Finding the target in search tasks using detection, localization, and identification responses

Visual search is a favourite paradigm for researchers interested in attention because of its ease of implementation and interpretation. Typically, visual search requires the participant to indicate the presence or absence of a prespecified target. Although the slope ratio for present versus absent trials is no longer considered to be indicative of whether search is serial or not, target detection remains a staple amongst studies examining theoretical and empirical aspects of attention. The current study sought to compare within subjects three tasks in a visual search paradigm, detection, localisation, and identification, using identical stimuli. Detection differed in both pattern of error rates and slope from both identification and localisation. Moreover, the slopes from identification and localisation were significantly correlated, whilst neither was significantly correlated with slopes from the detection task. These results suggest that researchers interested in using slopes to estimate search efficiency should use localisation or identification, rather than target detection.