How many surveys are needed to obtain valid results?
This is a complex question for which there is no simple answer. The number of surveys required generally depends upon two factors: • The precision required in the results: How precise do survey results need to be? For example, if a survey indicates that 68% of your customers like a new product or service, is it acceptable for that figure to be precise to within plus or minus 10%? [This means that between 58% and 78% of customers actually like the new product or service.] Is 7% acceptable, or is greater precision such as plus or minus 5%, or 3%, necessary? How the results will be applied generally determines the level of precision required, and the total number of surveys that need to be completed. • The number of different groups included in the study: Many survey results need to be examined in detail by calculating the results for several different groups. These might include separate results for men and women, older and younger age groups, or heavy and light product users. While the
