a little beef with a study we heard about a couple days ago, in which we learned that 8.5 percent of America's youth are addicted to video games. Langer has several problems with the study, one issue being that it was an opt-in online panel -- a "self-selected 'convenience sample'" -- rather than a probability sample (random sampling). It gets technical, but Langer's issue is that he's yet to hear a "reasonable theoretical justification for the calculation of sampling error with a convenience sample."
In the end, the author of the addiction study, Prof. Douglas Gentile (who is also the director of research for the National Institute of Media and the Family), wrote to Langer saying that he was unaware the data came from a convenience sampling. So, good news, everyone: either 8.5 percent of youths are addicted to video games ... or they just think they are. Yay, statistics!