• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Consult Clin Psychol. Author manuscript; available in PMC Jul 1, 2009.
Published in final edited form as:
PMCID: PMC2704891
NIHMSID: NIHMS110601

Dismantling Motivational Interviewing and Feedback for College Drinkers: A Randomized Clinical Trial

Abstract

Motivational Interviewing (MI) is a counseling style that has been shown to reduce heavy drinking among college students. To date, all studies of MI among college students have used a format that includes a feedback profile delivered in an MI style. This study was a dismantling trial of MI and feedback among heavy drinking college students. After an initial screen, 279 heavy drinking students were randomized to: 1) Web feedback only, 2) a single MI session without feedback, 3) a single MI session with feedback, or 4) assessment only. At 6 months, MI with feedback significantly reduced drinking, as compared with assessment only (effect size = .54), MI without feedback (effect size = .63) and feedback alone (effect size = .48). Neither MI alone nor feedback alone differed from assessment only. Neither sex, race/ethnicity, nor baseline severity of drinking moderated the effect of the intervention. Norm perceptions mediated the effect of the intervention on drinking. MI with feedback appears to be a robust intervention for reducing drinking, and may be mediated by changes in normative perceptions.

Keywords: motivational interviewing, feedback, college student, alcohol

Because of its widespread nature and serious consequences, reducing drinking among college students has become a major health objective in the U.S. (U.S. Department of Health and Human Services, 2000). In national surveys, approximately 80–85% of U.S. college students report consuming alcohol during the previous year (Johnston, O’Malley, Bachman, & Schulenberg, 2007; O'Malley & Johnston, 2002), and 40–45% report a heavy drinking episode during the two-week period prior to completing the survey (Johnston et al., 2007; O'Malley & Johnston, 2002; Wechsler, Lee, Kuo et al., 2002). Students who engage in heavy drinking are more likely than other students to experience a variety of problems. These problems include driving while under the influence, trouble with campus police, lower academic performance, engaging in unprotected or unplanned sexual activity, sexual assault, injury, and even death (Engs, Diebold, & Hanson, 1996; Hingson, Heeren, Winter, & Wechsler, 2005; Hingson, Heeren, Zakocs, Kopstein, & Wechsler, 2002; Wechsler, Davenport, Dowdall, Moeykens, & Castillo, 1994). In addition, the consequences of heavy drinking spread beyond the individual student. Non-heavy drinkers living on campuses with high rates of heavy drinking are at greater risk of experiencing physical and sexual assault, property damage, and disruption of sleep or study (Wechsler, Moeykens, Davenport, Castillo, & Hansen, 1995), while residents of the surrounding neighborhoods report more negative secondhand effects such as noise, litter, vandalism, and public intoxication (Wechsler, Lee, Hall, Wagenaar, & Lee, 2002).

Motivational interviewing with feedback (MIF) is one individual-level intervention that has garnered substantial support for reducing drinking among college students (Carey, Scott-Sheldon, Carey, & DeMartini, 2007; Larimer & Cronce, 2007; National Institute on Alcohol Abuse and Alcoholism, 2002). MIF can be conceptualized as two intertwined components: (1) Motivational interviewing (MI) and (2) personalized drinking feedback. MI is a directive, client-centered counseling style for eliciting behavior change by helping clients to explore and resolve ambivalence (Miller & Rollnick, 2002). The feedback component includes the presentation of information on personal drinking patterns, comparisons of the student’s drinking patterns to U.S. adult and college drinking norms, risk factors for heavy drinking, and negative consequences experienced as a result of heavy drinking.

There is also a small body of literature that suggests feedback alone may reduce drinking among college students (Agostinelli, Brown, & Miller, 1995; S. E. Collins, Carey, & Sliwinski, 2002; Larimer et al., 2007; Neighbors, Larimer, & Lewis, 2004; Walters, 2000; Walters, Bennett, & Miller, 2000; Walters, Vader, & Harris, 2007). In fact, two recent reviews have concluded that stand-alone feedback (e.g., web or mailed) may be as effective as in-person feedback among heavy drinking college students, at least in the short-term (Walters & Neighbors, 2005; White, 2006). The efficacy of stand-alone feedback interventions raises a basic question about the incremental value of the MI component in MIF interventions. Specifically, does MIF have an effect over feedback alone, delivered without personal contact? The question is of practical importance, because of the expense involved in training and supervising providers, concerns about student interest in seeking in-person help for drinking, and the time involved in administering in-person interventions (Larimer & Cronce, 2007; Walters & Neighbors, 2005).

The empirical literature comparing MIF and feedback-only interventions is sparse. One study found equivalent reductions in drinking after receiving feedback with or without an individual motivational interview (Murphy et al., 2004), whereas another study found a long term advantage for MIF compared to written feedback alone (White, Mun, Pugh, & Morgan, 2007). In addition, since MI interventions uniformly contain feedback, it is unclear whether MI alone (without the presence of a feedback profile) might be sufficient to reduce drinking. One small study (Juarez, Walters, Daugherty, & Radi, 2006) found that females reduced their drinking 8 weeks after receiving feedback, whether it was delivered by mail or via an in-person MI session, but there was no effect of an MI session that did not include feedback.

The present study was designed to dismantle MI and feedback, allowing us to test three questions simultaneously. First, we were interested in which intervention format(s) would reduce drinking over assessment only. This basic question adds to our knowledge of effective strategies for reducing college drinking. Second, we were interested in whether an in-person intervention (MI with feedback) would have an effect over feedback that was delivered without human contact. As previously mentioned, this question is important because of the additional cost involved with an in-person intervention. Third, we were interested in whether the inclusion of feedback would improve the effectiveness of MI. This question is important because it may tell us something about the mechanism of change in MI and feedback interventions, which to date, few studies have addressed. An alternative way of thinking about the second and third questions is to ask whether there is a synergistic effect of MI and feedback. In other words, is the combined effect of MI and FB greater than the effects of MI alone and of FB alone?

