Canadian Postsecondary Education Alcohol and Drug Use Survey: Technical notes

On this page

• Data source
• Eligibility criteria
• Weighting
• Limitations
• Data analysis and reporting
• Key definitions

Data source

This report uses data from the Canadian Postsecondary Education Alcohol and Drug Use Survey (CPADS). It includes data from the second and third cycle of the survey (see Table 1 below). To increase comparability between cycles, these results focus on 17- to 25-year-olds.

Table 1: Sample size of each cycle of the Canadian Postsecondary Education Alcohol and Drug Use Survey (CPADS)

School year	Ages recruited	Total sample size	Sample size included in analysis *	Number of participating institutions
2019–2020	17- to 25-year-olds	21,297	21,297	41
2021–2022	Ages 16+	40,931	31,643	45
2024–2025	Ages 16+	29,371	22,961	43

*: 17- to 25-year-olds

Eligibility criteria

School eligibility

For schools to be eligible, they had to:

1. have a registrar office
2. have more than 500 students
3. be a not-for-profit or private school
4. not offer exclusively online courses
5. be non-theological or non-military institutions

Student eligibility

For students to be eligible, they had to be at least 16 years old and studying online or in-person in Canada.

Weighting

This analysis used survey weights to ensure the data are representative of the student population at each school by sex and age. Results from each school were then further weighted to align with the overall student population of the regions captured in the survey.

Limitations

• Data availability differs by cycle:
  • 2019-20: No data were available from British Columbia, Saskatchewan, or the territories.
  • 2021-22: No data were available from the territories.
  • 2024-25: No data were available from the Northwest Territories or Newfoundland and Labrador.
• The survey title reveals that the survey is about substance use. This could have attracted respondents who are already more interested or knowledgeable about the topic, which might have affected the results.
• CPADS relies on self-reported data, which may contain some inaccuracies. People filling out the survey may misremember, make mistakes when entering information, or answer in ways they think are more socially acceptable.

Data analysis and reporting

• Estimates are based on students aged 17 to 25 years old.
• Due to rounding, percentages may not add up to exactly 100.
• Analysts used the modified Wilson method to construct confidence intervals.
• Analysts used the chi-square test to examine differences between cycles and sociodemographic groups.
• Responses from students in the same school tend to be similar, or “correlated”. To account for this and avoid underestimating variability, the analysis treated each school as a primary sampling unit and applied a cluster-based variance correction to reflect the complex survey design. The CPADS survey is not based on random sampling as schools and students choose whether to participate (“self-selection”). Because of this, the variance estimates should be interpreted with caution, as they may differ from the true variance.
• Important note: Estimates and variances may differ from earlier CPADS publications due to differences in denominators and the use of the cluster-based variance correction.

Key definitions

LGBTQ2+

An acronym representing diverse gender and sexual identities, including Lesbian, Gay, Bisexual, Transgender, Queer, and Two-Spirit people, with the "2" recognizing Indigenous Two-Spirit identities and the "+" including other non-binary, asexual, intersex, and pansexual individuals not explicitly mentioned.

ASSIST

An international group of researchers specialised in higher-risk substance use developed the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) for the World Health Organization (WHO) to help detect and manage substance use and related problems in primary and general medical care settings. The CPADS cannabis module adopts questions from ASSIST to screen respondents for higher-risk cannabis use. Respondents are categorised based on their consumption as follows:

1. Low risk of developing health and other problems (score of 0-3)
2. Moderate risk of developing health and other problems (score of 4-26)
3. High risk of developing health and other problems and indicative of a substance dependence (score of 27+)