DMQ-R

PR

OO

F

PR

OO

F

Fax +41 61 306 12 34E-Mail [email protected]

Eur Addict Res 404 DOI: 10.1159/0000XXXXX

Replication and Validation of the Drinking Motive Questionnaire Revised (DMQ-R, Cooper, 1994) Among Adolescents in Switzerland

Emmanuel Kuntsche

a Ronald Knibbe

b Gerhard Gmel

a Rutger Engels

c

a Research Department, Swiss Institute for the Prevention of Alcohol and Drug Problems, Lausanne ,

Switzerland; b Department of Medical Sociology, University of Maastricht, Maastricht , and

c Behavioural Science Institute, Radboud University Nijmegen, Nijmegen , The Netherlands

ness of the DMQ-R and its usefulness for assessing drink-ing motives among European adolescents from different cultural backgrounds.

Copyright © 2006 S. Karger AG, Basel

Introduction

Alcohol use is known to be determined by a large va-riety of factors that are found both within the individual (e.g. genetic disposition, personality characteristics, cog-nitions) and in his environment [e.g. factors on the level of the society, neighborhood, families, peer groups, drink-ing situations, see 1–3 , for reviews]. From both the con-ceptual point of view and in terms of prevention, the fac-tors most proximate to drinking are of strategic impor-tance. These factors are not only supposed to be more easily accessible for prevention efforts than most of the more distal ones, but also tend to refl ect or include distal factors such as culture, situation, or personality [4, 5] . Drinking motives are defi ned as the fi nal decision wheth-er to drink or not to drink and therefore the most proxi-mal factor for engaging in drinking [4–7] . In other words, drinking motives are the fi nal pathway to alcohol use, i.e. the gateway through which more distal infl uences, such as personality characteristics, are mediated [8–10] . On the other hand, however, drinking motives depend on

Key Words Drinking Motive Questionnaire Revised � Drinking motives � Measurement instrument � Alcohol use � Risky drinking

Abstract This paper has two aims: (1) to replicate the four-dimen-sional structure of the Drinking Motive QuestionnaireRevised (DMQ-R) in a national representative sample from Switzerland, and (2) to validate the relevance of distinguishing drinking motives by inspecting the rela-tions with expected consequences, alcohol use, and al-cohol-related and other problems. Confi rmatory factor analysis and linear structural equation models were es-timated based on answers of 5,617 8th to 10th graders (mean 15.1 years; SD = 0.95). The results confi rm the four-dimensional factor structure in general and among subpopulations defi ned by gender, age, and linguistic region. It could also be confi rmed that enhancement mo-tives followed by coping motives were strongly related to alcohol use and heavy drinking, whereas conformity motives were negatively related. Coping motives were related to problems independent of whether they were assessed as alcohol-related or not, while enhancement motives were associated solely with alcohol-related problems. In sum, the results demonstrate the robust-

European

AddictioncRe es ar h

Emmanuel KuntscheSwiss Institute for the Prevention of Alcohol and Drug Problems (SIPA)Research Department, PO Box 870CH–1001 Lausanne (Switzerland)Tel. +41 21 321 29 52, Fax +41 21 321 29 40, E-Mail [email protected]

© 2006 S. Karger AG, Basel1022–6877/06/0000–0000$23.50/0

Accessible online at:www.karger.com/ear

EAR404.indd 1EAR404.indd 1 16.12.2005 11:24:4216.12.2005 11:24:42

PR

OO

F

PR

OO

F

Kuntsche /Knibbe /Gmel /Engels

Eur Addict Res 404 2

past reinforcement from drinking [4, 5] and might also be used as justifi cation for a certain drinking style. From both a conceptual and practical point of view, the rele-vance of drinking motives for understanding adolescent drinking is probably underestimated, partly due to the fact that, until now, most studies using drinking motives have been restricted to North America [11] .

The central focus of this study is the measurement and relevance of drinking motives among adolescents in a multilingual European country. More specifi cally, it aims to replicate and validate a drinking motive instrument that was developed in the US and based on the motiva-tional model of alcohol use [4, 5] . This model assumes that drinking motives can be classifi ed along two underlying dimensions refl ecting the valence (positive or negative) and the source (internal or external) of the outcomes in-dividuals expect to achieve by drinking. Thus, individuals are supposed to drink to obtain positive outcomes (posi-tive reinforcement) or to avoid negative conse quences (negative reinforcement). In addition, they may be moti-vated by internal rewards such as enhancement of a de-sired internal emotional state or by external rewards such as social approval or acceptance. Crossing these two di-mensions results in four different drinking motive catego-ries: (a) internally generated, positive reinforcement mo-tives (drinking to enhance positive mood); (b) externally generated, positive reinforcement motives (drinking to obtain social rewards); (c) internally generated, negative reinforcement motives (drinking to reduce negative emo-tions), and (d) externally generated, negative reinforce-ment motives (drinking to avoid social rejection).

