Post-publication Peer Reviews to:
|
|
|
|
|||
|
Paul Misner, Sales Smartchive, LLC, Jim Kavadias, Bob Schultz, Jef Fieldman
Send letter to journal:
vpndude{at}vpndude.com Paul Misner, et al.
|
This study negates to filter out the socio-economic, and psychological backgrounds of the types of people who watch professional wrestling. Since viewer demographics are easily obtainable, (and partly used in this study) by Neilsen, one must question whether the goal of this study was to provide a truly scientific outcome.
By using bad science, the authors end up sabatoging what their own motives might have been, to not have children watch wrestling.
Let's do it right next time. Let's filter out the kids who watch WWE because there isn't a parent at home. Let's filter out all the kids who are violent because a parent beats them, and who watch WWE/
In this particular case, good science would have not required much more work.
Conflict of Interest:None declared |
|||
|
|
|||
|
Christopher J. Ferguson, Assistant Professor of Clinical and Forensic Psychology Texas A&M International University
Send letter to journal:
CJFerguson1111{at}aol.com Christopher J. Ferguson
|
This study purports to supply evidence that individuals who watch professional wrestling are more inclinced to engage in "date fighting" and other antisocial behaviors. It is unfortunate that the authors appear to be "naive" (as many authors in medical and social science are) as to the differences between "statistical" signifcance and "effect size". As "statistical significance" is highly sensitive to large sample sizes (as with the present study a sample of over 2000), rejecting or not rejecting the null-hypothesis is (in and of itself) arguably uninterpretable (see Cohen, 1994 for a discussion). A look at the various "statistically significant" relationships discussed in this study finds that the effect size (here denoted r2x100) ranges from essentially 0% to no higher than 3.6% Unfortunately such weak effect sizes (not atypical for media violence effects studies) are difficult to conclude as meaningful, particularly when the authors fail to report the confidence interval around the effect size (if such a confidence interval included 0% it should be concluded that no effect has been demonstrated). This study is an example of how statistics in social and medical sciences are in dire need for improvement. Conflict of Interest:None declared |
|||