treatment effects. Just as we may use a variety of statistical techniques and analyses to answer questions about these issues in a single study, the IDA framework provides a set of guidelines that may also be widely applied across a range of techniques and analyses using a pooled data analysis approach. These guidelines have evolved over time through efforts to conduct IDA with the goal of drawing inferences about substantive hypotheses at the level of the pooled analysis. These guidelines thus reflect ways of addressing challenges in accounting for between-study heterogeneity using IDA. Although in previous work we have discussed five sources of between-study heterogeneity that make IDA challenging (i.e., sampling, measurement, geographical region, history and design characteristics; Curran & Hussong, 2009), we now highlight two of these – sampling and measurement – and consider the implications of these sources of between-study heterogeneity for the ultimate goal of IDA, hypothesis testing.
