BackHome
Refresh: OffView Personal Schedule
Paper Summary
Share...

Direct link:

Converging Behavioral and Psychophysiological Measures: Evaluating the Effectiveness of Multimedia Learning Conditions With Dyslexic Learners

Sat, April 29, 2:45 to 4:15pm, Henry B. Gonzalez Convention Center, Floor: Meeting Room Level, Room 208

Abstract

The purpose of this study was to understand how dyslexic students learn in multimedia environments using behavioral and psychophysiological measures.

Research Questions
1. Which of four multimedia conditions results in the best learning performance?
2. What insights can EEG measures contribute regarding the process of learning?

This study integrates two widely accepted theories of learning, the Orton-Gillingham multisensory approach to teaching dyslexic students (Gillingham & Stillman, 1997) and the cognitive multimedia learning theory approach for designing instruction (Mayer, 2007). Both theories posit that students learn more effectively when information is presented through dual-channels of the working memory, the visual/auditory channel and the visual/pictorial channel. The study integrates these dual-coding tenets to explore how comprehension can be optimized for dyslexic learners.

The study employed four multimedia conditions: Text with Images Present (TP), Text with Images Absent (TA), Narration with Images Present (NP) and Narration with Images Absent (NA). Using an a priori power analysis, 18 participants were assigned to each condition (n = 72). Participants were screened using methods consistent with high-quality studies of dyslexic learners (Wiseheart et al., 2009). A previously validated multimedia narrative titled Discovering Australia was used (Ritzhaupt et al., 2011).

Data sources included:
- Learning performance measures including a multiple choice test (recognition) and a fill-in-the-blank assessment (cued recall).
- NIH Toolbox Cognitive Battery (to be reported elsewhere).
- Alpha event-related desynchronization percentage (ERD%) collected with a dry sensor wireless EEG headset.

Learning performance results revealed a significant difference on recall performance among the four conditions, F(3, 51) = 4.471, p < .05, eta = .208. Post hoc analyses indicated a difference between TP (M = 8.23) and NA (M = 4.42), p < 0.05. Significant difference was also found on recognition performance among the four conditions, F(3, 51) = 14.895, p < .05, eta = .467 with significant differences between TP (M = 6.13) and TA (M = 13.86), p < 0.05; TP (M = 6.13) and NP (M = 12.31), p < 0.05; TP (M = 6.13) and NA (M = 9.77), p < 0.05; TA ( = 13.86) and NA (= 9.77), p < 0.05.

Mean alpha ERD% among four groups was significantly different, F(3, 51) = 5.681, p < .05. Specifically, mean alpha ERD% in NP was higher than in any of the other three conditions, p < 0.05, indicating increased working memory load (Klimesch et al., 2005). Additionally, a positive relationship was found between mean alpha ERD% across the conditions and recognition performance (r = 0.31, p < 0.05).

While text with no images resulted in the best recognition performance, text with images produced best recall scores. Representational imagery appears to be useful for improving recall, which is cognitively more demanding than recognition. The condition one would expect to be most aligned with optimal information processing for dyslexic individuals – narration with images – produced the highest levels of working memory load and led to high recognition performance but low recall performance. These findings may indicate that our participants, dyslexic students at research universities, may have developed strategies for learning with traditional text.

Authors