In a survey of homemakers, the readership of Harper's was exaggerated and the readership of Modern Romance seemed much less than circulation figures indicated. Why would respondents incorrectly report their readership in this manner? Can you think of ways to avoid this bias?