Suchergebnisse

Die Ausführung ist eine Erläuterung zum Foliensatz "Allgemeines Interviewertraining für computerbasierte persönliche Befragungen" (s. https://doi.org/10.15465/gesis-sg_de_034).

Self-administered mixed-mode surveys are increasingly used as an alternative to face-to-face surveys for collecting data from the general population. However, little is known about how decisions regarding the incentive scheme and the mode-choice design jointly affect key outcomes such as response rates, net sample composition, and survey costs. To study this, we drew a probability sample of the residential population of the city of Mannheim, Germany (N = 2,980) and randomly assigned target persons to one of four incentive schemes (€0, €1, or €2 prepaid incentive on first contact, and €2 delayed prepaid incentive) and one of two mode-choice designs (concurrent or sequential [web-push]). Our results indicate that small prepaid monetary incentives work better in concurrent than in sequential designs. Moreover, a €2 prepaid incentive in a concurrent design proved particularly successful for older target persons, probably reinforcing their sense of trust and reciprocity, while also fitting better with their survey-mode preferences. Finally, a €2 delayed prepaid incentive in a sequential design primarily motivated target persons aged under 50 years. This combination of incentive scheme and mode-choice design also proved to be most cost-effective in that age group. Based on our results, we recommend using sampling frame information on age to address different age groups with different combinations of incentive scheme and mode-choice design. This may help to maximize response rates, achieve a balanced net sample composition, and minimize survey costs.

Concerns about interviewer effects in interviewer-mediated surveys have accompanied survey research for a long time. As interviewers are involved in nearly all aspects of the survey implementation process, they can affect almost all types of survey errors, including sampling error, nonresponse error, measurement error, and, to a lesser extent, error resulting from the coding and editing of survey responses. Building on the existing literature, this survey guideline provides an overview of interviewer effects and their estimation. It consists of two parts: first, an introductory text using the total survey error (TSE) paradigm as a theoretical framework to provide a general overview of interviewer effects; second, a brief introduction to calculating interviewer effects using multilevel analyses.