Factor structure of the Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES) in alcohol dependent outpatients

Abstracts

OBJECTIVE: The aim of this study was to investigate the reliability and factor structure of the Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES), version 8,¹ a 19-item self-reported instrument developed to measure readiness to change in alcohol-dependent alcoholics. METHODS: A Confirmatory Factor analysis of the SOCRATES was performed based on the factor structures previously demonstrated by Miller & Tonigan² and Maisto et al.³ in a sample with 326 alcohol-dependent outpatients. The questionnaire was translated into Portuguese, cross-culturally adapted and back-translated into English. During this process SOCRATES underwent some modifications to simplify some complex question formats. RESULTS: The analysis showed that two correlated factors provided the best fit for the data and that these were similar to Maisto et al.'s³ factors. CONCLUSIONS: There was less evidence to support a three-factor structure. The results are compared to previous studies and the reasons for discrepancies are discussed.

Alcoholism; Motivation; Factor analysis, statistical; Validation studies

OBJETIVO: O objetivo deste estudo foi o de investigar a confiabilidade e a estrutura fatorial da Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES), versão 8,¹ instrumento com 19 itens que mensura a prontidão para a mudança em dependentes de álcool. MÉTODOS: Uma análise fatorial confirmatória da SOCRATES foi realizada em uma amostra de 326 dependentes de álcool, tratados ambulatorialmente, tendo como base a estrutura fatorial demonstrada por Miller & Tonigan² e Maisto et al.³ O questionário foi traduzido e adaptado culturalmente para o idioma português, sendo posteriormente submetido ao procedimento da retradução para o idioma inglês. Durante esse procedimento, foram realizadas algumas modificações, visando a simplificar alguns itens que apresentaram formato complexo. RESULTADOS: As análises estatísticas mostraram a existência de dois fatores correlacionados que melhor exploraram o modelo, sendo este achado similar ao estudo de Maisto et al.³ CONCLUSÕES: Foi constatada menor evidência para o modelo de três fatores. Esses resultados são comparados com estudos prévios e as discrepâncias são discutidas neste artigo.

Alcoolismo; Motivação; Análise fatorial; Estudos de validação

ORIGINAL ARTICLE

Factor structure of the Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES) in alcohol dependent outpatients

Neliana Buzi Figlie^I; John Dunn^II; Ronaldo Laranjeira^I

^IUnidade de Pesquisa em Álcool e Drogas (UNIAD), Departamento de Psiquiatria, Universidade Federal de São Paulo, SP, Brasil

^IIRoyal Free & University College Medical School, London

^{Correspondence}

ABSTRACT

OBJECTIVE: The aim of this study was to investigate the reliability and factor structure of the Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES), version 8,¹ a 19-item self-reported instrument developed to measure readiness to change in alcohol-dependent alcoholics.

METHODS: A Confirmatory Factor analysis of the SOCRATES was performed based on the factor structures previously demonstrated by Miller & Tonigan² and Maisto et al.³ in a sample with 326 alcohol-dependent outpatients.

The questionnaire was translated into Portuguese, cross-culturally adapted and back-translated into English. During this process SOCRATES underwent some modifications to simplify some complex question formats.

RESULTS: The analysis showed that two correlated factors provided the best fit for the data and that these were similar to Maisto et al.'s³ factors.

CONCLUSIONS: There was less evidence to support a three-factor structure. The results are compared to previous studies and the reasons for discrepancies are discussed.

Keywords: Alcoholism. Motivation. Factor analysis, statistical. Validation studies.

Introduction

The Transtheoretical Model of Stages of Change, described by Prochaska and DiClemente,⁴ proposes a general and comprehensive explanation of the way people change their addictive behaviours through progressive and sequential steps (from pre-contemplation, to contemplation, determination, action and, finally, maintenance). Within this framework motivation manifests itself through stage-specific behaviours displayed by that patients exhibit.

