The New York Times did a special report on polling that was carried in papers ac

Question

The New York Times did a special report on polling that was carried in papers across the nation. The article pointed out how readily the results of a survey can be manipulated. Some features that can influence the results of a poll include the following: the number of possible responses, the phrasing of the question, the sampling techniques used (voluntary response or sample designed to be representative), the fact that words may mean different things to different people, the questions that precede the question of interest, and finally, the fact that respondents can offer opinions on issues they know nothing about.    (b) Consider this question: "Do you think fines for running stop signs should be doubled?" Do you think the response would be different if the question "Have you ever run a stop sign?" preceded the question about fines?                                        (c) Consider this question: "Do you watch too much television?" What do you think the responses would be if the only responses possible were yes or no?                         D)What do you think the responses would be if the possible responses were rarely, sometimes, or frequently?

Explanation / Answer

ANSWER

(b) Part

Answer is 'yes'

Justification:

If the respondent had indeed run a stop sign, it is very likely that he/she would respond in negative to doubling the penalty because he/she could as well be a victim of the doubling.

However, this lacuna can be overcome by analysing the response data in two different groups - one where the respondent had run a stop sign and the other where respondent had not run a stop sign. A Contigency Chi-square test of significance on the response data would guide on whether the conclusion can be drawn on the pooled data or has to be done separately.

(c) and (d) Parts.   

The word 'too much' is absolutely vague and hence the response data would not be reliable to draw any meaningful and realistic conclusion. (d) part would yield a better result.

A much better response generator would be to give 4 or 5 levels in terms of number of hours spent on TV viewing over a fixed period, say past 7 days or past 15 days etc.

Incidentally, what is supposed to be (a) part????

Related Questions

Navigate

• Browse (All)
• Browse T
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.