Based on these considerations, Cooper [7] developed the Drinking Motive Questionnaire Revised (DMQ-R) for use among adolescents that reliably measures the four drinking motive categories of enhancement (internal, positive), coping (internal, negative), social (external, positive), and conformity (external, negative) among 13- to 19-year olds in the USA. At present, the DMQ-R is the most often applied multidimensional instrument to assess drinking motives in the North America [for re-view, see 11] . To our knowledge, however, no study exists that has used the DMQ-R to assess drinking motives outside North America. However, as alcohol consump-tion and drinking patterns among adolescents vary wide-ly across cultures [e.g. 12] , it remains unclear if adoles-cent drinking motives follow the same patterns and are associated with the same alcohol outcomes as in the USA.

This papers has two aims: (1) to replicate the four-di-mensional structure of the DMQ-R in a national repre-

sentative sample from Switzerland, and (2) to validate the relevance of distinguishing drinking motives by inspect-ing the relations with expected consequences, alcohol use, and alcohol-related and other problems. A confi rmation of the DMQ-R factor structure and its potential to predict alcohol use and problems would confi rm the strategic im-portance of motives to understand adolescents’ drinking and to develop effective interventions to prevent harm caused by this behavior.

Apart from the confi rmation of the four-dimensional structure in a national representative sample, the present investigation seeks to determine the equivalence of the factor structure in different subpopulations. In her study among 13- to 19-year-olds in the USA, Cooper [7] found no differences in factor structure according to gender, age, and race. Likewise, we expect no structural differences according to gender and age in Switzerland. Additionally, the present sample of Swiss adolescents offers the oppor-tunity to study cultural differences in one country by comparing adolescents living in the German-speaking part with those living in the French- and Italian-speaking ones. Evidence of such a structural equivalence would indicate that the DMQ-R could be a sound and useful instrument to assess drinking motives in other European countries too.

The second aim was to validate the relevance of dis-tinguishing drinking motives by inspecting the relations with expected consequences, alcohol use, and problems attributed and not attributed to alcohol. Per defi nition, enhancement drinkers drink to enhance internal positive feelings, social drinkers drink to have fun in social con-text, and coping drinkers drink to alleviate problems. Ac-cordingly, to validate the relevance of drinking motives, it is hypothesized that enhancement motives are related to fun enhancement expectations, social motives to social improvement expectations, and coping motives to prob-lem-alleviating expectations [cf. 8, 9, 10, 13] .

Concerning alcohol use, the literature shows that the internally generated motives, enhancement and coping, are most strongly related to personality traits [enhance-ment with extraversion, coping with neuroticism, e.g. 14–17] . Thus, drinkers with internal motives will be more consistent across drinking situations in how they drink than drinkers with mainly external motives resulting in higher levels of alcohol consumption [7] . Since confor-mity drinkers drink only when they are motivated by the presence of drinking adolescents, they are supposed to have lower drinking levels than otherwise motivated drinkers. In this study, we hypothesize that enhancement and coping motives are more strongly associated with

EAR404.indd 2EAR404.indd 2 16.12.2005 11:25:1616.12.2005 11:25:16

PR

OO

F

PR

OO

F

DMQ-R Replication and Validation in Switzerland


usual frequency and quantity of alcohol use and to heavy drinking than social motives. We further expect that con-formity motives are negatively associated with these drinking indicators.

Since coping motives are defi ned by the decision to drink to cope with problems, we hypothesize that coping motives are related to problems in general [i.e. problems not attributed to alcohol such as poor academic perfor-mance, see e.g. 18] but also to alcohol-attributed prob-lems since defi cits in problem-focused coping have never been adequately addressed [9, 19] . Due to higher levels of alcohol use among enhancement drinkers, we expect enhancement motives to be related to alcohol-attributed problems such as involvement in violent acts and risky or regretted sexual intercourse.

Methods

Study Design Data from Swiss participation in the ‘European School Survey

Project on Alcohol and Drugs’, ESPAD [ 12] , which has been con-ducted every 4 years since 1995 in European countries, were used for the present analysis. In 2003, the Swiss Institute for the Preven-tion of Alcohol and Drug Problems (SIPA) and the Addiction Re-search Institute (ARI) for the fi rst time conducted the survey for Switzerland together with 34 other European countries.

