About Me

Professional Bio

I am a quantitative methodologist and PhD researcher in Social Data Science, affiliated with the German Institute of Employment Research (IAB) through the GradAB programme and the University of Mannheim. My research focuses on survey statistics and data quality, with a particular emphasis on mode effects, measurement error, and panel conditioning

I am motivated by the question of how well social and behavioral concepts can be measured. My aim is to strengthen the bridge between measurement, modeling, and substantive theory. My work is grounded in a philosophy of science perspective that values transparency, accountability, replicability, and the cumulative development of knowledge.

Ongoing Projects

  • Dissertation (obviously) - working title: Improving the Reliability of Panel Survey Data for Labor Market Research: A Methodological Perspective on Mode Effects, Measurement Error, and Conditioning Bias
    I am being supervised by Prof. Joe Sakshaug (IAB, LMU Munich) and Prof. Florian Keusch (University of Mannheim).

Furthermore, I am currently working on three interconnected projects as part of my PhD in Social Data Science:

  • The Impact of Mode Design on Panel Retention and Survey Costs: Insights from the LPP Employee Panel.
    Most research on mode effects focuses on cross-sectional surveys, but far less is known about the implications of shifting data collection modes within an ongoing panel. In this project, I study the consequences of transitioning from an interviewer-administered mode to a mixed-mode design in the LPP Employee Panel. I examine how this transition affects the panel’s representativeness, attrition dynamics, and response behavior over time. In addition, I analyze the long-term implications for survey costs, providing evidence on the trade-offs between data quality and budgetary constraints in large-scale panel operations.

  • Analyzing Measurement Mode Effects in the LPP Employee Panel.
    Survey responses are shaped not only by what is asked, but by how it is asked. In this project, I analyze how different data collection modes influence the quality of measurement in the LPP Employee Panel. By linking survey responses with administrative records, I estimate mode-specific biases and evaluate how they affect key labor market indicators. I also assess longitudinal measurement invariance to understand whether the transition to mixed-mode data collection produces systematic differences in response behavior across waves. This work provides empirical insights into the stability and comparability of measurements across survey modes.

  • Distinguishing Behavioral Change from Reporting Error: Analyzing Panel Conditioning in the IAB-OPAL Panel Study.
    Panel conditioning refers to the phenomenon in which repeated participation in a survey shapes respondents’ future attitudes, behaviors, or reporting patterns. Yet the empirical literature has rarely distinguished between actual behavioral change and changes in reporting. In this project, I use data from a high-frequency panel study to disentangle these two mechanisms. By utilizing an unfielded sample drawn using the same sampling procedure as the fielded sample, I can estimate the effect of panel conditioning on actual labor market behavior by comparing administrative records of these two samples. Comparing survey responses with administrative data gets me insights about how panel conditioning affects how respondents report their labor market behavior. Distinguishing these two mechanisms is crucial for accurately interpreting dynamics in longitudinal data.

What got me into Methods ands Statistics?

I became drawn to methods and statistics when I first encountered the so called “replication crisis” in psychology and the broader social sciences. It was quite frustrating to realize that many influential findings were difficult to reproduce. Beginning to understand how much empirical research rests on the foundations of measurement quality and methodological choices, the open science movement provided a response that resonated with me: clearer documentation, transparency, more rigorous analytical approaches and working on establishing best practices in empirical research. This shift in the field shaped my own path, motivating me to focus on data quality, inference, and the statistical principles that make research more credible and trustworthy.

Research interests

  • Measurement Quality

My foundational interest lies in how social and behavioral phenomena are translated by data. Measurement quality is central to this process as the validity, reliability, and comparability of indicators determine the credibility of findings. Poor measurement cannot be rescued even by the most sophisticated statistical models, while careful operationalization and validation of measurement instruments, while slow and tedious, enables clearer inference and better theory-building.

  • Longitudinal Data

My interest in measurement naturally extends to longitudinal data, where issues of consistency and comparability become even more critical. If we want to understand changes over time, we need to be able to distinguish between changes caused by developing behaviors or attitudes and inaccuracies in our measurements. Longitudinal designs highlight the dependence of temporal inferences on measurement decisions, reinforcing the idea that rigorous operationalization and data quality are prerequisites for studying social processes.

  • Meta Science

The replication crisis has made clear that many weaknesses in empirical research arise not only from (rash and unquestioned) statistical practices, but from fundamental issues of measurement - something that hasn’t gotten nearly enough attention. Meta science provides a lens to examine how research is conducted, how evidence accumulates (and what hinders it), and how institutional incentives shape methodological choices.

  • Philosophy of Statistics

Engaging with meta-scientific debates has led me toward deeper questions in the philosophy of statistics, which provides the conceptual tools for linking data to claims about the world. Issues of uncertainty, model adequacy, evidence, and inference are not purely technical - they are philosophical questions about how we reason and obtain new knowledge with data. Statistics is my chosen approach to tackling fundamental issues in the philosophy of science about the broader logic of empirical inquiry. Specifically, the social sciences deal with complex, content-dependent, and often reflexive constructs (that’s a tough one), which makes the challenges of measurement, explanation and causal inference uniquely demanding.