Pay Attention: Researchers Debunk Game Link to Concentration Study

Remember that study from earlier this week which intimated that playing videogames and watching television were linked with attention problems in children? Texas A&M researcher Christopher J Ferguson and T. Atilla Ceranoglu, from Harvard Medical School, saw the research and responded with a scathing (for research anyway) rebuttal.

Ferguson, who has challenged the work of Iowa State University’s Craig Anderson before, and Ceranoglu, who uses games to assist in psychotherapy treatment, submitted their response—entitled Poor Measurement, Poor Controls and Spurious Results in Swing et al.—to Pediatrics, which also published the original research.

If you couldn’t tell from the title, the authors penned the response in order to “make readers aware of limitations and omissions in the methodology and conclusions of the Swing et al (2010) study of video games and attention.”

Ferguson and Ceranoglu offer four distinct weaknesses in the research in Television and Video Game Exposure and the Development of Attention Problems:

1.  …the literature review overlooks a number of recent studies that contradict their views on the relationship between video games and aggression, as well as studies suggesting that video games are more likely to increase, not decrease, attention and cognition.

2.  …the authors puzzlingly failed to use any of the clinically well-validated measures of attention problems, such as the Child Behavior Checklist and the Conners Rating Scale. It’s also unclear why the authors relied solely on teacher reports, failing to include parent reports that arguably would be based on greater familiarity with the child, and again chose not to employ existing validated measures.

3.  …the authors make no attempt to control for other commonly measured relevant variables that may influence attention such as home environment, school quality, parent education, or poverty, genetic risk, etc. It is quite possible that any effects found are spurious, and would disappear in a better controlled study.

4.  …all standardized regression coefficients for children in the study are less than .10. This indicates that the overlap in variance between media use and attention is less than 1%. Even taking these findings a face value, these are weak effect sizes without practical significance, effectively no different from zero.

The researchers concluded that the findings in the original study were “unable to support the weight that Swing et al. (2010) attempt to place on them, and give no cause for concern to clinicians, educators or parents.

Tweet about this on TwitterShare on FacebookShare on Google+Share on RedditEmail this to someone

Leave a Reply