Survey of Electors Following the 42nd General Election
R.A. Malatest & Associates (Malatest) has administered and analyzed the Survey on the 42nd General Election on behalf of Elections Canada. Malatest has prepared this narrative report to provide a straightforward description of survey findings. This section describes the context in which the survey took place and provides a summary of the methodology used for survey administration and analysis, as well as a few notes to the reader on the terminology used throughout the report.
Elections Canada commissioned a 20-minute national public opinion survey to capture electors' perceptions of the Agency's services leading up to and during the 42nd general election. Specifically, the Survey on the 42nd General Election sought to gather information on electors' knowledge, experiences and attitudes with respect to Elections Canada's services, as well as electors' experiences with the electoral process in general. The survey questionnaire was revised to reflect administrative and legislative changes in the electoral process. The survey considered whether electors' perception of identification requirements has changed since the 41st general election.
The survey also sought feedback on Elections Canada's approaches to electors with disabilities. After extensive consultation with national and provincial disability organizations across Canada, Elections Canada launched its Advisory Group for Disability Issues in February 2014; these consultations led to changes for the 2015 general election, including communications around how, when and where people with disabilities can register and vote.
Prior to launching full survey administration, Malatest field-tested the survey on October 27 and 28, 2015, with 49 English-speaking respondents.Footnote 1 The field test measured how well the questionnaire performed in terms of survey length, flow of questions, and clarity of content, as well as how well the methodology performed in terms of recruiting participants, sampling Canadian electors, and recording data. After completing the field test and implementing minor modifications to the programmed questionnaire, Malatest began full survey administration. The survey was administered between late October 2015 and mid-January 2016. The survey averaged 17˝ minutes. Each record was attempted up to 10 times.
In total, Malatest collected data from 3,516 eligible voters, including at least 500 from each of four subgroups of interest to Elections Canada: Aboriginal electors, young adults aged 18 to 34,Footnote 2 electors with disabilities, and foreign-born electors. In the end, 2,823 completes were obtained from a random stratified sample and were used for most analyses presented in this document. The remaining 693 completes were non-random in that specifically targeted subgroups of interest and were used only when considering the impact of age (young adults versus adults aged 35 or older), place of birth (Canadian versus foreign-born), or Aboriginal status.Footnote 3 The Marketing Research and Intelligence Association response rate for this survey was 19.5%, while the margin of error for overall results is +/- 1.8% (19 times out of 20). Note that margins of error for subgroups will be greater.
At the conclusion of the survey, Malatest compiled all data from completed telephone interviews into an SPSS database. Malatest then cleaned and coded the data, and developed weights for analysis. Most findings in this report are based on the random stratified sample weighted by age and gender. Exceptions exist for the analyses by subgroups of interest, where due to oversampling the comparisons are not weighted.Footnote 4
For more detail on the methodology employed, see the associated Methodology Report under separate cover.
A limitation of surveys of this nature is that they tend to overrepresent respondents who indicated that they had voted. Among survey respondents, the self-reported turnout was 90%, while the actual turnout rate among all eligible voters was much lower; reported as 68.3% by Elections Canada.
Two factors may be responsible for these overestimated turnout rates. For one, people who vote are more likely than non-voters to participate in studies about voting. Second, some respondents may have demonstrated social desirability bias, in that they did not in fact vote but reported doing so to present themselves in a more positive light. The net result of these factors is twofold. For one, groups generally less likely to vote are less represented in the survey sample than others. Additionally, non-voters may be erroneously considered as voters.
Malatest made all reasonable efforts possible to garner the views of non-voters, including ensuring 500 voters for each subgroup among respondents, and weighting the final database by age and gender to improve the representativeness of the sample. Weighting by age in particular improves representation of non-voters in the random stratified sample, since young adults are both less likely to vote and less likely to participate in post-election evaluations.
2.4 Notes for Readers
Readers will need to understand a number of terms used throughout this report. The terms "electors," "respondents" and "eligible voters" are all used interchangeably to denote all survey participants. The term "voters" specifically denotes survey participants who reported voting in the 42nd general election.
Income groups are used for analysis of socio-economic demographics, and are referred to as "higher-income," "medium-income," and "lower-income." The specific breakdown used for these groups is as follows:
- Lower-income refers to respondents living in households with income of less than $40,000 ($40k<);
- Middle-income refers to respondents living in households with income between $40,000 and $80,000 ($40k–$80k); and
- Higher-income refers to respondents living in households with income of over $80,000 (>$80k).
It should be noted that throughout the report, income was used for comparison, while education was not. It has been found that there was significant overlap between these two socio-economic indicators, and income was perceived as a stronger driver than education. As a result, findings by income bracket are accompanied by similar findings by education attainment.
Age groups include "young adults" and "older adults." The specific breakdown used for these groups is:
- Young adults refer to respondents aged 18 to 34;Footnote 5 and
- Older adults refer to respondents aged 35 and older.
Aboriginal respondents refer to those who self-identified as being of First Nations, Inuit or Métis descent. Aboriginal electors living on-reserve were compared with those living off-reserve to determine whether there was a significant difference. A few significant differences between the two groups were detected. These few differences between Aboriginal electors living on- and off-reserve are discussed in the report.
A more expansive definition of persons with disabilities, based on the Canadian Survey on Disability, was used in the 2015 survey.Footnote 6 Similarly, electors with a disability are respondents who self-identified as having one of a number of conditions, including:
- Blind or visual impairment;
- Impaired co-ordination or dexterity;
- Deaf or hard of hearing;
- Impaired mobility;
- Speech impairment;
- Developmental or intellectual disability;
- Emotional/psychological/mental health condition;
- Chronic pain; and
- Any other conditions that respondents considered a disability.
Readers should note that tables may not add up to 100% due to rounding or due to respondents being able to provide more than one response.
Finally, comparison with the years 2008 and 2011 are supported by data from the Survey of Electors Following the 40th General Election report and the Survey of Electors Following the 41st General Election, both available on Elections Canada's website.Footnote 7, Footnote 8 These comparisons were made when possible.
Return to source of Footnote 1 The election was held on October 19.
Return to source of Footnote 2 Previous surveys of electors used the 18-to-24 age range to oversample the young adults subgroup.
Return to source of Footnote 3 Oversampling was not necessary for electors with disabilities, since enough respondents from that subgroup were reached through random stratified sample.
Return to source of Footnote 4 The use of weights is further described in the methodological report.
Return to source of Footnote 5 As previously mentioned, previous surveys used the 18–24 range to define young adults.
Return to source of Footnote 6 A survey by Statistics Canada. See www5.statcan.gc.ca/olc-cel/olc.action?objId=89-654-X&objType=2&lang=en&limit=0.
Return to source of Footnote 7 Elections Canada, Survey of Electors Following the 40th General Election, March 2008; www.elections.ca/res/rec/eval/40eval/survey-electors_e.pdf
Return to source of Footnote 8 Elections Canada, Survey of Electors Following the 41st General Election, October 2011; www.elections.ca/res/rec/eval/pes2011/elsvy/src_e.pdf