The Fish-Shark go/no-go task was administered on an IBM-compatible computer using E-Prime 1.2 (Psychological Software Tools, Pittsburgh, PA). Children used one button on an RB-530 button box (Cedrus Corporation, San Pedro, CA) to respond. Before beginning the task, children were instructed to press the button when they saw a fish, to catch it in their fishing net, but that they should not press the button when they saw a shark, because sharks would break through their net. In an initial training phase, children were presented with all fish and shark stimuli, and practiced pressing the button to catch fish and withholding the button press to avoid catching sharks. Finally, 40 test trials were administered (75% fish trials, 25% shark trials). On each trial, the stimulus (fish or shark) appeared on the screen for 1500 milliseconds or until the child pressed the button. On fish trials when the child correctly pressed the button, a picture of the fish caught in a net appeared on the screen for 1000 milliseconds, accompanied by a bubbling sound, indicating that the child “caught the fish.” On shark trials when the child erroneously pressed the button, a picture of the shark breaking through a net appeared, accompanied by the sound of a buzzer. No feedback was presented when the child did not press the button. There was a 1000 millisecond inter-stimulus interval between end of the previous trial stimulus or feedback and the onset of the next trial stimulus. Trials were block-randomized so that each block of eight trials included six fish trials and two shark trials (where one shark trial followed two fish, and the other shark trial followed four fish, in a manipulation of preceding trial context), and all exemplars (10 fish, three sharks) appeared with roughly equivalent frequency. Trials with responses faster than 200 milliseconds were eliminated from the analysis, as they were too quick to reflect responding to the current stimulus. Proportion correct was computed separately for go and no-go trials, and mean response times (RTs) were calculated for correct go trials. The proportion of hits (correct go trials) and false alarms (incorrect no-go trials) were used to calculate sensitivity (d′; the standardized difference between the hit rate and the false alarm rate, calculated by subtracting the z-score value of the hit rate right-tail p-value from the z-score value of the false alarm rate right-tail p-value; Macmillan & Creelman, 2005 ). The d′ sensitivity index is used routinely in the signal detection literature and reflects the degree to which a subject responds differentially to two classes of stimuli, where higher values reflect better discrimination.