Files in this item
|(no description provided)|
|Title:||Teaching Summarization Skills: A Comparison of Training Methods|
|Author(s):||Day, Jeanne D.|
|Department / Program:||Psychology|
|Degree Granting Institution:||University of Illinois at Urbana-Champaign|
|Abstract:||The ability to summarize is an important skill that is dependent on correctly identifying and concisely relating main ideas. Despite the fact that locating main ideas is a basic comprehension process, students, particularly poorer ones, experience major difficulties producing summaries. For example, in recent research, Brown and Day developed five rules that could be used to condense text material: (1) delete redundancies, (2) delete trivia, (3) substitute a superordinate term for a list of exemplars, (4) locate any topic sentences and (5) invent topic sentences for paragraphs that lack them. The last two of these rules involve the identification of main ideas at the paragraph level. When junior college students were asked to summarize two simple expository texts, they failed to use these rules well. In fact, they performed more like 7th graders than like university students. We, therefore, undertook to train a group of junior college students how to summarize.
In the first experiment, junior college students of two ability levels, average and poor writers, were given one of four types of training that varied in explicitness: (1) self-management--students were given general encouragement to write a good summary, to capture main ideas, etc., but they were not told rules for achieving this end; (2) rules alone--students were taught to employ the five rules of summarization; (3) rules plus--students were given the self-management training of group one and the rules training of group two but were left to integrate the two sets of information for themselves; and (4) rules integrated--students received instruction in the rules integrated with explicit training in the monitoring of these rules, e.g., students were shown how to check that each paragraph had a topic sentence either underlined or written in. In the second experiment, a third level of students, poor readers, were given the most explicit training and their improvement was compared to that of average and poor writers who had received the same instruction. The hypotheses were that even with training time held constant, more explicit instruction would result in greater improvement and that better students would improve more and would require less explicit instruction to do so.
The explicitness of the training did influence how much students improved. Students given only self-management instructions did not write significantly better summaries. But those subjects given even minimal training in the rules improved the quality of their summaries and they maintained that improvement over time. Within the three rule training groups, the effect of explicitness was also evident. On the two main idea identification rules, the conditions were ordered rules integrated > rules plus > rules alone. The chance probability of finding this ordering only once is .16. It was found repeatedly. Ability level did play an important role by affecting the amount of improvement students made and the degree of instruction necessary for them to show any gains, but it did not alter the ordering of the conditions. Only on easy rules did the ordering of the training conditions vary. Most students used the deletion rules efficiently on the pretest and, therefore, needed no instruction on them, while on the mechanical superordination rule, a brief explanation of how it worked produced immediate, efficient use.
Thus, ability level and the difficulty of the material being taught interact to determine how explicit training must be. On difficult concepts and with slower students, explicit training in strategies for accomplishing a task coupled with routines to insure the successful application of those strategies is clearly necessary.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1980.
|Date Available in IDEALS:||2014-12-13|