Abstract. The role that question asking and dialog play in vicarious learning environments (VLEs) was investigated using a pretest-posttest design. Data was collected from 66 participants among four different learning conditions (interactive, standard vicarious, deep question monolog, and deep question dialog) were used in the study. The data analysis revealed that participants in the deep question monolog and deep question dialog condition outperformed the standard vicarious condition. The implications of these findings on VLEs and e-learning environments are discussed.