Is the majority of white Americans really anti-black?

“Implicit bias” is catchphrase made largely from claims that the Implicit Association Test, or IAT, can measure people’s unconscious prejudice toward another race, the other sex, or any topic of interest.  The IAT was developed at Harvard University.  Amazingly, the majority of white Americans who have taken the IAT have been classified as anti-black.  Does this make sense?

The IAT is widely supported by academics in psychology, sociology, neuroscience, and social sciences.  Most, not all, academics bend over backwards with IAT data sets to make research claims about apparent race discriminatory behaviors in various fields — e.g., health care, education, employment and hiring, criminal justice.  Activists use IAT research claims to justify discriminatory and repressive Diversity, Inclusion, and Equity (DIE) policies at all levels of government and the private sector.

In work for the National Association of Scholars, we examined the statistical reproducibility of white-against-black IAT implicit bias claims.  The IAT is a picture/reaction measurement that can be easily taken on a computer.  Tested against zero — i.e., with no comparison against anything — as a group measurement, IAT results are claimed be statistically significant, not chance results.  However, to be a valid, IAT results must be positively correlated with real-world observations.

Oswald and colleagues had the good sense to doubt the validity of race IAT research claims.  They performed a meta-analysis of studies examining the predictive validity of IAT results against measures of real-world behaviors and judgments for a broad range of racial bias categories.  They showed that the IAT provided little insight into who will discriminate against whom — i.e., it doesn’t work in practice.

Their data set was publicly available for our use, something that most other academic researchers cannot say today.  We used statistical p-value plots to try to independently reproduce Oswald and colleagues’ research claim that the IAT does not work in practice.  If we can reproduce their results, their claim is true; if not, it can be considered false.

Our analysis showed that — in agreement with Oswald and colleagues — IAT results (implicit bias measures) poorly predicted real-world behaviors and judgments of whites against blacks.  Their claim is true.  We also observed that IAT results explained less than one percent of race variability.  Ninety-nine percent of differences between the races was due to factors other than implicit bias.

Oswald and colleagues concluded that the poor performance of the IAT is mostly consistent with a flawed instruments explanation (i.e., problems in the theories that motivated the development and use of these measurement).  That is an explanation we support.  Why is it always Harvard researchers who seem to get the science more wrong than right?

First, the good news.  The work of Oswald and colleagues, many others, and our statistical reproducibility tests reinforce that white-against-black IAT data are minimally or not at all related to real-world behaviors and judgments.  The majority of white Americans are not anti-black.

Implicit bias science — if it can be called a science — is messy and uncertain, and the existing research evidence does not prove current theories of racism.  Attempting to measure racial implicit bias with the IAT is not a pathway or answer to address racism.  It is a flawed psychological test.

Now the bad news.  DIE policies based on implicit bias theory are well entrenched in our society, even though they are at best useless and at worst actively harmful.

There is enough good science out there to think that the implicit bias monster should be dead.  But alas — DIE seemingly cannot be slain by science alone, only held back until the next wild research claim comes along.

There is no real way to kill bad science within the science world.  Some higher administration has to step in.  Maybe this is something the new Trump administration can help fix.

S. Stanley Young, Ph.D. is the CEO of CGStat in Raleigh, North Carolina and is director of the National Association of Scholars’ Shifting Sands Project.  Warren Kindzierski, Ph.D. is a retired college professor (public health) in St Albert, Alberta.

<p><em>Image: tomaszmichalkania via <a  data-cke-saved-href=

 

Image: tomaszmichalkania via Pixabay, Pixabay License.

If you experience technical problems, please write to helpdesk@americanthinker.com