Addictive Behaviors 39 (2014) 811–817
Contents lists available at ScienceDirect
Addictive Behaviors
Alcohol-related problems and life satisfaction predict motivation to change among mandated college students Andrea R. Diulio, Ian Cero, Tracy K. Witte, Christopher J. Correia ⁎ Auburn University, Department of Psychology, 226 Thach Hall, Auburn, AL 36849, United States
H I G H L I G H T S • • • •
Life satisfaction and specific alcohol problems predicted motivation to change. Relationship between personal problems and motivation depends on social problems. Abuse/dependence symptoms and poor life satisfaction associated with motivation The nature of the relationship varies across levels of severity.
a r t i c l e Keywords: College students Alcohol Motivation to change Life satisfaction
i n f o
a b s t r a c t The present study investigated the role specific types of alcohol-related problems and life satisfaction play in predicting motivation to change alcohol use. Participants were 548 college students mandated to complete a brief intervention following an alcohol-related policy violation. Using hierarchical multiple regression, we tested for the presence of interaction and quadratic effects on baseline data collected prior to the intervention. A significant interaction indicated that the relationship between a respondent's personal consequences and his/her motivation to change differs depending upon the level of concurrent social consequences. Additionally quadratic effects for abuse/dependence symptoms and life satisfaction were found. The quadratic probes suggest that abuse/dependence symptoms and poor life satisfaction are both positively associated with motivation to change for a majority of the sample; however, the nature of these relationships changes for participants with more extreme scores. Results support the utility of using a multidimensional measure of alcohol related problems and assessing non-linear relationships when assessing predictors of motivation to change. The results also suggest that the best strategies for increasing motivation may vary depending on the types of alcohol-related problems and level of life satisfaction the student is experiencing and highlight potential directions for future research. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction The majority of college students (64%) report consuming alcohol in the past month and nearly half (44%) report a binge episode during that time (Substance Abuse and Mental Health Services Administration, 2009). Among college students who drink regularly, nearly half (47%) report experiencing five or more alcohol-related problems in the past year (Wechsler, Davenport, Dowdall, Moeykens, & Castillo, 1994). Research has focused on which factors lead to motivation to decrease alcohol consumption among college students, so that such factors can be incorporated into intervention efforts. Research on motivation to change has been influenced by the transtheoretical model (TTM), a theory introduced to explain how individuals progress toward behavior change (Prochaska & DiClemente, 1982; Prochaska, DiClemente, & Norcorss, 1992). The TTM posits that motivation to change is an ⁎ Corresponding author. Tel.: +1 334 844 6480; fax: +1 334 844 4447. E-mail address: [email protected] (C.J. Correia). 0306-4603/$ – see front matter © 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.addbeh.2014.01.001
important element of change; motivation is predictive of outcome, and different processes of change are hypothesized to be more appropriate or effective depending on the level of motivation. Motivation to change is hypothesized to increase when the perceived costs of a behavior outweigh the benefits. Several studies have applied the TTM, and more specifically motivation to change, to college student drinking. Studies focusing on predictors of motivation to change have produced mixed results. For example, one study suggests that light drinkers who have experienced few problems report greater motivation to change (Barnett, Goldstein, Murphy, Colby, & Monti, 2006). However, a second study found that mandated students who experience more alcohol-related problems are more motivated to change their level of alcohol consumption (Shealy, Murphy, Borsari, & Correia, 2007); that study also indicated that motivation to change was related to low life satisfaction. Relatedly, life satisfaction has shown a consistent, negative relationship with alcohol-related problems (Molnar, Busseri, Perrier, & Sadava, 2009; Murphy, McDevitt-Murphy, & Barnett, 2005). These latter findings are
812
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
consistent with the TTM's assertion that motivation to change increases as the costs of alcohol use increase, with costs operationalized in these specific studies as increased negative alcohol-related consequences and decreased life satisfaction. The current study sought to further examine the relationship between alcohol-related problems, life satisfaction, and motivation to change patterns of alcohol use by replicating and extending the methodology used in Shealy et al. (2007). Two significant improvements have been implemented. First, Shealy et al. used single-factor measures to assess alcohol-related problems. However, several recent factor analytic studies using college student samples have shown that alcoholrelated problems can be categorized into a number of dimensions (e.g., Maddock, Laforge, Rossi, O'Hare, 2001; Martens, Neighbors, Dams-O'Connor, Lee, & Larimer, 2007; Read, Kahler, Strong, & Colder, 2006). For example, the Rutgers Alcohol Problem Index, a commonly used measure of alcohol problems among college students, consists of three subscales: abuse/dependence symptoms, personal consequences, and social consequences (Martens, Neighbors, Dams-O'Connor, Lee, & Larimer, 2007). In addition, research suggests that college students do not perceive various alcohol-related problems to be equally aversive and that some outcomes labeled as problems by researchers may be perceived as neutral or even positive by college students (Mallett, Bachrach, & Turrisi, 2008; Mallett, Lee, Neighbors, Larimer, & Turrisi, 2006). Mallett et al. (2008) suggest that a student's perception of an alcohol-related problem likely influences the relationship between drinking outcomes and motivation to change. Stated more generally, the relationships between alcohol-related problems and motivation to change patterns of alcohol consumption may vary across types or categories of consequences. Indeed, a study of adults enrolled in treatment for alcohol use disorders reported that motivation to change was positively related to social and interpersonal consequences but inversely related to physical consequences (DiClemente, Doyle, & Donovan, 2009). The current study will explore similar relationships in a college student sample by using a multidimensional model of alcohol-related problems. Further, whereas the Shealy et al. (2007) study investigated linear bivariate relationships among alcohol-related problems, life satisfaction, and motivation to change, the current study used a significantly larger sample than Shealy et al. (2007), which allowed for a more detailed analysis of the relationships among the variables. More specifically, interactions among alcohol-related problems and life satisfaction were probed to identify and describe moderation, and both linear and quadratic relationships were tested and described. Based on previous studies, we hypothesized that alcohol-related problems would be positively related to motivation to change, and that life satisfaction would be negatively related to motivation to change. The extant empirical literature does not allow for specific hypotheses regarding interactions and the potential for linear versus non-linear relationships. 2. Methods 2.1. Participants Participants were 548 students (78% male; M age = 20; 98% White) from a large southeastern university who were referred to participate in a brief alcohol intervention in response to a violation of the university's alcohol policy. All measures were collected as part of the intervention process, and the use of the data for the current research was approved by the university's Institutional Review Board. 2.2. Measures Daily Drinking Questionnaire (DDQ; Collins, Parks, & Marlatt, 1985) is an open-ended calendar in which participants report the average number of drinks they have consumed for each day of the week for the past 28 days, in addition to the amount of time they spent drinking on those
days. The DDQ was also used to measure the number of binge drinking episodes the student engaged in during the past 28 days. A binge episode was defined as five or more drinks for males and four or more drinks for females (Wechsler, Dowdall, Davenport, & Rimm, 1995). The DDQ has been shown to be reliable and valid among college students (Collins et al., 1985). In the current study the DDQ was used for descriptive purposes. Rutgers Alcohol Problems Index (RAPI, White & Labouvie, 1989) is a 23-item screening measure that assesses the frequency of alcohol related problems among adolescents and young adults in the past 28 days. Responses are rated on a 5-point Likert scale, from none (0) to over 10 times (4). Previous research with the RAPI has identified the scale as containing three distinct subscales: abuse/dependence symptoms (12 items; α = .65) consist of severe consequences reflecting tolerance, personal changes and familial problems (e.g., Had withdrawal symptoms, that is, felt sick because you stopped or cut down on drinking); personal consequences (7 items; α = .63) are negative outcomes that only affect the student drinking (e.g., Neglected your responsibilities); and social consequences (4 items; α = .68) consist of consequences that affect the drinker, as well as those around him or her (e.g., Got into fights, acted bad or did mean things; Martens et al., 2007). Readiness to Change Questionnaire (RTCQ; Rollnick, Heather, Gold, & Hall, 1992) is a 12-item questionnaire based on the transtheoretical model stages of change that measure motivation among alcohol users. All responses are rated on a 5-point Likert scale, from strongly disagree (1) to strongly agree (5). Principal component analysis has confirmed a three-factor structure corresponding to pre-contemplation, contemplation, and action among the 12 items (Rollnick et al., 1992). However, the current study treated the measure as a unitary scale that makes use of all 12 items. Previous studies have used the unitary scales with college students (Collins, Carey, & Otto, 2009; McNally & Palfai, 2001) and other populations of drinkers (Budd & Rollnick, 1996). The unitary scale had acceptable internal consistency with the current sample (α = .75). Temporal Satisfaction With Life Scale (TSWLS; Pavot, Diener, & Suh, 1998) is a self-report measure that was used to assess current life satisfaction. Participants endorsed their degree of life satisfaction on five items using a 7-point Likert scale that ranged from strongly disagree (1) to strongly agree (7). The TSWLS has demonstrated good psychometric properties among college populations (Murphy et al., 2005; Shealy et al., 2007). The internal consistency of this measure in the current study was adequate (α = .87). 2.3. Data editing A series of preliminary analyzes were run to screen for missing data as well as potential outliers. We began with an initial sample of 590, and 42 subjects had missing values on one or more variables of interest. Independent sample t-tests suggested that there were no statistically significant differences between subjects with missing values and subjects with no missing values. Therefore, the subjects with missing values were deleted from the sample. The final sample size of data analyzed was 548. There were no outliers on the variables of interest. 2.4. Data analytic procedure The present investigation was concerned with whether alcoholrelated problems and life satisfaction were related to motivation to change alcohol-related behaviors, whether those relationships were interactive, and whether those relationships were constant at all levels of the variables (i.e., linear). To address these questions, a hierarchical multiple regression approach was employed. Following guidance from Aiken and West (1991), all predictor variables were initially centered at their means, quadratic terms were then created by squaring each of these centered variables, and six interaction terms were created by computing the products of all possible centered predictor pairs.
