Appendix - Survey of Election Officers for the 44^th Federal General Election

1. Methodological Notes

Sample and Sampling

The sample for this survey was provided by Elections Canada. It was based on the EC database of election officers who worked during the 44th federal general election. The total number of unique records in the database was 192,777. (Election officers were included in the database more than once if they worked in different positions or at different types of polls or polling places.) A sampling frame was developed based on the proportions in the database. The sampling frame included oversamples for targeted subgroups of the population to ensure adequate sample sizes for analysis.

The tables below present the target sample sizes by characteristics of the population.

Pre-test

The questionnaire was pre-tested by telephone. To pre-test the telephone questionnaire, respondents were first administered the survey and then asked a series of short follow-up questions. The debriefing following the survey provided an opportunity for respondents to offer feedback on the questionnaire. The follow-up questions were:

• Overall, how would you rate the extent to which the questions in this survey were easy to understand?
• Please tell me why you gave this rating to the survey. What specific words or questions did you find difficult to understand?

In total, 20 pre-test interviews were conducted by telephone. Respondents were able to participate in the official language of their choice. The survey was pre-tested on November 30 and December 1, 2021. The pre-test interviews were digitally recorded and reviewed by Phoenix SPI team members and Elections Canada officials.

There were no significant problems in terms of design or respondents' comprehension of the questions. The only issue was questionnaire length. As a result of the pre-test, 21 questions were removed from the questionnaire.

There was no formal pre-test of the web version of the questionnaire.

Data Collection

All fieldwork was conducted using computer-assisted telephone interviewing (CATI) and computer-assisted web interviewing (CAWI) technology. In total, a stratified random sample of 4,168 election officers was surveyed between December 10, 2021, and January 14, 2022. Of the 4,168 surveys administered, 2,490 were completed by telephone and 1,678 were completed online. Based on a sample of this size, the overall results can be considered accurate to within ±1.5%, 19 times out of 20. The margins of error for sub-samples discussed in the report are larger. Seventy-seven percent (77%, unweighted) of the surveys were completed in English, and 23% (unweighted) were completed in French.

The following specifications applied to the telephone interviewing:

• Interviews averaged 17.2 minutes.
• Calling was conducted at different times of the day and the week to maximize the opportunity to establish contact.
• Up to 13 call-backs were attempted to reach potential respondents before a sample record was retired. The average number of call attempts was 2.1.
• Interviewers mentioned in their introduction that the study was sponsored by Elections Canada.
• There was no calling between December 23 and 27, 2021, or between December 31, 2021, and January 4, 2022.

The following specifications applied to the web interviewing:

• The average web completion time was 12.9 minutes.
• A total of 7,805 election officers were invited by email. The average number of email contacts was 1.6.
• The email invitation mentioned that the study was sponsored by Elections Canada.
• There were no emails sent between December 23 and 27, 2021, and between December 31, 2021, and January 4, 2022.

All survey respondents were informed that their participation was voluntary, and that information collected was protected under the authority of the Privacy Act. They were also informed that the anonymized database of all responses could be shared with researchers who collaborate with Elections Canada.

The data collection was conducted in accordance with the standards set out by industry associations as well as applicable federal legislation, including the Personal Information Protection and Electronic Documents Act, Canada's private-sector privacy law.

Response Rate

The following table presents information about the final call dispositions for this survey and calculation of the response rate:

The response rate formula is calculated as follows: [response rate = R/(U+IS+R)]. This means that the response rate is calculated as the number of responding units [R] divided by the number of unresolved [U] numbers plus in-scope [IS] non-responding households and individuals plus responding units [R].

Survey Weighting

The survey data were weighted to accurately reflect the distribution of election officers by region, type of position, type of poll and type of polling station. The table below shows the unweighted and weighted proportions for the variables used to create the weights:

• Previous
• Table of Contents
• Next