Birth order effects on intellectual performance show both positive and negative results. The confluence model reconciles these conflicting data by proving that these effects interact with the age of participants at testing, such that young children should show negative or no effects, whereas older individuals (past age 11 +/- 2 years) should show positive effects. Birth order studies strongly support this prediction. Some writers have claimed the apparent relation between birth order and intelligence is an artifact created by applying a cross-sectional analysis to data that should have been analyzed by comparing siblings within families. However, if siblings within the same family are compared all at the same time, their ages are necessarily different. As a result, birth order effects are confounded with age effects. Moreover, within-family data conceal patterns of aggregate effects that cross-sectional data reveal.