Individual Differences in Performance on Comprehension and Knowledge Tests With and Without Passages and Questions

Abstract Several studies showed that reading comprehension items, which use a multiple-choice format, were solved above chance level even when the participants received only the response alternatives. In the present study, we pursued two goals. First, we examined whether similar results apply to measures of crystallized intelligence. Second, we investigated the structural equation model structure of performance on the original and manipulated tests. Based on a sample of 210 young adults, we replicated the finding that, even if only response options remain, participants perform above chance level on reading comprehension tests and extended these findings to performance on declarative knowledge tests. We found stable individual differences in performance on the manipulated measures with latent variables capturing shared variance in performance. These factors were strongly related with latent factors capturing shared variance in performance on the original test versions suggesting that general intellectual abilities are critical for performance on manipulated measures.