Reader Comments
Post a new comment on this article
Post Your Discussion Comment
Please follow our guidelines for comments and review our competing interests policy. Comments that do not conform to our guidelines will be promptly removed and the user account disabled. The following must be avoided:
- Remarks that could be interpreted as allegations of misconduct
- Unsupported assertions or statements
- Inflammatory or insulting language
Thank You!
Thank you for taking the time to flag this posting; we review flagged postings on a regular basis.
closeHorribly low power
Posted by isumrl on 08 Jul 2013 at 16:01 GMT
I tend to like the pen drop paradigm. The real problem here that I can't believe reviewers missed is the horribly low power to find the effect. The best practices in this field (comparing violent and non-violent games) is to require at least 100 participants per cell, so 200 for just a V vs NV comparison (based on power analyses using meta-analytic effect size estimates). If you also want (as this article does) to consider more games, or characteristics of players, then the number of cells goes up accordingly. Thus, 64 participants is nowhere near enough to have sufficient power to find the effect. It's certainly possible that their results might hold up if they had used appropriately large samples, but it's also possible that they'd find something different. This is especially important when publishing null results - at a minimum, you should have had sufficient power to find the effect. Ultimately, this article poses some interesting questions, but it can't answer any of them with any confidence in the results with such a small sample.
RE: Horribly low power
morgantear replied to isumrl on 17 Jul 2013 at 23:43 GMT
Thanks for your comments. We too like the pen-drop paradigm as a simple measure with high fidelity, and attempted to replicate the original Greitemeyer and Osswald findings as a base from which to test some other questions related to violent video games. What followed was a number of surprising failures to replicate, surprising because; 1) several past studies found effects with the pen-drop task and similar cell sizes, and 2) we tried to create situations optimal for showing the effect. We think we address criticisms of power in the general discussion of the paper.
If failures to replicate effects are never reported (especially direct replications like ours) then we are only seeing a partial account of the true nature of the effect.