Two possible mechanisms of effect were examined: changes in perceived drinking norms and an increase in the use of protective drinking strategies. Of the five studies that have reported mediation analyses (one MIF study and four stand-alone feedback studies), three found support for changing normative perceptions (Borsari & Carey, 2000; Neighbors, et al., 2004; Walters, et al., 2007) and one found support for protective behaviors (Larimer et al., 2007) as a mediator of the intervention effect. Finally, consistent with previous work (Lewis, Neighbors, Oster-Aaland, Kirkeby, & Larimer, 2007; Neighbors & Larimer, 2002; Walters et al., 2007), we examined sex, race/ethnicity, drinking severity, and readiness to change as potential moderators. Moderator effects have been inconsistent in previous research, and thus no hypotheses were offered.

METHODS

Participants and Recruitment

Participants were recruited from a medium sized private university in the southern United States during the Fall of 2006 and Spring of 2007. Participants were at least 18 years old and reported at least one heavy-drinking episode (i.e., 5 or more drinks for men, 4 or more drinks for women, in a single episode) in the past 2 weeks. Participants were recruited using a variety of methods, including invitation emails to undergraduate psychology classes, brief presentations in undergraduate psychology and health courses, and flyers posted on campus. Approximately 1500 students were informed about the opportunity, leaving approximately 675 potentially eligible to participate in the study (assuming a 45% binge drinking rate). A power analysis indicated that 55 participants per condition would be sufficient to detect an effect size of .50 (Raudenbush & Liu, 2000, 2001). For their participation, students could receive $20 or psychology course extra credit at each assessment and for attending the in-person session (if so assigned). The project was approved by the Institutional Review Boards of both the University of Texas Health Science Center at Houston and the university from which the study participants were recruited.

A total of 279 students qualified for the study and agreed to participate. (An additional group of 75 students was recruited as part of a separate study question not addressed in this paper.) The sample was 64.2% female, 84.6% White, with a mean age of 19.8 years. Freshman made up 41.2% of the sample, sophomores were 21.1%, juniors were 21.9%, and seniors were 15.8%. Most students reported living in dorms/residence halls on campus (45.9%) or in off campus housing (37.6%).

Randomization, stratified by sex and heavy drinking frequency (i.e., one heavy episode in the past 2 weeks vs. more than one heavy episode), was completed automatically after the students entered their screening data. Participants then received an email directing them to the online consent and baseline assessment battery. The four study conditions included: 1) A personalized feedback report displayed on the screen (FBO; n=67); 2) A single session of MI without a personalized feedback report (MIO; n=70); 3) A single session of MI with a personalized feedback report (MIF; n=73); or 4) Assessment only (AO; n=69). Participants and counselors were not blinded to the group assignment. Figure 1 shows the flow of participants through the project.

Figure 1
Flow of Participants.

Measures

Measures were completed online at a baseline assessment, as well as at 3- and 6-month follow-up assessments. Outcome measures included alcohol consumption and alcohol-related problems; potential mediators included normative perceptions and protective behaviors; potential moderators included readiness to change, drinking severity, and demographic variables.

Alcohol consumption was assessed using a 7-day drinking calendar modified from the Daily Drinking Questionnaire (R. L. Collins, Parks, & Marlatt, 1985). Participants were asked to think about a typical week in the past month, and for each day, to estimate how many drinks they typically consumed on that day. To calculate peak blood alcohol concentration (BAC), participants also reported the number of standard drinks consumed and the duration of their heaviest drinking episode in the past month. We used this information, along with gender and weight, to calculate an estimated peak BAC.

Alcohol-related problems in the past 3 months were measured by the 23-item Rutgers Alcohol Problem Index (RAPI; White & Labouvie, 1989). The RAPI has been shown to have good reliability among college students (Borsari & Carey, 2000; Neal & Carey, 2004). Reliability in the present study was α=.87 for the total scale.

Normative drinking perceptions were measured by asking students to estimate the percent of U.S. college students of their gender who drank more than they did. Norm discrepancy was determined by subtracting participants’ estimates of what percent of students drank more than them, from the percent of students who actually drank more than them (based on national surveys).

Protective behaviors in the past 3 months were assessed by the 15-item Protective Behavioral Strategies Survey (PBSS; Martens et al., 2005). The PBSS asks respondents to identify different protective behaviors they might have used in the last 3 months while they were drinking. Reliability in the present study was α=.86 for the total scale.

Readiness to change was measured using the 12-item Readiness to Change Questionnaire (RTCQ; Rollnick, Heather, Gold, & Hall, 1992). The RTCQ contains three 4-item scales that assess levels of precontemplation, contemplation and action. The RTCQ has demonstrated adequate reliability and validity (Carey, Purnine, Maisto, & Carey, 1999; Heather, Rollnick, & Bell, 1993), though some have presented important limitations of existing measures of readiness to change among college students (Carey & Hester, 2005). Reliability in the present study was α=.64, .79 and .80 for the precontemplation, contemplation, and action subscales, respectively, and α=.86 for the total scale.

Drinking severity was measured using the Alcohol Use Disorders Identification Test (AUDIT; Saunders, Aasland, Babor, Delafuente, & Grant, 1993), which has shown adequate reliability and validity among college drinkers (Fleming, Barry, & Macdonald, 1991; Kokotailo et al., 2004; O'Hare & Sherrer, 1999).

