#academictwitter It's time to update how we teach about effect sizes & the policy relevance of research findings.

Here are some resources:

1) I'm hopeful my paper can serve as an accessible introduction to interpreting the importance of effect sizes.

https://journals.sagepub.com/doi/abs/10.3102/0013189X20912798?journalCode=edra
3) A picture is worth 1,000 words. We need to help students to create better visualizations of research evidence.

Explore work by @EdwardTufte

Or even simple visual guides like this from @Stata
https://www.stata.com/support/faqs/graphics/gph/stata-graphs/
4) We need to teach students about opportunity costs and unintended consequences to contextualize policies and programs.

This @J_HumanResource paper by Jepsen & Rivkin on reducing class sizes in CA is one of my favorite examples:

http://jhr.uwpress.org/content/44/1/223.refs
6) Let's encourage students to ask - can that program or policy be effective at scale?

This @IESResearch i3 report about attempts to scale promising education interventions is sobering

https://ies.ed.gov/ncee/pubs/20184013/pdf/20184013.pdf
7) We also have to reaffirm for students the value of the "gut-check."

Does an effect seem plausible & make common sense?

Is there a theory of action & causal chain of events that would explain the findings?

Check out @NoraEGordon & @clconaway new book for a practical guide:
10) Finally, we should build & refine empirical ES benchmarks for specific fields, contexts & outcomes

🚨 @DaveEvansPhD & @Yuan_Fei_ have done just this in a new review of interventions that impact educational outcomes in low- & middle-income countries.

https://drive.google.com/file/d/1HkaK76YJTWGdKWmJYQvVJ3kgW9ieVkO4/view
. . . and here is a paper that provides empirical ES benchmarks for metal health and prevention programs.

https://pubmed.ncbi.nlm.nih.gov/30136245/ 
You can follow @MatthewAKraft.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: