Ivan Ropovik just published a study in Intelligence looking at executive function in children. He used a number of PEBL tests, as well as several non-PEBL versions of tests for which PEBL has implementations, to look at the factor structure of executive control in children. One of their basic covariance (structural equation) model of the data is shown on the left.The basic results for this model show that executive function in this sample can be divided into several large bins: working memory (aka cognitive flexibility); ability to learn; selective attention, and inhibition. Importantly, this is probably the first published paper that used PEBL's 'Victoria' stroop task. Computerized stroop tasks are tricky because the interference tends to reverse for manual responses in comparison to verbal responses (naming is easier to do verbally, matching colors is easier to do non-verbally), and so one thing this did was establish a fairly strong correlation between the PEBL Victoria stroop and a verbal task (+.39), which was a relatively high correlation for this study (where inter-test correlations maxed out at around .6. The test was also modestly correlated with som other tasks (TMTA=.22, WCST-64 = -.18), but uncorrelated with a several as well related to ability to learn, showing some specificity of the task.
This paper joins a growing set of research studies done using PEBL on children. I find this approach is really important, because many cognitive-behavioral computer tasks were originally designed for adults, and may not translate well to children. This research shows it is possible to use several of the PEBL tests in a childhood sample.