1.	Surveys should be conducted annually. Data replicated over time are the most reliable and actionable.
2.	Surveys should be sent at about the same time every year.
3.	The survey sample must be sufficiently large to ensure accurate and actionable results.
4.	The station should be aware of the standard error for each question, based on the sample size and response, and avoid overreacting to responses that are statistically insignificant.
5.	The questions should be unbiased, quantitative, and designed by, or in conjunction with, an experienced research professional. Stations should develop trends by including questions that will be asked in each future survey.
6.	Open-ended questions on surveys are difficult to interpret and are generally ineffective. A station that requires answers to many open-ended questions should consider conducting focus groups, rather than including them on a survey.
7.	The survey results should be analyzed by a sophisticated relational database or a survey analysis software program.
8.	As with all research, the best and most useful results are achieved by tracking trends in the data over time. This means conducting annual surveys to get the most benefit from the data.
9.	While stations may choose to analyze the survey results themselves, many find that they receive the optimum benefit from a survey through an analysis of the results by a research professional outside the station. Often they can examine the results more objectively, use their experience with other stations to provide insight and comparisons, and view the station from a fresh perspective.

[ Overview Table of Contents | Previous Section | Next Chapter | Toolkit Home ]