What a Planning Task Actually Measures in Children

What the tests measured

A child sits at a touch screen. Coloured boxes appear, one of them conceals a blue token, and when the child finds it another is hidden and the search starts again. A box that has given up a token will never hold one again. Two scores come out of that task. Returns to an emptied box count as memory failures. The second, the strategy score, counts how often the child begins a fresh search from the same box, where consistency earns the better mark, and it belongs to the component the researchers labelled planning.

Beside it ran a tower task, in which discs are moved into a target arrangement in as few moves as possible, and a sheet on which the pencil alternates between numbers and letters. Jessica O’Brien and four colleagues at the Kennedy Krieger Institute and Johns Hopkins sorted fifteen such tasks into four components of executive function and published the results in Archives of Clinical Neuropsychology in 2010. Fifty-six children with ADHD, twenty-six of them girls, were compared with ninety typically developing children, forty-two of them girls, all aged eight to thirteen and matched on age and socioeconomic status.

Working memory separated the groups on all four of its tasks, at p values of .05 or below, and boys and girls with ADHD were affected alike (interaction p = .84). Response preparation proved narrower. Only the timed motor examination and the trial-to-trial variability of reaction time distinguished the ADHD group from controls, at p = .0002 and p = .02, while naming colour patches and scanning a page for printed numbers did not.

Among the girls the planning component separated ADHD from control at p = .0004, with a partial eta squared of 0.25; among the boys it produced nothing at all, at p = .77. A partial eta squared of that size puts a quarter of the variance in the girls’ scores down to group membership. The strategy score carried most of that weight, at p = .00006, while the tower task reached p = .04 and number-letter switching only p = .07, a trend rather than a result.

Motor overflow ran the other way and exposed how these comparisons work. Overflow is involuntary movement elsewhere in the body while one hand performs a deliberate movement, and girls with ADHD showed more of it than control girls (p = .0003) while boys with ADHD were indistinguishable from control boys (p = .94). The two ADHD groups did not differ from each other in overflow; their comparison groups did. Typically developing boys of this age overflow more than typically developing girls, so the boys with ADHD had further to travel before they stood out.

What it means for you and your child

A phrase like impaired planning in a report describes performance on tasks of this kind. For the girls in this sample, most of that impairment came from a single habit of search, whether a child adopts a consistent point of departure and holds to it. The finding is narrow, and its narrowness is what makes it usable. A child who begins each attempt wherever her hand happens to land is doing something specific and visible, at a desk as readily as at a screen.

Every score here was calculated against same-sex peers, and the overflow result shows what that does. Typically developing girls are ahead of typically developing boys on motor and executive tasks in this age range, which sets the comparison group for a daughter higher than the one a son is held to. The authors regard that asymmetry as the practical point of the study.

Subtype predicted nothing. The thirty-five children with the combined presentation and the twenty-one with the inattentive presentation were indistinguishable across all four components, at p values ranging from .17 to .89. Whatever a subtype label describes about a child’s behaviour at home and at school, it forecast no part of how these children performed at the table.

Average IQ sat slightly above the population mean, reading difficulty was screened out and most co-occurring conditions were excluded, which the authors acknowledge may have attenuated the deficits they measured, as may a sample too small to confirm that the subtypes really do perform alike. What they measured, they measured in an unusually clean group of children.

Boys at these ages showed no planning deficit, and the authors’ reading is that they develop these skills later and may grow into the difficulty during adolescence, an expectation rather than a finding, since no child in this study was followed forward. The girls’ result describes how a child went about looking for a token in a box, over two days of testing, in Baltimore, and it extends no further than that. A girl measured against girls can be plainly behind them while performing at the level of a boy nobody has raised a concern about. That arithmetic is fixed before the first task begins.