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
Lower-order centered predictors were entered first, all quadratic terms were entered in the second block, and the remaining interactions were added in the third and final block (Cortina, 1994). Significant effects were further probed. Unless otherwise specified, all reported are unstandardized. All covariates (i.e., non-significant or unrelated quadratic and interaction effects) were held at 0 during post-hoc probe procedures. Interactions were probed using the Johnson–Neyman technique, which identifies cut-points along the range of the moderating predictor (e.g., social problems) where the focal predictor (e.g., personal problems) transitions from significant to non-significant (Aiken & West, 1991). In this way, the technique is capable of identifying regions of a moderator where the focal predictor is significantly related to the outcome of interest and to what degree. This method is preferred over other approaches, as it avoids the need to choose an arbitrary set of points at which to probe an effect (Hayes & Matthes, 2009). Linear interaction probes were conducted using the MODPROBE (Hayes & Matthes, 2009) macro for SPSS (version 18). Quadratic effects were probed using a procedure that yields conceptually similar results to the Johnson–Neyman technique, though through a different mathematical approach. In a quadratic relationship, the slope of a given predictor changes depending on the level of that predictor, reaching exactly 0 at the inflect point (Aiken & West, 1991). Establishing a confidence interval for this point is mathematically equivalent to identifying the range along the predictor where the slope is not significant (Hirschberg & Lye, 2005). This fact was exploited to probe significant quadratic effects, thereby identifying regions of significance, similar to the Johnson–Neyman technique outlined above. Quadratic probes were conducted using the Java (version 7) code written by the second author (I.C.).
813
Table 2 Descriptive statistics and Pearson correlations. Variable
1
2
3
4
1. Abuse/dependence symptoms 2. Personal 3. Social 4. Life satisfaction 5. Readiness to change
.45* .43* −.29* .25*
.44* −.28* .22*
−.30* .17*
−.20*
Note: * = p b .01.
that, as social problems increased, there was a weaker relationship between personal problems and motivation to change (t[533] = −2.01, p = .045, R2block = .01). While this interaction term had only small incremental effect on R2, such indices are incomplete measures of the effect of an interaction on their own, and the magnitude of the change in slope must also be considered (Champoux & Peters, 1987). As shown in Fig. 1, the relationship between a respondent's personal problems and his/her motivation to change differs, depending upon the level of concurrent social problems. The Johnson–Neyman probe revealed that, for participants reporting below average social problems, personal problems were positively associated with motivation to change. However, the relationship between personal problems and motivation to change weakens as social problems increase, eventually becoming non-significant for participants at or above mean levels of social problems (cut-point = −.10; 53% of participants below cut-point). Specifically, social problem scores at the cut-point were associated with a 27% weaker slope of personal problems on motivation than social problem scores at the bottom of the observed spectrum (i.e., a .86-unit difference on the social problem subscale). 3.3. Quadratic effects
3. Results 3.1. Descriptive statistics Participants reported an average of five binge episodes over the past month and consuming 16 drinks per week. For abuse/dependence, personal, and social consequences on the RAPI, participants reported an average of one problem in each category for an average of three alcohol-related problems in the past-month. Descriptive statistics among alcohol-related problems, life satisfaction and motivation to change are presented in Table 1. All variables were significantly correlated with each other and in the direction hypothesized. Bivariate correlations are presented in Table 2. Alcohol-related problems were positively correlated with each other (r's ranging from .43 to .45) and to motivation to change (r's ranging from .17 to .25) and inversely correlated with life satisfaction (r's ranging from − .28 to − .30). Life satisfaction was inversely correlated with motivation to change (−.20). 3.2. Linear interactions Results of the hierarchical regression analyses are presented in Table 3. The overall model accounted for 14% of the variance in motivation to change, F(14, 533) = 5.95, p b .001. As depicted, there was a significant interaction between social problems and personal problems such
Table 1 Descriptive statistics for primary study variables.
Social Personal Abuse Quality of life Motivation
Mean
Std. dev.
Min.
Max.
Skewness
Kurtosis
0.96 1.30 1.40 29.34 6.33
1.46 1.78 2.26 4.44 6.87
0.00 0.00 0.00 9.00 −12.00
13.50 9.50 13.00 35.00 27.00
2.65 1.88 2.31 −1.25 −0.61
12.37 3.78 5.87 2.49 0.38
Regression analyses further revealed significant quadratic effects (R2block = .03) for abuse/dependence symptoms (t[533] = − 3.18, p = .01) and for life satisfaction (t[533] = − 2.52, p = .01). Both of the coefficients for these predictors were negative, indicating that an inverted-U shape characterizes their effects. Results from the probes of abuse and life satisfaction are depicted in Figs. 2 and 3, respectively. There are two regions of the curve where the relationship between abuse/dependence symptoms and motivation to change is significant. On the lower, less pathological end of the distribution, there is a positive association between abuse/dependence symptoms and motivation to change, with the slope being steepest for those with the least pathology. As implied by its negative regression coefficient, this effect weakens as pathology increases, eventually becoming non-significant (lower
Table 3 Unstandardized results from final regression block.
Lower-order terms Intercept Abuse Social Personal Life satisfaction Quadratic terms Abuse2 Social2 Personal2 Life satisfaction2 Interaction terms Abuse by social Abuse by personal Abuse by life satisfaction Social by personal Social by life satisfaction Personal by life satisfaction
B
(SE)
t
p
7.729 1.048 0.314 0.552 −0.302
−.428 −.242 −.315 −.293 −.082
18.050 4.330 1.000 1.880 −3.670
.000 .000 .320 .060 .000
−0.157 −0.012 −0.070 −0.023
−.050 −.074 −.079 −.009
−3.170 −0.160 −0.880 −2.520
.002 .871 .377 .012
0.084 0.157 0.006 −0.244 −0.043 0.037
−.093 −.085 −.033 −.122 −.045 −.047
0.910 1.860 0.170 −2.010 −0.950 0.800
.365 .064 .868 .045 .342 .424
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
2.0
Non-Sig. Region
Sig. Region 1.5 1.0 0.5 0.0 -0.5
Social = --0.10
Strength (slope) of Personal Problems
814
-1.0 -1.5 -2.0
-2
-1
0
1
2
3
4
5
Social Problems (mean centered) Fig. 1. Strength of the relationship between personal problems and motivation to change, across levels of social problems. Dotted lines represent the non-simultaneous 95% confidence bounds for the slope of personal problems, and the darkened line represents the region of social problems where the slope of personal problems is statistically significant.