Demographics included sex, age, year in school, race/ethnicity, athletic participation, fraternity/sorority membership, dating status (i.e., single, committed relationship, engaged/married), and residential status (i.e., on campus, off campus with friends, off campus with parents).

Intervention Procedures

Feedback

The personalized feedback was modified from the electronic-Check-Up to Go (e-CHUG; http://www.e-chug.com), a commercially-available feedback program. Using the information from a participant’s assessment, the feedback included: (1) a quantity/frequency summary of drinking behavior (e.g., standard drinks consumed in the last 30-days, estimated peak BAC, caloric intake), (2) comparison to U.S. adult and campus norms, (3) level of risk (e.g., AUDIT score, tolerance, estimated genetic risk), (4) estimated dollar amount and percent of income spent on alcohol, and (5) local referral resources. For those students in the FBO condition, the feedback form was displayed immediately on the computer screen after the participant completed the baseline assessment. Those in the MIF condition received their feedback profile during the MI session.

In-person sessions

The in-person sessions were delivered by two Ph.D. level counselors and five clinical psychology doctoral students. Each counselor completed 40 hours of MI training (including lecture, role play and practice) and submitted four practice tapes prior to seeing participants. To insure fidelity, counselors completed a checklist for each session and all sessions were videotaped for weekly supervision. Providers delivered both kinds of sessions (i.e., MIO and MIF).

The MIO sessions followed the stylistic elements as discussed in Miller and Rollnick (2002). Using the format outlined in Walters and Baer (2006), sessions included the following elements: (1) Orienting the participant to the session and the limits of confidentiality; (2) Exploring the participant’s drinking, including peak episodes and related problems; (3) Discussing ambivalence around drinking; (4) Utilizing the “Readiness Rulers” to elicit importance and confidence language; (5) Discussing change in the hypothetical or concrete; and if appropriate (6) Developing a plan for change. The counselor also provided the participant with a list of campus and community resources related to alcohol.

The MIF sessions were designed to be identical to the MIO sessions, with the exception of the additional personalized feedback profile. The feedback form used in the MIF sessions was identical to that used in the FBO condition. Participants in the MIF group were given a copy of their personal feedback report and a list of community resources related to alcohol.

To insure fidelity, a random subset of videotaped sessions (30 MIF and 30 MIO) was coded by two independent coders using the Motivational Interviewing Treatment Integrity Code 3.0 (MITI; Moyers, Martin, Manuel, & Miller, 2003). Sixteen tapes were double coded and intraclass correlation coefficients (ICC) were calculated to measure inter-rater reliability. The ICCs for the MITI counselor global scores were all in the “fair” category (evocation=0.48, collaboration=0.47, autonomy/support=0.45, direction=0.45, empathy=0.59, and global MI spirit=0.58; Cicchetti, 1994). The mean MITI counselor global scores for evocation, collaboration, autonomy/support, direction, empathy, and global MI spirit were at or above a beginning proficiency level. As a conservative test, separate t-tests were computed on each of the 5 MITI counselor global scores. Results indicated no statistically significant differences between the two conditions on any of these scores (all p>0.05).

Ninety percent of participants completed the 3-month follow-up and 86% completed the 6-month follow-up. Participants were not more likely to drop out of the study at either follow-up based on baseline characteristics or study condition. There were no adverse events or side effects reported in any of the intervention groups.

Preliminary Analyses and Data Analysis Strategy

The correlations at baseline between the three outcome measures were: drinks per week and peak BAC, .597; drinks per week and alcohol-related problems, .484; peak BAC and alcohol-related problems, .433. Because of concern about overlapping content between the three outcome measures, we used principal component analysis to create a composite drinking variable. At baseline, a first component accounted for 67% of the variance, reflecting roughly equal contributions from each of the three standardized drinking measures. This first component was used as the composite measure of drinking outcome. To preserve comparability of the composite measure over time, the coefficients from the baseline analysis were applied to the unstandardized measures at each follow-up assessment. Therefore, the composite measure at baseline is a standard score (with mean 0 and standard deviation 1) and is expressed in the same units at the 3-month and 6-month follow-ups.

The effects of the interventions on the composite drinking variable were analyzed using a mixed linear model, a type of multilevel model that allows for the use of partial data from subjects who did not participate in both follow-ups. Subject was a random effect, the intervention conditions were fixed between-subject effects, and time (baseline and the two follow-ups) was the within-subject effect. To address the questions described in the introduction, conditions were compared on the composite drinking measure. When differences between conditions emerged on this measure, comparisons were then made on each of the three specific outcome measures (drinks per week, peak BAC, and alcohol-related problems). All tests were conducted using a p value of .05.

Preliminary analyses showed significant quadratic components of some time paths; therefore time was analyzed as a categorical variable, represented by dummy variables for the two follow-up times. Effects of intervention on the outcomes, adjusted for their baseline levels, appeared as time by intervention interactions. To test for the synergistic effect of MI and FB, we parameterized the interventions via factorial structure, crossing MI and FB and testing for an interaction effect. In the mixed linear model, this interaction appeared as a MI*FB*Time interaction.

We assessed norm discrepancies and protective behaviors as potential mediators, following the procedures described by Baron and Kenney (1986): 1) we used regression to determine whether the intervention significantly affected the composite drinking measure; 2) we regressed the possible mediators on intervention to determine whether the intervention affected either or both of them; 3) in regressions of the composite drinking measure on the intervention and one potential mediator, we tested whether the latter affected the outcome, while holding intervention constant; and 4) in the same regressions as in step 3, we assessed change in the strength of the effect of intervention on the outcome compared to the regression in step 1; that is, whether the mediator accounted for some or all of the effect of the intervention.

