Language selector

Appendix A: Methodology

Page controls

Page content

The OHRC commissioned the Environics Research Group to do a public opinion survey on human rights in Ontario. The OHRC followed the Ontario Government procurement process for research services and the Environics Research Group was the successful vendor of record.

Environics conducted the survey between January 24 and February 2, 2017, and then provided the OHRC with cross-tabulation data tables and an analysis of findings along with the complete survey data file.

The cross-tabulation data tables and the complete survey data file are available on the OHRC’s website for download and additional research. The data tables are numbered and organized based on the survey questionnaire and the information in this report is endnoted accordingly. The information in this report relies on the data identified in the cross-tabulation tables as statistically significant by the Environics Research Group.

Survey method

Environics conducted the survey online using a sample of 1,501 people age 18 and older, drawn from a panel of over 100,000 members. The OHRC chose the online method to avoid a “social desirability” bias, where people tell telephone interviewers what they think is the more socially desirable response. In a more anonymous online survey, people can be more honest about how they feel on sensitive issues like human rights and discrimination. Online panel surveys are a widely accepted research methodology.

Target population

The survey’s objective was to gather data from a representative sample of Ontario residents (18 years and over) that was large enough to be extrapolated to the full population with a reasonable degree of confidence, and that would allow for analysis by important subgroups. A sample size of 1,500 can provide meaningful and statistically reliable results for important segments of the population, whether this is by region, gender, age or other relevant demographic characteristics such as education level, religious practices, LGBTQ identity, presence of a disability, and ethnicity. These factors (and others) were expected to be important differentiators in opinions about human rights in Ontario.

The survey’s sample size presents a challenge when considering another important factor: how discrimination may be unique or distinct when it occurs based on two or more grounds that are protected under the Code. The complete survey data file is available for researchers who wish to do a more in-depth intersectional analysis.

The results of an online panel survey cannot be quoted in terms of “margin of sampling error” in accordance with industry practices.[53] However, the final data were weighted so that the results accurately reflect opinions across strata proportionate to the actual population.

Respondent profile

Table 11: Regional distribution of survey respondents

  Sample Size
(Weighted)
Sample Size
(Unweighted)
Population*
East/Central Ontario 285 302 20%
Greater Toronto Area GTA 720 661 45%
City of Toronto 315 310 21%
Outer GTA 405 351 24%
South western Ontario 405 420 29%
Northern Ontario 90 118 6%
Total 1501 1501 100%
  * Population %
based on 2011 Census

 

Graph 5: Gender identity distribution of survey respondents

This pie graph shows that 52% of respondents identified men, and 48% identified as women.

 

 

Graph 6: Age distribution of survey respondents

This bar graph shows that 21% of respondents were 18 – 29; 25% were 30 – 34; 28% were 45 – 59; and 26% were 60 or older.

 

Graph 7: Education level of survey respondents

This bar graph shows that 15% of respondents had a high school diploma or less; 40% had college/some university; 30% had undergraduate university; and 15% had post graduate education.

 

Table 12: Household income of survey respondents

Earn <$40,000 22%
Earn $40k Ð<$60k 17%
Earn $60k - <$80k 17%
Earn $80k - <$100k 13%
Earn $100k - <$150k 19%
Earn $150k+ 13%

 

Graph 8: Birthplace of survey respondents

 This pie graph shows that 78% of respondents were born in Canada, and 22% were born outside of Canada.

Table 13: Race, ethnicity, religion and disability status of survey respondents
Group % of Ontario population* Expected n Actual n (unweighted)
Racialized/Ethnic groups  
Total racialized groups 25.9 389 363
South Asian 7.6 114 63
Chinese / South East Asian 5 75 161
Black 4.3 65 59
Indigenous 3.5 53 59
Religious groups  
Muslim 4.5 68 36
Hindu 2.9 44 18
Jewish 1.5 23 37
Sikh 1.4 21 11
Disability status  
Persons with a disability 15.5%** 233 332
  * Based on the 2011 National Household Survey

** Based on information from the Accessibility Directorate of Ontario

 

Questionnaire design

The OHRC based its survey questionnaire on one designed for the Quebec Human Rights Commission through a partnership with several universities. Some of the questions were modified for Ontario's context and the particular needs of the OHRC.

The survey asked respondents 19 questions about awareness of human rights and discrimination, attitudes towards groups and their accommodation needs, and experience with discrimination, as well as 17 demographic questions about level of education, income, labour force participation, creed, age, ethnicity/ancestry/place of birth, current postal code, disability, gender identity, sex-assigned at birth, and lived gender identity.[54]

The OHRC recognizes that the meaning and use of words related to people’s identities can evolve and change over time. Certain terms in the questionnaire were chosen to ensure common understanding across respondents. Some terminology and statements were intently provocative, to elicit the candid views of respondents. In this report, we have replaced some of those terms to respect how individuals and groups identify themselves today.

The survey was available in English and French. Respondents were offered the opportunity to communicate by telephone with the polling firm if they had any difficulty completing the survey online.


[53] As there is no source of random e-mail addresses, this survey used an online panel. Canada’s research industry association, the Marketing Research and Intelligence Association (MRIA), has recently issued a new “code of practice” for its members stating that because panel-based surveys are not based on random probability samples, their results cannot be quoted in terms of “margin of sampling error,” as is used for probability-based telephone surveys.

[54] Survey Q.E (gender identity), Q.H (sex assigned at birth) and Q.I (lived gender identity) are based on a multi-dimensional sex/gender measure recommended in: Transgender-inclusive measures of sex/gender for population surveys: Mixed-methods evaluation and recommendations. Greta R. Bauer, Jessica Braimoh, Ayden I. Scheim, Christoffer Dharma. Published: May 25, 2017https://doi.org/10.1371/journal.pone.0178043.

Attachments

Book Prev / Next Navigation