Survey Design for Associations

Inês Carvalho-Reyes, NANA's director of research, holds a PhD in survey methodology and leads the annual State of Associations survey along with NANA's compensation survey and quarterly data briefs. She teaches this course on the mechanics of a survey that produces data an association can compare year to year, work done with support from Fieldstone Data Partners, which fields the State of Associations survey.

Why changing a question's wording between survey years breaks the comparison, even when the intent stays the same, and how to phrase a question once and reuse it without drifting.

Calculating the response count a survey needs to support a valid finding for a given membership base, and why a low response rate can still produce usable data if the sample is representative of the whole membership.

Picking a rating scale, five-point, seven-point, or a simple yes/no, that measures what the survey actually needs and doesn't nudge a respondent toward a particular answer.

Publishing a survey's margin of error alongside its headline number, the practice Inês Carvalho-Reyes uses for every State of Associations release, so a reader can judge how much weight the finding can bear.

Research and data staff, and any association planning its first member survey or looking to make an existing one comparable across years, including chapter staff running their own local surveys. Bring a past survey instrument if one exists; the comparability module works directly from a real set of questions. Pairs with the Survey Design Guide.

• Write survey questions that stay comparable across multiple survey years

• Set a sample size that supports a valid finding for a given membership base

• Choose a response scale that avoids leading a member toward an answer

• Report a survey's margin of error alongside its headline findings