cut-point = 2.18; 88% of participants below cut-point).1 However, the probe also revealed a small group of individuals at the top of the distribution for whom abuse/dependence symptoms are actually negatively associated with motivation to change (upper cut-point = 6.55; 3% of participants above cut-point). This effect becomes stronger as abuse increases, with individuals at the maximum score of the distribution (i.e., a 4.72-point difference on the RAPI abuse subscale) having a 147% more negative slope than those at the upper cut-point. Results from the probe for life satisfaction revealed a largely similar trend. For people at the top end of the scale (i.e., people with the least pathology), there is a strong, negative association between life satisfaction and motivation to change. However, as life satisfaction diminishes, each additional unit-change on that predictor is associated with a weaker relationship with motivation, eventually reaching non-significance (cut-point = − 3.49; 83% participants above cut-point). From the maximum score to the cut-point (i.e., an 8.5-point difference on the life satisfaction scale), there was a 74% reduction in slope. 4. Discussion The current study is the first investigation to assess the relationships among specific types of alcohol related problems, life satisfaction and motivation to change among high-risk college students referred to a brief alcohol intervention. Bivariate relationships among the variables were in the hypothesized direction and replicated results reported by Shealy et al. (2007). The bivariate correlation between abuse/ dependence symptoms and motivation to change was the strongest among the three types of alcohol-related problems, and social problems had the weakest bivariate correlation with motivation to change. Additionally, life satisfaction had statistically significant negative correlations with all three alcohol-related problems and motivation to change. To investigate these relationships further linear interactions and quadratic effects were estimated. The increased sample size in the current study relative to Shealy et al. allowed us to conduct a more powerful set of analyses aimed at determining how alcohol-related problems and life satisfaction predict motivation to change. First, we found an interaction between personal consequences and social consequences. Specifically, the relationship 1 Note, because there were two significant regions of the curve, two separate cut points were detected by the post-hoc probes. The first cut point delineates the significant region on the lower end of the abuse spectrum from the non-significant region in the middle of that same spectrum. The second delineates the significant region at the very top of the abuse spectrum from the non-significant region in the middle. For simplicity, we refer to these as the lower cut-point and the upper cut-point, respectively. To help characterize our findings, we also present the percentage of participants that reported abuse levels on the significant side of each cut-point (i.e., percentage below lower cut point, percentage above upper cut-point).
between personal consequences and motivation to change weakens as endorsement of social consequences increases, eventually becoming non-significant near the mean level of social consequences in our sample. This finding suggests that the relationship between social and personal problems and motivation to change cannot be understood in isolation from one another. Results from the two quadratic probes also replicate and expand on the results of the Shealy et al. (2007) study. For a majority of the sample, motivation to change was positively associated with abuse/dependence symptoms and negatively associated with life satisfaction. However, results also suggest that these effects did not apply to the more severe participants in the sample. That is, students with very high abuse/ dependence symptoms and very low life satisfaction did not experience continued increases in motivation to change, and students with the most severe abuse/dependence symptoms (3% of the sample) actually reported decreased motivation to change. These results suggest that the tendency for abuse/dependence symptoms and low life satisfaction to galvanize students and motivate them to change reaches a limit, and students at the greatest risk of developing alcohol abuse or dependence may be demotivated. There is a high degree of comorbidity between alcohol use disorders and other mental health conditions (Lo, Monge, Howell, & Cheng, 2013). Students with especially high levels of abuse/ dependence symptoms and low life satisfaction may view drinking as a means to cope with negative affect. Previous studies have shown that alcohol can result in a brief attenuation of stress (e.g. Solomon, 1980; Baker et al., 2004). However, if intoxication is repeatedly paired with an immediate decrease in negative affect, the negative affect may come to function as an establishing operation of consuming alcohol, resulting in increased motivation to drink when experiencing negative affect (Baker et al., 2004). Alcohol's role as a strategy for coping with negative affect may become more salient than the reported negative consequences, thus disrupting the relationship between problems and motivation to change found among the majority of college students. The complex relationship between abuse/dependence symptoms, life satisfaction and motivation to change warrants additional research, and the role of negative affect and coping motives as potential moderating influences could further illuminate the relationship. 4.1. Limitations The cross-sectional design of the study allows for the measurement of statistical associations among variables. However, due to motivation to change, life satisfaction, and alcohol-related problems being simultaneously measured, course of endorsement for each variable cannot be determined. A follow-up study that collects longitudinal data will need to be performed to confirm the sequence in which the variables may impact each other, as well as to confirm the findings in a second unique sample of students. Regarding the values (cut-point) along the moderator where effects transition from significant to non-significant relationships, they are meant to be probabilistic and may be sample specific. Future studies need to be conducted to determine how reliable the identified values are across samples, and a likely outcome is that a range of values will emerge that center around the actual transition point. Additionally, as this study utilized mandated students, additional research is needed to determine if similar models can be applied to more general student samples. Furthermore, the internal consistency estimates for the RAPI subscales were relatively low (.63–.68). Similar internal consistency estimates were reported in the initial factor analytic study (.68–.75) despite relatively strong factor structures (Martens et al., 2007), suggesting that the individual RAPI items may be independently measuring unique facets of the underlying latent variables. The use of measures with low reliability can have the effect of flattening curves, rather than creating illusory ones, and can inflate the likelihood of Type II errors (Shepperd, 1991). Thus, the true effects may actually be stronger and steeper than what was observed in the current study. Lastly, some of the findings were discussed in the context of clinical theory
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
815
A 11
Non-Sig. Region
Sig. Region 10
8
Abuse/Dependence Symptoms = 6.55
Inflect Point Abuse/Dependence Symptoms = 2.18
Pred. Motivation
9
Sig. Region
7 6 5 4 -2
0
2
4
6
8
Abuse/Dependence Symptoms (mean centered)
B 3
Non-Sig. Region
Sig. Region 2
Inflect Point
1 0 -1 -2 -3
Abuse/Dependence Sympsoms = 6.55
Abuse/Dependence Symptoms = 2.18
Slope of Abuse
Sig. Region
-4 -2
0
2
4
6
8
Abuse/Dependence Symptoms (mean centered) Fig. 2. Predicted motivation to change (Panel A) and simple slope of abuse/dependence symptoms (Panel B) by levels of reported abuse/dependence symptoms. All other covariates are controlled at their means. Note that dotted lines in Panel B represent the non-simultaneous 95% confidence bounds for the slope.
and research on the relationships among life satisfaction, alcoholrelated consequences and motivation to change being related to drinking for social reasons and to cope with stress and comorbid conditions. However, this study cannot confirm those relationships, as data on drinking motives were not collected. The model needs to be further tested in studies that include assessment of drinking motives. 4.2. Clinical implications Considering the time restrictions associated with brief alcohol interventions, it is valuable to study what may be most effective in increasing motivation to change in the greatest percentage of students. A recent review of studies assessing the content of personalized feedback interventions reported that 60% included a component designed to engage students in a discussion about the alcohol-related consequences they have experienced (Miller et al., 2012). Our results suggest that such a discussion may indeed be a useful tool for increasing motivation to change. However, clinicians should be aware that the relationship between negative consequences and motivation is neither linear nor uniform across consequences. Therefore, there may not be a “one size fits all” approach to determining how to maximize the effect negative consequences have on the perceived pros and cons of alcohol use.