Miller⁵ defined motivation as the '"probability that a person will enter into, continue, and adhere to a specific change strategy'". Motivation is seen as a state of readiness or eagerness to change, which may fluctuate from one time or situation to another and can be understood as an internal state influenced by external factors. The concept of motivation has been extensively studied in the substance misuse field. ^6,7,8,5,9 For alcohol misuse, the patient’s stage allocation and readiness to change at the screening treatment entry have been shown to predict measures of treatment outcome, such as alcohol consumption and time for theto first drink.^10,11,12

Miller developeddrafted The Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES) with the aiming toof assessing motivation to change in problem drinkers. Initially Aa 32-item version was initially developed with the item content specifically focused on problem drinking. The aim was to measure the five principal stages of change: pre-contemplation, contemplation, determination, action and maintenance.^13,14 Since itsthe initial draft, the questionnaire has undergone extensive psychometric testing and has been substantially modified regarding in relation to item content and the total number of items. The latest version of the questionnaire has 19 items and has been shown to consistists of three orthogonal factors, which have been named Taking Steps, Recognition and Ambivalence.²

Recently Maisto et al.,³ has recently studied the factor structure of the 19-item SOCRATES in a sample of primary care patients and found only two factors. One factor was they called AMREC (a combination of ambivalence and recognition) and consisted of nine items and the second, Taking Steps, had six items and corresponded to the factor of the same name described by Miller and Tonigan.² A confirmatory factor analysis suggested that this two-factor structure fits the data better than the 3-factor structure proposed by Miller and Tonigan.² Another study with adolescents presenting for treatment of alcohol use disorder identified two factors called taking steps and recognition.¹⁵

The aim of the present study was to examine the reliability and factor structure of the Portuguese version of the SOCRATES 8¹ among alcohol- dependent patients and to investigate whether a two- or three-factor structure fits the data better.

Method

1. Setting

The study was performed at the São Paulo Hospital São Paulo/Federal University of São Paulo, a federally funded public teaching hospital. Two clinics were used: a gastroenterology clinic and a specializedst alcohol abuse/dependence treatment clinic. We deemedconsidered it important to havinge patients in the sample who couldmayight display very different levels of motivation as in relation to seeking help for their alcohol problem, hence the choice of the two settings. IThe inclusion criteria were as follows:

1) At the specialist alcohol abuse/dependence treatment clinic: all patients who sought help for alcohol- related problems and who scored mild, moderate or severe on the Short-Form Alcohol Dependence Data Questionnaire (SADD).¹⁶

2) At the gastroenterologicy clinic: all patients were screened with the AUDIT ¹⁷ and those with scoring positive score (score >8) were interviewed with the Short-Form Alcohol Dependence Data Questionnaire (SADD).¹⁶ Thereafter, the same criteria were applied.

During the 21-month study period, of the 336 patients who presented to the gastroenterologicy clinic, of whom 158 (47%) scored positive on the AUDIT.¹⁷

The exclusion criteria were: patients who abuseding of substances other than alcohol and patients who were confused. Women were also excluded from the study. As there were so few women presenting to either service, weit was thought better to exclude them as gendersex could have been a confounder in subsequent analyses.

2. Subjects

One hundred and fifty-one patients were interviewed from the gastroenterology clinic and 175 from the specializedst abuse/dependence alcohol treatment clinic were interviewed, giving a totalling of 326 subjects. Five patients refused to take part in the study. The interviews were conducted atin the outpatient clinics at the first appointment.

The demographic characteristics of the sample are shown in Table 1 and their patterns patterns of of alcohol consumption and biological markers are shown in Table 2.

3. Measures:

An experienced psychologist conducted the structured interviews were conducted by an experienced psychologist. Cards with answer options were used to facilitate patients' responses. The interview consisted of the following:

1) Demographic data: age, gendersex, race, marital status, schoolingeducation, occupation, and family income.

2) The Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES) Version 8¹: to measure the stages of change in relation to Recognition, Ambivalence and Taking Steps. The version 8 has a different numbering system for the items. For the sake of clarity we have used the original numbering system described by Miller & Tonnigan² for all three studies.

3) Short-Form Alcohol Dependence Data Questionnaire (SADD): this scale measures the severity of alcohol dependence (mild, moderate and severe). The Portuguese version was developed by Jorge and Masur.¹⁶

4) Pattern of alcohol consumption: it was documented using the interview schedule developed for the WHO/ISBRA Study on State and Trait Markers in Alcoholism.¹⁸ The questions concern the pattern of alcohol consumption regardingin relation to the type of drink, the quantity and frequency of use in the previouslast 30 days as well as the heaviest period of lifetime alcohol consumption. One unit of alcohol was considered astaken to be equivalent to 10ml of pure alcohol.

5) Biological markers: aspartate aminotransferase (AST), alanine aminotransferase (ALT) and gamma glutamyltransferase (GGT). In the gastroenterologicy clinic these tests were routine, thusso most patients have results, however, in the specializedst alcohol abuse/dependence clinic fewer patients had these tests thusso there are more missing values.

