Substance Abuse and Health: A Structural Equation Modeling Approach to Assess Latent Health Effects
| dc.contributor.advisor | Khan, Shahedul | |
| dc.contributor.advisor | Li, Longhai | |
| dc.contributor.committeeMember | Srinivasan, Raj | |
| dc.contributor.committeeMember | Guo, Xulin | |
| dc.creator | Islam, Naorin | |
| dc.date.accessioned | 2016-12-19T19:47:39Z | |
| dc.date.available | 2016-12-19T19:47:39Z | |
| dc.date.created | 2016-12 | |
| dc.date.issued | 2016-12-19 | |
| dc.date.submitted | December 2016 | |
| dc.date.updated | 2016-12-19T19:47:40Z | |
| dc.description.abstract | Background of the Study. Repeated use of a substance (alcohol or drug) may lead to mental and physical sickness, personality changes, insomnia, nausea, mood swings and other disturbances. The number of people addicted to alcohol and/or drug has been increasing every year at an alarming rate. Although the extent of abuse is not directly measurable (i.e., latent), statistical techniques allow us to describe such a hypothetical construct using available information. Objective. There are many factors potentially associated with substance abuse (e.g., smoking, education, cultural background). Although these variables are readily available in many studies, the cause (e.g., a measure of drug or alcohol abuse) is latent, with the observed variables being its manifestations. A measure of a latent health factor index could also be of particular interest. In this study, we investigate the effects of socio-demographic variables on substance (drug and alcohol) abuse and health in the Canadian population. In particular, the objective is to address the following questions: (a) What would be a reasonable hypothesis to explain causes of substance abusive behavior (i.e., cause and effect relationship)? (b) What model would adequately describe the cause-and-effect relationship between the observed variables and health and substance-related latent variables? (c) What covariates are significantly associated with alcohol and drug abusive environments and health status? Method. To describe the cause-and-effect relationship among substance abuse, health and socio-demographic variables, we consider structural equation modeling. One of the appealing features of this technique is that it provides a concise assessment of complex relationships. The idea is to formulate a hypothesis regarding such relationships based on prior knowledge about the problem at hand, and then evaluate this hypothesis using statistical techniques. The main goal is to develop a model/hypothesis which can adequately describe the interrelationships among these variables. Summary Results. The study is based on a survey conducted by Health Canada. We consider 2012 survey data for Saskatchewan and Manitoba, and then develop models to describe the complex relationships among three hypothetical constructs (drug and alcohol abusive environments and heath) and socio-demographic variables. One of the important findings of the study is that an increase in the severity of drug abusive environment may worsen the health of individuals. Another interesting finding is that smoking has no direct effect on health, but it may lead to an environment (alcohol or drug abusive) that could have negative impact on health. Based on our findings, we conclude that substance abuse may significantly deteriorate health. This research will provide policy-makers as well as the public with an understanding of the extent of impacts of substance abuse and relevant socio-demographic variables on health. | |
| dc.format.mimetype | application/pdf | |
| dc.identifier.uri | http://hdl.handle.net/10388/7632 | |
| dc.subject | Structural equation modeling | |
| dc.title | Substance Abuse and Health: A Structural Equation Modeling Approach to Assess Latent Health Effects | |
| dc.type | Thesis | |
| dc.type.material | text | |
| thesis.degree.department | Mathematics and Statistics | |
| thesis.degree.discipline | Mathematics | |
| thesis.degree.grantor | University of Saskatchewan | |
| thesis.degree.level | Masters | |
| thesis.degree.name | Master of Science (M.Sc.) |