Present data were collected by means of a paper-pencil ques-tionnaire which was sent to schools to be administered to each pu-pil in the relevant classes between the end of April and the end of June 2003. To avoid systematic dropouts, the exact date of the dis-tribution of the questionnaires was not communicated to the school boards ahead of time. Teachers who administered the question-naires in the classroom were advised only to respond to adolescents’ queries about the procedure and to guarantee the independent com-pletion of the questionnaire without interference from classmates. The time frame for fi lling out the questionnaires was one school lesson (about 45 min). According to the Helsinki Declaration [20] , the students could freely choose to participate and confi dentiality was ensured at all stages of the study.

Measures An interdisciplinary research group from the participating

countries developed the ESPAD questionnaire [12] including ex-pected personal consequences and usual frequency of alcohol use, 5+ drinking, and problems. Countries were free to include addi-tional questions to the ESPAD core questionnaire. In Switzerland, the Drinking Motive Questionnaire (DMQ-R) [7] and a question concerning the usual quantity of alcohol use was included. The re-sulting questionnaire was translated under the supervision of SIPA into the three languages most frequently spoken in Switzerland: German, French, and Italian.

Dinking Motives The Drinking Motive Questionnaire Revised [DMQ-R, 7] is a

20-item self-report measure of the relative frequency of drinking

for four conceptually and empirically distinct reason dimensions (i.e. enhancement, social, conformity, and coping motives). Par-ticipants were instructed to consider all the times they have drunk alcohol and to indicate how many o occasions they have drunk for each given motive. Each dimension consists of 5 items and is rated on a relative frequency scale ranging from ‘Never’ (coded as 1) to ‘Almost always’ (coded as 6). The exact wording of all items is given in table 1 .

Expected Personal Consequences of Alcohol Use The items stem from the original ESPAD questionnaire. The

question was ‘How likely is it that each of the following things would happen to you personally, if you drink alcohol?’ with the items ‘Have a lot of fun’, ‘Feel more friendly and outgoing’, and ‘Forget my problems’. The answer categories varied from ‘Very unlikely’ (coded as 1) to ‘Very likely’ (coded as 5).

Alcohol Use For the usual frequency of alcohol use, the questions asked the

number of drinking occasions in the last 30 days with answer cat-egories ‘0’, ‘1–2’, ‘3–5’, ‘6–9’, ‘10–19’, ‘20–39’ and ‘40 or more’. Midpoints of categories were used and 45 occasions for the upper category (40 times plus half range to mid-point of adjacent cate-gory).

Usual quantity of alcohol use assessed the total amount of stan-dard drinks of any alcoholic beverage (beer, wine, spirits, and alco-pops as examples) consumed at a typical occasion. The answer cat-egories were ‘less than 1 drink’, ‘1 drink’, ‘2 drinks’, ‘3 drinks’, ‘4 drinks’ and ‘5 or more drinks’. Midpoints of categories were used and 0.5 drinks for the lower and 5.5 drinks for the upper category (5 times plus half range to mid-point of adjacent category).

5+ Drinking . The question was ‘Think back once more over the last 30 days. How many times (if any) have you had fi ve or more drinks in a row?’ with the answer categories ‘None’, ‘1’, ‘2’, ‘3–5’, ‘6–9’ and ‘10 or more times’. Midpoints of categories were used and 11.25 occasions for the upper category (10 times plus half range to mid-point of adjacent category).

Drinking Problems (Alcohol-Attributed and Nonattributed in the ESPAD Core Questionnaire) Poor Academic Performance (Not Alcohol-Attributed). The stu-

dents were asked ‘Which of the following best describes your aver-age grade at the end of the last school term?’. The answers were coded 1 for the best possible grade to 6 for the worst and in this way enables an assessment of poor academic performance.

Violence and Sexual Intercourse (Alcohol-Attributed Questions). The question was ‘Have you ever had any of the following problems due to your alcohol consumption?’ The students could indicate if they ever experienced a ‘Scuffl e or fi ght’, ‘Damage to objects or clothing you owned’, and ‘Victimized by robbery or theft’. These three questions were added up to a violence summary score (value range: 0 to 3). The two questions ‘Engaged in sexual intercourse you regretted the next day’ and ‘Engaged in sexual intercourse without a condom’ were added to a risky sexual intercourse summary score (value range: 0 to 2).

