Project Canada 1975-80 Panel Study
DOI
10.17605/OSF.IO/KT2GJCitation
Bibby, R. W. (2021, March 26). Project Canada 1975-80 Panel Study.Summary
The Project Canada Research Program has been carried out from the University of Lethbridge. National surveys of adults 18 and over have been conducted in 1975, 1980, 1985, 1990, and 1995. Adult surveys in 2000 will complete the program. The goal has been to generate extensive information on life in Canada, with specific attention given to social issues, intergroup relations, and religion.Data File
Cases: 1056Variables: 617
Weight Variable: None
Data Collection
1975-1995Funded By
The 1975 survey was carried out for a cost of about $14,000 and had four major sources: the United Church of Canada ($2,000), the Canadian Broadcasting Corporation ($3,000), the Solicitor General of Canada ($5,000), and the University of Lethbridge ($4,000). In 1980, the panel portion of the survey was made possible by grants from the Social Sciences and Humanities Research Council of Canada ($10,000) and the United Church of Canada ($2,000). The second phase of Project Canada 1980, which involved filling the core out into a full national sample, cost approximately $8,000 and was funded primarily by the University of Lethbridge. Project Canada 1985 was funded completely by the Social Sciences and Humanities Research Council of Canada ($45,000); Project Canada 1990 and Project Canada 1995 were both funded by The Lilly Endowment, Inc. (about $65,000 each).Collection Procedures
All five of the adult surveys have made use of self-administered questionnaires and have been conducted by mail over approximately a four-month period. Questionnaires have ranged from eleven to twenty pages in length, and have included 300 to 400 variables. With minor variations, the procedures have involved (1) mailing the questionnaire with a front-page cover letter, (2) sending a follow-up postcard, and (3) mailing a second questionnaire.Sampling Procedures
A representative sample of about 1,100 cases is sufficient to claim a confidence level of 95% and a confidence interval of four percentage points when generalizing to the Canadian adult population. Size and representativeness are the two key criteria in being able to generalize with accuracy from a sample to a population. Considerable care therefore has been taken to ensure that both standards have been met.Concerning size, an interest in provincial comparisons resulted in 1,917 cases being gathered in 1975; 1,482 in 1980; 1,630 in 1985; 1,472 in 1990; and 1,765 in 1995. With respect to representativeness, the nation has been stratified by province (10) and community size (greater than 100,000, 99,999-10,000, less than 10,000), with the sample drawn proportionate to the populations involved. As resources have improved, the number of communities being drawn on has increased from 30 in 1975 to 43 in 1980, 104 in 1985, 145 in 1990, and 228 in 1995. Participants have been randomly selected using telephone directories. Discrepancies between the sample and population characteristics have been corrected by weighting for provincial and community size, along with gender and age. Each of the five samples has been weighted down to about 1,200 cases in order to minimize the use of large weight factors (i.e., three or more).
All of the samples are highly representative of the Canadian population. Samples of this size and composition, should be accurate within about four percentage points on most questionnaire items, 19 times in 20 similar surveys. Comparisons with similar Gallup poll items, for example, have consistently found this to be the case.
A major interest of the ongoing national surveys has been to monitor social change and stability. Each survey sample since 1980 has consisted of (a) a core of people who participated in the previous survey and (b) new participants, who are used to create a full national sample of about 1,500 cases. For example, while the first 1975 survey was a typical cross-sectional survey with 1,917 participants, the Project Canada 1980 sample of 1,482 people included 1,056 who also had been involved in 1975.
The 1995 sample of 1,765 cases comprised 816 people who participated in previous surveys and 949 new cases. Of the 816, 400 had participated in the 1975 survey. They made up the ongoing core who have participated in all the surveys (236) and a special panel supplement (164), which was obtained through our adding as many of the original 1975 participants as we could whom we had "lost" between 1975 and 1995.
Various panels can be constructed from the surveys according to the five-year interval desired (e.g., 1975-85, 1980-90, 1990-95). While no claim is being made that these panels are representative of all Canadians, they do provide intriguing and novel data on the attitudes, outlooks and behaviour of a core of Canadians over the last quarter of the 20th century. The panels can be weighted as deemed desirable by data users.
For national surveys, the Project Canada return rates have been relatively high - 52% in 1975, 65% in 1980, and about 60% in 1985, 1990 and 1995. We tend to hear from about 65% of the people who have participated previously and just over 50% of those being contacted for the first time - favourable to the seldom - reported cooperation rates of (at best) around 65% obtained with face-to-face and telephone interviews.