Alfredo Morabia1, Yan Zheng. 1. School of Earth and Environmental Sciences, Queens College - CUNY, Center for the Biology of Natural Systems, 163-03 Horace Harding Expressway, Flushing, NY 11365, USA. alfredo.morabia@qc.cuny.edu
Abstract
OBJECTIVES: Determine whether raffling off prizes to increase participation in surveys triggers multiple responses from the same participants. METHODS: All staff, instructors, and students of Queens College, New York, were invited to a 5 minute online survey about commuting in winter 2007. Entry in the raffle to win one of two iPODs was not conditional upon survey completion. RESULTS: Participation was 18.7 %, and 3913 eligible responses. Of the 183 (4.7 %) participants with two answers, responses were consistent, with weighted kappas ranging between 0.48 and 1.00, and Pearson r's between 0.81 and 1.0. Kappa's were lower for opinion than for factual questions. The duplicate responses came from people with characteristics similar to the full sample. Commuting modes included car (46 %), bus but no subway (28 %), and bus and subway (21 %). The median commuting time was 90 minutes round-trip, and was greater than 120 minutes for 1/4 of the respondents. CONCLUSION: Participation was low despite the raffle. The small percentage of duplicate answers, maybe prompted by the reminding email, had a negligible impact on the results.
OBJECTIVES: Determine whether raffling off prizes to increase participation in surveys triggers multiple responses from the same participants. METHODS: All staff, instructors, and students of Queens College, New York, were invited to a 5 minute online survey about commuting in winter 2007. Entry in the raffle to win one of two iPODs was not conditional upon survey completion. RESULTS: Participation was 18.7 %, and 3913 eligible responses. Of the 183 (4.7 %) participants with two answers, responses were consistent, with weighted kappas ranging between 0.48 and 1.00, and Pearson r's between 0.81 and 1.0. Kappa's were lower for opinion than for factual questions. The duplicate responses came from people with characteristics similar to the full sample. Commuting modes included car (46 %), bus but no subway (28 %), and bus and subway (21 %). The median commuting time was 90 minutes round-trip, and was greater than 120 minutes for 1/4 of the respondents. CONCLUSION: Participation was low despite the raffle. The small percentage of duplicate answers, maybe prompted by the reminding email, had a negligible impact on the results.
Authors: Alfredo Morabia; Fang Fang Zhang; Maya A Kappil; Janine Flory; Frank E Mirer; Regina M Santella; Mary Wolff; Steven B Markowitz Journal: Prev Med Date: 2012-01-31 Impact factor: 4.018