This experiment compares children's comprehension and problem solving with the same information presented in three different media formats: an embodied video game, a first-person video, and a print narrative. The embodied video game emphasizes interaction and causation, where the player moves the narrative forward by causing change through interaction. According to embodiment theorists, the ability to create knowledge is predicated upon the ability to identify and connect changes, and what causes change in events. Comprehension is measured in this study with the Event-Indexing Model, (EIM). Research on the EIM indicates that identification of causation is often highly correlated to identification of other elements of comprehension, including memory of time, space, objects, and intentions across events. This experiment examines whether media format, which emphasizes embodied interaction and identification of causation, improves comprehension and problem solving. In question 1, this experiment examines whether the embodied video game will lead to superior comprehension and problem solving outcomes compared to the same information presented in a video or a printed text. Question 2 compares comprehension and problem solving when the reading text condition follows playing the game and watching the video. The third question examines the role of causation, which is the ability to identify actions that create changes between narrative events in a text. This dissertation analyzes comprehension and problem outcomes across media: as an embodied video game, a video, or a printed text. Additionally, it examines reading performance across presentation order, and the importance of identification in situation model construction.