Statistical Analysis of Mental Health Factors

Abstract

This study aims to build a model that effectively predicts the presence of depression in 25,113 Canadian citizens ages 15 and over living in the ten provinces. Predictors were chosen for their potential to introduce significant stress into the lives of individuals. The tested variables obtained from the Canadian Community Health Survey (CCHS) of 2012 include current age, sex, marital status, BMI, frequency of alcohol consumption, presence of cancer, drug use, employment status, originating from Canada, and current school enrollment. These variables were subjected to a forward selection process in SPSS to fit a statistically significant binomial logistic regression model. The relationship between individual variables and depression was also examined through the Crosstabs function in SPSS. Results suggested that all chosen explanatory variables have a significant association with the response variable except presence of cancer, which also did not show significance in the final fitted model. Key Terms: Mood disorder, Depression, Mental Health, Canada

Discipline: Statistics

Faculty Mentor: Dr. Karen Buro