Suppose we surveyed 1006 Canadian adults and found that 41% believed that Canadian values were more right wing than 10 years previously.

a) A newspaper reporting these results stated, “Canadian opinion on this issue ranges from 33% agreement for 18-to-34-year-olds to 48% agreement for those over 54 years of age. These results are accurate to within plus or minus 3.1%, 19 times out of 20.” Comment on the ethics of this statement as it relates to the ASA Ethical Guidelines in Appendix C.

b) Ontario has 39% of the Canadian population, whereas British Columbia has only 13%. Should we survey three times as many people in Ontario as in British Columbia in order for the confidence intervals on our results to be of similar width in the two provinces?