Detail from "the second line," a painting by Bob Graham. For more about the artist, click here.

Online Program

A Choice in Mode: A Solution for Increasing Response Rates of Hard-to-Survey Populations?

*Marieke Haan, University of Groningen, Netherlands 
Yfke Ongena, University of Groningen, Netherlands 

Keywords: hard-to-survey populations, mixed-mode data collection, mode effects, nonresponse, tailoring and targeting

Young adults, dual earner families, ethnic minorities and city dwellers are known as hard-to-survey populations. Combining different modes of data collection in a targeted approach could be an effective solution to increase response rates and to improve representation of these subgroups of the population. Previous research has shown that respondents may have preferences for different approach techniques (Dillman, West and Clark, 1994) and different modes of responding (Groves and Kahn, 1979; De Leeuw, 2005).

We will present findings of an experiment within the European Social Survey, in which half of the respondents, approached face-to-face or by phone, can choose to be interviewed face-to-face, by phone or online. By offering the respondents a choice in mode, we expect higher response rates. However, since self-selection of respondents into different modes might increase measurement biases (Sakshaug, Yan and Tourangeau 2010), half of the sample will be randomly assigned to a mode. Enriched addressed based sampling was used to collect the data from these respondents, which was created by using databases with information on population characteristics in zip code areas. Forty municipalities of the Netherlands were selected (with variance in urbanization), and three other selection variables were used to oversample and reach hard-to-survey populations: newly build houses (likely to oversample full-time working couples, and people aged 15-34), low income neighborhoods (likely to oversample ethnic minorities and people aged 15-34) and a random selection of remaining zip codes.

Through this experiment we will know which kind of participation request for the survey and which data collection mode these groups prefer, and therefore it will be possible to develop a targeted mode-approach to lower nonresponse rates in the future.

ASA Meetings Department · 732 North Washington Street, Alexandria, VA 22314 · (703) 684-1221

Copyright © American Statistical Association.