| |
Abstract:
Specific goals: Two main explanations of Stroop interference
are the relative speed of processing and the connectionist model of
Cohen et al (1996). This work was designed to test predictions
derived from these accounts using the Number Stroop paradigm.
Methods: Two experiments investigated the effect of increasing the
numerical distance between the targets and distracters when either
reading or counting different numbers of arabic numerals.
Experiment 3 investigated the interfering effects on counting of
two distracters either bracketing the target response (four 3s and
5s) or both greater than or less than the target response (four 2s
and 3s or four 5s and 6s). Results: In experiments 1 and 2,
increasing the numerical distance between targets and distracters
when either reading or counting reduced interference. In other
words, reaction times to two 5s were less than to two 3s whether
you read the numeral or counted the number of numerals. Strikingly,
however, in experiment 3, interference was less when the two
distracters bracketed the target response than if the distracters
were either both greater or less than the target response.
Conclusions: The symmetrical interference of reading on counting
and counting on reading is incompatible with the speed of
processing account. However, the current results also require
integration of the computational models of number line (Dehaene
& Changeux, 1993) with those of Stroop before they can be
explained.
|
