OUP user menu

RELIABILITY AND VALIDITY OF THE DUTCH VERSION OF THE READINESS TO CHANGE QUESTIONNAIRE

L. Defuentes-Merillas, C. A. J. Dejong, G. M. Schippers
DOI: http://dx.doi.org/10.1093/alcalc/37.1.93 93-99 First published online: 1 January 2002

Abstract

— The aim of the present study was to evaluate the psychometric properties of the Dutch version of the Readiness to Change Questionnaire (RCQ-D). The subjects were 246 excessive drinkers admitted to an addiction treatment centre and 54 offenders convicted of an alcohol-related crime in The Netherlands. The factor structure of the RCQ-D for the two samples combined was found to be consistent with the three-factor structure established for the original RCQ. The reliability of the items for each scale was found to be satisfactory. Allocated stage of change showed significant differences between the different subsamples. As expected, the scale scores for adjacent stages of change showed significantly higher inter-correlations than the scale scores for non-adjacent stages. Additionally, the negatively formulated items from the pre-contemplation scale were reformulated positively and their internal consistency tested among the offender sample. The positively formulated pre-contemplation items showed a higher alpha value than the negatively formulated items. We therefore suggest that the positively formulated items should replace the negatively formulated ones.

INTRODUCTION

The stages of change model of Prochaska and Diclemente (1983) provides an understanding of addictive behaviour by describing the motivational stages of change which a person passes through in an attempt to resolve his/her addiction problem. The model introduces the possibility of using different intervention strategies at different stages of change, from ‘pre-contemplation’ to ‘contemplation’ and ‘action’ to ‘maintenance’ (Heather et al., 1991, 1993; Rollnick et al., 1992; Norman et al., 1998). Underlying the model is the assumption that the stage of change a person is in can be accurately assessed.

There have been several attempts to operationalize and measure the stages of change, such as the Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES, by Miller and Tonigan, 1996) and the University of Rhode Island Change Assessment Scale (URICA, by McConnaughy et al., 1983). Rollnick et al. (1992) empasized that the instrument to assess the stages of change should be short, easy to administer and suitable for use in applied settings such as public health clinics. On this basis, they developed a 12-item ‘Readiness to Change Questionnaire’ (RCQ) for excessive drinkers. The RCQ consists of three 4-item scales representing three stages of change: pre-contemplation (P), contemplation (C) and action (A). The questionnaire can be used to assess the stage of change for a client and is thus essential for deciding on the appropriate intervention for individuals with very different needs (Crittenden et al., 1998; Morera et al., 1998). The psychometric properties of the RCQ have been shown to be fairly good (Rollnick et al., 1992).

The Dutch version of the Readiness to Change Questionnaire (RCQ-D) was translated by the second and third authors. Since three of the pre-contemplation items from the original version of the questionnaire are formulated negatively, which is not very desirable from a psychometric point of view because these negative items can lead to confusion, it is generally recommended that they should be avoided (Edwards, 1957). For this reason, a positive version of the pre-contemplation scale was formulated and presented to the offender sample. Furthermore, abstinence has been added as a possible outcome to the Dutch version of the RCQ, while the original English version only refers to cutting down alcohol consumption. Such extension allowed us to use the questionnaire to assess readiness to change drinking behaviour for both total abstinence and controlled drinking as possible outcomes.

In this article, we will report on the psychometric properties of the RCQ-D administered to two samples of excessive drinkers. First, we will describe the adaptation and development of the Dutch questionnaire along with the data collection procedures. Second, we will report on the reliability and validity of the questionnaire. Finally, we will compare the internal consistency of the positively and negatively formulated pre-contemplation items of the RCQ-D. It is postulated that the positively formulated items from the pre-contemplation scale will discriminate better among excessive drinkers than the negatively formulated items.

