Relevance & Research Question

Telephone surveys have historically been a popular form of data collection in labor market research and continue to be used to this day. Yet, telephone surveys are confronted with many challenges, including imperfect coverage of the target population, low response rates, risk of nonresponse bias, and rising data collection costs. To address these challenges, many telephone surveys have shifted to online and mixed-mode data collection to reduce costs and minimize the risk of coverage and nonresponse biases. However, empirical evaluations of the intended effects of introducing online and mixed-mode data collection in ongoing telephone surveys are lacking.

Methods & Data

We address this research gap by analyzing a telephone employee survey in Germany, the Linked Personnel Panel (LPP), which experimentally introduced a sequential web-to-telephone mixed-mode design in the refreshment samples of the 4th and 5th waves of the panel. By utilizing administrative data available for the sampled individuals with and without known telephone numbers, we estimate the before-and-after effects of introducing the web mode on coverage and nonresponse rates and biases.

Results

We show that the LPP was affected by known telephone number coverage bias for various employee subgroups prior to introducing the web mode, though many of these biases were partially offset by nonresponse bias. Introducing the web-to-telephone design improved the response rate but increased total selection bias, on average, compared to the standard telephone single-mode design. This result was driven by larger nonresponse bias in the web-to-telephone design and partial offsetting of coverage and nonresponse biases in the telephone single-mode design. Significant cost savings (up to 50% per respondent) were evident in the web-to-telephone design.

Added Value

Using a unique experimental design we showed the potential for known telephone number coverage bias in telephone surveys. However, while introducing the web mode eliminates this coverage error, there is potential for other error trade-offs. For practitioners, this underscores the importance of carefully weighing the potential trade-offs between costs and multiple sources of error when designing a specific study.

Relevance & Research Question

The web survey mode offers a high degree of flexibility while also being highly cost-efficient. However, in the context of web surveys, potential coverage issues are repeatedly raised as a problem, as target respondents without internet access cannot participate in an internet-based survey. In addition, there are respondents with internet access for whom participating online would be possible from a formal perspective, but who do not want to participate via the internet. I want to know: Are the reasons for abstaining from participating online-based also influenced by internet-related characteristics, apart from simply having or not having internet access?
Methods & Data

I use data from the GESIS Panel.pop Population Sample, a probability-based self-administered mixed-mode panel of the German general population, surveyed via web and paper mode (push-to-web design with paper mode as alternative mode). I perform logistic regressions with mode selection as the dependent variable and various internet-related characteristics as independent variables (frequency of internet use, internet skills, variety of internet use, and number of internet-enabled devices). I put particular emphasis on the key challenge of identifying causal effect directions between internet-related characteristics to construct my models with appropriate control variables. For all regressors, I use not only the quasi-metric scale level frequently used in the literature, but I also examine different threshold levels through varying thresholding.
Results
Data analysis is ongoing. I am going to have results in January 2026.
Added Value

With my analyses, I want to shed light on (internet-related characteristics of) those individuals who are not reached by online-only. Knowledge about those not reached by online-only also contributes to evaluating a target group-oriented use of paper questionnaires when conducting surveys.

Relevance & Research Question
Longitudinal studies are transitioning from face-to-face to push-to-web designs for cost-efficiency and timeliness. However, for ageing populations, concerns persist about mode effects on data quality and measurement equivalence. The English Longitudinal Study of Ageing (ELSA) Wave 12 dress rehearsal provides empirical evidence addressing: Does push-to-web compromise data quality in aging surveys? What mode effects emerge comparing web-first to traditional face-to-face approaches?
Methods & Data
Wave 12 implemented sequential web-first then CAPI design for 375 individuals (June-August 2025). Web-first participants (n=260 households) received online invitations with CAPI follow-up for non-responders after 5 weeks. We assessed mode effects across multiple dimensions: item non-response, break-off patterns, scale reliability, occupation/industry coding quality (CASCOT confidence scores), and measurement consistency versus Wave 11 single-mode CAPI. Respondent feedback (n=31) and interviewer reports provided qualitative insights.
Results
Push-to-web showed mixed outcomes. Individual web completion reached 33.6%, but household completion was only 21.5%, requiring 78% of web-first households to receive CAPI follow-up. Web median duration (87.6 minutes) exceeded estimates (60 minutes), with break-offs concentrated at household grid and financial modules. However, web mode demonstrated improved data quality: item non-response was lower compared to CAPI Wave 11 CAPI; transformed web questions improved the data quality for job and accupation responses; scale reliabilities remained excellent (CASP-19 α=0.90). Mode preference split: 50.7% respondents preferred web for flexibility and self-pacing; 18.8% valued face-to-face interaction. Telephone-administered cognitive assessments proved challenging, with hearing difficulties affecting completion quality.
Added Value

Evidence from ELSA study provides critical evidence for longitudinal surveys implementing push-to-web among older populations. Key contributions: (1) household-level completion metrics are essential—individual response rates mask significant household non-completion, which has big implications on fielwork costs; (2) structured web design reduces measurement error in complex variables despite completion burdens; (3) sequential mixed-mode designs must balance mode-specific strengths—web excels at structured data while face-to-face remains valuable for cognitively demanding tasks and panel engagement; (4) realistic duration expectations and improved re-entry protocols are crucial. Findings demonstrate push-to-web viability for ageing populations when supported by tailored strategies and hybrid approaches for mode-sensitive content.