An experiment was conducted to compare the disruptive effects of irrelevant information (names of colours) superimposed on relevant information (bands of colours) with the effects of irrelevant information occupying the interval between stimuli. It was found that irrelevant information disrupted performance only when it occupied the interval between stimuli, suggesting that interference was primarily with the rehearsal process and not with the perceptual process. It made no difference whether die names of colours were the same as or different from the bands of colours. Further, control conditions suggested that colours can be kept-track of better than colour names. The results are discussed in terms of visual imagery and the spatial window model of keeping-track performance.
PREVIOUS STUDIES OF THE EFFECTS of irrelevant materials upon keeping-track performance have utilized a paradigm whereby the relevant stimuli were presented sequentially at an irregular rate (Karsh & Monty, 1972; Monty, Karsh, & Taub, 1967a, 19676). The irrelevant stimuli were either interspersed with a string of relevant stimuli in a manner that made the overall presentation rate regular, or they preceded each relevant symbol in a manner that enabled them to serve as a cue to alert S to the onset of the next relevant stimulus. These studies were consistent in indicating (1) that irrelevant stimuli (a green square or the letters U, V, & W) which were distinctly different from the letters that had to be recalled (Q, R, & S ) could be used to pace rehearsal and thus improve performance ( Monty, Karsh, & Taub, 1967a; 19676), and (2) that materials that were similar to the relevant materials interfere with the processing of the relevant information and thus degrade performance (Karsh & Monty, 1972). While these studies were also consistent in suggesting that the major effects of the irrelevant materials appear to be in the rehearsal and not in the encoding stage of processing, the generality of their findings as well as their•Requests for reprints should be sent to