SUBJECTS AND METHODS

Subject samples

Two subject samples completed the RCQ: patients receiving treatment at an addiction centre and offenders convicted for an alcohol-related crime. The first sample consisted of 246 patients at an addiction treatment centre (NOVADIC, of the Network for Addiction Treatment Services, The Netherlands). Their ages varied from 18 to 66 years (mean ± SD: 35.68 ± 9.79), and 79.3% were males. A total of 192 of the patients (78%) attended a detoxification programme and are referred to as the ‘detox subsample’ while 54 (22%) attended both a detoxification programme and a treatment programme and are therefore referred to as the ‘treatment subsample’. Patients met the DSM-IV (American Psychiatric Association, 1994) criteria for alcohol and/or other substance dependence (opiates, benzodiazepines, cocaine, cannabis, etc.).

The second subject sample consists of 54 offenders recruited at different court of justice offices in the following cities in The Netherlands: Groningen, Haarlem, Maastricht, Middelburg, Roermond and Rotterdam. These offenders were convicted for an alcohol-related crime, mostly for driving under the influence, and required to follow an alcohol education programme. Their ages varied from 16 to 40 years (mean ± SD: 22.5 ± 5.2); 93.8% of the offenders were male.

Instruments

The RCQ-D was designed to resemble the original RCQ. All of the items are judged along a 5-point scale ranging from ‘Strongly Disagree’ to ‘Strongly Agree’ and assigned a score ranging from –2 to +2.

The original RCQ was formulated with reduced drinking as a goal. However, we considered abstinence a possible outcome as well. This was done by extending items 2, 4, 8 and 12 with an explanatory sentence (see Table 1). Three of the pre-contemplation items from the original RCQ are formulated negatively (i.e. 1, 10 and 12). Since it is generally recommended to avoid negatively formulated items (Edwards, 1957), these items were positively reformulated as follows: ‘1. I think that I drink too much’; ‘10. It is necessary for me to think about changing my drinking’; ‘12. Drinking less alcohol would be useful for me’.

View this table:
Table 1.

Item loadings for two factors extracted using Varimax rotation with percentage variance accounted for by each using two samples combined

Items (English version)Factor I (41.6%) Pre (–), Con (+)Factor II (11%) Action
Pre, pre-contemplation; Con, contemplation. (P), pre-contemplation; (C), contemplation; (A), action.
Dutch translations of each item: 1. ‘Ik vind dat ik niet teveel drink’ (P). 2. ‘Ik probeer nu minder te drinken dan ik gewend ben’ (A) (‘minder’ kan ook zijn ‘niets’: Ik probeer niets meer te drinken). 3. ‘Ik geniet van een glas (alcohol) maar soms drink ik teveel’ (C). 4. ‘Soms denk ik dat ik minder zou moeten drinken’ (C) (‘minder’ kan ook zijn ‘niets’: Soms denk ik dat ik niets zou moeten drinken). 5. ‘Nadenken over mijn drinkgedrag is tijdverspilling’ (P). 6. ‘Ik heb sinds kort mijn drinkgewoontes veranderd’ (A). 7. ‘Iedereen kan gemakkelijk zeggen iets aan hun drinkwoontes te doen, maar ik doe er echt iets aan’ (A). 8. ‘Ik ben in de fase waarin ik zou moeten nadenken over minder alcohol drinken’ (C) (‘minder’ kan ook zijn ‘niets’: Ik ben in de fase waarin ik zou moeten nadenken niets meer te drinken). 9. ‘Mijn drinkgewoontes zijn soms problematisch’ (C). 10. ‘Het is voor mij niet nodig na te denken over het veranderen van mijn drinkgewoontes’ (P). 11. ‘Ik ben op dit moment bezig mijn drinkgewoontes te veranderen’ (A). 12. ‘Minder alcohol drinken zou voor mij geen zin hebben’ (P) (‘minder’ kan ook zijn ‘niets’: Niets meer alcohol drinken zou voor mij geen zin hebben).
1.I don't think I drink too much (P)–0.66–0.12
2.I am trying to drink less than I used to (A) (‘less’ can also be I am trying to drink less than I used to or nothing at all0.260.73
3.I enjoy my drinking, but sometimes I drink too much (C)0.270.03
4.Sometimes I think I should cut down on my drinking (C) (‘cut down’ can also be ‘stop’: Sometimes I think I should stop my drinking)0.690.32
5.It's a waste of time thinking about my drinking (P)–0.63–0.27
6.I have just recently changed my drinking habits (A)0.170.81
7.Anyone can talk about wanting to do something about drinking, but I am actually doing something about it (A)0.070.82
8.I am at the stage where I should think about drinking less alcohol (C) (‘less’ can also be ‘nothing’: I am at the stage where I should think about drinking less alcohol or nothing at all)0.660.31
9.My drinking is a problem sometimes (C)0.790.21
10.There is no need for me to think about changing my drinking (P)–0.68–0.28
11.I am actually changing my drinking habits right now (A)0.490.72
12.Drinking less alcohol would be pointless for me (P) (‘less’ can also be ‘nothing’: Drinking less alcohol or nothing at all would be pointless for me)–0.53–0.05

