This study investigates disparities in self-rated mental health among young adults aged 18-39 in Canada, across socio-demographic characteristics, using data from the 2017 General Social Survey, Public Microdata Files. After adjusting for age, household income, and other confounders, the results show that being female and single or divorced/ widowed were significantly associated with higher odds of reporting poor or fair mental health compared to males and those who were married or in common-law relationships, respectively.
In contrast, higher household income was protective against poor/fair mental health, with individuals in households earning $100k or more having significantly lower odds of reporting poor/fair mental health compared to those with a household income of less than $50k. Being an immigrant was also associated with lower odds of reporting poor or fair mental health compared to non-immigrants.
Regarding education, having a university degree was significantly associated with lower odds of poor/fair mental health compared to individuals with less than a high school education. No significant associations were found between visible minority status, middle household income, or educational attainment at the high school/trade/college level and poor/fair mental health.
The data used in this study comes from the 2017 General Social Survey (Cycle 31): Family, a Statistics Canada's Public Use Microdata Files (PUMFs). The data is obtained through ODESI, a service provided by the Ontario Council of University Libraries.
Access to ODESI is restricted to users with a DLI License, and it can be used for statistical and research purposes. The terms of the license can be found here.
McGill University's CAnD3 initiative has a license to use the data for training purposes. Those outside McGill University should not use the data provided through CAnD3's training activities for purposes unrelated to CAnD3 training. Those from another DLI institution should download the data using ODESI through their institution.
The codebook for this dataset is publicly available through ODESI here.
You can also access the codebook directly from Statistics Canada here.
- Statistics Canada. 2020. General Social Survey, Cycle 31, 2017 [Canada]: Family. See here.
| File | Description |
|---|---|
| Files.zip | Zip folder containing all project files. |
| Output Tables | Folder containing tables of sample characteristics, cross-tab, and logistic regression output. |
| Instructions for Reproducing Analysis.Rmd | RMarkdown file with instructions for reproducing the analysis. |
| Instructions for Reproducing Analysis.pdf | PDF version of the instructions for reproducing the analysis. |
| Code for Data Wrangling.R | R script for data recoding and handling missing cases. |
| Code for Descriptive Statistics.R | R script for generating descriptive statistics. |
| Code for Logistic Regression.R | R script for chi-square tests and logistic regression. |
| README.md | Project overview. |
| CAnD3 Reproducible.Rproj | R project file. |
For detailed instructions on how to reproduce the analysis, please refer to the Instructions for Reproducing Analysis.Rmd file available in this repository.The instructions file will guide you through running the R Scripts for data cleaning, descriptive statistics, and logistic regression analysis.
RandRStudio
tidyversesjPlotgtsummarybroom
To install the required R packages, run:
install.packages(c("tidyverse", "sjPlot", "gt", "gtsummary", "broom"))Operating System: Darwin 23.6.0 (macOS)Kernel Version: Darwin Kernel Version 23.6.0Memory: 16 GBProcessor: ARM64
- R Version: R version 4.4.1 (2024-06-14)