The Survey


Go directly to the online survey || Download the survey in PDF format


Introduction

The PASTS survey was designed to explore the importance ordinary Canadians ascribe to the past and history of their family, their country, and other aspects of their life; to inquire into the level of trust they have in various sources of historical information; and to document the way in which they engage with the past in their every day lives. The survey was designed to facilitate investigation of these issues by region as well as at the national level.


While the telephone interviews averaged 23 minutes, ten percent took 30 or more minutes to complete. Interviews were completed primarily at the Institute for Social Research (ISR) at York University in Toronto. About 15 percent of the interviews, representing most of the Quebec interviews, were completed by Jolicœur & Associés in Montreal.


Sample Design

A list of residential telephone numbers in Canada was used as a surrogate for a list of households in Canada. By randomly selecting telephone numbers we had a random selection of households. When there was more than one person 18 years of age or older living in a household, interviewers randomly selected the survey respondent by conducting the interview with the adult who would have the next birthday. The 2,000 interviews completed as part of the National Sample were allocated equally among five “regions” (Atlantic Canada, Quebec, Ontario, the Prairies and British Columbia). By allocating the same number of interviews in each region (400), we maximized our ability to make statistical comparisons between the regions.


Fortunately, Parks Canada also commissioned an additional 1,000 interviews in the metropolitan areas of Montreal, Toronto, Calgary, Edmonton, and Vancouver. Since the same questionnaire and respondent selection procedures were used for this metropolitan supplemental sample, the data have been combined with the national sample of 2,000 and a complete data file with 3,000 interviews has been prepared. Given the over representation of the smaller regions and the over representation of the five largest cities, it is necessary to weight the data before national estimates are made. (Information about the weights will be provided when the data sets are provided to users.)


Without having an enormous sample (and budget), no allocation of a sample to regions of the country is completely satisfactory. As with any Canada-wide survey, important groups such as First Nations that cross regional boundaries, recent immigrants who tend to cluster in big cities, and communities with long histories such as Acadians and many others important to understanding the role of the past in Canadians’ lives have limited representation. In order to obtain at least a glimpse into some of these communities, three supplemental samples of 100 interviews with Acadians in New Brunswick, 100 recent immigrants in the Region of Peel (immediately west of Toronto) and 100 members of First Nations in and around Saskatoon were completed. (For ease of analysis, these interviews have been added to the data file but they are not included in the weighted data used for national estimates.)

Web-Based Survey

A Web-based version of the survey that replicates the questions asked in the telephone survey has been developed. It is our hope that one of the uses of this version of the survey will be to collect interviews from other regions, groups and communities. We anticipate that the Web-based survey may also be a useful tool for educators. The survey can be viewed or completed by visiting the ISR Web site (http://www.isr.yorku.ca/projects/pasts1.html).


Questionnaire Design

A typical PASTS survey respondent answered about 70 questions. (The number of questions asked of each respondent varied somewhat as questions were asked or skipped depending on answers to previous questions). For most of the questions respondents either gave Yes/No answers or were asked to select a response from a list provided by the interviewer. For several key questions, however, respondents were encouraged to answer in their own words. (In the argot of the survey researcher, open-ended questions.) While many respondents were succinct, answering in a few words or a sentence, others gave longer and more complex answers. To ensure the answers to the open-ended questions were captured completely, they were audio taped and transcribed. About 90 percent of our respondents agreed to have their responses taped. When respondents did not agree to being taped, interviewers typed the gist of their responses to the open-ended questions directly into a computer file.


The questionnaire for the PASTS survey was developed by the research team over a number of months. Some of the questions are borrowed from, or inspired by, Rosenzweig and Thelen’s study in the United States Roy Rosenzweig and David Thelen, (Roy Rosenzweig and David Thelen, The Presence of the Past: popular uses of history in American life, Columbia University Press, 1988), and Hamilton and Ashton’s study in Australia, (Hamilton, Paula and Paul Ashton, Australians and the Past, Cultural History, 22, 2004). Other questions were designed specifically by the PASTS research team for the Canadian survey. Partners, advisory board members, and colleagues also provided useful suggestions. We developed at least twice as many questions as were used in the final questionnaire. In designing the questionnaire we were cognizant of the trade-off between survey length and willingness on the part of the public to participate in a survey. While a longer questionnaire would have allowed for coverage of some topics in greater depth, or the inclusion of more topics, it would have also reduced the proportion of Canadians willing to participate in the survey. We wanted to have as wide a cross section of Canadians as possible, including respondents with high and low levels of education, from families that had been in Canada for generations as well as more recent immigrants, the employed as well as those who did not work. A shorter survey allowed us to maximize our chances of getting average Canadians to compete the survey.


Six pretests of the questionnaire were competed between September 2006 and March 2007. Two of the pretests were conducted in French, the remaining four were in English. We completed four “think aloud” or cognitive group interviews where respondents were encouraged to explain their understanding of the intent of the questions and how they arrived at their answers. (Two of the cognitive tests were in English and two were in French.) Most of the study investigators were interviewed so they could evaluate the questionnaire from the viewpoint of a survey respondent. Investigators also listened in on pretest interviews and debriefed with the interviewers.


Data Collection

Computer Assisted Telephone Interviewing (CATI) techniques were used in the collection of the data. With CATI, a central computer delivers the questions to each interviewer’s work station and records the answers in a central data file. CATI automates the flow of the questionnaire, for example, skipping questions that are not appropriate based on answers to previous questions. This allows interviewers to focus on reading the questions, listening to the respondent, and entering responses correctly. We also used CATI to ameliorate any question order effects. For example, when asking respondents a series of questions about trustworthiness of various historical sources, some respondents used the answer to their first question as a benchmark and rated other sources relative to the ranking they gave for the source asked about in the initial question. In our survey, we used CATI to ensure that each source of historical information was found at the top, bottom and in the middle of the list of questions, thus removing order effects from overall ratings of trustworthiness.


Data collection for the national sample commenced in March 2007, and in September 2007 for the urban supplemental sample, and interviewing for these samples was completed in April 2008. Interviewing for the Aboriginal sample started in November 2007 and interviewing for the remaining two supplemental samples (Acadians and recent immigrants) started April 2008. Data collection for all three of these supplemental samples was completed by July 2008. In order to maximize the chances of getting a completed interview from each telephone number in the sample, call attempts were made during the day and the evening – for both week and weekend days. On average, 15 calls were made over many weeks before we gave up on the possibility of getting an interview. Efforts were made to “convert” refusals when respondents and/or households refused to participate when initially contacted by the interviewer. Using reverse directories, addresses were located for households in which a refusal was obtained and letters were sent to introduce the survey. Subsequent to the letter, an interviewer called and attempted to convince the potential respondents to complete the interview. Almost 20% or 550 of all the interviews were completed in households where the initial contact resulted in a refusal to participate in the interview.


The response rate, calculated as the number of completed interviews over the estimated number of eligible households contacted, was 55% for the national sample and 52% for the urban sample. The rates for the First Nations, the recent immigrant Peel sample and the Acadian samples were estimated at 65, 60 and 55 percent, respectively. In a time period of rapidly declining response rates, our rates compare favourably to those obtained in other surveys of this nature .

Data Availability

The research team is committed to posting a data file (as an SPSS file) for use by all interested parties when the project is completed.


David Northrup
Institute for Social Research
York University