Tuesday 14th July      Wednesday 15th July      Thursday 16th July      Friday 17th July     

Wednesday 15th July, 09:00 - 10:30 Room: O-202

How do interviewers influence data?
Convenor	Miss Claudia Karwath (Leibniz Institute for Educational Trajectories )
Coordinator 1	Miss Manja Attig (Leibniz Institute for Educational Trajectories)

Session Details

There are various modes of data collection, such as telephone, face-to-face or online. Some of these modes include the use of interviewers, which play a special role in the study organization. For example, they have to contact and motivate the respondent for participation. In this case, interviewers not only can influence the collecting data process (e.g., finding the right respondent) but furthermore they can affect data quality in different ways as well (e.g., influencing the response of the participants).

To organize studies, especially with different modes of data collection, it is important to involve possible effects of using interviewers.

Therefore, the following session focus on topics such as:

- Interviewer characteristics and their influence on data
- Interviewer-respondent interactions
- Effects of different modes of data collection

Paper Details

1. Social Desirability as interviewer effect between CAPI and CASI in Japan
Mr Isamu Sugino (Ochanomizu University)
Dr Daisuke Kobayashi (Jin-ai University)
Dr Aki Kaeriyama (Rikkyo University)

In Japan, the study on the mode-difference in the CAIs is not so many. We conducted the experimental CAI survey to investigate the mode effects, especially the influence of the presence of interviewers in 2014. The result shows CAPI respondents tend to make more socially desirable responses than CASI respondents. In the comparison of subjective status identification using IPWE to control non response errors, the mean value of the subjective social status on CAPI is significantly higher than that on CASI.
These results suggest that in the situation using CAI, social desirability bias(the effect of interviewer) is problematic.

2. Understanding dynamics of consent requests in surveys: Consent to biomarker data collection and administrative data linkage in the Health and Retirement Study
Ms Colleen Mcclain (Michigan Program in Survey Methodology)
Dr Sunghee Lee (Michigan Program in Survey Methodology)
Dr Jessica Faul (University of Michigan)

Collecting biomarkers and obtaining permission to link to administrative data within surveys holds significant promise, but introduces potential for nonresponse bias. We expand upon previous work by combining multiple data sources from the Health and Retirement Study, allowing for multifaceted exploration of consent dynamics in a face-to-face context. We first link requests for physical measurements, biomarker collection, and data linkage with respondent characteristics, moving beyond demographics to explore attributes such as personality that may correlate with consent. We then extend our focus on interviewers, incorporating results of behavior coding of the interviewer-respondent interaction and assessing interviewer characteristics

3. Financial Literacy and Interviewer Effects
Dr Tobias Schmidt (Deutsche Bundesbank)
Dr Panagiotta Tzamourani (Deutsche Bundesbank)
Professor Joachim Winter (LMU Munich)

In this paper we ask whether interviewer characteristics influence the answers to a standard set of financial literacy questions. We fear that interviewer effects can arise because of interviewers perhaps helping respondents to answer the literacy questions or some other obscure interaction between them. We use data from Germany’s wealth survey “The Panel on Household Finance (PHF)” to investigate this issue. Controlling for the usual respondent characteristics, we find that interviewer fixed effects explain a substantial fraction of the variance of the financial literacy score and interviewers’ age, gender and education have statistically significant effects on it.

4. Interviewer Effects in the CAPI-Study of the Early Childhood Cohort of the National Educational Panel Study (NEPS)
Miss Claudia Karwath (Leibniz Institute for Educational Trajectories)
Miss Manja Attig (Leibniz Institute for Educational Trajectories)
Mrs Hannes Kliem (Leibniz Institute for Educational Trajectories)

In the early childhood cohort of the National Educational Panel Study interviewers are involved in contacting the participants and collecting the data. Because of the age of the children (6 - 8 months) and the household situation, the interviewers are exposed to a special challenge, as they have to conduct the direct and indirect measurement assessments in a standardized way and in accordance with the administration rules as well as they need to handle up with the “household situation” (e.g., babies crying). We focus on interviewer effects, examine if interviewer characteristics influence the survey process and the collected data.