Sample and Missing Value Imputation Random cluster sampling was used, based on a list of all classes

of Swiss schools from 8th to 10th grade compiled by the Swiss Fed-eral Statistical Offi ce, where the classes served as the primary sam-

EAR404.indd 3EAR404.indd 3 16.12.2005 11:25:1616.12.2005 11:25:16

PR

OO

F

PR

OO

F



pling unit. An overall response rate of 83.1% was reached. The sample can be considered as representative for all 8th, 9th and 10th graders in public schools in the German-, French-, and Italian-speaking regions of Switzerland. Since drinking motives were ex-clusively assessed among drinkers, those who did not indicate at least one drinking occasion in the last 12 months (n = 1,415, 19.7%) were excluded. When a student did not answer one or two questions on drinking motives, the missing values were replaced by means of Markov Chain Monte Carlo (MCMC) estimates [21, 22] . MCMC is a simulation method in the Bayesian tradition. Starting with the prior distribution of observed values, random draws of missing values conditional on an item and individual parameters of ob-served values are taken. The individual missing value is imputed by randomly selecting a value from this conditional distribution for the given individual and his/her observed item response pattern. The procedure is iterative and uses Markov Chains until conver-gence of the posterior distribution. The advantage of such an im-putation method is that the information of observed values for an individual is taken into account, i.e. imputation is conditional on individuals that have the same response pattern on all but the miss-ing items. Thus, a maximum of information for an individual from other items of the drinking motive questionnaire was used. The program LISREL 8.51 [23] was applied for missing value imputa-tion.

All students who failed to answer more than two questions on drinking motives (n = 71; 1.2%) were excluded from the analysis. All students who failed to answer questions about expected per-sonal consequences and alcohol use and related consequences were also excluded (n = 90; 1.6%). The analyzed data consists of 5,617 12- to 18-year-old students (40.7% 8th graders, 45.9% 9th graders, and 13.4% 10th graders) of which 49.0% were boys and 70.9% came from the German-speaking part (22.3% French-speaking and 6.7% Italian-speaking). The mean age of the total sample was 15.1 years (SD = 0.95). More information about the ESPAD survey in Swit-zerland can be found in Gmel et al. [24] or in the relevant chapter of the international report [12] .

Statistical Analysis A confi rmatory factor analysis [CFA, 25, 26] was used to confi rm

the four-dimensional factor structure [enhancement, social, confor-mity, and coping motives, 7] in the present sample of 12- to 18-year-old students in Switzerland. Errors were allowed to correlate so as to compensate answer tendencies and sub-dimensional item rela-tions previously detected by an exploratory factor analysis (EFA) using a randomly selected sub-sample. To evaluate the overall mod-el fi t, the comparative fi t index (CFI), the non-normed fi t index (NNFI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR) were used.

Table 1. Results of a confi rmatory factor analysis (standardized item loadings), item means, inter-factor correlations, and internal con-sistencies

How often do you drink Enhancement Social Conformity Coping Means 8 SD

Because you like the feeling? 0.76 2.8281.55Because it’s exciting? 0.67 2.6181.49To get high? 0.67 2.2381.56Because it gives you a pleasant feeling? 0.81 2.6281.52Because it’s fun? 0.76 3.4181.75Because it helps you enjoy a party? 0.78 2.7281.58To be sociable? 0.39 1.5881.03Because it makes social gatherings more fun? 0.84 3.0181.68Because it improves parties and celebrations? 0.82 2.8881.70To celebrate a special occasion with friends? 0.57 3.8681.58Because your friends pressure you to drink? 0.56 1.3580.82So that others won’t kid you about not drinking? 0.67 1.3380.88You drink to fi t in with a group you like? 0.74 1.3280.85To be liked? 0.78 1.3580.86So you won’t feel left out? 0.82 1.3980.94To forget your worries? 0.79 1.9181.31Because it helps you when you feel depressed or nervous? 0.83 1.9181.30To cheer up when you’re in a bad mood? 0.82 2.0581.36Because you feel more self-confi dent or sure of yourself? 0.53 1.7781.21To forget about your problems? 0.83 1.9181.40Correlation with the factor ‘Social’ 0.92Correlation with the factor ‘Conformity’ 0.29 0.37Correlation with the factor ‘Coping’ 0.59 0.54 0.45Internal consistencies (Cronbach’s �) 0.85 0.82 0.87 0.88

All factor loadings are signifi cant at the 0.1% error level; model fi t: CFI = 0.920, NNFI = 0.897, RMSEA = 0.062, SRMR = 0.079.

EAR404.indd 4EAR404.indd 4 16.12.2005 11:25:1716.12.2005 11:25:17

PR

OO

F

PR

OO

F



The CFI and the NNFI are related to the total variance accounted by the model and values close to 1, e.g. higher than 0.90, are desired [26–28] . The RMSEA and the SRMR are related to the residual variance and values close to 0, e.g. smaller than 0.08, are desired [26, 28, 29] . Internal consistencies of each drinking motive dimen-sion were assessed by using Cronbach’s alpha.

