Chapter 1: A Template for Describing Individual Differences in Longitudinal Data, With Application to the Connection Between Learning and Ability

In an influential series of experiments published a century ago, Ballard (1913) reported that under some conditions children could recall a passage of poetry better after a short period of time and without additional rehearsal than they did immediately after learning the piece. For example, the mean percent of lines of a poem correctly recalled by one group of Ballard’s 12-year-old children over 7 days after the initial learning was 111, 117, 113, 112, 111, 99, and 94, where the percentages were scaled to be 100 on the last day of practice (p. 6). The demonstration of improved memory without rehearsal or review runs counter to the almost universal finding that retention of recently learned material declines quickly over time. Ballard coined the term reminiscence to describe unrehearsed improvement in memory for text typically after a delay. McGeoch (1935) traced the phenomenon back as far as 1903 to Alfred Binet, among others. The experiments provoked wide interest. Researchers later showed that reminiscence effects can be obtained using nonsense syllables, three-letter digits, consonants, abstract words, standard prose, and pictures of objects or of people. It was shown to occur in younger and older learners, and in depressed and nondepressed samples, in those with both high and low IQ. Eysenck and Frith (1977) studied reminiscence as a correlate of personality characteristics.