Question

The Nationwide Food Consumption Survey is the only survey on U.S household food consumption and individual diets. Done once a decade, it is the basis for decisions regrading food programs such as the multi billion dollar Food Stamps program. The survey conducted for the U.S Department of Agriculture in 1987-1988 did extensive interview that required from 800 minutes to over 5 hours to complete in exchange for a $2 reimbursement. As a result only about one-third of the people contacted participated. In addition, interviewers were inadequately trained, there was a high interviewer turnover, and the survey design was disrupted. The combination of low response rate and poor execution of the survey resulted in data with low credibility, prompting the General Accounting Office to question whether or not the data were usable.

a) What are the some potential sources of bias in this survey?

b) Explain how the problems you listed in (a) could influence the survey results.

Answer 1

1. Potential sources of bias in this survey are-

a) Inadequately trained interviewers

b) Low participation of the people

2. The inadequately trained interviewrs may result in poor execution of the survey as they are not trained they may ask questions in a different manner or they may record the responses incorrectly.

Secondly as low people particiapated in the survey, this means that the given sample is not a good representative of the population.. So this data should not be used.