For some negative consequences (i.e., personal and social), the relationship with motivation to change depends on other factors. Thus, interventions targeting personal alcohol-related consequences may be less effective if the individual is experiencing social consequences as well. For other negative consequences (i.e., low life satisfaction, abuse/dependence consequences), the relationship to motivation to change differs across levels of severity. Our results suggest that feedback on abuse/dependence symptoms and diminished life satisfaction may be therapeutic in increasing motivation to change for the majority of referred students. However, highlighting these symptoms may be demotivating for students with more severe impairment. As an alternative, or in addition to feedback interventions, students who report low life satisfaction and high abuse/dependence symptoms may benefit from a review of adaptive coping strategies. As noted, these students might be more likely to use alcohol to manage negative affect and experiences. If true, those students would fit a coping deficit model of alcohol use (Bandura, 1969); they may lack alternative methods of mitigating their distress and a discussion of adaptive and healthy coping strategies may be a more effective way of increasing their motivation to change. In conclusion, our results provide important information regarding the links between different types of alcohol-related problems and life satisfaction with motivation to change alcohol use among college
816
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
A
10
Non-Sig. Region
Sig. Region
Inflect Point
8
7
LS= -3.49
Pred. Motivation
9
6
5 -10
-8
-6
-4
-2
0
2
4
6
Life Satisfaction (mean centered)
B 0.6
Non-Sig. Region
Sig. Region
0.4
0.0 -0.2 -0.4
LS = -3.49
Slope of SL
0.2
-0.6 -0.8 -1.0 -10
-8
-6
-4
-2
0
2
4
6
Life Satisfaction (mean centered) Fig. 3. Predicted motivation to change (Panel A) and simple slope (Panel B) of Life Satisfaction (LS) by levels of reported LS. All other covariates are controlled at their means. Note that dotted lines in Panel B represent the non-simultaneous 95% confidence bounds for the slope.
students. The results inform the theoretical discussion on motivation to change and can be directly applied to clinical work with college student drinkers. Role of funding sources This study was not externally funded. Auburn University had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.
Contributors Diulio and Correia designed the study and Correia wrote the protocol. Diulio conducted literature searches, provided summaries of previous research studies, conducted the statistical analysis and wrote the first draft of the manuscript. Cero and Witte conducted additional statistical analyses and contributed to subsequent draft of the manuscript. All authors contributed to and have approved the final manuscript.
Conflict of interest All authors declare that they have no conflicts of interest.
References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage Publications. Bandura, A. (1969). Principles of behavior modification. New York: Holt, Rinehart & Winston.
Baker, T. B., Piper, M. E., McCarthy, D. E., Majeskie, M. R., & Fiore, M. C. (2004). Addiction motivation reformulated: An affective processing model of negative reinforcement. Psychological Review, 111, 33–51. Barnett, N.P., Goldstein, A. L., Murphy, J. G., Colby, S. M., & Monti, P.M. (2006). ‘I'll never drink like that again’: Characteristics of alcohol-related incidents and predictors of motivation to change in college students. Journal of Studies on Alcohol, 67, 754–763. Budd, R. J., & Rollnick, S. (1996). The structure of the Readiness to Change Questionnaire: A test of Prochaska & DiClemente's transtheoretical model. British Journal of Health Psychology, 1, 365–376. Champoux, J. E., & Peters, W. S. (1987). Form, effect size and power in moderated regression analysis. Journal of Occupational Psychology, 60, 243–255. Collins, S. E., Carey, K. B., & Otto, J. M. (2009). A new decisional balance measure of motivation to change among at-risk college drinkers. Psychology of Addictive Behaviors, 23, 464–471. Collins, R., Parks, G. A., & Marlatt, G. (1985). 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, 53, 189–200. Cortina, J. M. (1994). Interaction, nonlinearity, and multicollinearity: Implications for multiple regression. Journal of Management, 19, 915–922. DiClemente, C. C., Doyle, S. R., & Donovan, D. (2009). Predicting treatment seekers' readiness to change their drinking behavior in the COMBINE Study. Alcoholism: Clinical and Experimental Research, 33, 879–892. Hayes, A. F., & Matthes, J. (2009). Computational procedures for probing interactions in OLS and logistic regression: SPSS and SAS implementations. Behavior Research Methods, 41, 924–936. Hirschberg, J., & Lye, J. N. (2005). Inferences for the extremum of quadratic regression models. http://dx.doi.org/10.2139/ssrn.918626 (Available at SSRN 918626). Lo, C. C., Monge, A. N., Howell, R. J., & Cheng, T. C. (2013). The role of mental illness in alcohol abuse and prescription drug misuse: Gender-specific analysis of college students. Journal of Psychoactive Drugs, 45, 39–47.
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817 Maddock, J. E., Laforge, R. G., Rossi, J. S., & O'Hare, T. (2001). The College Alcohol Problems Scale. Addictive Behaviors, 26, 385–398. Mallett, K. A., Bachrach, R. L., & Turrisi, R. (2008). Are all negative consequences truly negative? Assessing variations among college students' perceptions of alcohol related consequences. Addictive Behaviors, 33, 1375–1381. Mallett, K. A., Lee, C. M., Neighbors, C., Larimer, M. E., & Turrisi, R. (2006). Do we learn from our mistakes? An examination of the impact of negative alcohol-related consequences on college students' drinking patterns and perceptions. Journal of Studies on Alcohol, 67, 269–276. Martens, M. P., Neighbors, C., Dams-O'Connor, K., Lee, C. M., & Larimer, M. E. (2007). The factor structure of a dichotomously scored Rutgers Alcohol Problems Index. Journal of Studies on Alcohol and Drugs, 68, 597–606. McNally, A.M., & Palfai, T. P. (2001). Negative emotional expectancies and readiness to change among college student binge drinkers. Addictive Behaviors, 26, 721–734. Miller, M. B., Leffingwell, T. R., Claborn, K., Meier, E., Walters, S., & Neighbors, C. (2012). Personalized feedback interventions for college alcohol misuse: An update of Walters & Neighbors (2005). Psychology of Addictive Behaviors. http://dx.doi.org/10.1037/ a0031174. Molnar, D. S., Busseri, M.A., Perrier, C. P. K., & Sadava, S. W. (2009). A longitudinal examination of alcohol use and subjective well-being in an undergraduate sample. Journal of Studies on Alcohol and Drugs, 70, 704–713. Murphy, J. G., McDevitt-Murphy, M. E., & Barnett, N.P. (2005). Drink and be merry? Gender, life satisfaction, and alcohol consumption among college students. Psychology of Addictive Behaviors, 19, 184–191. Pavot, W., Diener, E., & Suh, E. (1998). The temporal satisfaction with life scale. Journal of Personality Assessment, 701, 340–354. Prochaska, J. O., & Diclemente, C. C. (1982). Transtheoretical therapy: Toward a more integrative model of change. Psychotherapy: Theory, Research and Practice, 20, 161–173.
817
Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1992). In search of how people change: Applications to addictive behaviors. American Psychologist, 47, 1102–1114. Read, J. P., Kahler, C. W., Strong, D. R., & Colder, C. R. (2006). Development and preliminary validation of the young adult alcohol consequences questionnaire. Journal of Studies on Alcohol, 67, 169–177. Rollnick, S., Heather, N., Gold, R., & Hall, W. (1992). Development of a short ‘readiness to change’ questionnaire for use in brief, opportunistic interventions among excessive drinkers. British Journal of Addiction, 87, 743–754. Shealy, A. E., Murphy, J. G., Borsari, B., & Correia, C. J. (2007). Predictors of motivation to change alcohol use among referred college students. Addictive Behaviors, 32, 2358–2364. Shepperd, J. A. (1991). Cautions in assessing spurious “moderator effects”. Psychological Bulletin, 110, 315–317. Solomon, R. L. (1980). The opponent-process theory of acquired motivation. American Psychologist, 35, 691–712. Substance Abuse and Mental Health Services Administration (2009). Results from the 2009 National Survey on Drug Use and Health Summary of National Findings. Office of Applied Studies, NSDUH Series H-38A, HHS Publication No. SMA 10-4586 Findings. Rockville, MD (Retrieved from http://oas.samhsa.gov/nsduh/2k9nsduh/2k9Results.pdf) Wechsler, H., Davenport, A., Dowdall, G., Moeykens, B., & Castillo, S. (1994). Health and behavioral consequences of binge drinking in college: A national survey of students at 140 campuses. Journal of the American Medical Association, 272, 1672–1677. Wechsler, H., Dowdall, G., Davenport, A., & Rimm, E. B. (1995). A gender-specific measure of binge drinking among college students. American Journal of Public Health, 85, 982–985. White, H., & Labouvie, E. (1989). Towards an assessment of adolescent problem drinking. Journal of Studies on Alcohol, 50, 30–37.
