Sampling distributions are fundamental to an understanding of statistical inference, yet research shows that students in introductory statistics courses tend to have multiple misconceptions of this important concept. A common instructional method used to address these misconceptions is computer simulation, often preceded by hands-on simulation activities. However, the results on computer simulation activities' effects on student understanding of sampling distributions, and if hands-on simulation activities are necessary, are mixed. In this article, we describe an empirical intervention study in which each of eight discussion sections of an introductory statistics course at a large research university was assigned to one of two in-class activity sequences on sampling distributions: one consisting of computer simulation activities preceded by handson simulation using dice, cards, or tickets, and the other comprised of computer simulation alone with the same time-on-task. Using a longitudinal model of changes in standardized exam scores across three exams, we found significant evidence that students who took part in a hands-on activity before computer simulation had better improvement from the first midterm to the final exam, on average, compared to those who only did computer simulations. Supplementary materials for this article are available online.