Question

You are interested in conducting a survey at your school to determine if gender or ethnicity is related to perceptions about educational opportunity. The school has 1,800 students but you only have resources to interview 100. The student body is approximately equally divided between males and females. Twelve percent of the school's population is Asian American, 8% is African American, 10% is Native American, and 70% is Caucasian. You are interested in being able to generalize the results of your survey to the entire student body. Describe a technique for obtaining a sample of size 100 that would have a reasonable chance of producing results that could be generalized to the entire school. Suppose you sent the surveys out to classes and asked the students to complete them and return them to the main office. Would the returns be likely to be biased in any direction? If so, why? In what direction? Suppose you decide to conduct the survey by personal interview. Members of the student council, all of whom happen to be Caucasian, will conduct personal interviews with the selected students. One of the survey questions reads, "Do you agree or disagree with the following statement: All students, regardless of ethnicity, have equal opportunity to become part of student government." What potential bias is present in this question?

Answer 1

Answers

We can use cluster sampling here. In this case, the population is divided into sampling units or clusters, for example, zone within educational or geographical areas. Having divided the population into clusters, we can use simple random sampling to select the clusters to be examined. Consider Asian American as one cluster, African american as another cluster and so on. According to the size of the cluster randomly select individuals from each cluster for a total of 100 individuals.

Generally there would be bias in accordance to the differnce in ethnicity. Also if the selected students are in a younger, then the teachers opnion would play a very big role there.

The question itself is a source of bias and it seems to give an idea to the students to think in these lines. The question can be better worded to get a correct answer. Even if there is equal opportunity, some students may read the question and give a negative answer.