Baggett & Ehrenfeucht (1983)

Baggett, P. et Ehrenfeucht, A. (1983). Encoding and retaining information in the visuals and verbals of an educational movie. Educational Communication and Technology Journal, 31 (1), 23-32.

Le but de l'étude est de vérifier s'il y a compétition entre la communication visuelle et sonore lorsqu'on assiste à la projection d'un film.

Sujets. Quatre cent cinquante-neuf étudiants de la University of Colorado à Boulder et de la University of Denver participent à l'expérience. Les étudiants sont assignés aléatoirement à l'un des 17 traitements expérimentaux, les groupes variant entre 25 et 30 sujets.

Traitement. Le matériel expérimental est conçu à partir d'un film (11 minutes) intitulé Plant Traps: Insect Catchers of the Bog Jungle de Encyclopedia Britannica Films. Le matériel linguistique consiste en un texte d'une longueur de 1270 mots. Les 17 traitements expérimentaux découlent du combinatoire incomplet de trois dimensions (sur un total théorique de 24: 3 x 4 x 2), notamment:

Mesure. Le posttest est composé de 20 items de type vrai-faux, de 17 items à choix multiple et de 26 brèves questions (score maximum de 63 points).

Score (en %) au posttest

Test immédiat

Test différé

Aucune information

38,44

(6,38)

Transcription écrite seulement

56,69

(10,78)

45,14

(9,30)

Trame sonore seulement

53,46

(12,13)

42,19

(6,70)

Film muet seulement

47,89

(8,62)

45,09

(8,10)

Film muet après transcription écrite

66,25

(5,33)

57,33

(10,08)

Film muet après trame sonore

50,19

(12,02)

50,63

(10,52)

Film muet avant transcription écrite

66,73

(9,11)

57,30

(8,57)

Film muet avant trame sonore

60,93

(8,92)

52,77

(8,38)

Film en synchronie avec trame sonore

68,77

(6,25)

54,17

(7,92)

Les auteurs calculent une régression multiple pour identifier les facteurs qui peuvent prédire les scores au test. Une étude des résidus entre les scores estimés par cette régression et les scores obtenus réellement montre que les résultats des 17 combinaisons sont conformes à leur prédiction.

Discussion. À la lumière des résultats, les auteurs examinent plusieurs points.

" Questions 1a and 1b. There is no evidence for competition for resources between visuals and narration in the intact movie, or for an advantage in sequential presentations of doubling the study time. On the contrary, subjects could both encode and retain visual and narration information occurring simultaneously in the movie even better than they could such information occurring sequentially, even though the sequential information was studied twice as long. [...]

Question 2. In the sequential presentations, it does indeed matter whether the linguistic information is heard or read: Spoken narration and written text interact differently with visuals. In particular, written text can be studied before or after the visuals, and the effect is the same. [...]

Something very different happens for spoken narration at zero delay [= test immédiat]. When it is studied before the visuals, it is far inferior to when it is studied after the visuals. [...] This means that information in spoken linguistic material is encoded better when the visual material to which it is related is presented first, rather than second. When spoken linguistic material is presented before the visuals, the results are as poor as if the linguistic material were not presented at all. [...]

Question 3. A 1-week delay does indeed affect linguistic and visual material differently. [...] This study shows that humans are good at storing lots of verbal information [...] for a short time, but that less than half of it [...] lasts over a week. On the other hand, visual information, once encoded, is retained over a week. [...]

This study has shown that there is no competition for resources when related information is presented simultaneously in two media (visual and verbal / auditory). Therefore, synchronous visual and verbal / auditory input is an efficient way to present information. [...] in a show and tell presentation, one should not tell first and show second. To improve encoding and retention, one should either show and tell in synchrony, or show first and tell second " (p. 29-32).