Don't Go Too Easy On 'Em

The study I summarized here found that eliminating inference calls in a text significantly improved college students' free recall of the content.

But keep in mind that the free recall test in the experiment simply asked students to write down everything they could remember. Would we find improvements if students were asked to write a summary of the ideas in the text or to generalize from the information obtained? Is it possible that making a text less coherent might actually improve learning?

The study I shall introduce here and discuss in following posts (Kintsch, et. al, 1996) tries to get at this question. And even though its conclusions are as unsurprising as those in the Britton & Giilgoz study, it's worth a look. This is from the introduction to the study:

In order to optimize learning, should one make the comprehension process as easy as possible, or should one, as many educators insist, ensure that the learner participates actively and intentionally in the process of constructing the meaning of a text? The traditional view of learning and instructional design has long been governed by the notion that the learner's path should be made as trouble free as possible. Recent research on the acquisition of simple skills (e.g., Schmidt & Bjork, 1992), however, indicates that this approach does not always lead to optimal learning.

The Schmidt & Bjork article referenced above appears to agree with an earlier study by Gick and Holyoak (1987) that found that "some conditions that facilitate performance during training have negative effects on retention or learning-related tasks. Conversely, methods that slow the rate of learning during the training phase can have positive retention and transfer effects." Kintsch, et. al are interested in whether or not these effects can be found in cognitive processing.

Recent studies have demonstrated that revisions that increased the structural and explanatory coherence of texts resulted in substantial increases in recall among fifth-grade students (Beck, McKeown, Sinatra, & Loxterman, 1991) and college students (Beyer, 1987, 1990; Britton & Gulgoz, 1991). . . . Filling in this kind of information, which an author would normally expect readers to provide from their own background knowledge, proved to be very effective, particularly because the college students in their study lacked the domain knowledge to do so on their own . . . However, another effect that this text manipulation may have is to reduce the amount of active processing during reading. That is, easing the reader's burden of figuring out the meaning of the text could result in less effective learning.

Reference: Danielle McNamara, Eileen Kintsch, Nancy Butler Songer, Walter Kintsch (1996). Are Good Texts Always Better? Interactions of Text Coherence, Background Knowledge, and Levels of Understanding in Learning From Text Cognition and Instruction, 14 (1), 1-43 DOI: 10.1207/s1532690xci1401_1