Inconsistent flanker congruency effects across stimulus types and age groups: A cautionary tale

The flanker task is a common measure of selective attention and response competition across populations, age groups, and experiential contexts. Adapting it for different uses often involves changing methodological features that are rarely empirically compared with the previous design. This paper presents an example of how typical methodological changes can differentially elicit congruency effects across age groups. We compared two flanker tasks, using direction stimuli on a laptop versus color stimuli on a tablet, in young children (2–7 years; Experiment 1), older children (6–10 years; Experiment 2a), and adults (19–23 years; Experiment 2b). Young children showed the expected congruency effects in the direction task, and one year later a subset of the sample completed the color task, also showing congruency effects. Longitudinal comparisons showed no difference in the congruency effect across tasks, but nearly half of the sample was excluded due to high error rates. To avoid excluding children with few correct trials, we modified a new measure, signed residual time, to incorporate correctness and reaction time per trial. With the larger sample, this measure showed no difference in congruency effects across tasks. To compare these tasks when completed within the same session, we tested older children and young adults in both tasks and found congruency effects in the direction task but not the color task. These results raise concern that tasks adapted for young children may not perform comparably in other samples, and we caution researchers to anticipate this possibility when modifying cognitive tasks.