Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)



Committee Chair(s)

Elwin C. Nielsen


Elwin C. Nielsen


The purposes of this study were: (1) to determine which characteristics of male alcoholics are related to treatment success and length of stay in long-term treatment, and (2) to assess the efficacy of predicting outcome and length of stay on the basis of patient characteristics, The study was performed post hoc on 265 patients discharged from a 6- to 12-month residential alcoholism program on the grounds of a state hospital. After an extensive literature review, 19 predictor variables were selected which were most consistently related to outcome and length of stay in previous studies of shorter rehabilitation programs. Data for predictor variables were obtained from psychological testing, admission interview notes, anamneses, and symptom ratings, Treatment outcome at 6-month follow-up was assessed dichotomously (success - failure) and numerically (number of abstinence and social adjustment criteria met) based on questionnaire responses and second-hand information. Sample size varied across analyses, as cases were deleted for missing data.

Compared to patients who failed to benefit, treatment successes were less antisocial (p < .01) and reported fewer alcoholic withdrawal symptoms (p < .05, n = 131). The results also suggest that successes were less angry than failures (this variable reached significance in the analyses which were given the most consideration, and showed consistent trends in other analyses). Patient characteristics unrelated to treatment outcome were age, socioeconomic status, social stability, number of arrests, age at onset of drinking problem, problematic drinking by patients' parents, length of longest previous period of sobriety, number of previous alcoholism treatments, previous regular A.A. attendance, overall mental health, neuroticism, depression, obsessive compulsive traits, latent schizophrenia, IQ, and defensiveness. A four-variable discriminant function produced 70.23% correct classification of outcome (r = .34, p < .01, n = 131), but the 33.3% false negative rate raises a question about using the function as an acceptance criterion. The pattern of results implies that the longterm program fosters social integration, but does not overcome the effects of severe personality disorders or physical addictions.

Only IQ was related to length of stay, with more intelligent subjects remaining in treatment longer (p < .05, n = 233). Weak but statistically significant prediction of length of stay was obtained with a nine-variable regression equation (r = .34, p < .01, n = 199).



Included in

Psychology Commons