PD24-06: What Would It Take to Change Your Inference? Opening the Discourse About Causal Inferences to a Range of Stakeholders

Session Type: Professional Development Course

Abstract

The workshop promotes a discussion about the basis of evidence for educational decisions of policy and practice. By imagining the conditions necessary to change an inference in the concrete terms of how different the data would have to be or the properties of an omitted variable, this course can inform scientific discourse from a broad set of stakeholders familiar with conventional statistical techniques. Furthermore, the techniques make explicit the role of the researcher in choosing how to estimate and interpret quantitative analysis, and thus careful examination of the positionality of researchers and consumers of quantitative research. Specifically, those challenging racial injustice can engage in discourse on equal footing with those who have historically used quantitative analysis to enforce racial inequities.
In part I we will use Rubin’s causal model to interpret how much bias there must be to invalidate an inference in terms of replacing observed cases with counterfactual cases or cases from an unsampled population. This supports statements such as “XX% of the cases would have to be due to bias to invalidate the inference” In part II, we will quantify the robustness of causal inferences in terms of correlations associated with unobserved variables or in unsampled populations. This supports statements such as “an omitted variable would have to be correlated at qqq with both the predictor and outcome to change the inference.” Calculations for bivariate and multivariate analysis will be presented in the app: http://konfound-it.com as well as Stata and R.

Sub Unit

• Professional Development and Training Committee

Director

• Kenneth A. Frank, Michigan State University