To determine the equivalence of factor structure in different subgroups, the four-dimensional measurement model was fi rst es-timated freely between males and females, older (15–18 years) and younger students (12–14 years, median split), as well as between German-speaking and French/Italian-speaking students (since the latter are both minorities in Switzerland and have rather similar cultural backgrounds, these students were grouped together and compared with their German-speaking peers). Subsequently, the factor loadings were set as equal between the groups and the CFI, the NNFI, the RMSEA and the SRMR of the fi xed models were compared with the fi t indices obtained in the freely estimated models.

To confi rm the validity of the DMQ-R in the present sample of Swiss adolescents, three multivariate linear structural equation models (LSEM) were estimated. In the fi rst model, the three per-sonal expected consequences of alcohol use were the dependent vari-ables; the three alcohol use measures in the second model; and the three drinking problems in the third model. In all models, the latent variables of enhancement, social, conformity, and coping motives as well as gender and age were the independent variables.

EFA, CFA, and LSEM were performed using the software Mplus 3.11 [25] . This program has the advantage of directly adjust-ing the analysis for the sampling design effect of clusters (school classes).

Results

Replication of the Four-Dimensional Factor Structure The results of the Confi rmatory Factor Analysis re-

vealed high and homogenous item loadings on their spec-ifi ed factors ( table 1 ). One exception is the item ‘drinking to be sociable’ with a rather low item loading. The means of enhancement and social items were generally higher than those of conformity and coping. Lowest scores were found for the factor conformity. Exceptions are ‘drinking to be sociable’ with a rather low mean and drinking to cheer up when in a bad mood with a rather high mean. Items with the highest scores concerned drinking to cele-brate special occasions, because it is fun, and because it makes social gatherings more fun. The drinking motive factors are highly correlated. The highest correlations were found between enhancement and social motives followed by enhancement and coping. The fi t of the model turned out to be satisfactory with CFI and NNFI values stand-ing at around 0.90; the RMSEA value is below 0.08 and SRMR value is about 0.08. Internal consistencies of the drinking motive dimensions varied from � = 0.82 to � = 0.88.

To assess if the four-dimensional model of drinking motives is also valid for different subgroups, the confi r-matory factor analysis was estimated for gender and age groups and linguistic regions separately. Subsequently, the factor loadings in each model were fi xed to be equal between the groups. In all three conditions, allowing the parameters to vary between the groups did not increase the model fi t considerably ( table 2 ).

Relations to Expected Consequences and Alcohol Use and Related Problems For validation purposes, a linear structural equation

model was estimated with the four drinking motives as independent variables and different expected personal consequences of alcohol use as dependent variables. The results reveal that enhancement and social motives were positively related to the opinion that fun is likely to result from drinking whereas conformity and coping motives were negatively related to this opinion ( table 3 ). Social motives in particular were related to the opinion that feeling more friendly and outgoing is a likely consequence from drinking. Coping motives were positively related and conformity motives negatively related to the opinion that drinking alcohol is likely to help to forget problems. The explained variance of expected personal con-sequences varied from 19 to 44%.

To determine associations with different alcohol use measures, a second linear structural equation model was estimated. Enhancement and coping motives were posi-tively related to the usual frequency of alcohol intake. Enhancement and social motives were positively related and conformity motives negatively related to the usual quantity. The frequency of having fi ve drinks or more on single occasions was signifi cantly linked to enhancement

Table 2. Model fi t according to gender, age, and linguistic region

CFI NNFI RMSEA SRMR

Gender (boys vs. girls)Unconstrained model 0.915 0.896 0.063 0.081Fixed model 0.915 0.901 0.061 0.082

Age groups (12–14 vs. 15–18)Unconstrained model 0.917 0.898 0.062 0.081Fixed model 0.918 0.904 0.060 0.081

Linguistic region (German- vs. French/Italian-speaking)Unconstrained model 0.909 0.888 0.066 0.083Fixed model 0.907 0.892 0.065 0.084

EAR404.indd 5EAR404.indd 5 16.12.2005 11:25:1716.12.2005 11:25:17

PR

OO

F

PR

OO

F



and coping motives. The explained variance of alcohol use varied from 16 to 37%.

A third model was estimated for problems. In this model, only coping motives were positively related to all three problems. Enhancement motives were positively as-sociated with alcohol-related violence and sexual prob-lems. For social and conformity motives, no relation to drinking problems was found. The explained variance was between 5 and 7%.

