A second limitation of the study is that we were not able to answer the question of how much talk during shared-book reading is optimal for word learning or if certain children benefit from less talk. A third limitation is that the sample was drawn from a restricted geographic area – only two school districts. As a result,generalizing the study to broader populations of preschoolers maybe difficult. In addition, the sample size was small: 92 students clustered within 13 classrooms. Because of the small sample size,nonsignificant effects should not be over-interpreted. That is, they may not have been detected because of a lack of power. By the same token, we performed a large number of tests given such as mall sample and, in order to maintain power, did not adjust the Type I error rate to account for this. Our significant findings should,therefore, also be considered cautiously. Similarly, our inability to calculate proportion reduction in variance effect sizes, also likely asa result of the small sample size, is a further limitation of the study.