Contents lists available at ScienceDirect
Addictive Behaviors
Alcohol-related problems and life satisfaction predict motivation to change among mandated college students Andrea R. Diulio, Ian Cero, Tracy K. Witte, Christopher J. Correia ⁎ Auburn University, Department of Psychology, 226 Thach Hall, Auburn, AL 36849, United States
H I G H L I G H T S • • • •
Life satisfaction and specific alcohol problems predicted motivation to change. Relationship between personal problems and motivation depends on social problems. Abuse/dependence symptoms and poor life satisfaction associated with motivation The nature of the relationship varies across levels of severity.
a r t i c l e Keywords: College students Alcohol Motivation to change Life satisfaction
i n f o
a b s t r a c t The present study investigated the role specific types of alcohol-related problems and life satisfaction play in predicting motivation to change alcohol use. Participants were 548 college students mandated to complete a brief intervention following an alcohol-related policy violation. Using hierarchical multiple regression, we tested for the presence of interaction and quadratic effects on baseline data collected prior to the intervention. A significant interaction indicated that the relationship between a respondent's personal consequences and his/her motivation to change differs depending upon the level of concurrent social consequences. Additionally quadratic effects for abuse/dependence symptoms and life satisfaction were found. The quadratic probes suggest that abuse/dependence symptoms and poor life satisfaction are both positively associated with motivation to change for a majority of the sample; however, the nature of these relationships changes for participants with more extreme scores. Results support the utility of using a multidimensional measure of alcohol related problems and assessing non-linear relationships when assessing predictors of motivation to change. The results also suggest that the best strategies for increasing motivation may vary depending on the types of alcohol-related problems and level of life satisfaction the student is experiencing and highlight potential directions for future research. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction The majority of college students (64%) report consuming alcohol in the past month and nearly half (44%) report a binge episode during that time (Substance Abuse and Mental Health Services Administration, 2009). Among college students who drink regularly, nearly half (47%) report experiencing five or more alcohol-related problems in the past year (Wechsler, Davenport, Dowdall, Moeykens, & Castillo, 1994). Research has focused on which factors lead to motivation to decrease alcohol consumption among college students, so that such factors can be incorporated into intervention efforts. Research on motivation to change has been influenced by the transtheoretical model (TTM), a theory introduced to explain how individuals progress toward behavior change (Prochaska & DiClemente, 1982; Prochaska, DiClemente, & Norcorss, 1992). The TTM posits that motivation to change is an ⁎ Corresponding author. Tel.: +1 334 844 6480; fax: +1 334 844 4447. E-mail address: [email protected] (C.J. Correia). 0306-4603/$ – see front matter © 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.addbeh.2014.01.001
important element of change; motivation is predictive of outcome, and different processes of change are hypothesized to be more appropriate or effective depending on the level of motivation. Motivation to change is hypothesized to increase when the perceived costs of a behavior outweigh the benefits. Several studies have applied the TTM, and more specifically motivation to change, to college student drinking. Studies focusing on predictors of motivation to change have produced mixed results. For example, one study suggests that light drinkers who have experienced few problems report greater motivation to change (Barnett, Goldstein, Murphy, Colby, & Monti, 2006). However, a second study found that mandated students who experience more alcohol-related problems are more motivated to change their level of alcohol consumption (Shealy, Murphy, Borsari, & Correia, 2007); that study also indicated that motivation to change was related to low life satisfaction. Relatedly, life satisfaction has shown a consistent, negative relationship with alcohol-related problems (Molnar, Busseri, Perrier, & Sadava, 2009; Murphy, McDevitt-Murphy, & Barnett, 2005). These latter findings are
812
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
consistent with the TTM's assertion that motivation to change increases as the costs of alcohol use increase, with costs operationalized in these specific studies as increased negative alcohol-related consequences and decreased life satisfaction. The current study sought to further examine the relationship between alcohol-related problems, life satisfaction, and motivation to change patterns of alcohol use by replicating and extending the methodology used in Shealy et al. (2007). Two significant improvements have been implemented. First, Shealy et al. used single-factor measures to assess alcohol-related problems. However, several recent factor analytic studies using college student samples have shown that alcoholrelated problems can be categorized into a number of dimensions (e.g., Maddock, Laforge, Rossi, O'Hare, 2001; Martens, Neighbors, Dams-O'Connor, Lee, & Larimer, 2007; Read, Kahler, Strong, & Colder, 2006). For example, the Rutgers Alcohol Problem Index, a commonly used measure of alcohol problems among college students, consists of three subscales: abuse/dependence symptoms, personal consequences, and social consequences (Martens, Neighbors, Dams-O'Connor, Lee, & Larimer, 2007). In addition, research suggests that college students do not perceive various alcohol-related problems to be equally aversive and that some outcomes labeled as problems by researchers may be perceived as neutral or even positive by college students (Mallett, Bachrach, & Turrisi, 2008; Mallett, Lee, Neighbors, Larimer, & Turrisi, 2006). Mallett et al. (2008) suggest that a student's perception of an alcohol-related problem likely influences the relationship between drinking outcomes and motivation to change. Stated more generally, the relationships between alcohol-related problems and motivation to change patterns of alcohol consumption may vary across types or categories of consequences. Indeed, a study of adults enrolled in treatment for alcohol use disorders reported that motivation to change was positively related to social and interpersonal consequences but inversely related to physical consequences (DiClemente, Doyle, & Donovan, 2009). The current study will explore similar relationships in a college student sample by using a multidimensional model of alcohol-related problems. Further, whereas the Shealy et al. (2007) study investigated linear bivariate relationships among alcohol-related problems, life satisfaction, and motivation to change, the current study used a significantly larger sample than Shealy et al. (2007), which allowed for a more detailed analysis of the relationships among the variables. More specifically, interactions among alcohol-related problems and life satisfaction were probed to identify and describe moderation, and both linear and quadratic relationships were tested and described. Based on previous studies, we hypothesized that alcohol-related problems would be positively related to motivation to change, and that life satisfaction would be negatively related to motivation to change. The extant empirical literature does not allow for specific hypotheses regarding interactions and the potential for linear versus non-linear relationships. 2. Methods 2.1. Participants Participants were 548 students (78% male; M age = 20; 98% White) from a large southeastern university who were referred to participate in a brief alcohol intervention in response to a violation of the university's alcohol policy. All measures were collected as part of the intervention process, and the use of the data for the current research was approved by the university's Institutional Review Board. 2.2. Measures Daily Drinking Questionnaire (DDQ; Collins, Parks, & Marlatt, 1985) is an open-ended calendar in which participants report the average number of drinks they have consumed for each day of the week for the past 28 days, in addition to the amount of time they spent drinking on those
days. The DDQ was also used to measure the number of binge drinking episodes the student engaged in during the past 28 days. A binge episode was defined as five or more drinks for males and four or more drinks for females (Wechsler, Dowdall, Davenport, & Rimm, 1995). The DDQ has been shown to be reliable and valid among college students (Collins et al., 1985). In the current study the DDQ was used for descriptive purposes. Rutgers Alcohol Problems Index (RAPI, White & Labouvie, 1989) is a 23-item screening measure that assesses the frequency of alcohol related problems among adolescents and young adults in the past 28 days. Responses are rated on a 5-point Likert scale, from none (0) to over 10 times (4). Previous research with the RAPI has identified the scale as containing three distinct subscales: abuse/dependence symptoms (12 items; α = .65) consist of severe consequences reflecting tolerance, personal changes and familial problems (e.g., Had withdrawal symptoms, that is, felt sick because you stopped or cut down on drinking); personal consequences (7 items; α = .63) are negative outcomes that only affect the student drinking (e.g., Neglected your responsibilities); and social consequences (4 items; α = .68) consist of consequences that affect the drinker, as well as those around him or her (e.g., Got into fights, acted bad or did mean things; Martens et al., 2007). Readiness to Change Questionnaire (RTCQ; Rollnick, Heather, Gold, & Hall, 1992) is a 12-item questionnaire based on the transtheoretical model stages of change that measure motivation among alcohol users. All responses are rated on a 5-point Likert scale, from strongly disagree (1) to strongly agree (5). Principal component analysis has confirmed a three-factor structure corresponding to pre-contemplation, contemplation, and action among the 12 items (Rollnick et al., 1992). However, the current study treated the measure as a unitary scale that makes use of all 12 items. Previous studies have used the unitary scales with college students (Collins, Carey, & Otto, 2009; McNally & Palfai, 2001) and other populations of drinkers (Budd & Rollnick, 1996). The unitary scale had acceptable internal consistency with the current sample (α = .75). Temporal Satisfaction With Life Scale (TSWLS; Pavot, Diener, & Suh, 1998) is a self-report measure that was used to assess current life satisfaction. Participants endorsed their degree of life satisfaction on five items using a 7-point Likert scale that ranged from strongly disagree (1) to strongly agree (7). The TSWLS has demonstrated good psychometric properties among college populations (Murphy et al., 2005; Shealy et al., 2007). The internal consistency of this measure in the current study was adequate (α = .87). 2.3. Data editing A series of preliminary analyzes were run to screen for missing data as well as potential outliers. We began with an initial sample of 590, and 42 subjects had missing values on one or more variables of interest. Independent sample t-tests suggested that there were no statistically significant differences between subjects with missing values and subjects with no missing values. Therefore, the subjects with missing values were deleted from the sample. The final sample size of data analyzed was 548. There were no outliers on the variables of interest. 2.4. Data analytic procedure The present investigation was concerned with whether alcoholrelated problems and life satisfaction were related to motivation to change alcohol-related behaviors, whether those relationships were interactive, and whether those relationships were constant at all levels of the variables (i.e., linear). To address these questions, a hierarchical multiple regression approach was employed. Following guidance from Aiken and West (1991), all predictor variables were initially centered at their means, quadratic terms were then created by squaring each of these centered variables, and six interaction terms were created by computing the products of all possible centered predictor pairs.
