The transitory information effect occurs when learning is reduced as a result of information disappearing before the learner has time to adequately process it, or link it with new information. According to cognitive load theory such processing takes valuable working memory resources away from learning. This effect has been mainly investigated using instructional animations, and therefore the main motivation of this study was to investigate the impact of spoken explanatory text, a fundamental form of transitory information, rather than animations. This thesis investigated two strategies that were expected to reduce the harmful effects of transitory information; namely replacing spoken text with written text, and segmenting text into smaller sections. In a series of experiments, high school students (15-16 years old) learned about the 8 steps involved in passing a government bill (Experiments 1 and 2) and the economic cycle (Experiments 3, 4 and 5). In the first four experiments students were randomly allocated into a group according to a 2 (written vs. spoken) x 2 (segmented vs. continuous text) design. In the fifth experiment multimedia effects were examined using a 2 (written vs. spoken) x 2 (diagrams vs. no diagrams) design. The information provided in Experiments 3, 4 and 5 was less sequential and needed to be integrated across several segments, in contrast to Experiments 1 and 2 where the information was more sequential requiring less integration. Results from Experiments 1, 2, 4 and 5 indicated a significant advantage of written text over spoken text. However, a segmentation effect was only found in Experiments 1 and 2 (government bill), but not Experiments 3 and 4 (economic cycle). Hence, the segmentation strategy was not effective when information needed to be integrated across segments. In addition, Experiment 5 found that presenting two mediums (diagram plus text) was better than one (text alone), and observed a reverse modality effect where spoken text plus diagram was inferior to identical text with the same diagram. In conclusion, the results suggest that providing lengthy spoken text without adequate support is a disadvantage due to its transitory nature, whereas written text is a more robust alternative than segmentation.