The aim of the present studies was to examine the factor structure and psychometric properties of the A-RSAQ, a measure of primary appraisal (i.e., appraisal of general threat) of two common alcohol relapse risk situations. The findings from this study support a unitary factor structure, internal consistency, stability, and validity for each scale of the A-RSAQ. The component pattern for both samples is strong, and confirmation in a second sample strengthens the findings. Overall, the A-RSAQ appears to be a valid instrument for assessing threat appraisal presented by alcohol high-risk situations as perceived by abstinent alcohol abusers with ability to predict post-treatment drinking. Specifically, the A-RSAQ had excellent internal consistency reliability; adequate test-retest reliability; excellent construct, predictive, and factorial validity; good criterion-related validity; and acceptable discriminant validity, but more information on convergent validity is needed.
Appraisal of each of the risk situations had exceptionally high internal consistency and excellent temporal stability. The ratings of the two relapse risk situations shared 13% of variance in the first study and 48% in the second study, indicating some degree of independence in the appraisal of negative and social positive affect situation. This is consistent with the idea negative-affect and positive-affect based systems of motivation may involve different underlying systems (Baker et al., 1987
). Thus, the ability of an abstinent alcohol abuser to recognize the danger in one type of situation will not necessarily predict his/her ability to recognize risk in another major type of situation, and the threat of relapse posed by different types of risky situations needs to be addressed separately in treatment. This may suggest that more high-risk situations need to be included in the instrument. However, given the length of the instrument with two situations, the fact that these two situations are problematic for almost all alcoholics, and the predictive validity provided by these two situations, the current version of the A-RSAQ may be sufficient for clinical purposes.
The validity of the A-RSAQ was demonstrated in several ways. First, construct validity was supported by significant relationships between appraisal scores for each situation and three of the four Revised Ways of Coping Checklist coping scales, even after controlling variance due to appraisal of the other situation. This supports the use of the A-RSAQ as a measure of threat appraisal because the greater the appraised threat, the greater the likelihood that a coping strategy will be used. The minimal correlation with the Count Your Blessings coping scale may indicate that this coping method is not commonly used as a means to remain abstinent in high-risk situations. Appraisal scores for both situations were also significantly correlated with self-rated likelihood of drinking in the situation, providing evidence that the situations are perceived as threats to abstinence. The results suggest that the measure may be useful within treatment to identify individuals who do not adequately recognize the degree of threat inherent in these two common relapse situations. Individuals with low scores on one or both scales may need to learn to more accurately appraise the riskiness of these situations to their sobriety lest they fail to initiate coping strategies when risky situations occur.
Second, criterion-related validity was tested by examining the relationship between appraisal scores and the length of abstinence immediately preceding the time of first assessment. The hypothesis was supported by bivariate correlations. The relationship between length of past abstinence and the negative affect situation was no longer significant when the variance shared with the social positive affect situation was controlled for. This may indicate that the social positive affect situation remains threatening to overall well-being even with extended periods of abstinence, or that the negative affect situation appraisal does not add significant additional variance above that accounted for by the appraisal of positive social situations.
Third, discriminant validity was supported by demonstrating low correlations between appraisal scores and age and education, and no significant correlations with age of respondents when controlling for the other appraisal score and length of abstinence. However, when controlling for negative affect situation appraisal and length of abstinence, education remained significantly and negatively correlated with social positive affect situation appraisal scores. It is likely that older and more educated people are better able to handle social pressure situations and therefore perceive this type of situation as less threatening.
Convergent validity was tested by examining the relationship between appraisal scores and number of alcohol-related negative consequences ever experienced via the DrInC. Neither appraisal score was significantly related to DrInC scores, indicating that the current overall perceived threat of these high-risk situations is not related to the number of alcohol problems experienced in one’s lifetime. This finding might be due to the sample consisting of abstinent individuals. These consequences may have occurred at any time in the past. Situational appraisal is expected to predict increased future negative consequences due to increased drinking following appraisal in real life, not consequences preceding it. Also, alcohol problems are not good predictor of drinking outcomes (Adamson et al., 2009
). This suggests good discrimination between measures of problems and perceived threat.
Predictive validity also was demonstrated by ability to predict drinking during the 6 months after treatment. Higher threat appraisal on the social positive situation predicted fewer drinking days and fewer drinks on days that they drank and for the negative affect situation predicted fewer heavy drinking days. While negative and social positive affect situations may pose particular risks for drinking, those who appraise negative or positive affect situations as more threatening or harmful may be better able to mobilize their coping resources to manage or avoid these situations without drinking or without drinking heavily.
These validation results of the A-RSAQ are consistent with validation results of the cocaine RSAQ (Myers et al., 1996
). In both instruments, single-component solutions were found for each situation; both instruments were found to correlate with perceived relapse risk and to the coping scales, and the predictive validity of the cocaine RSAQ was supported for appraisal of the negative affect situation. Thus both instruments may have clinical utility in identifying individuals who need to learn to more accurately appraise the riskiness of situations in order to utilize coping skills.
Only one known scale is designed to assess relapse risk appraisal among adults treated for alcohol abuse. The Relapse Precipitants Inventory (RPI; Litman, 1986
; Litman et al., 1977
) was developed to assess perceived level of “danger to staying off drink” that each of 25 situations presented. Total RPI scores and scores on two of the three factors were shown to correlate significantly with subsequent relapse, indicating that the greater the number of situations rated as dangerous, especially for mood states and external events and euphoria, the more likely one was to have relapsed 6 or 15 months later. However, the number of dangerous situations is not conceptually the same as degree of threat seen in situations. RPI responses may represent the appraisal of ability to avoid drinking or self-efficacy and not the riskiness of the situation. Self-efficacy is predictive of drinking outcomes (Bandura, 1982
; Burling et al., 1989
; DiClemente et al., 1995
). This may explain why RPI results differ from those obtained in this current study and in Myers et al. (1996)
, both of which found that the higher the risk appraisal, the fewer days on which alcohol or cocaine was used during follow-up. Results with the RSAQ and A-RSAQ are both consistent with theories of the role of cognitive appraisal of risk (Lazarus and Folkman, 1984
In summary, the A-RSAQ is a psychometrically sound measure for assessing appraisal of threat posed by alcohol relapse-risk situations among abstinent adult alcohol abusers. The appraisal scores of the A-RSAQ correlate with relapse risk assessment for the situations presented, indicating that each situation presented a level of perceived relapse risk. The A-RSAQ is unique in that it is a measure of broad, general threat to individuals’ well-being in two high-risk situations. Results of the present study support the findings of Myers et al. (1996)
and provide additional support for the importance of appraisal of high-risk situations during abstinence. Clinically, these measures can be used to identify individuals who are not sufficiently aware of the degree of risk posed by common relapse situations.
One limitation of the present study is that the A-RSAQ was not validated against the only other known published alcohol relapse risk appraisal instrument, the RPI. However, since the RPI assesses number of risky situations rather than degree of risk appraisal, and since the RPI had limited validity data and used a statistical procedure that leads to sample-bound results, it is not clear that the RPI would be a better instrument for validation than the ones we used. A second limitation is that only two common high-risk situations were appraised, due to the length of assessment for each. It would be useful to determine whether adding several other situations would increase predictive validity without adding undue burden. Third, only sober community members and outpatients from one city were included. It would be useful to determine predictive validity in other populations and with alcoholics currently in residential treatment.