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
Lower-order centered predictors were entered first, all quadratic terms were entered in the second block, and the remaining interactions were added in the third and final block (Cortina, 1994). Significant effects were further probed. Unless otherwise specified, all reported are unstandardized. All covariates (i.e., non-significant or unrelated quadratic and interaction effects) were held at 0 during post-hoc probe procedures. Interactions were probed using the Johnson–Neyman technique, which identifies cut-points along the range of the moderating predictor (e.g., social problems) where the focal predictor (e.g., personal problems) transitions from significant to non-significant (Aiken & West, 1991). In this way, the technique is capable of identifying regions of a moderator where the focal predictor is significantly related to the outcome of interest and to what degree. This method is preferred over other approaches, as it avoids the need to choose an arbitrary set of points at which to probe an effect (Hayes & Matthes, 2009). Linear interaction probes were conducted using the MODPROBE (Hayes & Matthes, 2009) macro for SPSS (version 18). Quadratic effects were probed using a procedure that yields conceptually similar results to the Johnson–Neyman technique, though through a different mathematical approach. In a quadratic relationship, the slope of a given predictor changes depending on the level of that predictor, reaching exactly 0 at the inflect point (Aiken & West, 1991). Establishing a confidence interval for this point is mathematically equivalent to identifying the range along the predictor where the slope is not significant (Hirschberg & Lye, 2005). This fact was exploited to probe significant quadratic effects, thereby identifying regions of significance, similar to the Johnson–Neyman technique outlined above. Quadratic probes were conducted using the Java (version 7) code written by the second author (I.C.).
813
Table 2 Descriptive statistics and Pearson correlations. Variable
1
2
3
4
1. Abuse/dependence symptoms 2. Personal 3. Social 4. Life satisfaction 5. Readiness to change
.45* .43* −.29* .25*
.44* −.28* .22*
−.30* .17*
−.20*
Note: * = p b .01.
that, as social problems increased, there was a weaker relationship between personal problems and motivation to change (t[533] = −2.01, p = .045, R2block = .01). While this interaction term had only small incremental effect on R2, such indices are incomplete measures of the effect of an interaction on their own, and the magnitude of the change in slope must also be considered (Champoux & Peters, 1987). As shown in Fig. 1, the relationship between a respondent's personal problems and his/her motivation to change differs, depending upon the level of concurrent social problems. The Johnson–Neyman probe revealed that, for participants reporting below average social problems, personal problems were positively associated with motivation to change. However, the relationship between personal problems and motivation to change weakens as social problems increase, eventually becoming non-significant for participants at or above mean levels of social problems (cut-point = −.10; 53% of participants below cut-point). Specifically, social problem scores at the cut-point were associated with a 27% weaker slope of personal problems on motivation than social problem scores at the bottom of the observed spectrum (i.e., a .86-unit difference on the social problem subscale). 3.3. Quadratic effects
3. Results 3.1. Descriptive statistics Participants reported an average of five binge episodes over the past month and consuming 16 drinks per week. For abuse/dependence, personal, and social consequences on the RAPI, participants reported an average of one problem in each category for an average of three alcohol-related problems in the past-month. Descriptive statistics among alcohol-related problems, life satisfaction and motivation to change are presented in Table 1. All variables were significantly correlated with each other and in the direction hypothesized. Bivariate correlations are presented in Table 2. Alcohol-related problems were positively correlated with each other (r's ranging from .43 to .45) and to motivation to change (r's ranging from .17 to .25) and inversely correlated with life satisfaction (r's ranging from − .28 to − .30). Life satisfaction was inversely correlated with motivation to change (−.20). 3.2. Linear interactions Results of the hierarchical regression analyses are presented in Table 3. The overall model accounted for 14% of the variance in motivation to change, F(14, 533) = 5.95, p b .001. As depicted, there was a significant interaction between social problems and personal problems such
Table 1 Descriptive statistics for primary study variables.
Social Personal Abuse Quality of life Motivation
Mean
Std. dev.
Min.
Max.
Skewness
Kurtosis
0.96 1.30 1.40 29.34 6.33
1.46 1.78 2.26 4.44 6.87
0.00 0.00 0.00 9.00 −12.00
13.50 9.50 13.00 35.00 27.00
2.65 1.88 2.31 −1.25 −0.61
12.37 3.78 5.87 2.49 0.38
Regression analyses further revealed significant quadratic effects (R2block = .03) for abuse/dependence symptoms (t[533] = − 3.18, p = .01) and for life satisfaction (t[533] = − 2.52, p = .01). Both of the coefficients for these predictors were negative, indicating that an inverted-U shape characterizes their effects. Results from the probes of abuse and life satisfaction are depicted in Figs. 2 and 3, respectively. There are two regions of the curve where the relationship between abuse/dependence symptoms and motivation to change is significant. On the lower, less pathological end of the distribution, there is a positive association between abuse/dependence symptoms and motivation to change, with the slope being steepest for those with the least pathology. As implied by its negative regression coefficient, this effect weakens as pathology increases, eventually becoming non-significant (lower
Table 3 Unstandardized results from final regression block.
Lower-order terms Intercept Abuse Social Personal Life satisfaction Quadratic terms Abuse2 Social2 Personal2 Life satisfaction2 Interaction terms Abuse by social Abuse by personal Abuse by life satisfaction Social by personal Social by life satisfaction Personal by life satisfaction
B
(SE)
t
p
7.729 1.048 0.314 0.552 −0.302
−.428 −.242 −.315 −.293 −.082
18.050 4.330 1.000 1.880 −3.670
.000 .000 .320 .060 .000
−0.157 −0.012 −0.070 −0.023
−.050 −.074 −.079 −.009
−3.170 −0.160 −0.880 −2.520
.002 .871 .377 .012
0.084 0.157 0.006 −0.244 −0.043 0.037
−.093 −.085 −.033 −.122 −.045 −.047
0.910 1.860 0.170 −2.010 −0.950 0.800
.365 .064 .868 .045 .342 .424
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
2.0
Non-Sig. Region
Sig. Region 1.5 1.0 0.5 0.0 -0.5
Social = --0.10
Strength (slope) of Personal Problems
814
-1.0 -1.5 -2.0
-2
-1
0
1
2
3
4
5
Social Problems (mean centered) Fig. 1. Strength of the relationship between personal problems and motivation to change, across levels of social problems. Dotted lines represent the non-simultaneous 95% confidence bounds for the slope of personal problems, and the darkened line represents the region of social problems where the slope of personal problems is statistically significant.
