In partnership with the Public Health Agency of Canada’s Office of Behavioural Science (BeSciO), the Privy Council Office’s Impact and Innovation Unit (IIU) has launched a multi-wave public opinion research survey to better understand public health challenges in Canada. The 'Health, Attitudes, and Behavioural Insights Trends' (HABIT) survey, combines research methods from the field of behavioural science (BeSci) with robust policy analysis to help inform government policies, programs, and ongoing surveillance initatives. Data from this survey will provide a national perspective on cross-cutting Canadian public health outcomes and underlying public health behaviours, behavioural factors, social determinants
of health, and systems-level drivers.
The current project is the first wave of this cross-sectional study, gather data among people who reside in Canada (n = 2000) related to a range of topics central to the Government of Canada’s public health priorities. This survey will track a variety of public health-related behaviours and attitudes, providing an opportunity to analyze key factors predicting various public health outcomes as well as differences in those outcomes across various segments of the population. Sections of the survey include the following:
- General demographics & social determinants of health
- General health status
- Mental health and well-being
- Access to and use of various health services
- Substance use
- Health promotion
- Infectious disease prevention & management
- Views on public health system
- Perceived and experienced impacts of climate change on individual health
Pre-analysis plan: Is there a pre-analysis plan associated with this registration?
Refer to the data dictionary submission for a descritption of variables.
The analysis will be run in R. There will be two syntax files:
- Data cleaning: Rename variables, remove outliers, compute index variables for all question sets, and perform data reduction to create new variables (i.e., using an average composite or principal component scores).
- Statistical analysis: Implement the following analysis plan.
- Generate descriptive statistics for all variables (i.e., central tendency, distribution, and count), with and without splitting across primary socio-demographic characteristics.
- Visualize variables across primary socio-demographic splits.
- Use correlations and regressions to examine relationships across variables, with and without controlling for primary socio-demographic variables. Unless otherwise noted, regression models will control for socio-demographic variables, marked as “demographics”.
- Use appropriate statistical tests (e.g., chi-square, ANOVA) to identify significant differences across primary socio-demographic characteristics on key outcome variables.
- Where appropriate, data reduction techniques (e.g., PCA, Factor Analysis) will be applied on a select group of survey questions to best represent underlying psychological contructs.
- A bayesian model averaging approach will be used for exploratory analysis to identify variables with the highest posterior probability as significant predictors for dependent variables of interest.
- Use cluster analysis (e.g., Latent class, K-means) to identify audience segments related to public health.
Who is behind the project?
What is the project about?
26 November 2023