Preprints are early versions of research articles that have not been peer reviewed. They should not be regarded as conclusive and should not be reported in news media as established information.
xSUBMIT EV over IV 7.27.2022 submitted to ADVANCEb.pdf (441.39 kB)
Download file

Towards the Unification of External and Internal Validity: An Empirical Approach

Download (441.39 kB)
posted on 2022-08-26, 18:05 authored by Andrew Jaciw

We evaluate the role of selection bias in causal generalization. We compare levels of bias in estimates of average causal impacts from experiments and from non-experiments. To demonstrate we use data from 79 sites in the multisite trial of Tennessee's Student Teacher Achievement Ratio (STAR), and we use results obtained during a randomized experiment of the Alabama Math Science and Technology Initiaitve (AMSTI).    




Declaration of conflicts of interest


Corresponding author email

Lead author country

  • United States

Lead author job role

  • Independent researcher

Lead author institution

Empirical Education Inc.

Ethics statement

The work makes use of data from two experiments. The experiments complied with relevant ethical standards for human subjects protections.

Terms agreed

  • Yes, I agree to Advance terms

Comments (2)

Log in to write your comment here...

Agreed with this. Click here

Agreed terms and conditions name style