A Logit Model of the Probability of Divorce

Abstract

The rationale behind this study was to investigate divorce that is relatively under-researched in Canada. In addition, it was thought that a reasonably sophisticated statistical program, logistic regression, may be usefid in predicting the probability of divorce among Albertans. Thc theory behind the research was that a combination of structural, social processional and social psychological factors were predictive of divorce likelihood. The results of the logistic regression revealed that there were present structural, social processional and social psychological factors that predicted divorce probability. The structural variables (geographical location, religious affiliation and gender roles), social processional (women's participation in work) and social psychological (attitudes to divorce) factors were all significantly related as to the probability of divorce. The paper concludes with a discussion of these results.