When a sample is constructed in this way, the magazine is relying on their readers to send the survey back. The readers can choose whether or not to do so. Thus, the reader are self-selecting themselves into this study, which results in a self-selected sample.

There are many drawbacks to a self-selected sample. First, not enough respondents could return the survey, so analysis of the responses may not draw useful conclusions. Second, the responses may be biased towards the fringes of the distribution of opinions rather than an even spread.

The answer is therefore b. a self-selected sample.