| This dissertation examined whether visual word recognition involves serial or parallel processing, using the Stroop effect as an instrument of investigation. Twenty-five undergraduate students (18-49 years) read lists of Stroop stimuli consisting of nonsense syllables (ZFM, AICF, HXFFJ, QFIIKP), backwards traditional color words (DER, EULB, NEERG, WOLLEY), non-traditional color words (TAN, PINK, WHITE, VIOLET) and traditional color words (RED, BLUE, GREEN, YELLOW), which had either All, First, Central, or Last letters presented in red, blue, green, or yellow. After a 2 factor, repeated measures ANOVA revealed significance for word-context (p < .001) and position of coloring (p = .029), Bonferonni analyses indicated significant differences in Stroop interference for word-context lists having all letters colored (p < .001), replicating a previously reported finding in the literature. However, this word-context effect was attenuated by coloring specific positions since some differences reduced to nonsignificant levels. When position effects (i.e., differences in Stroop interference for each word-context based on position of coloring) were themselves examined, interference differences were found for traditional lists (p < .001; replicating another literature finding) and backwards lists (p < .001), but not for nonsense syllable and non-traditional lists. The latter three findings are new since they have not been previously reported in the literature. Two major conclusions emerged from this data: (1) a new hypothesis to explain the traditional Stroop effect (and remaining data) dubbed double checking was derived; and (2) whether words or non-words involve serial (i.e., position effects) or parallel (i.e., no position effects) processing, seems dependent on strategic attention control. Additionally, strategic attention control itself, seems dependent on an interaction between stimulus characteristics and task requirements. Recommendations for future research concerning the double check hypothesis and another phenomenon known as inhibition of return (believed to parallel some of the data in the study) are provided. |
