Weaker than you might imagine: Determining imageability effects on word recognition

Imageability – the ease of generating a mental image for word has been commonly used as predictor recognition but its effects are highly variable across literature, raising questions about robustness and stability construct. We compared six existing imageability norms in their ability to predict RT accuracy lexical decision naming thousands words. Results showed that, when sensorimotor sources variance were partialled out, predicted little unique performance effect sizes varied greatly between norms. Further analysis suggested that such heterogenous likely due inconsistent strategies how participants interpret rate norming studies, despite consistent instructions. Our findings suggest does not reliably facilitate ratings should be with caution research.