; 12.5%), two or three times per week (n = 12; 25%), once per week (n = 3; 6.3%), or monthly or less (n = 7; 14.6%). The Timeline Followback interview (TLFB; Sobell and Sobell, 1996) was used to assess drinking frequency and quantity for the 90 days prior to study enrollment and at all follow-up interviews. During the follow-ups, AA-NA attendance data were collected using the TLFB, and percent days attending AA-NA was calculated for each month. Involvement in AA-NA during follow-up was assessed using the Recovery Interview (Morgenstern et al., 1996). Twelve-step participation expectancies Eleven scales were initially constructed using rational criteria to cover domains that might enhance or detract from motivation and willingness to be involved in AA or NA. Four items, all worded in the first person, were constructed for each scale. Five of these scales focused on expected benefits of 12-step involvement: social support (items focus on the positive social aspects of AA-NA), structured time (items focus on AA-NA meetings providing positive activity), increased motivation (items focus on AA-NA enhancing or maintaining motivation to stay sober), skill learning (items focus on learning more about how to stay sober), and positive emotional reactions (items focus on positive emotional experiences related to AA-NA attendance such as social acceptance and hope). Three scales assessed potential concerns about 12-step involvement: negative emotional reactions (items focus on negative emotional responses to attendance), social concerns (items focus on negative social aspects of AA-NA), and spirituality concerns (items address concerns about spiritual aspects of AA-NA as well as reverse-scored potential benefits of spirituality). Two scales focused on barriers to involvement: program barriers (items focus on central aspects of the program with which the participant is uncomfortable or disagrees) and attendance barriers (items focus on access and availability of meetings and competing commitments). Finally, the social influences scale assessed the extent to which patients had received encouragement to attend AA-NA and had heard positive things about the program from others. At baseline, the 44 initial items were presented in the same random order for all participants. All responses to the items were on a 6-point scale from 1 = strongly disagree to 6 = strongly agree. The top of the questionnaire read: “The following statements reflect opinions that some people have about getting involved in Alcoholics Anonymous (AA) or Narcotics Anonymous (NA). Please indicate how much you agree with each statement by circling ‘strongly disagree,’ ‘disagree,’ ‘tend to disagree,’ ‘tend to agree,’ ‘agree,’ or ‘strongly agree.’”
540
JOURNAL OF STUDIES ON ALCOHOL / JULY 2006
Analysis As a first step in the analysis, we calculated Cronbach’s alpha (i.e., internal consistency) for each subscale of the 12-step participation expectancies measure. We required α to exceed .60. If the presence of any one item on a scale was detracting from the internal consistency of that scale, the item was removed. If the removal of one item did not yield an acceptable alpha, we considered the scale unreli-
able and dropped it from further consideration. We then examined the correlations among the scales and conducted an exploratory principal axis common factor analysis to determine whether the scales could be combined meaningfully into a higher-order total score. To assess concurrent validity, we correlated the scales with AA-NA experience and goals for AA-NA attendance. For predictive validity, we tested the measure’s ability to predict both AA-NA attendance and involvement after discharge.
