How do we account for missing and unobserved data?: Diminishing the impacts of hidden bias through novel statistical approaches

Session Type: Flash Talk Session

Abstract

Accurately assessing and theorizing about child development requires unbiased data, yet bias occurs through missing, unobserved, or not directly measured data. How can we evaluate and control for bias that is not measured? These papers present statistical approaches to address missing, unobserved, or not directly measured data and highlight how results and conclusions can differ based on the statistical approach used.

Sub Unit

• Methods, History, Theory

Individual Presentations

• Addressing Systematic Missing Data in Causally Interpretable Meta-Analysis - Presenting Author: David Barker, Brown University; Ruofan Bie, Brown University; Jon Steingrimsson, Brown University

• Measurement Invariance of the Strength and Difficulties Questionnaire in Early Childhood - Presenting Author: Alyssa R Palmer, University of Minnesota - Twin Cities; Ann S. Masten, University of Minnesota - Twin Cities; Daniel Berry, University of Minnesota - Twin Cities