Question

Exercise 5.3 For each of the following opinion polls, (1) determine whether the poll results offer strong support for the pollster’s conclusion, and, if they don’t, (2) specify the source of the problem (sample too small, unrepresentative sample, or nonrandom sampling). Assume that the conducting of each survey is free of technical errors such as mistakes in data processing or improper polling interviews.

1. An Internet site featuring national and world news asks visitors to participate in the site’s “instant daily poll” of hot topics. The current polling question is: “Should the words ‘under God’ be stricken from the Pledge of Allegiance if its recitation is required of public school children?” Twelve thousand people visit the site on the day that the poll is taken. Of those, seven thousand answer no to the question. The site concludes that 58 percent of Americans oppose modifying the Pledge.

3.A prominent sociologist wants to determine the sexual attitudes of women aged twenty-five to forty-five. The main question to be explored is whether heterosexual women in this age group feel satisfied with their partners’ sexual performance. The sociologist interviews two hundred of her friends who belong to the target group. She also asks two hundred of her female colleagues at her college to complete and return a survey asking the key question. She gets seventy-eight completed surveys back from women in the target group. She finds that 75 percent of all the interviewees say that they are not satisfied with their partners’ performance. She concludes that most heterosexual women aged twenty-five to forty-five aren’t happy with the sexual performance of their partners.

5.A national women’s magazine publishes a questionnaire on sexual harassment in the workplace. Respondents are asked to complete the questionnaire and mail it in to the magazine. The magazine receives over twenty thousand completed questionnaires in the mail. Sixty-two percent of the respondents say that they’ve been sexually harassed at work. The magazine reports that most women have been sexually harassed at work.

Answer 1

Opinion Poll 1: No, with regard to the opinion poll on the Pledge of Allegiance, the poll results do not offer strong support to the pollster's conclusion because some poll organizations utilize internet surveys. This particular opinion poll was also taken from an internet site.

In internet surveys, a sample is taken from a large group of volunteers, and the consequential reactions are measured to reflect the demographics of the populace of interest. Contrary to this, the prevalent net polls draw on whoever wishes to take part, rather than a scientific-methodical sample of the population, so random people participate in these internet polls, they are, therefore, not normally considered professional and the result doesn't seem accurate.

Opinion Poll 3: No, with regard to the opinion poll on the sexual attitudes of women aged 25 to 45, the poll results do not offer strong support to the pollster's conclusion because of three reasons. Firstly, the group size who actually responded to the poll was too small, as many did not respond to the poll. If people do not respond to a poll, then the poll samples may not be considered as representative samples from the population. This is indicative that it is possible that the characteristics of the people who have declined to be interviewed may be different from those who actually gave the interviews. This means that the actual sample taken is biased in some ways. Furthermore, 78 women are too small a group to generalize on the entire group of 25 to 45-year-old women. Lastly, the participants were friends of the interviewer, and not randomly taken participants from the general public.

Opinion Poll 5: NO. With regard to the opinion poll on sexual harassment at work, the poll results do not offer strong support to the pollster's conclusion because, first of all, 20000 women is not large enough a number to generalize the results with the colossal number of women who go out to work. So, in this case the sample size is insufficient for this poll to substantiate the results effectively. Furthermore, if the percentage of women found to be sexually harassed at work was exceeding the given 62% and was a higher figure, ultimately the results could be said to be a stronger, more convincing figure. But again, that was not the case either. So the Pollsters conclusion could not be supported by these results.

3: