| Abstract: | Over the past two decades or so, survey research findings on religious activities and beliefs have been confusing, with the results of researchers and private pollsters frequently differing from those of Statistics Canada. In this research note, the authors use four national omnibus surveys conducted in 2012 to explore the extent to which such differences are due to measurement variations. Finding few noteworthy differences, they proceed to focus on the samples being used, and draw on illustrative data from 2012 and 2015 to argue that the differences in findings may in large part be due to differences in samples. The resolution of sampling problems, they acknowledge, is extremely difficult, but nonetheless is a goal that has to be pursued. |
