The logistic regression analysis of psychiatric data

Area of Science:

• Statistics
• Psychiatry
• Psychotherapy Research

Background:

• Binary dependent variables are common in clinical research.
• Analyzing these variables requires specialized statistical methods.
• Existing methods like ordinary multiple regression have limitations with binary data.

Purpose of the Study:

• To present logistic regression as a suitable statistical method for binary dependent variables.
• To illustrate the application of logistic regression using data from the DSM-III field trials.
• To demonstrate its capability in handling both categorical and continuous independent variables.

Main Methods:

• Logistic regression analysis was employed.
• The dependent variable was treatment type (behaviorally- vs. psychoanalytically-oriented psychotherapy).
• Patient and clinician characteristics served as independent variables.

Main Results:

• Logistic regression successfully analyzed the effects of independent variables on the binary treatment choice.
• The method produced estimated probabilities constrained between 0 and 1.
• The odds ratio was defined and used as a measure of association.

Conclusions:

• Logistic regression is a robust statistical tool for analyzing binary outcomes in clinical and research settings.
• It provides interpretable results, including probabilities and odds ratios.
• The method is applicable to diverse independent variables, enhancing its utility in psychological research.