3.1 - Process of translation and cultural adaptation of the SOCRATES

The scale was translated by two researchers with a good command mastering of the English language, resulting in two translations that were each tested with a group of ten alcoholics, with the aiming to of discovering which form was most easily understood by subjects. This process resulted in a second draft which . This version was reapplied into another group of ten alcoholics, who were also asked to explain the meaning of the questions ands well as to answer them. The objectiveaim was to check that both patient and interviewer agreed on what the questions actually meant. For each item respondents use a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).

More complex questions had to be simplified as patients tended to answer only one part of the question without reference to the second half. For example, question 20: '"I have made some changes in my drinking and I want some help to keep from going back to the way I used to drink.'" Patients tended to answer this question either by agreeing or disagreeing that they had made changes to their drinking or agreeing or disagreeing that they wanted help to prevent returning keep them from going back tto the way how they used to drink. To further complicate the issuematters, if the patients had not made any changes into his or hertheir drinking, then by definition theyhe or she could not answer the question. Our solution to this problem was to split the question into its chronological and sequential components, first: '"I have made changes to my drinking habit.'" If the patients answeredsaid that they had not made any changes, then they were automatically assigned to the strongly disagree category. If they answered in the affirmatively, the second component of the question was put to them: '"I want help to keep from going back to the way I used to drink'", to which they could chose respond from strongly disagree to strongly agree. It can be argued that this is not a satisfactory solution, as patients may still want help to stop drinking, even if they do not want the specific type of help stated in the question. Short of excluding these three questions altogether, we felt that it was the best compromise. The same procedure was followed for questions 14 and 15. On the other hand, it is important to note that this procedure can damage the comparison betweenof the results inobtained from theof original version and those from the Portuguese version.

The third version of the questionnaire was then submitted put before to an "ad hoc committee" composed byof a Brazilian psychologist working in the addiction field with a good understanding of English, a native Brazilian professionalersonnative working in addictions who did not speak English, an English psychiatrist working in addictions who was fluent in Portuguese and an English psychologist working in the addictions with some understanding of Portuguese. The committee studied each question in turn looking at both the original question in English, and at its various translations into Portuguese. Eventually a consensus was reached as to which version was most easily understood by the average Brazilian patient whilst maintaining the original concept of the question.

The ad hoc committee's version of the questionnaire was applied with another group of alcohol- dependent patients, souch that each question was thus asked toof five different patients. The aim was fFor the patients the aim was that they to gaive their understanding of what the questions meant and to answer themit. These interviews were tape-recorded. Minor modifications were performedmade after each interview before the final draft ¹⁹. The Portuguese version of this final draft is presented in ^{Appendix 1}. The back-translation was performed by an American English teacher without reference to the original version of the questionnaire.

4. Ethical Considerations:

The study had been approved by the Ethics Committee for Medical Research of the Federal University of São Paulo, Brazil. All subjects received research information on the research; they signed an informed consent form prior to participating and were assuredguaranteed anonymity and confidentiality.

5. Statistical Analysis:

The first step was to investigate the factor structure and construct validity of the 19-item SOCRATES questionnaire. This was done using two first- order confirmatory factor analyses, the first modelled on the three independent factor structure described by Miller & Tonigan,² and the second on the two independent factors described by Maisto et al.³ The reliability of the factors was measured by looking at their internal consistency, using Cronbach's alpha, and item analysis was based on the item to total score correlation. The correlation between the factors and the results of the item analyses suggested the existence of a factor structure based on two non-independent factors.

Using the statistical package SAS, the covariance generated by CALIS (Covariance Analysis of Linear Structural Equations) was used to test confirmatory factor analysis models, estimate parameters and test the appropriateness of structural equation models using covariance structural analysis and maximum likelihood estimation. The fit of the different models was assessed using four indices: 1) the Goodness Fit Index (GFI) which ranges from 0 to 1, with 0.90 or higher being evidence of a model with a good fit;²⁰ 2) the Adjusted Goodness-of-Fit Indices (AGFI) refers to an adjusted GFI for degrees of freedom in the model and also ranges from 0 to 1;²¹ 3) a Chi-squared to degrees of freedom ratio (X²/df) of less than 2;²² 4) a standardized root-mean-square residual (RMSR) less than 0.05;²² and 5) a root-mean-square error of approximation (RMSEA) of less than 0.08.²³

Results

Confirmatory factor analysis: Tthe two- and three-factor models

To evaluate the 3-factor structure proposed by Miller & Tonigan² and the 2-factor structure shown by Maisto et al.³ we used confirmatory factor analysis (CFA) based on these models with independent errors and factors.

