Question: Sample Evaluation. Your political consulting group, Electem Inc., is assisting a local political candidate, Alice Evans, to determine the likelihood of her election to office during the upcoming campaign. She would like to have an extremely high level of confidence (95 percent) that she would receive more than a majority of the ballots cast. After conducting an extensive survey, you have determined that 53 percent of voters would prefer Evans with a precision of 5 percent and a corresponding reliability of 95 percent. In determining this information, your notes of the sampling process revealed the following information:
1. The district that she would represent is composed of eight neighborhoods. You randomly chose four of these neighborhoods and, within each, had workers poll voters (using a door-to-door technique) from the first 25 households that responded.
2. Your workers canvassed the neighborhoods during the day from 12 p.m. to 3 p.m.
3. Your workers asked voters to respond to the following question: "Do you support Alice Evans during the upcoming election?"
4. In one neighborhood (a community with restricted access to nonresidents), your workers could not obtain a sample of 25 voters using a door-to-door polling technique. As a result, they used a telephone survey to obtain the 25 responses for this neighborhood.
5. Some of your workers observed that, at homes indicating they would support Evans for office, campaign signs for her chief rival were present. The workers were surprised that these same households indicated they would support Evans.
6. George Clinton, one of your workers, indicated that he misunderstood the survey instructions, which indicated that a voter could indicate "yes," "no," or "undecided." George thought that he was required to obtain a "yes" or "no" answer; when voters indicated they were undecided, he pressed them to make a decision (which they ultimately did). 7. Billy Bush, another worker, did not verify the names of the individuals to whom he was speaking.
Required: a. Based on the sample results, without considering any of the information included in (1)-(7), what would you advise Evans about her electability?
b. How would your advice to her change if she established the following levels of reliability (assume that the sample estimate of 53 percent is unchanged)?
1. Reliability = 99 percent, precision = 8 percent.
2. Reliability = 90 percent, precision = 2 percent.
c. Summarize why your advice to Evans would change based on the reliability and precision noted in (b). d. For each of the issues included in your sampling notes, indicate how that issue could affect the advice you provide to Evans and her ability to rely on the sample evidence.