TABLE 1. Internal consistencies, sample means, and standard deviations for each subscale of the 12-Step Participation Expectancies Questionnaire and its associated items (n = 48) Subscales and associated items Social support (α = .83) Going to AA/NA meetings is a good way to meet “sober” friends. People at AA/NA could give me a lot of support. I don’t think people at AA/NA could be of any help to me.a Getting a sponsor through AA/NA would help me in my recovery. Structured time (α = .82) Going to AA/NA meetings can help me use some of my free time. Going to AA/NA meetings would keep me out of situations where I might be tempted to drink. I have much better things to do with my time than go to AA/NA meetings.a Going to AA/NA meetings would give me something to look forward to. Increased motivation (α = .82) Going to AA/NA meetings would help me remember why I want to stay sober. Going to AA/NA meetings would motivate me to stay sober. I would feel inspired to stay sober by seeing people at AA/NA who have been successful. Whether or not I go to AA/NA meetings will not affect how I feel about drinking.a Skill learning (α = .80) Through AA or NA, I could learn some useful skills to help me stay sober. I could learn a lot by working on the Twelve Steps of AA or NA. I don’t see how AA or NA could teach me anything new about recovery.a I could learn a lot by hearing about other people’s experiences in getting sober. Positive emotional reactions (α = .87) I think that AA/NA meetings could be uplifting. Being part of AA/NA would make me feel more hopeful. I would feel proud to be an AA or NA member. Negative emotional reactions (α = .66) I would get bored easily at AA/NA meetings. I would feel embarrassed going to an AA/NA meeting. Going to AA or NA would depress me. I would feel very nervous going to an AA/NA meeting. Social concerns (α = .64) I would not want to speak in front of a group at an AA/NA meeting. I would not want people at AA or NA to know about my personal problems. I think I would fit in well with most of the people who go to AA/NA.a I do not think I would like the people I meet at AA/NA. I don’t want people at AA or NA telling me how I should lead my life. I don’t want to hear other people talk about their problems at AA/NA meetings. Spirituality concerns (α = .66) I like that AA and NA are “spiritual” programs.a I feel very uncomfortable with the religious (or spiritual) aspects of AA/NA. In AA/NA meetings, there’s too much talk about spirituality and “Higher Powers” for me. I think that prayer or meditation could be very helpful in my recovery.a Attendance barriers (α = .60) I don’t have enough time to attend AA/NA meetings. There are plenty of AA/NA meetings in my area that I could go to.a It would be hard for me to get transportation to AA/NA meetings. Social influences (α = .70) Many people have encouraged me to go to AA or NA. I don’t know many people who were helped by AA or NA.a I know a number of people who really like the AA/NA program. Total expectancies score (the mean of subscales with reverse scoring for negative scales)
Mean (SD) 5.0 (0.7) 5.1 (0.9) 4.9 (0.8) 1.9 (0.9) 4.9 (1.1) 4.6 (1.0) 4.5 (1.3) 4.6 (1.3) 2.2 (1.2) 4.5 (1.3) 4.8 (0.8) 5.0 (0.8) 5.1 (.0.9) 5.0 (0.8) 2.7 (1.6) 4.8 (0.8) 4.9 (0.8) 5.0 (0.8) 2.2 (1.1) 4.6 (1.2) 4.7 (1.0) 4.7 (1.2) 4.8 (1.1) 4.7 (1.2) 2.8 (1.0) 3.0 (1.5) 2.0 (1.1) 2.9 (1.5) 3.3 (1.7) 2.8 (0.8) 2.4 (1.8) 2.2 (1.5) 3.7 (0.9) 2.7 (1.1) 3.4 (1.7) 3.2 (1.2) 2.5 (1.0) 4.5 (1.4) 2.4 (1.4) 3.4 (1.7) 5.2 (1.0) 2.3 (0.9) 2.2 (1.0) 5.1 (0.9) 2.8 (1.6) 4.7 (1.0) 4.8 (1.3) 2.2 (1.2) 4.7 (1.4) 4.6 (0.7)
Notes: Response options for the items range from 1 = strongly disagree to 6 = strongly agree. aItem is reverse scored when creating the subscale mean.
KAHLER ET AL.
(0.54), and the largest were for social support (0.90) and positive emotional reactions (0.89). Based on these results, we created a 12-step participation expectancies total score by reverse scoring the negative emotional reactions, social concerns, spirituality concerns, and attendance barriers scales and taking the mean of all 10 subscales (mean = 4.6 [0.7]). Correlations between this composite score and each subscale are presented in Table 2.
Results Internal consistency analyses revealed that two scales had alphas below the desired cut-off: social concerns (α = .53) and program barriers (α = .33). Within program barriers, two items focused on social aspects of the program. Given the scale’s low alpha, we combined these two socially focused items with the social concerns items. For this revised six-item social concerns subscale, α = .64 with no item detracting from internal consistency. Three scales had one item that detracted from the scale’s alpha but had adequate internal consistency when that one item was removed. The remaining six four-item scales all had adequate internal consistency, with α ranging from .66 to .83. Each retained item, along with its mean and standard deviation, are presented Table 1, along with the means, standard deviations, and alphas for each subscale. Participants, on average, tended to agree or agreed with positive aspects of AA-NA and tended to disagree or disagreed with negative aspects. The correlations among the subscales are presented in Table 2. The five scales tapping potential benefits of 12step participation were highly positively correlated with each other and with the social influences scale. The three scales measuring negative aspects of 12-step participation were negatively correlated with the benefit-related scales and positively correlated with each other and with the attendance barriers scale. All but four correlations among subscales were significant, with three of those four involving spirituality concerns. Results of a principal factors analysis of the 10 subscales indicated a very strong unidimensional structure accounting for 83.8% of the common variance with an eigenvalue of 5.25. The second factor had an eigenvalue of only 0.51. The smallest loadings on the first factor in absolute terms were for spirituality concerns (-0.45) and social influences
TABLE 2.
541
Validity analyses As shown in the bottom of Table 2, most subscales and the composite score showed modest correlations with AANA experience, with 7 of 11 being significant. All correlations but one with AA-NA attendance goals were larger than .35 and significant at p < .01. We used mixed model analyses to test the effect of 12step participation expectancies on AA-NA attendance over time, with a square-root transformation to correct positive skewness. Percent days attending AA-NA ranged from 41.2 (38.0) in Month 1 to 29.4 (34.7) in Month 6. We included in the analysis the 45 participants who provided at least 4 months of outcome data and examined only the 12-step expectancies total score given that testing the effects of each subscale would inflate risk of Type I error. The main effect of expectancies on AA-NA attendance was significant (B = 1.81, SE = 0.59, p = .004), and the Expectancies × Time interaction was nonsignificant (p > .80), indicating that expectancies predicted the overall level of attendance during follow-up but not changes in attendance over time. Expectancies also significantly predicted AA-NA involvement as measured by the Recovery Interview (B = 0.47, SE = 0.20, p = .02). The total expectancies score did not predict either percent days abstinent or drinks per drinking day during follow-up, all p > .70.
Correlations among 12-step participation expectancies subscales (n = 48)
Measure 1. Social support 2. Structured time 3. Increased motivation 4. Skill learning 5. Positive emotional reactions 6. Negative emotional reactions 7. Social concerns 8. Spirituality concerns 9. Attendance barriers 10. Social influences 11. 12-step expectancies total score Concurrent measures Prior AA-NA experience AA-NA attendance goals
1
2
3
4
5
6
7
8
9
.– .73 .79 .58 .86 -.53 -.64 -.23 -.48 .57 .87
.– .74 .58 .75 -.39 -.68 -.36 -.41 .34 .82
.– .67 .74 -.46 -.61 -.51 -.41 .45 .87
.– .51 -.35 -.60 -.36 -.31 .35 .73
.– -.56 -.73 -.32 -.39 .54 .89
.– .57 .26 .29 -.31 -.66
.– .32 .47 -.25 -.79
.– .33 -.17 -.53
.32 .59
.39 .58
.41 .53
.28 .27
.40 .74
-.17 -.37
-.24 -.46
-.35 -.36
Note: Correlations greater than .28 or less than -.28 are significant at p < .05.
10
11
.– -.44 -.52
.– .60
.–
-.45 -.39
.27 .45
.42 .65
542
JOURNAL OF STUDIES ON ALCOHOL / JULY 2006 Discussion
This study represents a first step in developing a measure of attitudes and expectancies regarding participation in mutual-help groups, the 12-Step Participation Expectancies Questionnaire (TSPEQ). Our findings suggest that the TSPEQ may be a useful assessment tool. Subscales for a variety of expected benefits and barriers as well as expected concerns and social incentives to 12-step participation were found to be internally consistent. Factor analyses indicated that subscales loaded on a higher order attitudinal disposition toward 12-step participation rather than to a larger number of dimensions, suggesting that a more brief assessment of attitudes toward AA-NA involvement may be feasible. Aggregating the subscales produced a total score that had good concurrent and predictive validity. However, factor analyses on the item level could not be conducted given the small sample. Such analyses may identify whether additional dimensions exist within the TSPEQ and whether a higher-order structure is supported in which items load on specific subscales, which in turn load on a single factor. The sample used in this initial study was limited to patients in alcohol detoxification who were participating in a clinical trial. Patients reported relatively high levels of intention to participate in AA-NA. A sample with greater diversity of intentions might increase variability in responses and reduce potential ceiling or floor effects. Although the TSPEQ is relatively comprehensive, the measure did not contain items addressing medication usage. There may be explicit or implicit opposition to the use of general psychotropic or anti-relapse/craving medications in 12-step fellowships (Tonigan and Kelly, 2004). In future work, we plan to develop an