In order to test the concurrent and discriminant validity of the RCQ-D, several instruments were used. Motivation to change was assessed in the offender sample with a single motivation for change question: ‘At this moment, are you considering stopping drinking or significantly cutting down on your drinking?’ The subjects could choose one of five alternative answers ranging from ‘absolutely not’ to ‘I have succeeded in stopping or reducing my drinking’. The subjects are then allocated to the pre-contemplation stage when choosing options 1 or 2; to the contemplation stage when choosing options 3 or 4; and to the action stage when choosing option 5.

The mental health status of all the patients was assessed using the Symptom Check List-90 (SCL-90; Derogatis et al., 1973), a self-report questionnaire with nine scales: Somatization, Obsessive–Compulsiveness, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation and Psychoticism. The Dutch version of the SCL-90 has been shown to have good psychometric properties (Arrindell and Ettema, 1986).

The Interpersonal Check List-R (DeJong et al., 1991) was used to measure interpersonal aspects of personality. The Dutch version consists of 128 items distributed across 16 scales constituting a circle and it has been shown to have good psychometric properties. The vertical axis represents Dominance vs Submission, whereas the horizontal axis represents Affiliation vs Hostility.

Perceptions of the therapeutic relationship were measured using the Helping Alliance Questionnaire (HAQ-R; Alexander and Luborsky, 1984), which is a self-report questionnaire with 11 items. The items are grouped into two subscales: Co-operation and Helpfulness. The HAQ has been found to have good psychometric properties in a Dutch population of patients with substance-dependence problems (DeWeert-van Oene et al., 1999).

The SCL-90, the ICL-R and the HAQ were administered during the routine testing of the patients, who gave their written consent for use of the data from these questionnaires for the present study, and also agreed to complete the RCQ-D. The medical Ethical Board had no objections to the procedures followed in this study.

Procedure

The RCQ was translated into Dutch according to the methodology published by Hunt et al. (1991). Two experts in the addiction field (the second and third authors) independently translated the original English questionnaire. Two independent translators then translated the two Dutch versions of the questionnaire back into English. These two versions were compared and differences were discussed. Finally a consensus version of the Dutch translation was established. Comments on this version of the questionnaire were gathered from care providers, and pilot testing of the questionnaire with patients showed it to have acceptable readability and good comprehension.

Data analysis

The psychometric properties of the RCQ-D were examined in four phases. First, an exploratory analysis of the structure of the questionnaire itself was undertaken. A principal components analysis was performed on the two samples combined together, in order to examine the similarities and differences in the factor structures of the Dutch and English versions. Second, the internal consistency was established by calculating the Cronbach's alpha coefficient for each of the 4-item stage of change scales. Third, the RCQ-D was subjected to a validity test, by comparing the scores for different categories of substance-dependent patients. The correlations between the RCQ-D scores and some other measures of psychosocial adjustment were used to determine the discriminant validity of the questionnaire. Finally, in order to improve the questionnaire accuracy, the negatively formulated items on the pre-contemplation scale were positively reformulated and the reliability of both the negatively and positively formulated versions were tested among the offender sample.

