Technical Specifications

• An online survey was administered to 3,400 Canadians aged 18 and older. In previous waves of this research, the primary method of data collection was telephone. While this year’s data is not directly comparable to that of previous years, migrating the tracking survey online allows for better measurement of aided brand awareness and facilitates direct comparisons with data from the U.S. Environmental Protection Agency’s ENERGY STAR survey.
• The survey averaged 15 minutes to complete, and the fieldwork was conducted between December 1 and 21, 2018.
• The sample quotas and survey completions were as follows:

• The sample design took into consideration the need to test awareness of province-specific advertising campaigns in Nova Scotia, New Brunswick and Alberta.
• The survey questionnaire was programmed using computer-assisted web interviewing (CAWI) technology. The programming was tested for skip logic by the initial programmer and by a second senior programmer. The script was then tested through random data generation.
• Following survey best practices, the questionnaire was pre-tested in advance of the fieldwork to ensure that it measured what it was intended to measure. There were 10 completions in each official language. No issues or concerns arose during the pre-test. In addition to the pre-test, the survey programming was tested using a random number generator and the output verified prior to the launch of the fieldwork. Following the fieldwork, the data were cleaned and checked using SPSS syntax. The review assessed response ranges and the length of time taken to complete the survey to identify any respondent who took an unreasonably short time answering and thus evidently “straightlined” responses. Any cases flagged for data quality were replaced prior to the weighting and tabulation of the data.
• The sample was drawn from Research Now’s panel of 450,000 online Canadians. The results of the online survey cannot be generalized to the population because the survey sample is a non-probability sample.
• Panellists were rewarded for taking part in the survey per the panel’s incentive program, which is structured to reflect the length of survey and the nature of the sample.
• Survey data have been weighted by region, age and gender to ensure that results are reflective of the Canadian population. Population figures from Statistics Canada – 2016 Census were used to construct the weights.
• Because the sample is based on those who initially self-selected for participation in the panel, no estimates of sampling error can be calculated.
• Sponsorship of the study was revealed (i.e., the Government of Canada).
• The estimated response rate was 17%, calculated using the formula outlined in the Standards for the Conduct of Government of Canada Public Opinion Research (Online Surveys).

• All steps of the project complied with market research industry standards and with the Standards for the Conduct of Government of Canada Public Opinion Research.

The potential for non-response bias was assessed by comparing the characteristics of respondents through unweighted and weighted data. The characteristics evaluated were gender and age. The table below presents a profile of the survey sample, unweighted and then weighted to match the population.

As is typically the case for general population telephone surveys, older individuals (those aged 55+) are overrepresented and younger individuals (those aged under 55) are underrepresented in the survey sample.