For the 3-factor model all items had significant weight in the corresponding dimensions – as shown in Table 3. The chi-squared of this model was 850.25, d.f.=152, p<0.0001. The indices of fit were: GFI=0.793, AGFI=0.7417, RMSR= 0.203, and RMSEA=0.1187, which show an less than ideal fit between the proposed 3-factor model and the current data.

For the 2-factor model (AMREC: Ambivalence – Recognition and Taking Steps) all items had significant weight in the corresponding dimensions (Table 3). The chi-square of this model was 335.72, df=90, p<0.0001). The indices of fit were: GFI=0.8866, AGFI=0.8488, RMSR= 0.1351, and RMSEA=0.0915. All these indices suggest a better fit for the 2-factor model as proposed by Maisto et al.³

Reliability and item analysis of the 2- and 3-factor models

For Miller & Tonigan's² 3-factor model, the items making up factor 1 (Recognition) have good reliability with item-total correlations of over 0.51, with the exception ofor item 1. For factor 2 (Ambivalence), the correlations are higher than 0.50 with the exception ofor item 12. For factor 3 (Taking Steps) the correlations are lower still but still aboveover 0.3, with the exception ofor item 15. For Maisto's ³ 2-factor model, factor 1 (AMREC) has good reliability with item to total correlations highergreater than 0.54 except for item 12. The items on factor 2 (Taking Steps) have correlations highergreater than 0.41 with the except forion of item 145. These results suggest that the items with low correlaetions should be excluded from the analysis – Table 3.

Correlation of the factors

To explore the relationship between the factors based on the two models, we studied the intra- and inter-factor correlations. The intra-factor correlations show whether the factors within each model are independent of one from the another. The inter-factor correlations show the relationship between the two models. For the 3-factor model there is a strong correlation between Recognition and Ambivalence. For the 2-factor model the correlations are much lower (Table 4).

The inter-factor correlations between the two models show that Recognition and Ambivalence are both strongly correlated with AMREC as are the Taking Steps factors from each model.

These results suggest that the three factors described by Miller & Tonigan² and two factors (AMREC and Taking Steps) proposed by Maisto et al.³ are not independent, indicating that a new factor structure should be investigated.

These results suggest that the three factors described by Miller & Tonigan² and two factors (AMREC and Taking Steps) proposed by Maisto et al.³ are not independent, indicating that a new factor structure should be investigated.

Item analysis, reliability and CFA for a model with two correlated factors

An item analysis and Cronbach's alpha were used with the data modelled on the combined first and second factors described by Maisto et al.³ The Cronbach's alpha was 0.846 and there were low correlations for items 1 (0.259) and 12 (0.2274), showing that these items aredo not form part of the factor. These items were excluded and the data were reanalyzed. This corrected factor 1 had 9 items and was very similar to Maisto et al.'s.³ AMREC factor, except for the fact that it incorporated item 17 and excluded item 12. It had item to total correlations greater than 0.55 and an alpha of 0.89. Item 15 was thus excluded from factor 2. This corrected factor 2 had 7 items and is similar to Maisto et al.'s³ second factor, except for the fact that it incorporates item 20. We named our first factor AmRec and our second factor Taking Steps.

To evaluate the structure of these two non-independent factors a further CFA was performed. The hypothesis of the existence of this new two-factor structure is partially supported by the results of the analysis. All the items have significant weight (p<0.001) in the corresponding dimensions (as shown in Table 6). The chi-squared of this model was 407.22 (d.f.=103, p<0.0001), (X² to d.f. ratio = 407.22/103 > 2) which shows that there are differences between the proposed model and the data. The indices of fit were: GFI = 0.869, AGFI = 0.827, RMSR = 0.091 and RMSEA = 0.0856. These results show that the model displays a good but imperfect fit. The correlation between the factors was 0.364 (p<0.01), which is statistically significant.