RESULTS

Structure of the RCQ-D

A principal components analysis with Varimax rotation was executed on the scores for the 12 items from the RCQ-D for the two samples combined. This produced a clear two-factor structure with three clearly distinguishable scales (pre-contemplation, contemplation and action). The two identified factors accounted for 52.6% of the total variance; one of the factors was bi-dimensional. The item loadings for these two factors were comparable to the item loadings in the English version of the questionnaire, which were found to account for 68.6% of the total variance (Rollnick et al., 1992), and also comparable to the item loadings in the RCQ-Spanish version, which accounted for 57.4% of the total variance (Rodríguez-Martos et al., 2000). The bi-dimensional factor grouped the pre-contemplation items at its negative pole and the contemplation items at its positive pole. The other factor was uni-dimensional and encompassed all of the action items. The item loadings for the two factors are presented in Table 1. Our factor structure differs from the structures detected by Rollnick et al. (1992) and Rodríguez-Martos et al. (2000); they found three uni-dimensional factors corresponding to the three stages of change (pre-contemplation, contemplation and action). In contrast, our analysis reveals a two factor structure in which one of the factors is bi-dimensional and the other is uni-directional.

The mean scores and SD for the items measuring the extent to which the subjects endorsed one of the three stages of change are shown in Table 2 and were found to be as follows: pre-contemplation (items 1, 5, 10, 12: 8.38 ± 3.84); contemplation (items 3, 4, 8, 9: 14.90 ± 4.2); and action (items 6, 2, 7, 11: 14.57 ± 4.5). Thus, the mean scores show low endorsement of the pre-contemplation items and higher endorsement of the other items (see Table 2).

View this table:
Table 2.

Means, SD, Cronbach's alpha coefficients and Pearson correlation coefficients for the Dutch Readiness to Change Questionnaire (RCQ-D) scales

MeanSDAlphaPreConAction
*P < 0.01.
RCQ-D scores are based on the 4-item scale scores for three scales for the two subject samples combined (246 patients and 54 offenders samples, n = 300).
Pre, pre-contemplation; Con, contemplation.
Pre8.383.840.68–0.574*–0.492*
Con14.904.20.700.520*
Action14.574.50.81

Relationships among the scale scores

Similar to Rollnick et al. (1992), the product-moment correlation coefficients were calculated between the scores for the three scales. This information allowed us to test a prediction derived from McConnaughy et al. (1983), namely that the correlations between adjacent scales (i.e. between pre-contemplation and contemplation and between contemplation and action) would be higher than the correlations between non-adjacent scales (i.e. between pre-contemplation and action). The correlations between the adjacent scales were found to be r(P × C) = –0.57, r(C × A) = 0.52, and between the non-adjacent scales r(P × A) = –0.49 (see Table 2). The product-moment correlations reported by Rollnick et al. (1992) were: r(P × C) = –0.53, r(C × A) = 0.57 and r(P × A) = –0.36. Differences between correlations were significant (P × C vs P × A, t = –1.83, df = 297, P < 0.05; C × A vs P × A, t = 11.95, df = 297, P < 0.05; both one-tailed tests).

Reliability aspects

The internal consistency of the different RCQ-D scales was established by calculating the Cronbach's alpha coefficients for the 4-item scales representing the different stages of change. The alpha values were as follows: pre-contemplation, α = 0.68; contemplation, α = 0.70; action, α = 0.81. This shows that the item scores can reasonably be regarded as constituting a scale in each case (see Table 2). Rollnick et al. (1992) reported somewhat higher Cronbach's alpha coefficients for the original RCQ scales (pre-contemplation, α = 0.73; contemplation, α = 0.80; action, α = 0.85), whereas Rodríguez-Martos et al. (2000) reported somewhat lower Cronbach's alpha coefficients in their RCQ-Spanish version.

