|
Abstract:
|
Statistics courses regularly caution against belief in the "gambler's fallacy" or the "hot hand fallacy". For years there has been little to no statistical evidence that such phenomena exist (hence the "fallacy"). However, recent research (Miller and Sanjurjo, 2014, 2015) suggests that previous analyses of such phenomena have been subject to a bias, and after correcting for this bias, there is in fact evidence that the effects are real. In this talk, we present results of classroom investigations relating to recent research on conditional dependence of outcomes in sequential data on streaks of past outcomes. We also discuss implications for coverage of the gamblers' (or hot hand) fallacy in statistical literacy courses.
|
