FREQUENCY AND VALENCE EFFECTS WITH WORD STIMULI FROM LARGE, EMPIRICALLY DERIVED STIMULUS SETS

Three experiments were conducted in search of valence effects on the p rocessing of words expected on the basis of Kitayama's (1990) model in volving attention, word frequency, and lexical access. In Experiment 1 , the effects of word previews using a large set of words were not in accordance with those found by Kitayama (1990), who used a very restri cted set. Instead, effects were in accordance with what would be expec ted from word frequency alone, although these effects also appeared wh en frequency was controlled. In Experiment 2, lexical decision accurac y patterns were consistent with predictions based upon word frequency, but not word valence, and disappeared when frequency was controlled. In Experiment 3, valenced word duration estimation patterns were those expected on the basis of word frequency alone; again, the effects wer e absent when frequency was controlled. A previous study, using the sa me words used currently, found word valence effects when moods were in duced, and other studies using mood induction have found valence effec ts with similar word sets. These results, in combination with the pres ent ones, suggest that mood induction may be required for word valence effects consistently to emerge.