It is not just memory: propositional thinking influences performance on the autobiographical IAT

Area of Science:

• Psychological assessment
• Implicit social cognition
• Substance use research

Background:

• The autobiographical Implicit Association Test (aIAT) is designed to detect concealed autobiographical events.
• Previous studies reported high accuracy for aIAT, but a cocaine user identification attempt yielded high false positives.
• This study investigated reasons for the inaccuracy of the cocaine aIAT.

Purpose of the Study:

• To explore the reasons behind the inaccuracy of the autobiographical Implicit Association Test (aIAT) for identifying cocaine users.
• To compare the accuracy of two different category label versions of the cocaine aIAT.
• To examine the influence of category labels on aIAT performance and its relation to self-concept and attitudes.

Main Methods:

• Two versions of the cocaine aIAT were created using descriptive ('guilty/innocent') and self-referenced ('as if you were/were not') category labels.
• Forty-one cocaine-abstinent participants, verified by hair analysis for 12-month abstinence, were randomly assigned to a condition.
• Participants completed bespoke implicit/explicit cocaine attitude measures and self-esteem assessments.

Main Results:

• The self-referenced aIAT version demonstrated low accuracy (19%), while the 'guilty/innocent' label version achieved 65% accuracy.
• The self-referenced aIAT results significantly correlated with self-concept measures.
• aIAT outcomes were not influenced by participants' attitudes toward cocaine users.

Conclusions:

• Category labels critically influence aIAT accuracy, indicating the involvement of propositional knowledge and self-concept.
• The aIAT may not reliably access implicit memory when specific memories are absent.
• The aIAT is not recommended for drug use detection; further research on its mechanisms and potential is warranted.