cut-point = 2.18; 88% of participants below cut-point).1 However, the probe also revealed a small group of individuals at the top of the distribution for whom abuse/dependence symptoms are actually negatively associated with motivation to change (upper cut-point = 6.55; 3% of participants above cut-point). This effect becomes stronger as abuse increases, with individuals at the maximum score of the distribution (i.e., a 4.72-point difference on the RAPI abuse subscale) having a 147% more negative slope than those at the upper cut-point. Results from the probe for life satisfaction revealed a largely similar trend. For people at the top end of the scale (i.e., people with the least pathology), there is a strong, negative association between life satisfaction and motivation to change. However, as life satisfaction diminishes, each additional unit-change on that predictor is associated with a weaker relationship with motivation, eventually reaching non-significance (cut-point = − 3.49; 83% participants above cut-point). From the maximum score to the cut-point (i.e., an 8.5-point difference on the life satisfaction scale), there was a 74% reduction in slope. 4. Discussion The current study is the first investigation to assess the relationships among specific types of alcohol related problems, life satisfaction and motivation to change among high-risk college students referred to a brief alcohol intervention. Bivariate relationships among the variables were in the hypothesized direction and replicated results reported by Shealy et al. (2007). The bivariate correlation between abuse/ dependence symptoms and motivation to change was the strongest among the three types of alcohol-related problems, and social problems had the weakest bivariate correlation with motivation to change. Additionally, life satisfaction had statistically significant negative correlations with all three alcohol-related problems and motivation to change. To investigate these relationships further linear interactions and quadratic effects were estimated. The increased sample size in the current study relative to Shealy et al. allowed us to conduct a more powerful set of analyses aimed at determining how alcohol-related problems and life satisfaction predict motivation to change. First, we found an interaction between personal consequences and social consequences. Specifically, the relationship 1 Note, because there were two significant regions of the curve, two separate cut points were detected by the post-hoc probes. The first cut point delineates the significant region on the lower end of the abuse spectrum from the non-significant region in the middle of that same spectrum. The second delineates the significant region at the very top of the abuse spectrum from the non-significant region in the middle. For simplicity, we refer to these as the lower cut-point and the upper cut-point, respectively. To help characterize our findings, we also present the percentage of participants that reported abuse levels on the significant side of each cut-point (i.e., percentage below lower cut point, percentage above upper cut-point).
between personal consequences and motivation to change weakens as endorsement of social consequences increases, eventually becoming non-significant near the mean level of social consequences in our sample. This finding suggests that the relationship between social and personal problems and motivation to change cannot be understood in isolation from one another. Results from the two quadratic probes also replicate and expand on the results of the Shealy et al. (2007) study. For a majority of the sample, motivation to change was positively associated with abuse/dependence symptoms and negatively associated with life satisfaction. However, results also suggest that these effects did not apply to the more severe participants in the sample. That is, students with very high abuse/ dependence symptoms and very low life satisfaction did not experience continued increases in motivation to change, and students with the most severe abuse/dependence symptoms (3% of the sample) actually reported decreased motivation to change. These results suggest that the tendency for abuse/dependence symptoms and low life satisfaction to galvanize students and motivate them to change reaches a limit, and students at the greatest risk of developing alcohol abuse or dependence may be demotivated. There is a high degree of comorbidity between alcohol use disorders and other mental health conditions (Lo, Monge, Howell, & Cheng, 2013). Students with especially high levels of abuse/ dependence symptoms and low life satisfaction may view drinking as a means to cope with negative affect. Previous studies have shown that alcohol can result in a brief attenuation of stress (e.g. Solomon, 1980; Baker et al., 2004). However, if intoxication is repeatedly paired with an immediate decrease in negative affect, the negative affect may come to function as an establishing operation of consuming alcohol, resulting in increased motivation to drink when experiencing negative affect (Baker et al., 2004). Alcohol's role as a strategy for coping with negative affect may become more salient than the reported negative consequences, thus disrupting the relationship between problems and motivation to change found among the majority of college students. The complex relationship between abuse/dependence symptoms, life satisfaction and motivation to change warrants additional research, and the role of negative affect and coping motives as potential moderating influences could further illuminate the relationship. 4.1. Limitations The cross-sectional design of the study allows for the measurement of statistical associations among variables. However, due to motivation to change, life satisfaction, and alcohol-related problems being simultaneously measured, course of endorsement for each variable cannot be determined. A follow-up study that collects longitudinal data will need to be performed to confirm the sequence in which the variables may impact each other, as well as to confirm the findings in a second unique sample of students. Regarding the values (cut-point) along the moderator where effects transition from significant to non-significant relationships, they are meant to be probabilistic and may be sample specific. Future studies need to be conducted to determine how reliable the identified values are across samples, and a likely outcome is that a range of values will emerge that center around the actual transition point. Additionally, as this study utilized mandated students, additional research is needed to determine if similar models can be applied to more general student samples. Furthermore, the internal consistency estimates for the RAPI subscales were relatively low (.63–.68). Similar internal consistency estimates were reported in the initial factor analytic study (.68–.75) despite relatively strong factor structures (Martens et al., 2007), suggesting that the individual RAPI items may be independently measuring unique facets of the underlying latent variables. The use of measures with low reliability can have the effect of flattening curves, rather than creating illusory ones, and can inflate the likelihood of Type II errors (Shepperd, 1991). Thus, the true effects may actually be stronger and steeper than what was observed in the current study. Lastly, some of the findings were discussed in the context of clinical theory
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
815
A 11
Non-Sig. Region
Sig. Region 10
8
Abuse/Dependence Symptoms = 6.55
Inflect Point Abuse/Dependence Symptoms = 2.18
Pred. Motivation
9
Sig. Region
7 6 5 4 -2
0
2
4
6
8
Abuse/Dependence Symptoms (mean centered)
B 3
Non-Sig. Region
Sig. Region 2
Inflect Point
1 0 -1 -2 -3
Abuse/Dependence Sympsoms = 6.55
Abuse/Dependence Symptoms = 2.18
Slope of Abuse
Sig. Region
-4 -2
0
2
4
6
8
Abuse/Dependence Symptoms (mean centered) Fig. 2. Predicted motivation to change (Panel A) and simple slope of abuse/dependence symptoms (Panel B) by levels of reported abuse/dependence symptoms. All other covariates are controlled at their means. Note that dotted lines in Panel B represent the non-simultaneous 95% confidence bounds for the slope.
and research on the relationships among life satisfaction, alcoholrelated consequences and motivation to change being related to drinking for social reasons and to cope with stress and comorbid conditions. However, this study cannot confirm those relationships, as data on drinking motives were not collected. The model needs to be further tested in studies that include assessment of drinking motives. 4.2. Clinical implications Considering the time restrictions associated with brief alcohol interventions, it is valuable to study what may be most effective in increasing motivation to change in the greatest percentage of students. A recent review of studies assessing the content of personalized feedback interventions reported that 60% included a component designed to engage students in a discussion about the alcohol-related consequences they have experienced (Miller et al., 2012). Our results suggest that such a discussion may indeed be a useful tool for increasing motivation to change. However, clinicians should be aware that the relationship between negative consequences and motivation is neither linear nor uniform across consequences. Therefore, there may not be a “one size fits all” approach to determining how to maximize the effect negative consequences have on the perceived pros and cons of alcohol use.
