CEREBRAL ASYMMETRY DURING VISUAL PATTERN RECOGNITION: A DYNAMIC MODEL.

摘要

The current study explored pattern and size of asymmetry during spatial frequency recognition. Effects of frequency, sex, and stimulus familiarization on processing and of asymmetry size and frequency range on proficiency were examined.;A LVF advantage emerged at F1 (1.4 cpd) and F3 (2.3 cpd); a RVF advantage, at F2 (1.7 cpd) and F4 (2.8 cpd). This pattern varied with familiarization. The percent correct and d' data showed significant Frequency x Visual Field interactions. These results suggest an effect of frequency on direction of advantage largely determined by sensitivity and dependent on familiarization and Frequency x Task Difficulty.;Sex differences emerged suggesting bases of male superiority in visuospatial, and female superiority in facial, processing frequently reported: The former may be due to males' using higher criteria to RVF presentations and having greater sensitivity to LVF presentations than females; the latter, to females' using the most efficient processing mode for a given stage of familiarization when processing 2-3 cpd, the most salient frequencies of facial stimuli.;The VFAI data and correlations between size of advantage and absolute accuracy suggest that solving difficult tasks entails quantitatively and qualitatively different processing than easy ones: The former use more unilateral activity than and the opposite hemisphere to the latter within the same frequency range. Post-hoc analyses of the significant frequency effects in the percent correct and d' data suggest that task difficulty is more salient than frequency range in determining recognition accuracy.;Twenty male and female dextral young adults performed a S/D recognition task to checkerboards (1-3.9 cpd). Each stimulus pair was successively presented parafoveally for 150 ms in the RVF or LVF. Absolute accuracy (percent correct), sensitivity (d'), and relative accuracy (visual field advantage index (VFAI)) measured recognition.;Thus, hemispheric specialization for spatial frequency recognition exists as a function of stimulus, subject, and task characteristics; familiarization; and dependent variable.;The current experiment is compared with Previc's (1982) and Sergent's (1982a). Modification of the spatial frequency model of hemispheric specialization (Sergent, 1982a), emergent processing level and underlying neural mechanisms of such specialization, and future research are discussed.