Question

A survey based on a random sample of 176 Hope students investigated whether there was evidence of a relationship between Gender (Male, Female) and Political Party Affiliation (Republican, not republican) in the population of all Hope students. Of 79 female students, 63.3% (50/79) were republican. Of 97 male students 67.0% (65/97) were republican.

42. (5 points) Comment on the population and how well-represented it is by the sample and why/why not.

43. (5 points) Comment on whether a cause-effect conclusion is possible and why/why not

Answer 1

42. 176is perhaps small sample size, compared to the population of Hope students. So probably the sample is not well represented. However, this size is okay for any pilot studies or to draw preliminary estimates/ inferences before starting the main study.

43. We cannot infer a cause-effect relationship from this. At most we can test whether there is relation between gender and political party affilations or not. Suppose our test shows that males are more likely to be Republicians. This doesn't guarantee a cause-effect relationship. That is, we can't say that being male 'causes' the person be a republician. It's just that more males are republicians. There might be some confounding variables at play which went unaccounted.