Correlation with total alcohol consumption in the previouslast 30 days

In order tTo examine the validity of the different models, we calculated the correlation between the scores obtained on each factor with total alcohol consumption in the previouslast 30 days. As this measure had an asymmetric distribution, a logarithmic transformation was performed. With our data modelled on Miller & Tonigan's ² three-factor structure, the Pearson's correlations were as follows: Ambivalence (r = 0.240, p<0.001), Recognition (r = 0.211, p<0.001) and Taking Steps (-0.178, p<0.001). For the model based on Maisto et al.'s³ two-factor model the correlations were: AMREC (r = 0.219, p<0.001) and Taking Steps (r = -0.142, p<0.01).

Discussion

The Transtheoretical Stages of Change Model has been questioned and attempts to measure the five stages as distinct entities have met withfound limited success.^24,25,26,27 The aim of this paper was to investigate the factor structure of the Portuguese version of SOCRATES. Miller & Tonigan² claim that SOCRATES has a 3-factor structure, consisting of Ambivalence, Recognition and Taking Steps. We found that a two non-independent factor structure fit the data better than a three-factor structure. The two factors were composed byof items groupeding around two distinct concepts that we named AmRec (a combination of Ambivalence and Recognition) and Taking Steps, similar to the results obtained by Maisto et al.³

One rResearch with college students²⁸ gives modest support for the SOCRATES having a 3-factor structure. In this study the authors first used confirmatory factor analysis with the 19-item SOCRATES questionnaire and found that the 3-factor model fit the data moderately well. They then removed the 3 items with the weakest cross-loadings and repeated the analysis. The 16-item version performed better than the full version.

Maisto et al.^3,15 found two factors that were almost identical to ours. These authors cite unpublished work suggesting that the ambivalence factor found by Miller & Tonigan (1996) may be unstable. They quote that Dermen et al.²⁹ failed to find significant factor loadings for the ambivalence items of the SOCRATES in a sample of patients from an inpatient substance abuse treatment centrer and that Busby and Parker³⁰ found the ambivalence scale to have poor internal consistency.

The reasons for the discrepancy between the factor structure reported by Miller and Tonigan² and that found in subsequent studies are diverse. First, the instrument in its present format may be unable to distinguish between ambivalence and recognition because the item content of the questionnaire is inadequate to make such a distinction. Our translation process has identified ambiguity in several items and subjects reported that certain items were difficult to answer. With better and more precise questions to characterize ambivalence, it might be possible to refine and modify the instrument to make a clearer distinction between this factor and recognition.

A second explanation is that ambivalence and recognition are indeed part of the same phenomenon. From one's day-to-day clinical experience, it may one canbe envisaged patients who recognize that they have a drink problem but are ambivalent about the need to stop drinking, to change their lifestyle or to follow a prescribed treatment package. At the early stages of treatment ambivalence and recognition are ordinarilycommonly seen together. Our factor analysis suggests that there is substantially more similaritycommonality between these concepts than there is uniqueness, since the sets of questions designed to measure them do not vary independently but are actuallyin fact highly correlated on a single factor.

Maisto et al.³ propose an alternative explanation for their failure to replicate Miller & Tonigan's² 3-factor solution, in particular their failure to find an ambivalence factor. In Miller & Tonigan's² study, the patients were from secondary care services and had considerably higher levels of alcohol use and problem severity of a nature and degree that would be easily recognized by most observers. They suggest that if an individuals isare still uncertain about whether or not tnotheyhe or she havinges a drinking problem including regardingin the face of alalcohol dependence and secondary complications, then other factors may be at play (e.g. pathological denial or cognitive impairment) which might be reflected in an ambivalence factor. However, our sample also came from secondary treatment services, most patients were moderately to severely dependent and those from the gastroenterology clinic had already developed physical complications of alcohol misuse, but we were still unable to find convincing evidence of a separate ambivalence factor.

It is possible that these different findings may in part be due to different factor analytic procedures. The study by Miller & Tonigan ² used factor analysis with principal components analysis with orthogonal and non-orthogonal rotation and with a criterion of eingenvalues >1.0. Maisto et al. ³ used principal components analysis with orthogonal rotation to extract factors with eigenvalues >1.0 followed by a confirmatory factor analysis. They also used goodness of fit analysis to compare their 2-factor model with Miller & Tonigan's ² 3-factor solution and found the 2-factor solution fit the data better.