Sex, race/ethnicity, readiness to change, and drinking severity were tested as potential moderators. Each hypothesized moderator was tested by including a moderator by intervention condition interaction in the previous analyses.

RESULTS

At baseline, the 279 participants reported consuming an average of 15.5 standard drinks per week, with a mean peak BAC of .165 mg% in the past month. There were no significant differences between the groups in terms of sex, race/ethnicity, drinks per week, peak BAC, negative consequences, the composite drinking measure, or protective behaviors (bivariate associations of each variable with intervention group, Pearson chi-square for sex and race/ethnicity, one-way ANOVA for other measures, all p’s >.10). The MIO sessions had a mean length of 40:18 minutes and the MIF sessions had a mean length of 50:09 minutes. Counselors varied significantly in their mean session length (F(1,6)=3.087, p=.008), but neither counselor assignment nor session length predicted changes in client drinking when holding intervention group constant. Table 1 shows the mean drinks per week, peak BAC, and alcohol-related problems at baseline, 3 and 6 months.

Table 1
Mean drinks per week, peak BAC, and alcohol-related problems by condition at baseline, 3 and 6 months.

Efficacy of the Intervention Conditions

All analyses were conducted using an intent-to-treat model. Thus, participants who completed the baseline assessment and were assigned to one of the in-person interventions (MIO or MIF), but who did not attend the in-person session, were included in the analyses.

Comparisons of the three interventions with assessment only (the first basic question) are shown in Table 2. Relative to AO, the MIF intervention significantly affected changes in the composite measure from baseline to 6 months (t(275)=−3.28, p=0.001, effect size=0.54). This difference was also significant for each of the three outcomes. In terms of the magnitude of differences at the 6-month follow-up, participants in the MIF group reported consuming 5.26 drinks per week less than participants in the AO group (t(275)=−2.63, p=0.01, effect size of 0.41), when adjusting for baseline levels. Participants in the MIF group reported a peak BAC that was 0.039 mg% less than participants in the AO group (t(275)=−2.32, p=0.02, effect size = 0.37), when adjusting for baseline levels. Finally, participants in the MIF group reported a mean alcohol problem score that was 2.32 points lower than participants in the AO group (t(275)= −2.35, p=0.02 effect size = 0.43), when adjusting for baseline levels. Neither MIO nor FBO differed from AO.

Table 2
Effects of FBO, MIO, and MIF vs. AO on drinking behavior.

Comparisons of MIF with FBO (the second basic question) are shown in Table 3. MIF lowered the composite measure significantly more than FBO at both 3 months and 6 months, relative to baseline. Follow-up tests indicated a similar finding for drinks per week at 3 months and 6 months, and for peak BAC at 6 months.

Table 3
Effects of MIO and FBO vs. MIF, on drinking behavior.

The final question concerned the additional contribution of feedback to a motivational interviewing intervention. MIF was significantly more efficacious than MIO at both 3 months and 6 months for the composite measure. Follow-up tests indicated a similar pattern of results for drinks per week and peak BAC at 3 and 6 months, and for alcohol-related problems at 6 months (see Table 3).

An additional way to test for differences between groups is in terms of an interaction effect. Using a factorial structure involving main effects for FB and MI and their interaction, the interaction effect was significant for the composite measure at both 3 months (t(275)=−2.14, p=.03, effect size=−.52) and 6 months (t(275)=−2.07, p=.04, effect size=−.57). The interactions were also significant for drinks per week at both 3 months (t(275)=−3.09, p=.002, effect size=−.75) and 6 months (t(275)= −2.11, p=.04, effect size=−.60) and for peak BAC at 6 months (t(275)=−2.11, p=.04, effect size=−.50). This suggests a synergistic effect of MIF, where its effect was greater than the sum of the effects of feedback alone and of MI alone.

The primary analyses were repeated, adjusting for the following baseline covariates: sex, frequency of binge drinking in the 2 weeks prior to baseline, race/ethnicity (nonwhite vs. white), weight, year in school, housing (dormitory, fraternity/sorority, other), participation in sports, smoking status, relationship status and drinking severity. Adding the covariates did not change the pattern of results.

Mediators and Moderators of Intervention Efficacy

Following the procedures described by Baron and Kenny (1986), 1) the MIF intervention significantly affected the composite drinking measure; 2) the intervention reduced norm discrepancies at 6 months by 17%, with the MIF group becoming more accurate in their norm estimates (i.e., smaller discrepancies) (p<0.001); 3) lower norm discrepancies were associated with lower composite score (p=0.006); and 4) adjusting for norm discrepancies reduced the magnitude of the intervention effect by 45% on the composite measure. The pattern of results was significant for peak BAC. Adjusting for norm discrepancies reduced the magnitude of the intervention on peak BAC by 79%. Protective behavior, which was only weakly related to the intervention and to the outcomes, did not mediate the intervention effect.

Sex, race/ethnicity, readiness to change, and drinking severity were tested as potential moderators of the intervention effect on drinking at 6 months by adding interaction effects to the model. Results showed no moderation effects for sex, race/ethnicity, readiness to change, or AUDIT score on any of the drinking variables (all p’s >.05).

DISCUSSION

MI with feedback has been found to reduce drinking in college student populations. However, the separate and collective effects of MI and feedback have not been previously examined in large clinical trials. This study was designed to answer three questions: whether any intervention would show a reduction over assessment only, whether an in-person feedback intervention would reduce drinking over a feedback intervention delivered without human contact, and whether MI’s effectiveness would be enhanced through the inclusion of a feedback profile. With regard to the first question, we found that at 6 months MIF reduced drinking over AO. The other intervention formats (MIO and FBO) did not show an effect over AO. With regard to our second question, we found that at 6 months MIF reduced drinking over FBO. With regard to our third question, we found that at 6 months MIF reduced drinking over MIO. The effects were consistent for the composite drinking measure and drinks per week, and mostly consistent for peak BAC. The results for alcohol-related problems were weaker and more inconsistent.

