In other words, what's important is the time of day you hear/see/read something, not when you try and remember it.
- information learned in the morning shows better immediate retention, but worse long-term retention
- short-term memory appears to improve as arousal levels fall
Three experiments investigated whether the time of day had an effect on short-term or long-term memory.
In the first experiment, the material used was a factual article from a New Scientist magazine. Short-term memory (as measured by 10 multi-choice questions) was best if the article had been read at 8am, and lowest if it had been read at 8pm. Surprisingly, there was a slight, short-lived improvement after lunch (during the post-lunch dip in arousal level), and another one after 8pm (at a similar dip in arousal). Long-term memory (as measured by performance on a category instance task) was apparently not affected by time of day. Nor was reading speed.
In the second experiment, the subjects listened to a story, at either 9am or 3pm. Their recall was tested immediately and again a week later. It was found that short-term recall was better if the story was heard at 9am but long-term recall was better if it had been heard at 3pm. It didn't seem to matter whether testing occurred at 9am or 3pm, nor did it matter whether the test occurred at the same time of day as the story was heard.
In the third experiment, the subjects were shift workers. The subjects, who were nurses, were shown a ten minute film on the use of radium therapy. The times used were more extreme this time: 8.30pm and 4am. Long-term recall was tested at four weeks. Long-term recall was consistently better if the film had been seen at 8.30pm than if it had been seen at 4am, but there was no effect on immediate recall. However, in the group least adjusted to shift work (part-time nurses and those on their first night shift), short-term recall was better if the film had been seen at 4am, while among the most adjusted group the reverse was true (short-term recall was best if the film had been seen at 8.30pm). Again, the time of testing made no difference.
Overall then, the experiments found that the time at which the information was presented consistently influenced immediate and delayed retention in opposite directions. It is not clear why there should be a differential effect. There was no evidence that retrieval efficiency was affected by the time of day.
An interesting implication of this work is that the recommendation from studies early this century, that academic work is better taught in the morning and physical subjects in the afternoon (based on findings that immediate memory was better in mid-morning, and perceptual-motor activity in the afternoon), may have been ill-founded.