Discussion

The fi rst aim of the present study was to confi rm the structure of the DMQ-R in a sample of adolescents out-side North America. The results of the confi rmatory fac-tor analysis revealed that the specifi ed four-factor model had an acceptable model fi t [e.g. 26–29] . Additionally, each drinking motive dimension had a good-to-excellent internal consistency, comparable to those reported in the Cooper study [7] . Also on the level of how strong items load on the motive dimensions, there is a remarkable sim-ilarity between the two studies. For example, in both stud-ies the items ‘makes social gatherings more fun’ and ‘im-proves parties and celebrations’ had the highest loadings on the factor ‘social motives’ and the item ‘be sociable’ had the lowest. Furthermore, the inter-factor correlations were similar. Both in our study and in that by Cooper,

the highest correlation was found between enhancement and social motives and the lowest between enhancement and conformity motives.

The present study also aimed to determine the struc-tural equivalence in different subgroups. When fi xing the factor loadings between the subgroups no substantial dif-ferences in the different fi t indices emerged. Since the NNFI takes degrees of freedom into account, the gain of degrees of freedom when fi xing the factor loadings be-tween subgroups resulted in a very slight increase in NNFI values. Taken together, the four-dimensional struc-ture of the DMQ-R was found to be invariant not only among males and females and among older and younger adolescents but also among adolescents living in both the German-speaking part and those living in the French- and Italian-speaking parts. This not only confi rms the results Cooper found among adolescents in the USA more than 10 years ago but also demonstrates the robustness of the DMQ-R and its usefulness to assess drinking motives among other European adolescents from different cul-tural backgrounds.

Concerning the validation of drinking motives, the re-sults confi rm that enhancement but not conformity or coping motives are related to fun enhancement expecta-tions. Social motives are especially related to social im-provement expectations, while coping but not conformi-ty motives are related to problem-alleviating expectations [cf. 8–10, 13] . Equally consistent are the results that en-

Table 3. Drinking motives as predictors of expected personal consequences, alcohol use, and drinking problems (standardized betas of the three multiple-multivariate linear structural equation models estimated)

Enhancement Social Conformity Coping R2, %

Expected personal consequences1

Have a lot of fun –0.43*** –0.32*** –0.12*** –0.08*** 43.9Feel more friendly and outgoing –0.01 –0.35*** –0.07*** –0.09*** 19.1Forget my problems –0.02 –0.06 –0.12*** –0.64*** 37.4Alcohol use2

Usual frequency –0.25*** –0.06 –0.06** –0.12*** 16.4Usual quantity –0.34*** –0.25*** –0.15*** –0.06** 37.55+ drinking –0.37*** –0.00 –0.05 –0.15*** 23.0Drinking problems3

Academic4 –0.05 –0.09 –0.03 –0.19*** 5.0Violence5 –0.17** –0.02 –0.00 –0.12*** 7.1Sexual5 –0.19** –0.06 –0.01 –0.12*** 5.1

All models were adjusted for gender and age; model fi t: 1 CFI = 0.914, RMSEA = 0.058; 2 CFI = 0.918, RMSEA = 0.056; 3 CFI = 0.916, RMSEA = 0.054; 4 non-alcohol attributed; 5 alcohol attributed. * p < 0.05, ** p < 0.01, *** p < 0.00.

EAR404.indd 6EAR404.indd 6 16.12.2005 11:25:1716.12.2005 11:25:17

PR

OO

F

PR

OO

F



hancement motives are related to all drinking measures used in the study, that conformity motives are negatively related to frequency and quantity of drinking, and that coping motives were particularly related to 5+ drinking [cf. 7, 9, 10] . These results are not only consistent with previous research but also confi rm the relevance of dis-tinguishing these four categories of drinking motives in predicting expected consequences and different alcohol use patterns.

In this study, we distinguished between alcohol-attrib-uted and nonattributed problems. The results confi rm that coping motives were related to problems indepen-dent of whether they were assessed to be alcohol related or not [7, 9, 10, 18, 19] . It might be the case that adoles-cents who have problems in general, such as poor aca-demic performance and who drink to cope with these problems, end up with other problems such as involve-ment in violent acts and risky or regretted sexual inter-course due to their alcohol consumption because defi cits in problem-focused coping have never been adequately addressed [9, 19] . However, due to the cross-sectional data used in the present study, such a causal succession in time could not be addressed. Additionally, the results show that enhancement motives were related to alcohol-attributed problems but not to nonattributed ones. It seems that enhancement drinkers do not have problems in general, but have problems related to their high levels of alcohol consumption; this is confi rmed by the strong association between enhancement motives and the differ-ent alcohol use measures.

