Institution: Karolinska Institute
Researcher(s): Thorell, L B, Lindqvist S, Bergman S, Bohlin G, Klingberg T.
Research: Cogmed JM
Published: Developmental Science, December 2008
Funding: This study was supported by grants from the Vårdal Foundation and the Swedish Council for Working Life and Social Research to the first author and by grants from the Knut and Alice Wallenberg Foundation and the Bank of Sweden Tercentenary Foundation to the last author. Training program, investigator training, and technical support provided by Cogmed. No funding provided by Cogmed.
Working memory (WM) – the ability to retain and manipulate information in mind for short periods of time – and inhibition control – the ability to refrain from acting impulsively – are important executive functions that are related to academic achievement and ADHD. As the importance of executive functions has become more widely known, there has been growing interest in whether these functions can be improved with cognitive training. Of particular interest is whether training could enhance these critical skills in young children and thus prevent difficulties in early academic achievement that may have otherwise occurred.
This question was the focus of the study reviewed here. Participants were 63 four to five year olds attending four different preschools in Sweden. Children were randomly assigned to one of four experimental conditions: Working Memory Training (WMT), Inhibitory Control Training (ICT), an active control condition (ACT), and a passive control condition (PCT). These were all normally developing children, none of whom had ADHD or any other known psychiatric diagnosis.
WMT was provided in daily 15 minute sessions over a five-week period. The focus was on training visuo-spatial WM based on research indicating that this is more strongly associated with ADHD than verbal working memory. During training a number of visual stimuli were presented sequentially on the computer screen; children had to remember their location and order and demonstrate this by clicking with the mouse in the correct locations. Task difficulty was manipulated by increasing the number of stimuli that had to be remembered and was adjusted after each trial according to whether the child had answered correctly. This ‘adpative’ training ensured that children were constantly challenged by tasks that were closely suited to their WM capacity.
Children in the ICT group were trained on various tasks designed to train inhibitory control. For example, on one task children had to respond quickly when a certain stimulus (e.g. fruit) flashed on the screen but not to respond when any other stimulus appeared. On another task, children responded whenever a certain stimulus (e.g. fish) was presented except when it was immediately followed by a stop-signal (e.g. fruit). Thus, on some trials children needed to respond quickly (by hitting a certain key) to be correct and on others they had to inhibit responding. Difficulty level was manipulated by decreasing the time that children were allowed to respond on trials when a response was required. As in WMT, training was conducted 15 minutes per day over five weeks.
The ACT condition had children play commercially available video games that were judged to place low demands on WM and inhibitory control. This condition provided an equivalent amount of time engaged in a computer-game task and allowed researchers to test whether tasks specifically designed to train important executive functions yielded better results. Children in the PCT condition did not participate in any computer activity and merely participated in the pre-post assessments.
Pre- and post-testing was conducted by an experimenter who was blind to the group assignment of each child. Eight different measures that assessed a variety cognitive functions, i.e., visuo-spatial working memory, verbal working memory, inhibitory control, and attention, were administered. These measures were all different from what children had been exposed to in training. These were all lab-based assessments and behavioral ratings from either parents and teachers were not obtained.
The researchers first examined children’s performance on the training tasks improved over the five-week training. This was done by comparing performance during the second and third training day to their performance on the last two days. For both WMT and ICT, results indicated significant improvement on the tasks children were trained on.
The more important question is whether children who received WMT and ICT improved more on non-trained tasks than control children. To examine this, improvement on each measure was calculated by subtracting the pre-test score from the post-test score and the improvement made by children in the different groups was compared. (Because there were no differences in how children in the two control groups performed, they were combined into a single group for these analyses).
Children who received WMT showed greater improvement on the non-trained tasked of visuo-spatial WM and verbal WM than control children. The magnitude of these differences would be considered large by conventional standards, i.e., effect sizes greater than .80. They also showed greater improvement on computerized tests of visual and auditory attention; the effect sizes here were moderate. Generalization to improvements in inhibitory control were not found.
Although children in the ICT group improved during training on those tasks they were trained on, there was no evidence that this generalized to any of the non-trained tasks. Thus, unlike children who received WMT, these children showed no more improvement on any measures than comparison children.
Summary and Implications
Results from this carefully designed study indicate that WM can be improved in preschool children by five weeks of WM training. In addition to gains on non-trained measures of WM, this training also showed evidence of generalizing to improvements in attention. On the other hand, training of inhibitory control, another important executive function, showed no evidence of generalization even though improvement on the trained tasks had occurred.
The large effects obtained for WM training on working memory and the moderate effects found for attention suggest that this type of training could perhaps be helpful in early intervention with children who have deficits in these areas. It should be noted, however, that participants in this study were all typically developing children and that the impact of training on preschoolers with WM and/or attention deficits remains to be demonstrated. The current study also did not examine whether the benefit evident immediately after training ended would persist. Although similar training of working memory has been shown to reduce inattentive symptoms in older children with ADHD, and to yield gains that persist beyond six months, this may not be the case with younger children.
While the above cautions are important to keep in mind, results from this study are extremely promising and represent the first demonstration that WM training can benefit young children. Future studies should examine the impact of such training on preschoolers with WM and attention deficits, and test for the duration of these effects. Such work will be necessary to better understand the potential of WMT in improving young children’s academic and behavioral functioning.