Implicit Association Test Predictive Validity Meta-Analysis

T. Andrew Poehlman authored a meta-analysis reviewing 122 research reports (184 independent samples, 14,900 subjects) on the predictive validity of Implicit Association Test (IAT) measures. The analysis found an average correlation of .274 for predicting behavioral, judgment, and physiological measures. It also compared IAT measures to explicit self-report measures, which had an average correlation of .361 but greater variability.

Use Cases

• Analyzing the predictive power of implicit measures based on correlation data from 184 samples
• Comparing implicit and explicit measure validity based on effect size comparisons for 156 samples
• Investigating incremental validity of IAT and self-report measures based on intercorrelation analysis
• Studying predictive validity for socially sensitive topics based on findings on impression management

Strengths

• Analysis based on 122 research reports aggregating 184 independent samples
• Includes data from 14,900 subjects providing a substantial sample size
• Provides specific average correlation figures (.274 for IAT, .361 for self-report) for predictive validity

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• Data may reflect temporal/source bias inherent to the meta-analysis methodology

Provenance

Source

T. Andrew Poehlman

Collection Method

Meta-analysis of 122 existing research reports