1. Based on the background information, describe briefly how samples of students are obtained from each university. Speficially, address:
(a) How are students (from a given university) invited to take part in the survey? (Name two possible ways.)
(b) What must actually happen after a student receives an invitation for that student's opinions to be included in the sample/data? That is, what must the student do with the invitation? (Think carefully about this. This is important!)
2. In the newspaper article provided in the summary, the author indicates that the results of these rankings are not "scientific."
Consider a sample of students from a single university only (for example, a sample of ISU students, NOT considering how UNIVERSITIES are chosen) used to make inferences about the population of students from that university. Also consider the process by which data is gathered.
Using what you know from statistics and utilizing statistical terms, describe what is meant by that statement. That is, what is wrong with this sample or what is wrong with how the sample data was gathered?
3. Consider a sample of students from a single university, say for example, IU, that is used to make inferences about the population of students at that university (again, say IU).
Comment on possible ways that selection bias could occur in a sample of IU students. Answer these questions:
(a) What is selection bias?
(b) What about how the sample for a university (e.g. IU) is obtained could likely lead to selection bias?
(c) Give an example of how this might occur.
4. Again consider a sample of students from a single university, say, for example, Clemson students, that is used to make inferences about the population of students from that university (again, say Clemson).
Comment on any possible or probable response bias in the data, opinions and behaviors recorded by the sample of Clemson students. Answer these questions:
(a) What is response bias?
(b) What is implied from the article about how it may have occurred in this sample?
5. What would be wrong with using the statistics from the sample of students from a given university to make inferences about the opinions and behaviors of the population of all students at that university? Answer these questions:
(a) What is wrong with the sample? and
(b) What would it do to the statistics (and therefore the estimates) for the population?