Allocation of subjects to a stage of change

Allocation of the subjects to a particular stage of change was initially based on the highest raw score obtained for the three scales. As Heather et al. (1991) recommended in the case of a tie between two scale scores, the one representing the furthermost stage of change should be selected on the grounds that this represents the furthermost point in the change process. Of the 300 subjects who completed the RCQ-D, 38 (or 12.7%) were allocated to the pre-contemplation stage, 114 (or 38%) to the contemplation stage and 148 (or 49.3%) to the action stage. There were 39 ties in the data: one between scores representing pre-contemplation and contemplation, seven between pre-contemplation and action and 31 between contemplation and action. This meant that one subject was allocated to the contemplation stage and 38 to the action stage on the basis of a tied score.

Following Rollnick et al.'s (1992) methodology, score profiles were calculated to investigate the degree to which a strictly interpreted stage model was adhered to. This means that a person's readiness to change should be found to reside in a single stage of change at the point of assessment and that the scores for each subject should show an elevation on only one scale and concomitantly lower scores on the two other scales representing other stages of change. Frequencies for all possible profiles of positive (+) versus negative or zero (–) are presented in Table 3.

View this table:
Table 3.

Frequencies and percentages of different scale profiles for the two samples separately and together

PatientsOffendersCombined
ProfileFrequency%Frequency%Frequency%
The plus and minus signs in every profile correspond with the positive or negative score in the pre-contemplation, contemplation and action respectively. For example, profile D scores positive in pre-contemplation and negative in both contemplation and action.
A (+ + +)41.611.95 1.7
B (+ + –)10.411.92 0.7
C (+ – +)20.835.65 1.7
D (+ – –)135.31629.629 9.7
E(– + +)16567.11425.917959.7
F (– + –)2811.4814.83612
G (– – +)176.923.719 6.3
H (– – –)166.5916.725 8.3
Totals24654300

The profiles adhering to a strict stage model are those labelled D, F and G. Taken together, these profiles account for 84 (or 28%) of the observed score patterns. The pre-contemplation (pattern D = 9.7%) and contemplation (pattern F = 12%) profiles (n = 65) were clearly more prevalent than the action (pattern G = 6.3%). The most common profile was pattern E (n = 179), which was found to account for 59.7% of the subject profiles. This pattern presents a negative or zero score for the pre-contemplation scale and positive scores for the other two adjacent scales (contemplation and action).

Validity aspects

To check the validity of the RCQ-D, the total sample was divided into three groups according to recruitment status: detox subsample, treatment subsample (both from the patient sample) and offenders sample. Separate factor analyses with Varimax rotation were conducted for the three groups and found to show, first, that the items for the three groups were explained by the same factors (i.e., pre-contemplation, contemplation and action). This was particularly clear for the largest group, the detox patients: factor 1 (pre-contemplation) explained 36.31% of the variance, factor 2 (action) explained 12.44% of the variance and factor 3 (contemplation) explained 9.22% of the variance. This three-factor structure resembles that proposed by Rollnick et al. (1992).

It was also postulated that the allocation of the subjects to a particular stage of change using the RCQ-D would differ, depending on the group. As predicted, the offender group reported more pre-contemplation (37%) than both groups of patients, the detox subsample (7.8%) and the treatment subsample (5.5%). The treatment subsample was found to have more subjects in the action stage of change (63%) than both the detox subsample (51.6%) and the offender sample (27.8%). The means, SD and sample sizes for the three groups classified according to the RCQ-D scales are presented in Table 4. Separate analyses of variance (ANOVA) on the means for the different groups showed significant differences for all of the RCQ-D scales: pre-contemplation [F(2, 319) = 30.8, P < 0.001]; contemplation [F(2, 314) = 13.6, P < 0.001]; and action [F(2, 309) = 21.5, P < 0.001]. Post hoc multiple comparisons to determine exactly which of the group means differed significantly from the others showed the offender sample to differ significantly from both the detox subsample and the treatment subsample for all of the RCQ-D scales.

