Relevance & Research Question:

In this contribution we will report on the 2025 recruitment of new respondents for the GIP. This includes a general overview of the recruitment procedure and results, unexpected obstacles we encountered along the way, lessons learned, and things to look out for when handling sample recruitment in-house. As we observe a general trend of moving more processes in-house among academic survey projects, often to reduce costs, this contribution will be of interest for a wide audience of survey practitioners who (plan to) handle their own recruitment of respondents.

Methods & Data:

In September and October 2025, we sent out 7,000 invitations to prospective new respondents for the GIP. About 2/3 went to a random sample from the population registers of 135 municipalities drawn by GESIS and 1/3 to addresses sampled from a commercial database. We report on the process of handling the recruitment of a probability sample in-house, obstacles we encountered and possible solutions to these, and recruitment results such as response rates.

Results:

During the recruitment process, we encountered a number of obstacles, such as uncooperative municipalities, coordinating the printing and sending of invitation letters and reminders in-house, print quality, handling the prepaid cash incentives, and a higher-than-expected number of invitations being returned by the postal service as undeliverable, presumably due to incorrect addresses, particularly in the commercial address database sample. While the fieldwork is still ongoing, preliminary results indicate that about 1,900 recruitment interviews have been started and about 1,100 to 1,200 new respondents will be recruited into the panel, while about 800 of the 7,000 invitations have been returned undeliverable.

Added Value:

We provide a hands-on report with concrete advice for conducting a survey recruitment and avoiding potential obstacles and costly errors. In addition, we report recent information on the quality of obtained address data, response rates, and recruitment success for a probability online panel in the ever-changing survey climate in Germany, which will be of interest to any practitioners planning the recruitment of a probability sample.

Relevance & Research Question

Panel consent is the permission given by respondents to be re-contacted for future panel waves. The lower the panel consent rate, the larger the initial panel attrition and the higher the risk of panel consent bias, which threatens data quality. This study investigates whether incentives for panel consent and repeated requests for panel consent can increase panel consent rates.

Methods & Data
We conducted an experiment in the first wave of the new online panel survey "International Mobility Panel of Migrants in Germany" (IMPa) on two innovative design features: (1) A repeated request for panel consent within the questionnaire: Respondents who do not provide consent at the first request are immediately followed up with a second request to reconsider their decision. (2) Incentivizing panel consent: Depending on the randomized experimental group, respondents are (not) offered an additional incentive (5€) for providing panel consent at the first and / or second request. In group 1, no incentive for panel consent is offered at neither request. In group 2, respondents are offered an incentive for providing panel consent at both requests. In group 3, panel non-consenters of the first request are offered an incentive for panel consent at the second request.

Results

Our analyses (N≈41.000) show that 1) panel consent rates at the first request are higher if incentives are offered; 2) the second request significantly increases the cumulated panel consent rate; 3) the second-request effect is highest for group 3, where incentives are offered only at the second request; 4) the cumulative panel consent rate is highest for groups 2 and 3. Regarding the cost effectiveness, group 3 leads to panel consent rates as high as for group 2, while requiring costs per panel consenting respondent is close to group 1. We will also provide results on how the experimental design affects wave 2 response rates.

Added Value

This paper introduces and assesses two innovative survey design features to maximize panel consent rates. Furthermore, we analyse the costs associated with each design and derive recommendations for panel consent request designs.

Relevance & Research Question:

In recent years, probability-based mixed-mode panels have become a common tool in empirical research. Most panels rely on continuous, regular surveys that ask participants about core topics at fixed intervals (in a similar way to classic longitudinal panels). In addition, some of these panels offer additional or ad hoc studies. These studies allow internal or external researchers to conduct additional and in-depth surveys. The Robert Koch Institute (RKI) Panel 'Health in Germany', set up in 2024, operates exactly in line with this logic. However, for research purposes, the question arises as to what extent additional studies (at irregular intervals on different topics) influence the non-response of further (regular) surveys. In this presentation we will present first findings within the RKI Panel.

Methods & Data:

The analysis draws on online survey response metrics from panel waves of the RKI Panel, complemented by information on invitations to and participation in an ad hoc survey conducted between the regular waves. A comparison is made between two randomly selected groups of panel members invited versus not invited to the analyzed ad hoc survey. Multivariate models control for sociodemographic characteristics to isolate potential effects attributable to the ad hoc survey.

Results:

Preliminary findings show that participants who were invited to the ad hoc survey exhibit marginally lower sub-group response rates for the following wave. However, controlling for sociodemographic characteristics in a multivariate model, we see no substantial effect regarding potential non-response for participants of the ad hoc survey. Thus, participants show stable response patterns overall. Data collection is still ongoing. Further results on more ad hoc surveys and following waves will be presented at the conference.

Added Value:

The findings contribute to an emerging evidence on the effects of additional survey burden in mixed-mode probability panels. By investigating invitation effects, the study offers practical insights for panel management, especially regarding contact frequency and respondent burden in newly established panels such as the RKI Panel.