Another interesting point concerns the differences be-tween enhancement and social motives. Although both dimensions are highly correlated, they are indeed differ-ently associated with most of the dependent variables in-vestigated. This was the case because in multiple regres-sion models all independent variables are mutually ad-justed. In other words, the coeffi cients indicate the increase e.g. in alcohol use or in drinking problems, as-suming all other variables are equal. Thus, accounting for a particular level of social motives, enhancement motives were associated with 5+ drinking. On the other hand, when a particular level of enhancement motives is fac-tored in, no association with social motives can be shown. However, in bivariate analyses (results not presented), social motives were additionally related to drinking fre-quency, 5+ drinking and alcohol-attributed problems be-cause enhancement motives are ‘hidden’ due to the high correlation of both. This demonstrates the importance of analyzing the different motive dimensions in a mutually adjusted way to arrive at the specifi c effect of a particular

motive dimension on alcohol use, drinking problems and other variables.

To conclude, the present study provides evidence for the usefulness of the DMQ-R among adolescents in Swit-zerland and similar drinking cultures in Europe. The re-sults not only confi rmed the four-dimensional factor structure in general but also among boys and girls, among 12- to 14-year-olds and 15- to 18-year-olds, and among adolescents living in the German-speaking part and those living in the French- and Italian-speaking parts of Swit-zerland. Additionally, the results are very consistent with those obtained in a sample of adolescents in the USA more than 10 years ago [7] . This makes us optimistic that the DMQ-R, which is the most frequently applied multi-dimensional questionnaire to assess drinking motives among adolescents in North America [for a review, see 11] , is a valid and reliable instrument to assess adolescent drinking motives in other European countries.

Apart from their implication for research, the present results should be important for prevention. The motiva-tional model of alcohol use that was originally developed to understand goals and incentives among alcoholics [5] was also shown to be useful to improve the motivation-al structure of young people and to reduce their sub-stance use via motivational counseling [30] . Current de-velopments in prevention and early intervention further include attention to the interplay between personality and motivational factors underlying risky alcohol use in adolescents [31] . The present validation of the DMQ-R in a multi-linguistic sample is therefore supposed to con-tribute to identify adolescents at risk from alcohol prob-lems in Europe and to design preventive strategies ac-cordingly.

Acknowledgments

The Swiss ESPAD study was funded by the Swiss Federal Offi ce of Public Health (Grant No. 02.000537), SIPA, and ARI. Addi-tional funds for the analysis of drinking motives were obtained from the Swiss National Science Foundation (contract number 101511-105481/1). The authors would like to thank Esther Grich-ting and Jürgen Rehm for their participation in the Swiss ESPAD project.

EAR404.indd 7EAR404.indd 7 16.12.2005 11:25:1716.12.2005 11:25:17

PR

OO

F

PR

OO

F



References

1 Ham LS, Hope DA: College students and prob-lematic drinking: a review of the literature. Clin Psychol Rev 2003; 23: 719–759.

2 Hawkins JD, Catalano RF, Miller JY: Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: implications for substance abuse prevention. Psychol Bull 1992; 112: 64–105.

3 Kuntsche EN, Rehm J, Gmel G: Characteris-tics of binge drinkers in Europe. Soc Sci Med 2004; 59: 113–127.

4 Cox WM, Klinger E: A motivational model of alcohol use. J Abnorm Psychol 1988; 97: 168–180.

5 Cox WM, Klinger E: Incentive motivation, af-fective change, and alcohol use: a model; in Cox WM (ed): Why People Drink – Parameters of Alcohol as a Reinforcer. New York, Gardner Press, 1990, pp 291–314.

6 Carpenter KM, Hasin DS: Reasons for drink-ing alcohol: relationships with DSM-IV alco-hol diagnoses and alcohol consumption in a Community Sample. Psychol Addict Behav 1998; 12: 168–184.

7 Cooper ML: Motivations for alcohol use among adolescents: development and valida-tion of a four-factor-model. Psychol Assessm 1994; 6: 117–128.

8 Catanzaro SJ, Laurent J: Perceived family sup-port, negative mood regulation expectancies, coping, and adolescent alcohol use: evidence of mediation and moderation effects. Addict Be-hav 2004; 29: 1779–1797.

9 Cooper ML, Frone MR, Russell MA, Mudar P: Drinking to regulate positive and negative emotions: a motivational model of alcohol use. J Pers Soc Psychol 1995; 69: 990–1005.

10 Cronin CJ: Reasons for drinking versus out-come expectancies in the prediction of college student drinking. Subst Use Misuse 1997; 32:

1287–1311.

