Geoff's paper with colleague Moto Yamaguchi published in Journal of Experimental Psychology (JEP): General, one of the top 5 international journals in experimental psychology. The piece is on the measurement of implicit racial/ethnic bias, which is most often measured using the Implicit Association Test (or IAT).

Not just well known in psychology but in popular culture. The IAT test uses speed of response (and error rate) to measure associative connections between target concepts (e.g. black/white) and evaluative concepts (good/bad). It’s very easy and quick to use and has now been employed millions of times to asses implicit racial bias.

In the piece the pair dissect the test and demonstrate that it relies on explicit categorisation into racial/ethnic categories and for that reason is not truly implicit. Geoff & Moto go on to devise a new measure, a multiattribute version of the IAT, or m-IAT, to show what happens when you make the racial/ethnic categorisation truly implicit. The paper is quite technical but it’s a game changer!