Pictorial encoding and testing impact recognition memory

The cognitive theory of multimedia learning (CTML) and the picture superiority theory (PST) were tested in multiple experiments using a recognition memory paradigm. A three-phase paradigm was used in which participants first studied information via pictures plus narration, pictures alone, or narration alone. During this study phase, participants saw a PowerPoint presentation containing 48 photos, 48 narrated statements, or both concerning a man walking through a house and interacting with various household objects. Second, participants received conflicting information via pictures plus narration, pictures alone, or narration alone. Finally, participants took a written or pictorial recognition test. According to PST, studying pictures primes the learner to use two representational systems (verbal and non-verbal/imagery) which results in more retrieval routes to memory representations compared to studying using only a verbal representation system (Paivio, 1971). The PST predicted that studying pictures should result in higher percent correct scores on recognition memory tests than studying words, and that there should be no differences in the treatment effect for studying pictures plus narration versus pictures alone. In contrast, the CTML predicted that the treatment effect should be greater for studying pictures plus narration versus narration alone compared to pictures alone versus narration alone. Integrating the verbal information (narration) and the non verbal/imagery information (pictures) should promote generative learning and deep processing (Mayer, 2001). Results showed the treatment effect of studying pictures and narration versus narration alone (d=1.00) was greater than the treatment effect of studying pictures versus narration alone ( d=-.34) on a written test. This effect was greater on a pictorial test (P+N v. N: d=2.13, P v. N: d=1.33). Additionally, studying pictures plus narration led to higher memory test scores than studying pictures alone on both written (d=1.00) and pictorial tests (d=.40). These results support the cognitive theory of multimedia learning and do not support the picture superiority theory. Consistent with the CTML, studying pictures plus narration results in dual code processing using both verbal and non-verbal (imagery) codes, and integrating this information into one representation creates a robust memory representation that enables successful retrieval on verbal (written) tests and non-verbal/imagery based (pictorial) tests.