Date of Award

2018

Degree Type

Thesis

Degree Name

Master of Science in Statistics

Department

Computer Science and Statistics

First Advisor

Prabhani Kuruppumullage Don

Abstract

Statistics, and other quantitative courses, are an essential part of general education and training in many fields. However, these courses routinely lead to stress and anxiety in many students. Students often turn to substance use to manage feelings of stress and anxiety, which has major public health concerns due to the short-term and long-term risks of substance use. This thesis explored the relationship between quantitative anxiety, general forms of anxiety, and alcohol use during the semester in which students were enrolled in an introductory quantitative methods course. Model training using the Spring 2018 data (n = 119) suggested a final model of general anxiety, quantitative anxiety, time, an anxiety by time interaction, and random intercepts, where quantitative anxiety and general anxiety were both positive predictors of alcohol use. The R2 value for the fixed effects was 0.12, and for the fixed and random effects was 0.65. The model was tested on a second independent sample (from Fall 2017, n = 19) where the mean squared error was 0.38, suggesting the model performed well in a second sample as the error term was close to zero. Examination of mean trajectory plots of the predicted versus actual values for alcohol use in the Fall 2017 dataset showed the model did well predicting some but not all patterns of change. These findings were further con- firmed by fitting a piecewise latent growth curve model to the Spring 2018 dataset, which showed great fit to the data. The results of this study suggest that quantitative anxiety is positive predictor of alcohol use even when taking into account general forms of anxiety. While this study does have several limitations, the results merit further investigation to explore the relationship between quantitative attitudes and substance use. Future studies should aim to replicate the findings of the current study while addressing the limitations, and assess if quantitative attitudes are related to the use of other substances common on college campuses.

Available for download on Saturday, July 27, 2019

Share

COinS