|
Article Excerpt
Recent studies of children with poor attention, as rated by teachers, have consistently shown that these children make more errors (false alarms to non-targets) when searching for targets in a cluttered display on a computer screen than children rated as having good attention (Cornish, Wilding, & Hollis, 2006; Wilding, 2003; Wilding & Burke, 2006; Wilding, Munir, & Cornish, 2001;). Thus they tend to respond by clicking the computer mouse on non-targets or background, or in some cases repeat responses on already located targets. However, time taken for each correct response (and also for each error response) did not differ significantly between the two groups. Manly et al. (2001) using the Skysearch task in the Test of Everyday Attention for Children (TEA-Ch) also found no difference in time between groups differing in attentional ability. In the above computerized visual search task, the difference in error rate between groups was more reliable in more difficult versions of the search task, which required a difficult discrimination between targets and foils or alternation between two different targets. However, children with poor attention showed no tendency to adjust their speed less appropriately than children with good attention on these more difficult tasks compared with the simpler ones (Cornish et al., 2006), and hence there was no indication that the increased difference in errors was due to more impulsive responding by these children in the more difficult conditions.
Impulsivity is a commonly accepted component of attention deficit hyperactivity disorder (ADHD) and it is frequently implied that in consequence children diagnosed with ADHD will make errors due to premature responding without adequate analysis of the stimulus situation. However, the above results offer no support for this assumption in that the two groups did not differ in speed of responding, but did differ in accuracy. Furthermore, time and error rate were not correlated in these studies and these measures were related to different measures of individual differences (time was related to measures of general ability and accuracy to attentional ability; see Wilding, 2005, for a review). Even though the children in the poor attention group were not formally diagnosed with ADHD, the rating scale that was used incorporated standard criteria for this condition and included both attention and hyperactivity ratings, which were highly correlated in the samples tested. Though the studies of Wilding et al. (2001), Wilding (2003) and Wilding and Burke (2006) did not select children with extreme impairments on the distribution of these ratings, Cornish et al. (2006) compared children from the top and bottom 5% of the distribution on the rating scale they employed and obtained essentially the same results as in the other studies. It is reasonable, therefore, to conclude that the results are generalizable to children with a clinical degree of impairment in attention.
In fact, the available evidence that children with high impulsivity make errors due to premature responding is not strong and there is no clear evidence that children with ADHD perform in this way. In many studies such children have been found to respond more slowly than children with good attention. Though there is more reliable evidence that such children are intolerant of requirements to delay responses, this may well depend on different mechanisms. Their response time is also more variable than that of children with good attention (Kuntsi, Oosterlaan, & Stevenson, 2001).
In a complex study, Sergeant and Scholten (1985) tested three groups (with only eight participants in each group), overactive and distractible, normally active and distractible, and normally active and attentive (in more widely used terminology these equate to groups with attention deficit disorder with hyperactivity, attention deficit disorder without hyperactivity and a control group). There were three instruction/incentive conditions given to each group: no instructions always given first, speed instructions and accuracy instructions, given in a balanced order. Displays were presented consisting of two, three or four letters and a decision was required as to whether a target letter was or was not present. Analyses were carried out on response times only. Errors varied appropriately between conditions and did not differ significantly between groups.
Though the precise implications of the complex pattern of findings that the authors report are unclear, they certainly did not demonstrate fast inaccurate responding as the main source of poor performance in the overactive-distractible group. This group demonstrated no inclination to perform quickly in the speed condition (cf Stevens, Boydstun, Dykman, Peters, & Sinton, 1967; Stevens, Stover, & Backus, 1970), suggesting some problem in control systems (presumably in the frontal lobes) responsible for adjusting response strategies to match task requirements. Furthermore, examination of error rates at different response times showed that, while both the distractible and control groups showed a very pronounced trade-off of speed for accuracy, there was almost no sign of such a trend in the overactive-distractible group, who produced almost as many errors on slower responses as on faster ones. This demonstrates that many errors in this group were not due to impulsive responding.
Van der Meere, Gunning, and Stemerdink (1996), using a similar scanning task, presented one or two targets followed by a display of four items. The probability of a target occurring in the display was .5 or .25 in different conditions. There were no differences in scanning speed between the ADHD and control groups (indexed by the slope of response time against memory load), nor any differences in the effects of target probability on speed or errors. The ADHD groups were slower and less accurate overall, but the authors concluded that inefficiency in these groups was not due to inefficient processing or favouring speed against accuracy (i.e. impulsive responding) or poor ability to switch set when the less frequent response was required in the condition with low target probability. They suggested that the differences were due to delayed motor processing, but do not elaborate on this proposal nor suggest...
|