Norm perceptions mediated the effect of the intervention, with participants in the MIF condition becoming more accurate in their normative drinking estimates, and changes in norm perceptions being linked to changes in drinking behavior. In contrast to a previous study (Larimer et al., 2007), we did not find that protective behaviors mediated the intervention effect. However, our interventions did not specifically target protective behaviors, which may limit the usefulness of this finding. In addition, we did not find that sex, race/ethnicity, readiness to change, or baseline drinking severity moderated the effect of the intervention. The present study was strengthened by the use of a randomized prospective design, well validated measures, a rigorous training and supervision sequence, and two commercially-available intervention formats. Our 6 month retention rate (86%) was excellent relative to similar studies (Lewis et al., 2007; Neighbors & Larimer, 2002; Walters et al., 2007; White et al., 2007).

Our findings must also be viewed in light of several study limitations. First, our sample contained a disproportionate number of female and White students, although our study population did reflect roughly the composition of the institution. This may limit the generalizability of our findings to other kinds of students or other institutions. Second, we relied on self-report outcome data, though we used procedures to minimize the chance of biased reporting, including using well-validated measures and offering assurances of confidentiality. Research suggests that self-reports are generally more valid under these conditions (Williams, Aitken, & Malin, 1985; Wolber, Carne, & Alexander, 1990). Third, our present results cannot account for the effect of the assessment on drinking. A number of researchers have speculated that assessment interviews can contribute to clinical outcome (Clifford & Maisto, 2000; Kypri, Langley, Saunders, & Cashell-Smith, 2007; Ogborne & Annis, 1988; Project MATCH Research Group, 1998; L. C. Sobell & Sobell, 1981; M. B. Sobell, Brochu, Sobell, Roy, & Stephens, 1987), though only a handful of studies have experimentally tested this hypothesis. There were also limitations related to our choice of intervention formats. For instance, the MIO and MIF conditions varied in contact time because of the addition of the feedback component. Although this project approximated a dismantling study, for practical reasons we did not employ a true dismantling design. For instance, the feedback format varied (i.e., online vs. face-to-face) and we did not include a condition that received face-to-face feedback without the MI segment of the session. Our selection of experimental conditions was based on feasibility concerns, as well as how MI and feedback interventions are typically used in practice. Because our FBO condition was delivered on the Internet, we were also not able to verify whether students in fact read their feedback profiles.

Despite these limitations, our results have three main implications. First, our finding about the effectiveness of the typical MI format (MI with feedback) supports the existing literature on the effectiveness of this intervention. Our study also suggests that the inclusion of both an in-person MI session and a feedback profile is more potent than either feedback alone or MI alone in this population. In fact, our findings suggest that MI and feedback have a synergistic effect when used together. The effect size of the MIF/AO contrast was similar to those reported in other brief interventions for college drinkers (see Carey et al., 2007 for a review). In practical terms, the 50-minute MIF intervention reduced 6 month drinking and drinking-related problems by nearly 30%, when compared to assessment only. The generalizability of our findings is strengthened by the use of a commercially available feedback (http://www.e-chug.com) and published intervention protocol (Walters & Baer, 2006).

Second, our findings on the lack of effectiveness of stand-alone feedback help to qualify previous research. Although several studies have found that mailed or web-based feedback can effectively reduce drinking, most have been limited to a relatively small sample and short follow-up (Walters & Neighbors, 2005). Studies that follow students out to 6 or 12 months have reported mixed results, with some showing an effect (Larimer et al., 2007; Neighbors & Larimer, 2002) and others not showing an effect (S. E. Collins et al., 2002). Studies that have used multiple assessment windows often find that the effect of feedback fades in the long-term. For instance, Collins, Carey & Sliwinski (2002) reported reductions in drinking at 6 weeks as a result of receiving mailed feedback, though the differences were no longer present at 6 months. Likewise, Walters, Vader and Harris (2007) found that web-feedback reduced drinking over control at 8 weeks, but the differences between groups were no longer evident at 16 weeks. Thus, it is possible that our first follow-up assessment (3 months) missed any short-term effect that the feedback might have had.

Finally, our mediator and moderator findings are mostly consistent with previous research. Change in perceived norms has been the most consistent mediator of MI and feedback interventions (Borsari & Carey, 2000; Neighbors et al., 2004; Walters & Neighbors, 2005). This is congruent with the body of research that suggests that college drinkers tend to overestimate how much others are drinking (Borsari & Carey, 2001) and misperceive the prevailing norms toward drinking and drunkenness (Perkins & Berkowitz, 1986). In this study, the average baseline student estimated that they were at the 57th percentile of U.S. college students, when they were actually at the 83rd percentile. Arnett (2000) described a period of “emerging adulthood” when young people strongly rely on the behavior of their peers to judge whether or not their own behavior is acceptable. Thus, it makes sense that changing normative perceptions might be a strong motivator to make changes in drinking. However, it is also important to note that both the FBO and MIF conditions contained identical feedback information. The fact that norm perceptions mediated the effects of MIF on drinking behavior, but FBO failed to have an effect on drinking behavior, may suggest that the MI portion of the interview helps to amplify the effect of the feedback, perhaps by eliciting verbal commitments to change based on the discrepancy of the normative information. Absent a provider, it might be possible to increase the salience of the normative information, for instance, by having students participate in an interactive feedback program, or actually calculate their own feedback.

