The owner of a large company is conducting a survey about job satisfaction, including questions about salary, hours, stress, and other conditions. There are about eight hundred employees in the company, including one hundred executive positions, six hundred fifty middle-management positions, and fifty custodial positions. The owner wants to include about ten percent of his employees in the survey sample.

Identify a sampling method that might produce a biased sample. Explain why using the method for this survey would be a poor choice.

Question

The owner of a large company is conducting a survey about job satisfaction, including questions about salary, hours, stress, and other conditions. There are about eight hundred employees in the company, including one hundred executive positions, six hundred fifty middle-management positions, and fifty custodial positions. The owner wants to include about ten percent of his employees in the survey sample.

Identify a sampling method that might produce a biased sample. Explain why using the method for this survey would be a poor choice.

blues · Answer

The sample should be drawn proportionally from all levels of the staff. It would be no good to sample a disproportionally large number of executives, for example - they might be happier with their jobs than the others (higher salaries, better benefits) or they might be less happy with their jobs than the others (stuck behind a desk all day, unhappy with corporate environment, etc). The same arguments against disproportional sampling could be applied to any of the other groups.

For one example.