ESRA 2027 sessions by theme

Back to Overview of Sessions

Session Details

Longitudinal data are valued for their ability to capture individual stability and change over time and to strengthen analyses of causal relationships. Researchers and policy-makers rely on longitudinal data to understand societal developments and inform policy decisions. Yet this key advantage also creates a fundamental measurement challenge: observed differences between waves may reflect not only substantive change, but also measurement error. Distinguishing true change from measurement-induced change is therefore essential for drawing valid conclusions about change over time.
Measurement error is a long-standing concern in survey research, and repeated measurement introduces additional complexities. Errors may be correlated across waves, inflating observed stability, while random response error can create fluctuations that do not reflect true change. Measurement properties may change over time because respondents interpret questions differently, questionnaire wording or survey modes change, or repeated participation affects response behavior through panel conditioning. Such changes can violate measurement invariance and complicate comparisons across waves and groups. Recall and other reporting errors may further distort observed patterns of change.
These forms of measurement error can bias estimates of stability and change, individual trajectories, and associations over time, thereby affecting substantive conclusions and policy-relevant findings. Understanding when such errors arise and how they can be assessed, modeled, or corrected remains an important methodological challenge.
This session invites contributions on the identification, assessment, and consequences of measurement error in longitudinal data. Contributions may examine the effects of measurement error on estimates of stability and change, correlated errors across waves, measurement invariance over time, panel conditioning, and changes in measurement conditions-such as survey mode or questionnaire design-that affect comparability across waves. Studies evaluating methods for assessing, modeling, and correcting measurement error, as well as work demonstrating its consequences for substantive longitudinal conclusions, are especially encouraged.