11 Kuntsche EN, Knibbe RA, Gmel G, Engels RCME: Why do young people drink? A review of drinking motives. Clin Psychol Rev 2005;

25: available http://dx.doi.org/doi:10.1016/j.cpr.2005.06.002 .

12 Hibell B, Andersson B, Bjarnason T, Ahlström S, Balakireva O, Kokkevi A, Morgan M: The ESPAD Report 2003 – Alcohol and Other Drug Use Among Students in 35 European Countries. Stockholm, The Swedish Council for Information on Alcohol and Other Drugs, CAN Council of Europe, Co-operation Group to Combat Drug Abuse and Illicit Traffi cking in Drugs (Pompidou Group), 2004.

13 Read JP, Wood MD, Kahler CW, Maddock JE, Palfai TP: Examining the role of drinking mo-tives in college student alcohol use and prob-lems. Psychol Addict Behav 2003; 17: 13–23.

14 Cooper ML, Agocha VB, Sheldon MS: A mo-tivational perspective on risky behaviors: the role of personality and affect regulatory pro-cesses. J Pers 2000; 68: 1058–1088.

15 Loukas A, Krull JL, Chassin L, Carle AC: The relation of personality to alcohol abuse/depen-dence in a high-risk sample. J Pers 2000; 68:

1153–1175. 16 Stewart SH, Devine H: Relations between per-

sonality and drinking motives in young adults. Pers Individ Diff 2000; 29: 495–511.

17 Stewart SH, Loughlin HL, Rhyno E: Internal drinking motives mediate personality domain – drinking relations in young adults. Pers Indi-vid Diff 2001; 30: 271–286.

18 Windle M, Windle RC: Coping strategies, drinking motives, and stressful life events among middle adolescents: associations with emotional and behavioral problems and with academic functioning. J Abnorm Psychol 1996; 105: 551–560.

19 Kassel JD, Jackson SI, Unrod M: Generalized expectancies for negative mood regulation and problem drinking among college students. J Stud Alcohol 2000; 61: 332–340.

20 World Medical Association (WMA): World Medical Association Declaration of Helsinki. Ethical Principles for Medical Research In-volving Human Subjects. Ferney-Voltaire, WMA, 2002.

21 Congdon P: Applied Bayesian Modelling. Chichester, Wiley, 2003.

22 Hox JJ (ed): Multilevel Analysis: Techniques and Applications. Mahwah, Lawrence Erl-baum Associates, 2002.

23 Jöreskog KG, Sörbom D: LISREL Software Version 8.51. Chicago, Scientifi c Software In-ternational, 2001.

24 Gmel G, Rehm J, Kuntsche EN, Wicki M, Grichting E: Das European School Survey Project on Alcohol and Other Drugs (ESPAD) in der Schweiz. Wichtigste Ergebnisse 2003 und aktuelle Empfehlungen. Lausanne/Zürich, Schweizerische Fachstelle für Alkohol- und an-dere Drogenprobleme/Institut für Suchtfor-schung. Also available at http://www.sfa-ispa.ch/DocUpload/ESPAD.pdf , 2004.

25 Muthén LK, Muthén BO: Mplus: Statistical Analysis with Latent Variables – User’s Guide. Los Angeles, Muthén & Muthén, 2004.

26 Ullman JB: Structural equation modeling; in Tabachnick BG, Fidell LS (eds): Using Multi-variate Statistics. Boston, Allyn & Bacon, 2001, pp 653–771.

27 Bentler PM, Bonett DG: Signifi cance tests and goodness of fi t in the analysis of covariance structures. Psychol Bull 1980; 88: 588–606.

28 Hu L, Bentler PM: Cut-off criteria for fi t in-dexes in covariance structure analysis: conven-tional criteria versus new alternatives. Struct Equat Mod 1999; 6: 1–55.

29 Browne MW, Cudeck R: Alternative ways of assessing model fi t; in Bollen KA, Long JS (eds): Testing Structural Models. Newbury Park, Sage, 1993.

30 Cox WM, Klinger E: Motivational structure: relationships with substance use and processes of change. Addict Behav 2002; 27: 925–940.

31 Stewart SH, Conrod PJ, Marlatt GA, Comeau MN, Thush C, Krank M: New developments in prevention and early intervention for alco-hol abuse in youths. Alcohol Clin Exp Res 2005; 29: 278–286.

EAR404.indd 8EAR404.indd 8 16.12.2005 11:25:1716.12.2005 11:25:17

DMQ-R

Documents

Transcript of DMQ-R