The removal of 3 items of low reliability (questions 1, 12 and 15) may have had some effect on our factor analysis. Miller & Tonigan² used the 20-item SOCRATES but eliminated one item (question 6) due to its low factor loading and difficulty in asking the question in a follow-up setting. Maisto³ excluded item 17 due to a clerical error. They then excluded items 1, 15 and 20. The three items that we removed werecame from the Taking Steps and Recognition sections of the questionnaire.

Our study used a Portuguese version of the SOCRATES, therefore, the process of translating the questionnaire needs to be examined as this could also have affected the performance of the instrument. A poor translation may affect the internal consistency, reliability and validity of an instrument originally developed in another language and culture.^31,32 We followed a meticulous process to ensure that a semantically equivalent, reliable and valid version was produced. This involved alternative translations of the questionnaire being produced, a bilingual multi-professional committee that examined each question in turn, taped interviews with patients and extensive piloting. Patients were not only asked to answer the questions but also to say what they thought the questions meant. Psychometric analysis showed that our version had good internal consistency and reliability. Our 2-factor model is consistent with the findings of other researchers, although of the three items that we excluded due tobecause of their low inter-item correlations, only one was also excluded by Maisto et al.³ (item 1). We did resort to changing the structure of three questions, which may have altered the performance of the questionnaire. We felt these changes were necessary to enable all patients to answer all the questions, but our solution was not entirely satisfactory.

Sutton²⁶ has questioned Miller & Tonigan's² findings and criticizeds them for not reporting the full correlation matrix. He noteds that the pre-contemplation and determination subscales were negatively correlated (-0.70 in the outpatient sample and –0.62 in the aftercare sample) and that the Taking Steps and maintenance Maintenance subscales were positively correlated (0.69 and 0.56), and that other correlations were modest. He interpreteds these results as suggesting a lack of distinction between precontemplation and determination and between action and maintenance.

Conclusion

Further work is necessary to investigate the factor structure of the SOCRATES in different clinical and cultural populations, mainly outside Anglo Saxon countries. Future research is needed to understand the motivation aspects that the SOCRATES does measure and it will show how this scale could be improved.

Further work is necessary to investigate the factor structure of the SOCRATES in different clinical and cultural populations. But if neither the five stages of change nor the 3-factor solution couldan be consistently verified, then either the Stages of Change model, or the instruments designed to measure it, may need to be reconsidered.

References

Sponsor: Fundação de Amparo à Pesquisa do Estado de São Paulo - FAPESP.