additional scale to assess whether individuals believe that AA-NA members will disapprove of their use of psychotropic medication. It also would be useful to examine patient characteristics that predict attitudes toward 12-step participation and changes in attitudes toward 12-step participation over time. The TSPEQ developed in this study was designed to survey common attitudes that patients with alcohol dependence may have about participating in 12-step mutual-help groups. In addition to providing a global index of patients’ attitudes toward such groups, the TSPEQ can provide potentially useful information regarding specific facilitators and barriers to AA-NA participation that can be discussed when patients are considering 12-step mutual-help group involvement. Better understanding of specific factors affecting attitudes toward 12-step participation ultimately may help guide clinical practice. At a minimum, the assessment
of expectancies for 12-step participation may provide a relatively robust predictor of the level of future participation. Future research that builds upon this initial work can help to refine further a psychometrically sound assessment of attitudes regarding AA-NA involvement. References EMRICK, C.D., TONIGAN, J.S., MONTGOMERY, H., AND LITTLE, L. Alcoholics Anonymous: What is currently known? In: MCCRADY, B.S. AND MILLER, W.R. (Eds.) Research on Alcoholics Anonymous: Opportunities and Alternatives, Piscataway, NJ: Rutgers Center of Alcohol Studies, 1993, pp. 41-76. HUMPHREYS, K. Clinicians’ referral and matching of substance abuse patients to self-help groups after treatment. Psychiat. Serv. 48: 14451449, 1997. JANIS, I.L. AND MANN, L. Decision Making: A Psychological Analysis of Conflict, Choice, and Commitment, New York: Free Press, 1977. KAHLER, C.W., READ, J.P., STUART, G.L., RAMSEY, S.E., MCCRADY, B.S., AND BROWN, R.A. Motivational enhancement for 12-step involvement among patients undergoing alcohol detoxification. J. Cons. Clin. Psychol. 72: 736-741, 2004. KELLY, J.F. Self-help for substance-use disorders: History, effectiveness, knowledge gaps and research opportunities. Clin. Psychol. Rev. 23: 639-663, 2003 KELLY, J.F. AND MOOS, R. Dropout from 12-step self-help groups: Prevalence, predictors and counteracting treatment influences. J. Subst. Abuse Treat. 24: 241-250, 2003. KELLY, J.F., MYERS, M.G., AND BROWN, S.A. A multivariate process model of adolescent 12-step attendance and substance use outcome following inpatient treatment. Psychol. Addict. Behav. 14: 376-389, 2000. MORGENSTERN, J., KAHLER, C.W., FREY, R.M., AND LABOUVIE, E. Modeling therapeutic response to 12-step treatment: Optimal responders, nonresponders and partial responders. J. Subst. Abuse 8: 45-59, 1996. MORGENSTERN, J., LABOUVIE, E., MCCRADY, B.S., KAHLER, C.W., AND FREY, R.M. Affiliation with Alcoholics Anonymous after treatment: A study of its therapeutic effects and mechanisms of action. J. Cons. Clin. Psychol. 65: 768-777, 1997. SKINNER, H.A. AND ALLEN, B.A. Alcohol dependence syndrome: Measurement and validation. J. Abnorm. Psychol. 91: 199-209, 1982. SOBELL, L.C. AND SOBELL, M.B. Timeline FollowBack: A Calendar Method for Assessing Alcohol and Drug Use: User’s Guide, Toronto, Canada: Addiction Research Foundation, 1996. TIMKO, C., MOOS, R.H., FINNEY, J.W., AND LESAR, M.D. Long-term outcomes of alcohol use disorders: Comparing untreated individuals with those in Alcoholics Anonymous and formal treatment. J. Stud. Alcohol 61: 529-540, 2000. TONIGAN, J.S., CONNORS, G.J., AND MILLER, W.R. Alcoholics Anonymous Involvement (AAI) scale: Reliability and norms. Psychol. Addict. Behav. 10: 75-80, 1996. TONIGAN, J.S., CONNORS, G.J., AND MILLER, W.R. Participation and involvement in Alcoholics Anonymous. In: BABOR, T.F. AND DEL BOCA, F.K. (Eds.) Treatment Matching in Alcoholism, New York: Cambridge Univ. Press, 2003, pp. 184-204. TONIGAN, J.S. AND KELLY, J.F. Beliefs about AA and the use of medications: A comparison of three groups of AA-exposed alcohol dependent persons. Alcsm Treat. Q. 22 (2): 67-78, 2004.
