Question

Pete is conducting a survey to determine his customers’ overall satisfaction about the quality of his company’s products. He sends out surveys to the 5 customers who have purchased the largest number of items over the past year. Are his results likely to be representative of the population he is trying to analyze?

Answer 1

No.

When trying to analyze an overall satisfaction with Quality of service. The number of population he is analyzing should not be limited to a certain factor. This would cause a narrow data information gather and would not represent the entire population. The surveys should be based randomly to avoid data bias/filters and actual results would be realistic.

Answer 2

No.
Because, when to do some sort of analysis (such as this one), you need to take (for example) a RANDOM SAMPLE from the POPULATION of the problem that is being analyzed. In this example (problem), Pete wants to evaluate OVERALL satisfaction of customers, so he should NOT send the surveys ONLY to the customers who have purchased the LARGEST number of items, but to the randomly selected customers, in order to obtain REPRESENTATIVE results of the OVERALL satisfaction. If he sends the surveys only to the customers who have bought the largest number of items, he will obtain VERY HIGH satisfaction of customers, as results, of course, and this will not be representative results.