ESRA 2019 Programme at a Glance


Methodological Issues in Surveying Older People 2

Session Organisers Mr Jan-Lucas Schanze (GESIS - Leibniz Institute for the Social Sciences (Dept. Survey Design and Methodology))
Dr Emanuela Sala (Department of Sociology and Social Research, University of Milano Bicocca)
Dr Annette Scherpenzeel (SHARE – Survey of Health, Ageing and Retirement in Europe, Munich Center for the Economics of Aging (MEA) )
Professor Wander van der Vaart (University of Humanistic Studies, Utrecht)
Ms Daniele Zaccaria (Department of Sociology and Social Research, University of Milano Bicocca)
TimeWednesday 17th July, 16:30 - 17:30
Room D16

The fast ageing processes which European societies are going through lead to a situation that has never existed before. Many of the current pension systems in European countries will in the long term be unsustainable, the costs of health care are steadily rising and intergenerational cohesion seems to be threatened. The key to meet these manifold challenges is knowledge. National and international surveys that target on or cover the ageing population groups can deliver this knowledge but face special methodological issues. For example:
• strong systematic nonresponse related to health problems
• cognitive impairments affecting data quality
• restricted use of modes (online, mobile)
• the need of specific interviewer skills and training
• conversion of gate keepers who refuse / hesitate to let an elderly person be interviewed
• use of proxy interviews or 'triad interviews' and consequences for data quality.
• undercoverage of the institutionalized elderly
• tracing which panel members died
With this outline, we invite papers addressing any of the topics mentioned above or other issues of surveying the elderly, as well as papers presenting potential survey tools that might help to improve data collection among the elderly and reduce sources or error.

Keywords: ageing, elderly, coverage, bias, health, cognition, data quality, gate keepers, proxy, traid interviews, institutionalized population

Interviewer Effects in Face-to-Face Interviews Among Older Respondents

Professor Geert Loosveldt (KU Leuven) - Presenting Author
Dr Koen Beullens (KU Leuven)
Dr Caroline Vandenplas (KU Leuven)

Properly answering questions during a survey interviews requires that respondents have sufficient communicative and cognitive abilities to proceed the different phase of the answering process. There are indications that these skills are related to age and that older adults cope differently with survey questions than younger respondents. This can influence the interaction with older and younger respondents in a different way. We will test whether interviewing older adults leads to increased interviewer-respondent interaction and whether the intensified interaction has an impact on the interviewer variance and increases the intra interviewer correlations (IICs) among older respondents.
Data collected in 13 countries during round 7 of the European Social survey for which primary sample unit information is available are used in the analysis.
First we indirectly assess response behavior in the different age groups. For that, four interviewer assessments of different aspect of response behavior (e.g. asking for clarification, not understanding the questions, reluctance to answer) and interview duration are used. Next we calculate the intra interviewer correlation for 72 substantive questions within 7 age groups in each country taking into account the PSU information. This analysis results in 6552 IICs that are nested within questions and countries. Subsequently all the IICs are used as dependent variable in a cross-classified multilevel model with age groups as fixed independent variables and country and questions as random effects. This approach allows an evaluation of the direct effect of age on the interviewer effects which is necessary to answer the question whether we obtain higher interviewer effects among older respondents.
The results show that there is empirical support that face-to-face survey interviews with older respondents (71 and above) are more complex and results in higher IICs.


Who Interviews Whom? Interviewer Effects and Consequences for Survey Measures in Nursing Home Populations

Mr Patrick Kutschar (Institute of Nursing Science, Paracelsus Medical University Salzburg) - Presenting Author
Professor Martin Weichbold (Department of Sociology, Paris Lodron University Salzburg)

Face-to-face surveys are applied frequently when conducting research in elderly populations. Interviewers (INT) play a decisive role for the data quality. Interviewer characteristics may affect measurement processes and influence results. Such threats are of special concern in surveys of nursing home residents (NHR). This study uses data from the c-RCT „PIASMA“, analyses, if NH-intern and extern interviewers ‘produce’ different responses, assesses consequences for the precision of estimates, and determines, if the study outcomes vary by interviewer characteristics.

509 NHR with up to moderate cognitive impairment were examined by 32 INT (CAPI) about pain, depression and quality of life. Interviewer variance was operationalized using coefficients of variation (ρINT). Each items' uncorrected confidence interval (CI95%) was compared to the adjusted CI95% (deftINT). Interviewer biases (gender, NH-intern/extern) concerning the study outcomes were analyzed applying ANCOVA.

Interviewer heterogeneity accounts for 21.3% of total item variance. Variation is significantly higher for internal INT (ρ=0.28) than for external INT (ρ=0.15). Considerable losses in precision are observed when CIs are not adjusted (e.g. maximum pain: CI [5.9-7.0] vs. deftCI [5.4-7.5]). ANCOVA shows interaction effects between INT-gender and affiliation (p≈0.002; ɲ²=7.6%): NHR interviewed by internal males (M=3.8) and by external females (M=3.9) report the lowest pain on average. Highest pain scores are reported when NHR were examined by internal females INT (M=5.1) and external male INT (M=4.9).

Outcomes reported by NHR are influenced by whom they were interviewed. Our findings give rise to potential risks of imprecise and biased survey estimates in nursing home populations. Interviewer gender and affiliation may evoke gender-specific and authority-related interpretative patterns and meaning attributions which affect the question-answer-process. Interviewer effects seem to threaten the survey data quality in institutionalized elderly populations.


How to Measure Couples’ Transitions into Retirement

Dr Laura Konzelmann (Federal Institute for Population Research (BiB)) - Presenting Author

Against the background of population ageing and pressing issues on the pension system, scholars pay increasingly attention to working careers of older people. To broaden the understanding of prolonged and shortened working careers and the transition from working life to retirement it is essential to take into account the partnership context.
With our study TOP (“Transitions and Old Age Potential)” we shed light on the dyadic nature of retirement decisions in Germany. TOP is a CATI-design panel study (1st wave in 2013, 2nd wave in 2015/16, 3rd wave scheduled in the beginning of 2019) that focuses on birth cohorts between 1942 and 1958. In the 3rd wave the respondents’ partners are included for the first time. We want to find out more about the causes and dynamics of retiring together or apart. To this end, a new survey module for our panelists is implemented in the upcoming wave which also serves as a separate survey for the panelists’ partners (CORE – Couple Retirement).
We are faced with some challenges, such as ensuring comparability between both subsamples (anchor/partner) regarding the central classification of retirement/employment status, which is measured by a combination of objective and subjective traits. By doing so, we tap into the “blurry” status between active working life and retirement that characterizes this part of the lifecycle for many people, who consider themselves as already retired although they lack one or more of the often used traits of retirement (certain age, receiving pension benefits, etc.). Comparability is further complicated by the sequential order of the subsamples in the field. Another aspect is the gatekeeper function of the anchor which might lead to biases in the partner sample.
In our presentation we want to focus on the dyadic aspect in TOP and discuss challenges and problems in surveying older couples with the audience.