Past studies have reported inconsistent moderator findings. With few exceptions, outcomes have been similar across sex, family history of alcohol abuse, fraternity/sorority membership, participation in sports, and motivation to change (S. E. Collins et al., 2002; Marlatt et al., 1998; Murphy et al., 2004; Walters et al., 2007). Larimer et al. (2007) found that neither sex nor drinking severity moderated the effect of a feedback intervention, though a more conservative analysis suggested that women benefited more from the intervention. There is also evidence that MI and feedback interventions may be more effective among heavier drinkers (Murphy et al., 2001) and those who drink for more social reasons (Neighbors et al., 2004). In both cases, it would seem to make sense that normative information might have a greater effect on those who are more interested in social comparison or those who are heavier drinkers (and hence have a more extreme profile). However, another study (Carey, Henson, Carey, & Maisto, 2007) found that strength of social comparison predicted changes even in the absence of an intervention. Contrary to expectations, students who used social cues to guide their behavior maintained or even increased their consumption, relative to other students.

It is curious that we found a stronger intervention effect for MIF at 6 months than at 3 months. It is possible that our findings were influenced by the semester variation in drinking that is typical of college drinkers (Del Boca, Darkes, Greenbaum, & Goldman, 2004). Since two-thirds of our sample was recruited in the Fall, the 3 month follow-up was more likely to occur in January or February (a lower, and potentially more homogenous drinking period), while the 6 month follow-up was more likely to occur during April or May (a higher, and potentially more heterogeneous drinking period that includes Spring Break). Similarly, although we did not find significant differences between groups at baseline, it is still possible that the MIF group, which scored generally higher on baseline drinking measures, had the most room to decrease.

Given the body of evidence in support of MI and other feedback-based interventions, future research might consider ways to best disseminate such findings. While the present study suggests one intervention format, there may be other models that are better suited to how interventions are conducted in practice. Most intervention studies have used a format similar to ours; however, a 45 minute counseling intervention may still be too awkward and lengthy for other contexts in which heavy drinking students are typically encountered (e.g., judicial settings, residence life, student health). In fact, there has been far less research on brief interventions delivered by non-specialists, even though the bulk of interactions around alcohol may happen outside of formal alcohol treatment contexts. Our study suggests that one direction for such efforts might be through the inclusion of feedback in brief interventions. The feedback portion of our MIF session typically lasted 10–15 minutes, but because of the presence of additional MI components, it is unclear specifically what the contribution of the feedback is to the overall session. One way the feedback contribution might be better clarified is through an analysis of in-session client language; specifically, the extent to which feedback changes the nature of client speech and how in-session speech is related to treatment outcome. In an analysis of session tapes from Project MATCH, Moyers et al. (2007) found that MI-consistent therapist behaviors were more likely to be followed by client talk in support of change, whereas MI-inconsistent therapist behaviors were more likely to be followed by client talk in support of the status quo. However, beyond this general support for MI practice, there is little information on the extent to which feedback might be able to elicit interest or readiness to change. Further research on therapeutic mechanisms of change may help to identify more efficient ways of conducting treatment interactions.

Acknowledgments

This project was supported by R01 AA016005-01 funded by the National Institute of Alcohol Abuse and Alcoholism.

Footnotes

Publisher's Disclaimer: The following manuscript is the final accepted manuscript. It has not been subjected to the final copyediting, fact-checking, and proofreading required for formal publication. It is not the definitive, publisher-authenticated version. The American Psychological Association and its Council of Editors disclaim any responsibility or liabilities for errors or omissions of this manuscript version, any version derived from this manuscript by NIH, or other third parties. The published version is available at http://www.apa.org/journals/ccp/.

Contributor Information

Scott T. Walters, University of Texas School of Public Health.

Amanda M. Vader, University of Texas School of Public Health.

T. Robert Harris, University of Texas School of Public Health.

Craig A. Field, University of Texas at Austin.

Ernest N. Jouriles, Southern Methodist University.

