Stroop interference is the result of comparable, not of differential processing speeds of two stimulus dimensions

On a digit-counting Stroop task, processing of the slower, nonverbal, i.e., number, dimension was slowed further by a large-number set (6 to 9), as compared with a small-number set (1 to 4). In the task, neutral symbols or conflicting digits were arranged on a horizontal line (e.g., @, 444) and on two separate sheets. Each sheet contained 120 stimulus arrays. Subjects counted out loud the number of symbols or digits in each array, and their counting times for each sheet were recorded. 23 subjects received the small-number set while 21 received the large-number set. It was found that counting a large number of symbols took significantly longer time (by 162 sec. per 120 stimulus arrays) than counting a small number of symbols. Moreover, interference was nonexistent (2 msec, per stimulus array) when a large number of conflicting digits were counted but was of a typical magnitude (110 msec, per stimulus array) when a small number of conflicting digits were counted. This suggests that Stroop interference is better explained as the result of comparable, not of differential, processing speeds of the two stimulus dimensions. Implications for the cause and the locus of Stroop interference are discussed.