Resolving empirical inconsistencies concerning priming, frequency, and nonword foils in lexical decision.

Two inconsistencies in the priming literature were investigated. Schuberth and Eimas (1977) reported that semantic priming and frequency have additive effects on RTs in lexical decision tasks, whereas Becker (1979) reported that the same two factors interact. Furthermore, Shulman and Davison (1977) reported greater priming given legal vs. illegal nonword foils, whereas Stone and Van Orden (1989) found no difference in the same situation. The present study investigated these inconsistencies by manipulating nonword foil lexicality (i.e., the similarity of nonword foils to words), semantic priming, and word frequency in two lexical decision experiments. In a facilitation dominant, controlled priming experiment (i.e., long prime-target onset asynchronies and a high proportion of related prime s), the priming benefit (related vs. neutral primes) and the significant interaction of priming benefit and word frequency were unaffected by nonword foil lexicality. Given pronounceable nonword foils, there were no priming costs (unrelated vs. neutral primes). However, given illegal nonword foils, words preceded by unrelated primes were recognized more rapidly than words preceded by neutral primes (a "reversed" cost). In an automatic priming experiment (i.e., brief prime-target onset asynchronies and a lower proportion of related primes), there was no significant interaction of priming benefit and frequency. Implications for resolving the two empirical inconsistencies are discussed.