REFERENCES

  • Agostinelli G, Brown JM, Miller WR. Effects of normative feedback on consumption among heavy drinking college students. Journal of Drug Education. 1995;25:31–40. [PubMed]
  • Arnett JJ. Emerging adulthood. A theory of development from the late teens through the twenties. The American Psychologist. 2000;55:469–480. [PubMed]
  • Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology. 1986;51:1173–1182. [PubMed]
  • Borsari B, Carey KB. Effects of a brief motivational intervention with college student drinkers. Journal of Consulting and Clinical Psychology. 2000;68:728–733. [PubMed]
  • Borsari B, Carey KB. Peer influences on college drinking: A review of the research. Journal of Substance Abuse. 2001;13:391–424. [PubMed]
  • Carey KB, Henson JA, Carey MP, Maisto SA. Which heavy drinking college students benefit from a brief motivational intervention? Journal of Consulting and Clinical Psychology. 2007;75(4):663–669. [PMC free article] [PubMed]
  • Carey KB, Hester RK. Rethinking readiness-to-change for at-risk college drinkers. The Addictions Newsletter. 2005;12(2):6–8.
  • Carey KB, Purnine DM, Maisto SA, Carey MP. Assessing readiness to change substance abuse: A critical review of instruments. Clinical Psychology: Science and Practice. 1999;6:245–266.
  • Carey KB, Scott-Sheldon LA, Carey MP, DeMartini KS. Individual-level interventions to reduce college student drinking: A meta-analytic review. Addictive Behaviors. 2007;32:2469–2494. [PMC free article] [PubMed]
  • Cicchetti DV. Guidelines, criteria, and rules of thumb for evaluating normal and standardized assessment instruments in psychology. Psychological Assessment. 1994;6:284–290.
  • Clifford P, Maisto S. Subject reactivity effects and alcohol treatment outcome research. Journal of Studies on Alcohol. 2000;61:787–793. [PubMed]
  • Collins RL, Parks GA, Marlatt GA. Social determinants of alcohol consumption: The effects of social interaction and model status on the self-administration of alcohol. Journal of Consulting and Clinical Psychology. 1985;53:189–200. [PubMed]
  • Collins SE, Carey KB, Sliwinski MJ. Mailed personalized normative feedback as a brief intervention for at-risk college drinkers. Journal of Studies on Alcohol. 2002;63:559–567. [PubMed]
  • Del Boca FK, Darkes J, Greenbaum PE, Goldman MS. Up close and personal: Temporal variability in the drinking of individual college students during their first year. Journal of Consulting and Clinical Psychology. 2004;72(2):155–164. [PubMed]
  • Engs RC, Diebold BA, Hanson DJ. The drinking patterns and problems of a national sample of college students, 1994. Journal of Alcohol and Drug Education. 1996;41:13–33.
  • Fleming MF, Barry KL, Macdonald R. The Alcohol Use Disorders Identification Test (AUDIT) in a college sample. International Journal of the Addictions. 1991;26:1173–1185. [PubMed]
  • Heather N, Rollnick S, Bell A. Predictive validity of the Readiness to Change Questionnaire. Addiction. 1993;88:1667–1677. [PubMed]
  • Hingson R, Heeren T, Winter M, Wechsler H. Magnitude of alcohol-related mortality and morbidity among U.S. college students ages 18–24: Changes from 1998 to 2001. Annual Review of Public Health. 2005;26:259–279. [PubMed]
  • Hingson R, Heeren T, Zakocs RC, Kopstein A, Wechsler H. Magnitude of alcohol-related mortality and morbidity among U.S. college students ages 18–24. Journal of Studies on Alcohol. 2002;63:136–144. [PubMed]
  • Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE. Monitoring the Future national survey results on drug use, 1975–2006: Volume II, College students and adults ages 19–45 (NIH Publication No. 07-6206) Bethesda, MD: National Institute on Drug Abuse; 2007.
  • Juarez P, Walters ST, Daugherty M, Radi C. A randomized trial of motivational interviewing and feedback with heavy drinking college students. Journal of Drug Education. 2006;36:233–246. [PubMed]
  • Kokotailo PK, Egan J, Gangnon R, Brown D, Mundt M, Fleming M. Validity of the alcohol use disorders identification test in college students. Alcoholism, Clinical and Experimental Research. 2004;28:914–920. [PubMed]
  • Kypri K, Langley JD, Saunders JB, Cashell-Smith ML. Assessment may conceal therapeutic benefit: findings from a randomized controlled trial for hazardous drinking. Addiction. 2007;102:62–70. [PubMed]
  • Larimer ME, Cronce JM. Identification, prevention, and treatment revisited: individual-focused college drinking prevention strategies 1999–2006. Addictive Behaviors. 2007;32:2439–2468. [PubMed]
  • Larimer ME, Lee CM, Kilmer JR, Fabiano PM, Stark CB, Geisner IM, et al. Personalized mailed feedback for college drinking prevention: A randomized clinical trial. Journal of Consulting and Clinical Psychology. 2007;75:285–293. [PMC free article] [PubMed]
  • Lewis MA, Neighbors C, Oster-Aaland L, Kirkeby BS, Larimer ME. Indicated prevention for incoming freshmen: Personalized normative feedback and high-risk drinking. Addictive Behaviors. 2007;32:2495–2508. [PMC free article] [PubMed]
  • Marlatt GA, Baer JS, Kivlahan DR, Dimeff LA, Larimer ME, Quigley LA, et al. Screening and brief intervention for high-risk college student drinkers: Results from a 2-year follow-up assessment. Journal of Consulting and Clinical Psychology. 1998;66:604–615. [PubMed]
  • Martens MP, Ferrier AG, Sheehy MJ, Corbett K, Anderson DA, Simmons A. Development of the protective behavioral strategies survey. Journal of Studies on Alcohol. 2005;66:698–705. [PubMed]
  • Miller WR, Rollnick S. Motivational interviewing: Preparing people for change. 2nd ed. New York: Guilford Press; 2002.
  • Moyers TB, Martin T, Christopher PJ, Houck JM, Tonigan JS, Amrhein PC. Client language as a mediator of motivational interviewing efficacy: Where is the evidence? Alcoholism, Clinical and Experimental Research. 2007;31:40s–47s. [PubMed]
  • Moyers TB, Martin T, Manuel JK, Miller WR. The Motivational Interviewing Treatment Integrity (MITI) code. Unpublished manuscript. 2003
  • Murphy JG, Benson TA, Vuchinich RE, Deskins MM, Eakin D, Flood AM, et al. A comparison of personalized feedback for college student drinkers delivered with and without a motivational interview. Journal of Studies on Alcohol. 2004;65:200–203. [PubMed]
