Abstract
The finding of longer reaction time prior to longer-lasting responses (e.g., Klapp, 1977a; Klapp, Wyatt, & Lingo, 1974) was not replicated by Kerr (1979) in a statistically powerful attempt. The present experiment confirms that this effect is not reliable when duration knowledge of results is similar to that used by Kerr, although a reliable long-short effect was obtained with other knowledge of results arrangements. This unexpected sensitivity to knowledge of results suggests that previous results from this paradigm must be interpreted with caution.