October 2023 Friday’s e-links: Research Replication

Starting our October e-links, we focus on watching out for flawed (exceedingly believable) conclusions from auspicious behavioral economists.

What We Can Learn From a Child and a Marshmallow

Since nursery school children first took the marshmallow test in 1972, researchers have revised the experiment and its meaning.

Why We Needed to Retake the Marshmallow Test

After doing their own version of the marshmallow test, researchers from NYU and UC Irvine told us to look at the value of delayed gratification somewhat differently.

Undigestible Food (Studies)

Because of research replication issues and questionable data, many academic papers need more rigor from their authors and skepticism from their readers.

Why the Twitter Predictor Was About Much More Than the Dow

Although one seven year old study indicates that Twitter sentiment could predict the Dow, a research replication attempt says no.