The use of respondent incentives in PIAAC: the field test experiment in Germany
| Autor: | Martin, Silke, Helmschrott, Susanne, Rammstedt, Beatrice |
|---|---|
| Rok vydání: | 2014 |
| Předmět: |
Sozialwissenschaften
Soziologie Social sciences sociology anthropology PIAAC Erhebungstechniken und Analysetechniken der Sozialwissenschaften Methods and Techniques of Data Collection and Data Analysis Statistical Methods Computer Methods Antwortverhalten Anreizsystem Datengewinnung Interview Umfrageforschung Bundesrepublik Deutschland response behavior incentive system data capture interview survey research Federal Republic of Germany |
| Zdroj: | Methods, data, analyses : a journal for quantitative methods and survey methodology (mda), 8, 2, 223-242, The implementation of PIAAC in Germany |
| Druh dokumentu: | Zeitschriftenartikel<br />journal article |
| ISSN: | 2190-4936 |
| DOI: | 10.12758/mda.2014.009 |
| Popis: | "In PIAAC, each participating country was required to attain a response rate of at least 50 percent, as long as evidence was provided that there was either no or only low nonresponse bias in the data. Achieving 50 percent is a challenge for face-to-face surveys in most Western countries and also in Germany. Previous research showed that the use of incentives is an effective tool to increase response rates in different kinds of surveys. However, incentives may have differential effects on certain socio-demographic groups, because the perceived benefits of an incentive are subjective. To assess the effects of incentives on response rate and nonresponse bias, an experiment with three incentive treatments (10 Euro-coin, 25 Euro and 50 Euro in cash) was implemented in the German PIAAC field test. Results show that response rates increased as the incentive increased. With regard to nonresponse bias, the results are less explicit. According to logistic regressions, the main factors for participation in the 50 Euro condition are age, citizenship, and municipality size and in the 25 Euro condition, only municipality size. Bivariate analyses put these results into perspective. For all treatment groups, a low potential for bias is visible, and there is no statistical evidence that response distributions of the realized sample across treatments are different." (author's abstract) |
| Databáze: | SSOAR – Social Science Open Access Repository |
| Externí odkaz: |
|