View this table:
Table 4.

Means, SD and sample sizes for the three subject groups classified according to the Dutch Readiness to Change Questionnaire (RCQ-D) scales

DetoxificationTreatmentOffender
ParameterPreConActionPreConActionPreConAction
Pre, pre-contemplation; Con, contemplation.
Mean 7.77 15.43 15.32 7.4415.3515.1711.812.311.09
SD 3.53 4.12 4.29 3.04 3.03 4.01 4.07 4.44 4.54
n 192192192545454545454

To determine whether the allocations to particular stages of change overlapped with the results of some other instruments, measuring therapeutic relation and psychopathology, the relations between the RCQ-D scores and the HAQ, the ICL and the SCL-90 were calculated. As predicted, none of the RCQ-D scales correlated significantly with any of the subscales of the ICL-R, the SCL-90 or the HAQ, with the exception of the RCQ-D action stage, which correlated moderately with the co-operation scale of the HAQ (r = 0.17; P < 0.02).

The correspondence between the RCQ-D allocations and motivation to change, assessed using a single motivation for change question with the offender sample, was examined next. In response to this single question with five alternative answers, 22 subjects were allocated to the pre-contemplation stage, 22 to the contemplation stage and four to the action stage. Six of the 54 offenders did not answer to this motivation to change question. ANOVA were performed with the stage of change selected using this question as the independent variable and the raw scale scores for the three stages of change derived from the RCQ-D as the dependent variables. It was postulated that subjects allocated to a given stage of change by this motivation to change question would score higher on the corresponding RCQ-D scale. The means and SD for the three RCQ-D scales and the revised RCQ-D pre-contemplation scale—positively reformulated—are presented in Table 5. The results confirm the hypothesis for all of the RCQ-D scales, including the revised pre-contemplation scale: pre-contemplation (F = 3.85, df = 47, P < 0.001); contemplation (F = 4.15, df = 47, P < 0.001); action (F = 4.14, df = 47, P < 0.001); and revised pre-contemplation (F = 4.44, df = 47, P < 0.001). The results in Table 5 show that subjects categorized in the pre-contemplation stage obtained higher mean scores on both of the pre-contemplation scales than on the other scales. Similarly, subjects categorized in the contemplation stage obtained a higher mean score on the contemplation scale than on the other RCQ-D scales and subjects categorized in the action stage obtained a higher mean score on the action scale than on the other scales.

View this table:
Table 5.

Means and SD for the three Dutch Readiness to Change Questionnaire (RCQ-D) scales and the revised pre-contemplation scale according to subjects categorized for motivation to change for the offender sample

StagePre-contemplation scaleContemplation scaleAction scalePre-contemplation scale, revised
Pre-contemplation, this scale (n = 22)
    Mean3.09–2.91–4.23 4.9
    SD2.91 1.95 2.97 2.86
Contemplation, this scale (n = 22)
    Mean–2.95 2.86 0.77–2.41
    SD 2.03 3.88 3.46 2.48
Action, this scale (n = 4)
    Mean–2.75 3.5 5–3.75
    SD 2.28 3.20 2.74 3.19

Item substitution

To examine whether the negatively formulated items 1, 10 and 12, based on the original English pre-contemplation scale, could be replaced by positively formulated items, the reliability of both the negatively and positively formulated versions of the RCQ-D was tested. The Cronbach's alpha coefficient for the positively formulated pre-contemplation scale was found to be somewhat higher (α = 0.83) than the Cronbach's alpha coefficient for the negatively formulated pre-contemplation scale (α = 0.81).

