“Language Testing” published a really wonderful article last week about a study of the impact that audio-visual modes of presentation have on rater scores of speaking samples. In this study, raters were given speaking samples with and without video. According to the article, “raters were significantly more likely to score comprehensibility higher when the presentation mode was audio-visual.” This was true in cases when the raters were familiar with the speaker’s accent and also when they were not.
There is, by the way, quite a bit of discussion in the article about accent familiarity, which is worth reading if you aren’t familiar with arguments about how that impacts raters.
The authors conclude that “the study provides evidence that Accent Familiarity and Presentation Mode interact to potentially impact Comprehensibility ratings. Presentation Mode, as a separate variable, independent of Familiarity, is a unique potential source of variation. The results indicate that the audio-visual Presentation Mode has a stronger effect than Accent Familiarity on Comprehensibility ratings. The results suggest that semi-direct tests with audio-only presentation and audio-visual presentation should be evaluated to ascertain how Accent Familiarity and Presentation Mode affect Comprehensibility ratings.”
You’ll need some kind of institutional access (or a full wallet) to read this one.