  • Murphy JG, Duchnick JJ, Vuchinich RE, Davison JW, Karg RS, Olson AM, et al. Relative efficacy of a brief motivational intervention for college student drinkers. Psychology of Addictive Behaviors. 2001;15:373–379. [PubMed]
  • National Institute on Alcohol Abuse and Alcoholism. A call to action: Changing the culture of drinking at US colleges. Bethesda, MD: US Department of Health and Human Services; 2002.
  • Neal DJ, Carey KB. Developing discrepancy within self-regulation theory: Use of personalized normative feedback and personal strivings with heavy-drinking college students. Addictive Behaviors. 2004;29:281–297. [PubMed]
  • Neighbors C, Larimer ME. Efficacy of computer delivered normative feedback in reducing alcohol consumption among high risk drinkers; Paper presented at the annual convention of the Association for the Advancement of Behavior Therapy; Reno, NV. 2002.
  • Neighbors C, Larimer ME, Lewis MA. Targeting misperceptions of descriptive drinking norms: Efficacy of a computer-delivered personalized normative feedback intervention. Journal of Consulting and Clinical Psychology. 2004;72:434–447. [PubMed]
  • O'Hare T, Sherrer MV. Validating the Alcohol Use Disorder Identification Test with college first-offenders. Journal of Substance Abuse Treatment. 1999;17:113–119. [PubMed]
  • O'Malley PM, Johnston LD. Epidemiology of alcohol and other drug use among American college students. Journal of Studies on Alcohol. Supplement. 2002;14:23–39. [PubMed]
  • Ogborne AC, Annis HM. The reactive effects of follow-up assessment procedures: An experimental study. Addictive Behaviors. 1988;13:123–129. [PubMed]
  • Perkins HW, Berkowitz AD. Perceiving the community norms of alcohol use among students: Some research implications for campus alcohol education programming. The International Journal of the Addictions. 1986;21:961–976. [PubMed]
  • Project MATCH Research Group. Matching alcoholism treatments to client heterogeneity: Project MATCH three-year drinking outcomes. Alcoholism, Clinical and Experimental Research. 1998;22:1300–1311. [PubMed]
  • Raudenbush S, Liu X-F. Statistical power and optimal design for multisite randomized trials. Psychological Methods. 2000;5:199–213. [PubMed]
  • Raudenbush S, Liu X-F. Effects of study duration, frequency of observation, and sample size on power in studies of group differences in polynomial change. Psychological Methods. 2001;6:387–401. [PubMed]
  • Rollnick S, Heather N, Gold R, Hall W. Development of a short 'readiness to change' questionnaire for use in brief, opportunistic interventions among excessive drinkers. British Journal of Addiction. 1992;87:743–754. [PubMed]
  • Saunders JB, Aasland OG, Babor TF, Delafuente JR, Grant M. Development of the Alcohol-use Disorders Identification Test (Audit): WHO collaborative project on early detection of persons with harmful alcohol consumption. Addiction. 1993;88(6):791–804. [PubMed]
  • Sobell LC, Sobell MB. Frequent follow-up as data gathering and continued care with alcoholics. International journal on addictions. 1981;16:1077–1086. [PubMed]
  • Sobell MB, Brochu S, Sobell LC, Roy J, Stephens JA. Alcohol treatment outcome evaluation methodology: State of the art 1980–1984. Addictive Behaviors. 1987;12:113–128. [PubMed]
  • U.S. Department of Health and Human Services. Healthy People 2010: Understanding and improving health. 2nd ed. Washington, DC: U.S. Government Printing Office; 2000.
  • Walters ST. In praise of feedback: An effective intervention for college students who are heavy drinkers. Journal of American College Health. 2000;48:235–238. [PubMed]
  • Walters ST, Baer JS. Talking with college students about alcohol: Motivational strategies for reducing abuse. New York: The Guilford Press; 2006.
  • Walters ST, Bennett ME, Miller JH. Reducing alcohol use in college students: A controlled trial of two brief interventions. Journal of Drug Education. 2000;30:361–372. [PubMed]
  • Walters ST, Neighbors C. Feedback interventions for college alcohol misuse: What, why and for whom? Addictive Behaviors. 2005;30:1168–1182. [PMC free article] [PubMed]
  • Walters ST, Vader AM, Harris TR. A controlled trial of web-based feedback for heavy drinking college students. Prevention Science. 2007;8:83–88. [PubMed]
  • Wechsler H, Davenport A, Dowdall G, Moeykens B, Castillo S. Health and behavioral consequences of binge drinking in college. A national survey of students at 140 campuses. JAMA. 1994;272:1672–1677. [PubMed]
  • Wechsler H, Lee JE, Hall J, Wagenaar AC, Lee H. Secondhand effects of student alcohol use reported by neighbors of colleges: The role of alcohol outlets. Social Science & Medicine. 2002;55:425–435. [PubMed]
  • Wechsler H, Lee JE, Kuo M, Seibring M, Nelson TF, Lee H. Trends in college binge drinking during a period of increased prevention efforts: Findings from four Harvard School of Public Health College Alcohol Study surveys:1993–2001. Journal of American College Health. 2002;50:203–217. [PubMed]
  • Wechsler H, Moeykens B, Davenport A, Castillo S, Hansen J. The adverse impact of heavy episodic drinkers on other college students. Journal of Studies on Alcohol. 1995;56:628–634. [PubMed]
  • White HR. Reduction of alcohol-related harm on United States college campuses: The use of personal feedback interventions. International Journal of Drug Policy. 2006;17:310–319.
  • White HR, Labouvie EW. Towards the assessment of adolescent problem drinking. Journal of Studies on Alcohol. 1989;50:30–37. [PubMed]
  • White HR, Mun EY, Pugh L, Morgan TJ. Long-term effects of brief substance use interventions for mandated college students: Sleeper effects of an in-person personal feedback intervention. Alcoholism, Clinical and Experimental Research. 2007;31:1380–1391. [PubMed]
  • Williams GD, Aitken SS, Malin H. Reliability of self-reported alcohol consumption in a general population survey. Journal of Studies on Alcohol. 1985;46:223–227. [PubMed]
  • Wolber G, Carne WF, Alexander R. The validity of self-reported abstinence and quality sobriety following chemical dependency treatment. The International Journal of the Addictions. 1990;25:495–513. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...