Performances de mémorisation en fonction de la modalité d'encodage au cours des 24 heures chez des travailleurs postés

International audienceA serial probe recognition task was used to explore the effects of modality (visual vs. auditory) on memorization during the 24-hour day. The aim was to determine whether performance would vary with the level of vigilance, either in quantity (number of items remembered) or in quality (memorization strategies). In each of the six sessions run on different days at 4-hourly intervals (at 3, 7, 11, 15 19 and 23 hours), volunteer shift workers first completed Thayer’s (1989) questionnaire of subjective alertness and then performed the memory task. For each trial, six common nouns were presented one by one, followed by a probe that was or not issued from the list. The modality of word presentation varied randomly between encoding (list of words) and restitution (probe), allowing an investigation into intra and inter-modal word recognition to be carried out. Overall recognition rates did vary according to modality at encoding, but not at restitution. Auditory list presentation resulted in a significant superiority in the morning when vigilance was lowest, thereby favouring the hypothesis of quantitative variations in memory performance during the 24-hour day. Even though overall recognition rates did not vary across test periods for either modality, a recency effect occurred at all delays following auditory list presentation, and in the afternoon following the visual list presentation. Variations in the shape of serial position curves across the 24-hour day were interpreted as changes in the strategies spontaneously adopted by subjects whilst completing the memory task. The auditory superiority effect demonstrated in this field study is discussed in the light of the findings reported in laboratory studies, i.e. the chronological coding of auditory input, the persistence of the acoustic code and phonological recoding of visual input. Most importantly, the effect was robust and stable across the day, suggesting a number of potential applications, particularly during the night when shift workers’ alertness is at its lowest.Les effets de la modalité de présentation (auditive vs visuelle) sur la mémorisation sont explorés chez des travailleurs postés à 3, 7, 11, 15, 19 et 23 heures dans une tâche de reconnaissance sérielle. Une supériorité d’un encodage auditif des mots est mise en évidence tout au long du nycthémère et expliquée par les particularités fonctionnelles de ce canal perceptif et du traitement des informations auditives visuelles en mémoire à court terme. Lors d’une présentation visuelle des mêmes mots, l’allure des courbes sérielles de mémorisation fluctue parallèlement au niveau de vigilance des opérateurs. En l’absence de différence du nombre total de mots reconnus lorsqu’ils sont présentés auditivement, ces résultats plaident en faveur de variations circadiennes des stratégies de traitement que les opérateurs adopteraient spontanément dans ce type de tâche. L’absence de fluctuations circadiennes des performances de reconnaissance lors d’un encodage auditif des mots, suggère des applications potentielles de ce format de présentation pour certains types de messages et cela, en particulier la nuit, lorsque le niveau de vigilance diminue