For some negative consequences (i.e., personal and social), the relationship with motivation to change depends on other factors. Thus, interventions targeting personal alcohol-related consequences may be less effective if the individual is experiencing social consequences as well. For other negative consequences (i.e., low life satisfaction, abuse/dependence consequences), the relationship to motivation to change differs across levels of severity. Our results suggest that feedback on abuse/dependence symptoms and diminished life satisfaction may be therapeutic in increasing motivation to change for the majority of referred students. However, highlighting these symptoms may be demotivating for students with more severe impairment. As an alternative, or in addition to feedback interventions, students who report low life satisfaction and high abuse/dependence symptoms may benefit from a review of adaptive coping strategies. As noted, these students might be more likely to use alcohol to manage negative affect and experiences. If true, those students would fit a coping deficit model of alcohol use (Bandura, 1969); they may lack alternative methods of mitigating their distress and a discussion of adaptive and healthy coping strategies may be a more effective way of increasing their motivation to change. In conclusion, our results provide important information regarding the links between different types of alcohol-related problems and life satisfaction with motivation to change alcohol use among college
816
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817
A
10
Non-Sig. Region
Sig. Region
Inflect Point
8
7
LS= -3.49
Pred. Motivation
9
6
5 -10
-8
-6
-4
-2
0
2
4
6
Life Satisfaction (mean centered)
B 0.6
Non-Sig. Region
Sig. Region
0.4
0.0 -0.2 -0.4
LS = -3.49
Slope of SL
0.2
-0.6 -0.8 -1.0 -10
-8
-6
-4
-2
0
2
4
6
Life Satisfaction (mean centered) Fig. 3. Predicted motivation to change (Panel A) and simple slope (Panel B) of Life Satisfaction (LS) by levels of reported LS. All other covariates are controlled at their means. Note that dotted lines in Panel B represent the non-simultaneous 95% confidence bounds for the slope.
students. The results inform the theoretical discussion on motivation to change and can be directly applied to clinical work with college student drinkers. Role of funding sources This study was not externally funded. Auburn University had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.
Contributors Diulio and Correia designed the study and Correia wrote the protocol. Diulio conducted literature searches, provided summaries of previous research studies, conducted the statistical analysis and wrote the first draft of the manuscript. Cero and Witte conducted additional statistical analyses and contributed to subsequent draft of the manuscript. All authors contributed to and have approved the final manuscript.
Conflict of interest All authors declare that they have no conflicts of interest.
References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage Publications. Bandura, A. (1969). Principles of behavior modification. New York: Holt, Rinehart & Winston.
Baker, T. B., Piper, M. E., McCarthy, D. E., Majeskie, M. R., & Fiore, M. C. (2004). Addiction motivation reformulated: An affective processing model of negative reinforcement. Psychological Review, 111, 33–51. Barnett, N.P., Goldstein, A. L., Murphy, J. G., Colby, S. M., & Monti, P.M. (2006). ‘I'll never drink like that again’: Characteristics of alcohol-related incidents and predictors of motivation to change in college students. Journal of Studies on Alcohol, 67, 754–763. Budd, R. J., & Rollnick, S. (1996). The structure of the Readiness to Change Questionnaire: A test of Prochaska & DiClemente's transtheoretical model. British Journal of Health Psychology, 1, 365–376. Champoux, J. E., & Peters, W. S. (1987). Form, effect size and power in moderated regression analysis. Journal of Occupational Psychology, 60, 243–255. Collins, S. E., Carey, K. B., & Otto, J. M. (2009). A new decisional balance measure of motivation to change among at-risk college drinkers. Psychology of Addictive Behaviors, 23, 464–471. Collins, R., Parks, G. A., & Marlatt, G. (1985). 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, 53, 189–200. Cortina, J. M. (1994). Interaction, nonlinearity, and multicollinearity: Implications for multiple regression. Journal of Management, 19, 915–922. DiClemente, C. C., Doyle, S. R., & Donovan, D. (2009). Predicting treatment seekers' readiness to change their drinking behavior in the COMBINE Study. Alcoholism: Clinical and Experimental Research, 33, 879–892. Hayes, A. F., & Matthes, J. (2009). Computational procedures for probing interactions in OLS and logistic regression: SPSS and SAS implementations. Behavior Research Methods, 41, 924–936. Hirschberg, J., & Lye, J. N. (2005). Inferences for the extremum of quadratic regression models. http://dx.doi.org/10.2139/ssrn.918626 (Available at SSRN 918626). Lo, C. C., Monge, A. N., Howell, R. J., & Cheng, T. C. (2013). The role of mental illness in alcohol abuse and prescription drug misuse: Gender-specific analysis of college students. Journal of Psychoactive Drugs, 45, 39–47.
A.R. Diulio et al. / Addictive Behaviors 39 (2014) 811–817 Maddock, J. E., Laforge, R. G., Rossi, J. S., & O'Hare, T. (2001). The College Alcohol Problems Scale. Addictive Behaviors, 26, 385–398. Mallett, K. A., Bachrach, R. L., & Turrisi, R. (2008). Are all negative consequences truly negative? Assessing variations among college students' perceptions of alcohol related consequences. Addictive Behaviors, 33, 1375–1381. Mallett, K. A., Lee, C. M., Neighbors, C., Larimer, M. E., & Turrisi, R. (2006). Do we learn from our mistakes? An examination of the impact of negative alcohol-related consequences on college students' drinking patterns and perceptions. Journal of Studies on Alcohol, 67, 269–276. Martens, M. P., Neighbors, C., Dams-O'Connor, K., Lee, C. M., & Larimer, M. E. (2007). The factor structure of a dichotomously scored Rutgers Alcohol Problems Index. Journal of Studies on Alcohol and Drugs, 68, 597–606. McNally, A.M., & Palfai, T. P. (2001). Negative emotional expectancies and readiness to change among college student binge drinkers. Addictive Behaviors, 26, 721–734. Miller, M. B., Leffingwell, T. R., Claborn, K., Meier, E., Walters, S., & Neighbors, C. (2012). Personalized feedback interventions for college alcohol misuse: An update of Walters & Neighbors (2005). Psychology of Addictive Behaviors. http://dx.doi.org/10.1037/ a0031174. Molnar, D. S., Busseri, M.A., Perrier, C. P. K., & Sadava, S. W. (2009). A longitudinal examination of alcohol use and subjective well-being in an undergraduate sample. Journal of Studies on Alcohol and Drugs, 70, 704–713. Murphy, J. G., McDevitt-Murphy, M. E., & Barnett, N.P. (2005). Drink and be merry? Gender, life satisfaction, and alcohol consumption among college students. Psychology of Addictive Behaviors, 19, 184–191. Pavot, W., Diener, E., & Suh, E. (1998). The temporal satisfaction with life scale. Journal of Personality Assessment, 701, 340–354. Prochaska, J. O., & Diclemente, C. C. (1982). Transtheoretical therapy: Toward a more integrative model of change. Psychotherapy: Theory, Research and Practice, 20, 161–173.
817
Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1992). In search of how people change: Applications to addictive behaviors. American Psychologist, 47, 1102–1114. Read, J. P., Kahler, C. W., Strong, D. R., & Colder, C. R. (2006). Development and preliminary validation of the young adult alcohol consequences questionnaire. Journal of Studies on Alcohol, 67, 169–177. Rollnick, S., Heather, N., Gold, R., & Hall, W. (1992). Development of a short ‘readiness to change’ questionnaire for use in brief, opportunistic interventions among excessive drinkers. British Journal of Addiction, 87, 743–754. Shealy, A. E., Murphy, J. G., Borsari, B., & Correia, C. J. (2007). Predictors of motivation to change alcohol use among referred college students. Addictive Behaviors, 32, 2358–2364. Shepperd, J. A. (1991). Cautions in assessing spurious “moderator effects”. Psychological Bulletin, 110, 315–317. Solomon, R. L. (1980). The opponent-process theory of acquired motivation. American Psychologist, 35, 691–712. Substance Abuse and Mental Health Services Administration (2009). Results from the 2009 National Survey on Drug Use and Health Summary of National Findings. Office of Applied Studies, NSDUH Series H-38A, HHS Publication No. SMA 10-4586 Findings. Rockville, MD (Retrieved from http://oas.samhsa.gov/nsduh/2k9nsduh/2k9Results.pdf) Wechsler, H., Davenport, A., Dowdall, G., Moeykens, B., & Castillo, S. (1994). Health and behavioral consequences of binge drinking in college: A national survey of students at 140 campuses. Journal of the American Medical Association, 272, 1672–1677. Wechsler, H., Dowdall, G., Davenport, A., & Rimm, E. B. (1995). A gender-specific measure of binge drinking among college students. American Journal of Public Health, 85, 982–985. White, H., & Labouvie, E. (1989). Towards an assessment of adolescent problem drinking. Journal of Studies on Alcohol, 50, 30–37.
