Mental Health
Can the IAT show if a woman is thinking of harming herself?
Assessment of Self-Injurious Thoughts Using a Behavioral Test. Matthew K. Nock and Mahzarin R. Banaji sought to determine if the IAT could be of use in addressing a rather difficult, painful issue: Nonsuicidal self-injury such as cutting. Attempting to figure out whether a patient is going to harm him- or (usually) her- self through an interview process has proven unsuccessful. There are a variety of quite powerful reasons for patients to lie, chief amongst them being that they fear institutionalization if they report a desire to injure themselves. By testing a group of young women who had already committed self-injury against a group of non-cutters, the researchers demonstrated the IAT was very accurate at determining which group a given test subject was from Nock and Banaji devised a two-part IAT that would measure someone’s associations with cutting. The first test was focused on “identity”, determining if a subject associated cutting with themselves. The subject was shown a picture of either skin that had been cut or uncut skin. Then on the left side of the screen the word “cutting” was displayed and on the right, the word “no cutting” was displayed. For one test block, the word “me” was displayed under “cutting” and “not me” was displayed under “not cutting”, for the other test block they were flipped. A second similar test was focused on attitudes and evaluations of cutting by determining if the subject associated cutting with positive or negative words.
The test subjects were adolescent females. The experimental group had committed some act of non-suicidal self-injury in the last year, while the control group had not. Those who had committed self-injurious acts were dramatically more likely to associate cutting with themselves in the identity test. And although both the self-injury and the comparison group had negative associations with self-injury, the comparison group of non-injurious adolescents had far greater negative associations with cutting than the group of girls who had injured themselves in the last year.
The authors are careful to note in the conclusions that this test did not predict whether a specific teenage girl was more likely to deliberately injure her-self in the future. The experiment instead worked backwards, using knowledge already obtained to figure out whether or not the IAT matched it. Further testing is needed to determine the predictive value of the IAT in this area, but the dramatic differences between the self-injurious group and the control group in both tests shows that the IAT may be a far better measure than interviews for determining whether or not people have a predilection to self harm.
