Last chance! Hybrid ICPSR course (July 8-12, 2024): “Sensitivity Analysis: Quantifying the Robustness of Inferences to Alternative Factors or data"

[Ken Frank]

Hope you enjoy the July 4th holiday week for folks in the US. 

 

I will be teaching an ICPSR Summer Workshop, "Sensitivity Analysis: Quantifying the Robustness of Inferences to Alternative Factors or Data," July 8-12, 2024. Hybrid (virtual, FtF in Ann Arbor, MI, with asynchronous access to materials).  See more details below.

 

 

Most social scientists need to consider the unobserved conditions that could change an inference. This workshop is about the statistical tools (R,  Stata or the konfound-it app) to do just that, quantifying the strength of evidence to inform policy, practice, or general social science.

 

The workshop is a full 20 instruction hours of instruction including opportunities to adapt to specific models such as mediation, moderation, diff in diff, logistic regression, hazard/cox regression, multilevel models, differential attrition from treatment groups, etc.

 

Go here to register. Group discounts available as well.

 

Ken

 

 

Sensitivity Analysis: Quantifying the Robustness of Inferences to Alternative Factors or Data," July 8-12, 2024

The phrase “But have you controlled for …” is fundamental to social science but can also create a quandary. Even after controlling for the most likely alternative explanations for an inferred effect there may be some alternative explanation that cannot be ruled out with observed data.  The first response is to develop the best models that maximally leverage the available data. After that, sensitivity analyses can inform discourse about an inference by quantifying the unobserved conditions necessary to change the inference.

Specifically, in this course participants will learn how to generate statements such as “An omitted variable would have to be correlated at ___ with the predictor of interest and with the outcome to change the inference.” Or “To invalidate the inference, __% of the data would have to be replaced with counterfactual cases for which the treatment had no effect.” Because these statements express sensitivity in terms of correlations or cases they have wide accessibility. Rooted in the foundations of the general linear model and potential outcomes, the techniques can be adapted to a range of analyses, including logistic regression, propensity-based approaches, and multilevel models. As a result, they can broadly facilitate discourse about inferences among researchers who seek to make an inference, challengers of that inference, policymakers and clinicians.

Participants in this sensitivity course will learn statistical conceptualization, application, and software including the pkonfound commands in R (Stata also available) as well as the app http://konfound-it.com. Participants should be familiar with statistical control as in the general linear model (i.e., regression). Participants should have their preferred statistical software installed on their computer, and optimally to have identified one or more statistical inferences (e.g., published, straight from recent output) for which they would like to quantify the robustness.