This article describes some results on the effect of the preceding stimulus on performance in an intensity‐identification experiment with feedback. Based on previous research and our own subjective experiences as listeners, we had expected that sensitivity would increase when the intensities of the preceding and current stimuli were close together. The results of our experiments show, however, that this is not the case: sensitivity is independent of the preceding stimulus. The fact that the response variance tends to decrease when the intensities of the preceding and current stimuli are close together is found to be caused solely by sequential effects in response bias. These findings create an interesting puzzle concerning the mechanism by which feedback improves sensitivity in identification. Apparently, the most obvious hypothesis—namely, that the feedback enables the previous stimulus to be used as a standard for judging the current stimulus—is false.