Testing the Efficacy of Brain Training: a Comment on Owen, et al. (2010)
Oliver W. Hill1 and Zewelanji Serpell1
The recent study by Owen, et al. (2010) and the subsequent news article by Katsnelson (2010), both published in Nature, have caused a firestorm of attention in the popular media. Katsnelson’s provocative headline, “No gain from brain training,” has led to a number of hasty conclusions about the ineffectiveness of cognitive training that need to be addressed. We are members of a team of investigators on a three-year study funded by the National Science Foundation (NSF) to test the efficacy of cognitive training on the mathematics performance of inner-city students. While we are just beginning the second year of this study and only have preliminary data, these data coupled with several other (as yet) unpublished studies have produced effects that are robust enough to warrant mention, given the recent attention to this topic.
Critiques of the Owen, et al. study have already been made regarding the inadequacy of their treatment, both in terms of frequency, and intensity. They used relatively short (10 minutes) and infrequent (3 times per week for six weeks) training sessions. In contrast, one commercial cognitive training organization that we have investigated in the U.S., LearningRx, administers training in five one-hour intensive one-on-one sessions for 15 to 30 weeks. Training procedures consist of tasks that emphasize auditory or visual processes and that require attention and reasoning throughout the training. Students are trained to develop appropriate strategies to complete the tasks through the structured experience provided by the training procedures. The tasks that comprise the training are organized in a progressively more challenging manner and are designed to target one or more specific cognitive skills. The tasks repeatedly make demands on processing abilities and progressively increase those demands. It is the progressive intensity of the tasks that functions as the mechanism through which the cognitive functions are trained.
This organization has fifty-five centers in twenty-five states, and several thousand students complete the program each year, which produces a wealth of data. Several unpublished studies on these data (e.g., Marachi, 2006, Luckey, 2009) have reported robust changes in scores on cognitive measures, such as the Woodcock-Johnson Tests of Cognitive Abilities, in relatively short periods of training. In our NSF study, which is using an on-line version of the LearningRx procedure, preliminary data analysis indicates both improvement in cognitive assessment scores, and transfer of training to mathematics performance. In a pilot study conducted during our first year of training with a sample of 151 students (ages 12-17), we were able to achieve significant gains in measures of processing speed, working memory, visual memory tasks, auditory memory tasks, short-term memory and long-term memory, with robust effect sizes (eta2) ranging from .091 to .332 (Hill, Serpell & Turner, 2010). Interestingly, we also found that before training, the most significant cognitive predictor in mathematics performance was processing speed (Beta=.351, spc2=.06), while after training, the most significant cognitive predictor of mathematics performance was the measure of logic and reasoning (Beta=.391, spc2=.129). In the final phase of our study, we are using f-MRI to look at the recruitment of new cortical areas in the processing of mathematics that these changes suggest.
Final conclusions on the efficacy of cognitive training programs await the results of large sample size, randomized control studies using these types of more intensive training programs, such as our current study. In the meantime, the literature should reflect that hasty conclusions should not be drawn from Owen, et al. (2010).
References
Hill, O., Serpell, Z. & Turner, J. (2010). Transfer of Cognitive Skills Training to Mathematics Performance in Minority Students. Poster present at the Annual Convention of the Association for Psychological Science, Boston, MA, May, 2010.
Katsnelson, A. (2010). No gain from brain training. Nature, 464, p. 1111.
Luckey, A. (2009). Cognitive and academic gains as a result of cognitive training.
Unpublished dissertation, Arizona State University.
Marachi, R. (2006). Statistical analysis of cognitive change with LearningRx training
procedures. Unpublished manuscript, Department of Child & Adolescent Development, California State University — Northridge.
Owen, A., Hampshire, A., Grahn, J. Stenton, R. Dajani, S. Burns, A., Howard, R., &
Ballard, C. (2010). Putting brain training to the test. Nature advance online publication, doi:10.1038/nature09042 (20 April 2010), pp. 1-4.
1 Department of Psychology, Virginia State University, Petersburg, Virginia 23806, USA. This research is supported by Award# HRD-0714874 from the National Science Foundation.
