just sent this out to one of the guys at pwinsider in regards to the Pediatrics journal study linking wrestling viewership to violent behavior. Dunno if they'll post it or not, but here it is for you to read, if interested.
The point in question (contains links to the study .pdf and a refutation of the research):
http://www.wwe.com/inside/news/durantstudyrefute
[quote name='me']I'm not certain if this is any interest to you at all, but reading the WWE Press Release, the research in question, and the Media Life citation of a refutation, I have to admit that I can't stand by the WWE this time around.
First, Robert Thompson, head of Syracuse University’s Center for the Study of Popular Televsion, claims that the study only identifies correlation (a relationship between two variables, in this case watching TV wrestling and antisocial/aggressive behavior). Well, that's certainly true of a correlation, but Thompson makes claims that stand in the face of all longitudinal research (research conducted at various points in time over a long period). A correlation is, yes, independent of cause-and-effect, but that is not the case in longitudinal research. When something is measured at point 1 (in this case, watching wrestling), while another at point 2 (antisocial/aggressive behavior), then the cause-effect relationship is less contentious, because it is a logical fallacy to suggest that what was measured at time period 2 could cause what was measured at time period 1. Wrestling viewing was measured in the Fall of 1999, while the aggressive "fighting" measures were taken in Spring of 2000; it is *impossible* to suggest that something that happened in 2000 can affect something that happened in 1999. Thus, any correlation can be imputed to mean causation, since there is a statistically significant relationship, and the time ordering only goes in one direction. This correlation *could* be contended for falsely imposing causal order if measured simultaneously. Since it was not, there are no reasonable grounds for Mr. Thompson's statements. In short, correlation can be used as causation in longitudinal analysis, but hesitantly so. I'm not familiar with Robert Thompson, but for him to suggest otherwise is mistaken and misleading at best, and intentionally deceptive and irresponsible at worst.
Second, a refutation must tackle the merits of the study, so to produce a press release merely smarming the paper as "junk science" is an attempt to prevent otherwise uninformed people from reading the research in question. It is professionally irresponsible and poor form to simply attack a study for the results it produces, just as it is to praise studies that produce results one finds desirable. If you want to challenge the methodology of the paper in question, then it would be pertinent to ask why the authors focused professional wrestling; after all, when it comes to tv programming labelled harmful or detrimental to youth affectation, the boundaries are seemingly limitless. To focus strictly on professional wrestling, then, doesn't use proper controls in its research.
Third, Pediatrics is, without fail, the most highly regarded journal of its kind, and its impact factor (which measures circulation, citations of previous studies, etc.) is the highest amongst other pediatric medical journals. It is also a peer-reviewed journal, so Mr. McMahon incorrectly criticized it on that end.
Fourth, while I understand the ligitious-crazed history of the WWE, their claim of legal action is peculiar, and without merit. I question whether or not anyone in the company even read the study, simply because the study found this correlation to exist amongst females, but not males. To quote the paper, "Among males, the variables that had been found to be significantly associated with watching wrestling for the cross-sectional sample in the fall of 1999 were not significant when analyzed longitudinally. Among females, the frequency that they reported watching wrestling on TV in the fall of 1999 was significantly associated with the frequency of engaging in several indicators of fighting and weapon carrying in the spring of 2000 (Table 6)" (e269). So, the study is, more or less, much ado about nothing. There is *NO* significant correlation with males, who are the unarguable majority demographic of WWE and professional wrestling program. Even considering the correlation with females, the r-values listed in Table 6 of the study suggest that, if watching wrestling does have an impact on fighting/aggressiveness, it is minimal at best. The kinds of correlation (r) values listed are small enough to question why they would be published, as the unexplained variance (1 - r) leads one to the following conclusions: of those who carried weapons to school, 88% of the probability of that is unexplained by wrestling viewership (1 - .12); of those who fought at school, 86% of the probability is unexplained by wrestling viewership (1 - .14) and so on. In short, there is far too little explained by this study, and what does explain fighting/aggressive behavior, statistically, only applies to females.
While I would argue that, by the conclusion, the study offers more non-findings that acquit wrestling viewership from violent/antisocial tendencies (for males and partially for females), it is inappropriate for the WWE and Robert Thompson to attempt arguing with researchers, who clearly they cannot engage is competent discourse with. If that were the case, they would shrug off the non-findings, and consider the findings for females practically irrelevant. Instead, they will, regretfully, bring more attention to themselves, in a negative light, for choosing to publish a "refutation" of peer-reviewed scholarly works. The study is fairly shallow, lacks relevant control variables, and seems overreaching in its conclusion, but instead of pointing out methodological flaws, the WWE has resorted to name calling and citations of a scholar who seems deprived of methodological knowledge.[/quote]