A fresh look on semantic priming effects

The research presented in this thesis examined the assumption of automaticity of semantic priming effects, through the manipulation of the proportion of related prime-target pairs (relatednesss proportion, RP), prime visibility (masked vs. unmasked), and type of task (lexical decision vs. semantic categorization). In addition to the analysis of mean RT, the effect of these three manipulations on the RT distribution was also examined. Responses to words (targets) are faster when they are preceded by a word (the prime) related in meaning compared to when the preceding word is semantically unrelated (e.g., RT for hawk-EAGLE < table-EAGLE). This semantic priming effect is widely believed to be automatic when the time between the onset of the prime and target is short, generally less than 250 ms, and is generally explained in terms of automatic spreading activation. The research presented in this thesis examined the assumption of automaticity of semantic priming effects, through the manipulation of the proportion of related prime-target pairs (relatedness proportion, RP), prime visibility (masked vs. unmasked), and type of task (lexical decision vs. semantic categorization). In addition to the analysis of mean RT, the effect of these three manipulations on the RT distribution was also examined. Contrary to the assumption that semantic priming at a short SOA is automatic, all three manipulations impacted on the size of the semantic priming effect and produced different RT distribution patterns. These findings are used to reconsider the notion of automatic spreading activation in explaining semantic priming effects at short SOAs. An alternative view in which semantic priming effects are explained in terms of task-­‐dependent processes is proposed.