DISCUSSION

Psychometric properties

The data show that the psychometric properties of the RCQ-D closely resemble those of the original English questionnaire. The factor structure of the RCQ-D for the two subject samples combined in the present study was found to be consistent with the three-factor structure previously established for the RCQ. The reliability of the items constituting the different scales was also found to be satisfactory. As expected, scales on adjacent stages of change showed significantly higher inter-correlations than scores on non-adjacent stages.

The stages of change allocated to the subjects showed significant differences across the different groups. As predicted, the treated subsample was found to have more subjects in the action stage, when compared to the detox subsample and the offender sample. These results clearly reflect the stages of change model: subjects further along with treatment are more likely to be in the action stage of change than subjects not as far along with treatment or receiving less intensive treatment. Only 28% of subject profiles were consistent with the strictly interpreted stages of change model. However, if we consider the meaning of E profile (– + +) as representing a preparation stage of change (Heather et al., 1993) or a decision-making profile (McConnaughy et al., 1983), then 87.7% of profiles will be consistent with the stages of change model.

The factor structure of the RCQ-D differs slightly from the factor structure reported by Rollnick et al. (1992) for the original English RCQ, which produced three uni-dimensional factors reflecting three stages of change: pre-contemplation, contemplation and action. In contrast, our analysis revealed a two-factor structure with a single bi-dimensional factor and a single uni-dimensional factor. The bi-dimensional factor grouped the pre-contemplation items at its negative pole and the contemplation items at its positive pole. The uni-dimensional factor grouped the action items together. Nevertheless, the factor structures for the Dutch and English versions of the questionnaire were highly similar and the three stages of change were clearly seen.

The results of the present study also confirm McConnaughy et al.'s (1983) prediction that the correlations between adjacent scales will be higher than the correlations between non-adjacent scales. The correlations involving the pre-contemplation scale tended to be negative, due to decreasing scores on the contemplation and action scales reflecting an increasing readiness to change. The Pearson correlations for the three scales exhibited the same pattern observed for other subject populations (McConnaughy et al., 1983; Rollnick et al., 1992).

The lack of correlations between the RCQ-D scores and the HAQ, SCL-90 and ICL scores was expected, because none of the questionnaires is specifically aimed at the assessment of motivation for behavioural change. Additionally, these correlations show that the RCQ-D is not measuring the same concepts as those instruments.

Changes introduced to the Dutch RCQ

The positively formulated items from the pre-contemplation scale of the RCQ-D were found to have equal or even better reliability. Since positively formulated items are easier to comprehend, we suggest the use of this RCQ-D version in the future.

Treatment version

During the preparation of this manuscript, Heather et al. (1999) published a treatment version of the RCQ(TV). In this version of the questionnaire, items were reformulated to include both abstinence and reduced drinking as possible goals. In addition, it was attempted to implement subscales for Preparation and Maintenance, but these efforts did not succeed. Patients do not appear to differentiate the Action stage from these stages.

Since the Dutch version of the RCQ already included both abstinence and reduced drinking as possible goals of treatment, the RCQ-D may also be used as the Dutch treatment version of the questionnaire, as Heather et al. (1999) recommended.

The RCQ-D appears to be an appropriate instrument for assessment prior to treatment entry and assessment during treatment. As Heather et al. (1999) have pointed out, the most relevant use of the questionnaire is to distinguish between clients who are ready to change their drinking behaviour, and can therefore be offered skills-based and other action-oriented treatment procedures more or less immediately, from those who are not ready to change and are probably in need of further motivational counselling. In addition to providing a convenient measure of a drinker's stage of change, the RCQ-D is both quick and easy to administer (in the form of self-completion) and quick and easy to score. In other words, the questionnaire is suitable for use in any clinical settings.

Footnotes

  • * Author to whom correspondence should be addressed at: Amsterdam Institute for Addiction Research, Keizersgracht 818, 1017 EE Amsterdam, The Netherlands.

REFERENCES

View Abstract