Mar
11
2015
A new CcaM study evaluated existing measures of TV and game violence exposure. Results indicate that one form (direct estimates) are most appropriate for media violence research, and highlight how we can improve our measurement in the future.
High quality measurement is key for high quality science. If we hope to know how media violence may influence youth, for example, we need to make sure our measure of media violence is sound. In this study, CcaM researchers compared the most common ways of measuring media violence to identify which were valid measurement tools. Specifically, three measures of violent game and TV exposure were evaluated:
(1) direct estimates, which ask respondents to report how often and how long they watch TV shows and play games that contain violence;
(2) user-rated favorites, which ask respondents for their three favorite TV shows and games, how often they watch/play these titles, and how much violence they contain; and
(3) agency-rated favorites, which uses the respondent’s favorite titles, but these are coded by the researcher for presence of violence.
Each of these three measures was evaluated by looking at the stability of the measure over a four-month period (test-retest reliability), by looking at their relationship with coded media diaries (criterion validity), and by looking at their relationship with aggressive behavior and gender (construct validity).
Results showed that all three of these measures were reliable and valid for game violence exposure. For television violence exposure, however, only the direct estimates were reliable and valid. This indicates that researchers who are interested in measuring media violence exposure should rely on direct estimates, as this was the only measure that worked for television and game violence exposure.
These findings were obtained using a sample of 238 Dutch early adolescents between 10 and 14 years old. These adolescents completed two surveys with the three exposure questions, as well as two media diaries. The titles reported as favorite media, and the titles reported in the media diaries, were content coded for presence of violent content using the Kijkwijzer and PEGI rating systems.
This study is the first to evaluate several measures of exposure to violence in games and on TV. Although effects of media violence exposure on aggression have been studied for decades, there is surprisingly little work that has evaluated these exposure measures. Such evaluations are important, because the quality of measurement affects researchers’ ability to detect and interpret effects. If measures are unreliable, this may result in smaller or even null effects. If measures are invalid, this means it is difficult to interpret any effects found. Knowledge about the quality of current measures will help researchers to evaluate past research, and improve their future work. The paper also identifies several ways in which researchers can further improve these measures.
The study was conducted by a team of CcaM researchers: Karin Fikkers, Jessica Taylor Piotrowski, and Patti Valkenburg. Click here to read the article “Assessing the Reliability and Validity of Television and Game Violence Exposure Measures”, published in Communication Research.
If you would like more information about this article, please contact the corresponding author – Karin Fikkers – via email at