作者: Richard M. Shiffrin , Aita Salasoo , Timothy C. Feustel
DOI: 10.1037//0096-3445.112.3.309
关键词:
摘要: SUMMARY The repetition effect refers to the finding that speed and accuracy of naming a visually presented word is enhanced by single prior presentation word. A new technique was developed study this phenomenon: visual signal-to-noise ratio printed item in field masks slowly increased. When interest, increase ceased at predetermined time; when latency continued until could be named. Experiment 1 tested validity against more traditional threshold measure ease identification, which for brief exposure, followed mask. performance levels first were equated two techniques both words nonwords, equal slightly stronger than nonwords. In la psychometric functions presentations obtained, giving as function final exposure duration. large interaction seen with yielding superior technique, but reverse true 2 version used: difference in'the latencies necessary nonword identification found additive due repeated presentations. Taken together, results experiments suggest separate contributions lexical status effect. used groups word-nonword was, if anything, increased mixed lists nonwords 3. This result rules out certain guessing bias interpretations differences. Experiments 1,2, 5, lag between repetitions had most small nonsignificant on latency. However, rec-, ognition performance. 3, shifting case identical items produced very decrease effect, suggesting minimal role low-level physical features 4, orthographic similarity (i.e., spelling overlap) previously not sharing common morpheme studied. (sometimes significant) facilitation such observed. suggests letter-name clusters play some model outlines relative episodic traces particular events, uriitized representations semantic memory, identification. unitization characterizes missing plays prominent model. Specifically, attributed presence memory are assumed uniformly