Received 21.10.2003

Accepted 04.07.2003

• 1. Miller WR. SOCRATES The Stages of Change Readiness and Treatment Eagerness Scale (Version 8). Albuquerque (NM): Univ New Mexico; 1995.
• 2. Miller WR, Tonigan JS. Assessing Drinkers' Motivation for change: The Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES). Psychiatry Addict Behav 1996;10(2):81-9.
• 3. Maisto AS, Conigliaro J, McNeil M, Kraemer K, O'Connor M, Kelley ME. Factor structure in a sample of primary care patients. Addict Behav 1999; 24(6): 879-92.
• 4. DiClemente CC, Prochaska JO, Velicer WF, Fairhurst S, Rossi JS, Velaquez M. The process of smoking cessation: an analysis of the precontemplation, contemplation and preparation stages of change. J Cons Clin Psychology 1991;59:295-304.
• 5. Miller WR. Motivation for treatment: a review with special emphasis on alcoholism. Psychological Bull 1985;98:84-107.
• 6. Connors GJ, Donovan DM, DiClemente CC. Substance Abuse Treatment and Stages of Change. New York (NY): The Guilford Press; 2001.
• 7. Tucker JA, Donovan DM, Marlatt GA. Changing Addictive Behavior - Bridging Clinical and Public Health Strategies. New York: The Guilford Press; 1999.
• 8. Miller WR, Rollnick S. Motivational Interviewing preparing people to change addictive behaviour. New York: The Guilford Press; 1991.
• 9. Miller WR. Motivational interviewing with problem drinkers. Behav Psychotherapy 1983;11:147-72.
• 10. Heather N, Rollnick S, Bell A. Predictive validity of the Readiness to Change Questionnaire. Addiction 1993;88:1667-77.
• 11. Buddy R, Rollnick S. (1996) The structure of the Readiness to Change questionnaire: a test of Prochaska and DiClemente's transtheoretical model. Br J of Health Psychology 1994;1:365-76.
• 12. McMahon J, Jones BT. Post-treatment abstinence survivorship and motivation for recovery: the predictive validity of the readiness to change (RCQ) and negative alcohol expectancy (NAEQ) questionnaires. Addiction Research 1996;4:161-76.
• 13. Miller WR, Meyers RJ, Hester RK., Delaney H, Montgomery HA, Abbott PJ. The effectiveness of the community reinforcement approach: Preliminary findings of a clinical trial. Paper presented at the 5^th international conference on Treatment of Addictive Behaviours; 1990; Sydney, Australia.
• 14. Miller WR, Tonigan JS, Montgomery HA, Abbott PJ, Meyers RJ, Hester RK, Delaney H. Assessment of client motivation for change: preliminary validation of the SOCRATES instrument. Paper presented at the annual meeting of the Association for Advancement of Behavior Therapy; 1990; San Francisco, CA.
• 15. Maisto SA, Chung TA, Cornelius JR, Martin CS. Factor structure of the SOCRATES in a clinical sample of adolescents. Psychology Addict Behav 2003;17(2):98-107.
• 16. Jorge MR, Masur J. The use of the Short-Form Alcohol Dependence Data Questionnaire (SADD) in Brazilian Alcoholic Patients. Br J Addiction 1985;80:301-5.
• 17. Méndez EB. A Brazilian version of the AUDIT (Alcohol Use Disorders Identification Test). Unpublished master dissertation. Universidade Federal de Pelotas Social Medicine Department Post-Graduate Program; 1999.
• 18. Tabakoff B, Dongier M. The WHO/ISBRA Study on State and Trait Markers in Alcoholism: Progress Report. Alcohol Clin Exp Research 1996;20(8):23-7.
• 19. Figlie, NB. Motivation in alcoholic outpatient in specialised alcoholism and gastroenterology clinical treatment program. Unpublished master dissertation. Universidade Federal de Săo Paulo Psychiatry Post-Graduate Program; 1999.
• 20. Byrne BM. Structural Equation Modelling with EQS and EQS/Windows: Basic Concepts, Applications, and Programming. Thousand Oaks (CA): Sage; 1994.
• 21. Pedhazur EJ, Schmelkin LP. Measurement, Design and Analysis: An integrated approach. New Jersey (NJ): Lawrence Erlbaum Associates Inc., Publishers; 1991.
• 22. Tabachnick BG, Fidell LS. Using Multivariate Statistics. 3^rd ed. New York: Harper Collins; 1996.
• 23. Arbuckle JL. Amos User's Guide (Version 3.6). Chicago (IL): Smallwaters Corporation; 1997.
• 24. Davidson R. Editorial and commentaries: "Prochaska and DiClementes model of change: a case-study?" Br Jf Addiction 1992;87:21-835.
• 25. Sutton S. Can "stages of change" provide guidance in the treatment of addictions? A critical examination of Prochaska and DiClemente's model. In: Psychological Treatments and the Addictions. Cambridge (UK): Cambridge University Press; 1996. p. 189-205.
• 26. Sutton S. Back to the drawing board? A review of applications of the transtheoretical model to substance use. Addiction 2001;96:175-86.
• 27. Sutton S. Assessing stage of change in current and former smokers. Addiction 2002;97:1171-82.
• 28. Vik PW, Culbertson KA, Sellers K. Readiness to change among heavy-drinking college students. J Stud Alcohol 2000;61:674-80.
• 29. Dermen K, Kousty J, Connors GJ, Czarnecki D. SOCRATES scores of alcoholic patients: Factor structure and relationship to pretreatment characteristics. Paper presented at the Annual Meeting of the Association for the Advancement of Behavior Therapy; 1997; Miami, FL.
• 30. Busby LD, Parker JD. Norms and reliability for the stages of change and treatment eagerness scale (SOCRATES) obtained from drinkers accepted for inpatient treatment. Poster session presented at the Annual Meeting of the Association for the Advancement of Behavior Therapy; 1997; Miami, FL.
• 31. Perneger TV, Leplege A, Etter JF. Cross-cultural adaptation of a psychometric instrument: two methods compared. J Clin Epidem 1999;52(11):1037-46.
• 32. Lee JA, More SJ, Cotiwan BS. Problems of translating a questionnaire in a cross-cultural setting. Prev Vet Medicine 1999;41:187-94.

Appendix 1

Publication Dates

• Publication in this collection
  27 Oct 2004
• Date of issue
  June 2004

History

• Accepted
  04 July 2003
• Received
  21 Oct 2003