Surveying Political Activists: The Effectiveness of a Mixed Mode Survey Design

This article reports on the used of a mixed mode survey – mail and internet – to study 2004 New Mexico Democratic presidential caucus voters. Potential respondents were sent a letter asking them to answer on‐line or to request a mail survey. We tested for sample representativeness, possible differences between survey modes, and the effectiveness of reminder postcards on survey response. We find that the mixed mode survey does a relatively good job of representing the sample, but that web responders were younger, whiter, wealthier, and more educated. We find no difference by survey mode for answers to our substantive survey questions once we control for demographics. Postcards were very effective in increasing the response rate. In addition, we find that respondents had a difficult time understanding how to enter the survey web address on their web browser and suggest that information on browser use be included in the initial contact letter and follow‐up postcards.