Solution-Type of nonprobability sample | Homework Help

1.”Four out of five doctors prefer Fred’s Aspirin.” This is an example of a type of nonprobability sample called a judgment sample. In a judgment sample, you get the opinions of preselected experts in the subject matter. Unfortunately, no matter how well informed the experts are, like with all nonprobability samples, the results cannot be generalized to the population. In aIDition, we’re not dealing with exhaustive categories. The doctors prefer Fred’s Aspirin to what? There are many different kinds of aspirin sold throughout the country and the world. Some doctors may not even like aspirin, maybe preferring ibuprofen. Also, some doctors may not even have access to Fred’s Aspirin in their area. There’s no way that four out of every five doctors prefers Fred’s Aspirin. Who picks the doctors and how? Is it just one sample of five doctors or many?

Are the doctors from one area or many? Etc., etc.

There are many examples of this kind of misleading statement:

Four out of five dentists prefer Trident gum.

Four out of five dentists recommend Crest toothpaste.

Four out of five parents spank their children. (According to parents.com)

2.

First, the statement does not state where these doctors are selected from. Could the doctors selected work for Fred”s Aspirin or be sponsored by Fred”s Aspirin?

Coverage error would be selection bias and may be unethical if the survey is more favorable to the survey”s sponsor (Fred”s Aspirin). It is possible that they are not including doctors that are not affiliated with Fred”s Aspirin.

There would also be a nonresponse error if Fred”s Aspirin deliberately avoided surveying people who would potentially not reflect positive survey results. This would also be unethical. There would also be a sampling error if Fred”s Aspirin did not disclose sample size and margin of error in an effort to promote their own product in the viewpoint they want to advertise.

Measurement error can include

1) leading questions promoting their viewpoint,

2) Hawthorne effect from a biased interviewer,

3) participant knowingly giving false responses.