The effects were higher for pre-lesson questioning (d = 0.94) and post-lesson questioning (d = 0.86), compared to questions interspersed during the lesson (d = 0.52); when the questionings were delayed (d = 0.72) compared to immediate (d = 0.54); and where there was teacher modeling (d = 0.69) compared to none (d = 0.47).

