No basis in bias science

News that the Australian Taxation Office has been running unconscious bias training courses raises the question: why are taxpayers footing the bill for a potentially flawed psychological test?

The course uses the Harvard Implicit Association Test, which employs image and word association to determine the level of ‘unconscious bias’ an individual has towards those of a different race, sex, and so on. My colleague Jeremy Sammut highlighted the socially destructive nature of this test, but the origins themselves are equally disturbing.

The IAT was introduced into the scientific literature in 1998 by researchers Anthony Greenwald, Debbie McGhee and Jordan Schwartz. However, not only does the test suffer a replicability problem — meaning that some of the results have not been successfully replicated — a number of psychologists have come out and challenged its efficacy.

News that the Australian Taxation Office has been running unconscious bias training courses raises the question: why are taxpayers footing the bill for a potentially flawed psychological test?

The course uses the Harvard Implicit Association Test, which employs image and word association to determine the level of ‘unconscious bias’ an individual has towards those of a different race, sex, and so on. My colleague Jeremy Sammut highlighted the socially destructive nature of this test, but the origins themselves are equally disturbing.

The IAT was introduced into the scientific literature in 1998 by researchers Anthony Greenwald, Debbie McGhee and Jordan Schwartz. However, not only does the test suffer a replicability problem — meaning that some of the results have not been successfully replicated — a number of psychologists have come out and challenged its efficacy.

A 2009 report by psychology professor Hart Blanton demonstrates the evidence between IAT scores and real-world behaviour is virtually non-existent. A Kirwan Institute Study on implicit bias found such tests can be damaging because the range of responses is limited. And a paper published by Gregory Mitchell and Philip Tetlock argue the claims made by proponents of the IAT are exaggerated, and the test fails to consider alternative factors that could influence an individual’s responses.

After the IAT was introduced in 1998, many private companies such as McDonalds and Google started teaching their employees about unconscious bias. But now, in the era of diversity bureaucracy, the adoption of pseudo-scientific programs that place feelings over facts has sadly also become the new norm for taxpayer-funded institutions.

The Australian Public Service Commission dedicates a page to ‘unconscious bias.’ The Queensland Government claims the IAT can be used to bring awareness to organisational and individual biases. And many more government agencies now cite ‘unconscious bias’ in their diversity programs.

The idea that a government agency would want to test the unconscious thoughts of its employees and try to change them, is disturbing enough. But when a test is this flawed, it is also an egregious waste of taxpayer money.

Monica Wilkie is a policy analyst at The Centre for Independent Studies.