Indiana University Northwest
There has been much literature on analogies in concept development (Brown & Clement, 1989; Grandgenett & Thompson, 1991; Gentner, 1983; Gick & Holyoak, 1980; Sternberg, 1977). Most all writing seems to focus on discovering or developing the best possible analogy for a given comparison. A good analogy seems to accurately or completely represent critical relationships and concepts while less important attributes are considered irrelevant. Most such literature typically falls short of providing practical information for the average classroom teacher on how to actually use analogies as an instructional tool.
It makes sense that students could benefit from complete and accurate
mental models. Some analogies have been reported as highly representative
of relationships found in their target domain (Curtis, 1988; Halpern, 1987).
While there are a few computing analogies commonly accepted, others might
be considered extreme, dissimilar or even useless (Figure 1.). It
seems appropriate to prefer a perfect analogy, a mental model which perfectly
represents the target concept. It has been assumed that a really
good analogy would better serve conceptualization.
|Figure 1. Some analogies rated as poor or low quality
A group of 85 educational computing analogies (Galloway, 1992, b.) were rated by three separate experts: a. one who is employed as a microcomputer operator and had completed the education computer literacy course, b. a secondary science teacher who uses microcomputers as a personal/professional management tool and to regularly augment instructional experiences, and c. a secondary computer science teacher who also teaches adult education courses for the state's Department of Education.
There were analogies on everything from fundamental concepts (command, data, file, program, etc.) to software and other elements of computer use (word processor, spreadsheet, input/output, variable, etc.). Analogies were rated on a 5 to 1 scale based on whether analogies where considered highly representative of the target domain or were considered a very poor representation.
The 16 "best" analogies and 16 "worst" analogies were selected on the basis of that rating and were then used separately in the instruction of two groups of beginning computing students (11 subjects per group). While the two groups were available because of their enrollment in the computer literacy course, this treatment was randomly assigned to the two groups. Group A received exposure to the best analogies and group B to the worst or poorest of the analogies.
Both groups were taught by the same instructor, with the same content sequence and the same performance expectations. The analogies, dichotomized between the two groups, were spread out across the course to coincide with instructional units and related activities within the course (word processing, BASIC programming, graphics, database, spreadsheet, etc.). Analogies were well developed to include a discussion of both similarities and differences between the comparison and target domains. Students participated directly in the discussions as the analogical comparisons were explored and analyzed.
Data was collected through a pre/post multiple choice test on computing concepts (45 questions). Also, a personal journal was used for students to turn in short essay responses to five specific questions throughout the course that called for students to fully explain in detail. In each case, students were told that their task was to help the reader reach a complete and accurate understanding. Essay responses were blind scored with group assignments hidden from the reviewer.
Finally, the final exam used in the course was included. It served
as a measure of overall computer literacy including both knowledge of skills
and procedures and conceptual understanding.
While pretest data confirmed an extremely high degree of beginning similarity
between the two groups, posttest data show that the two groups remained
nearly identical at the end of the course. Table 1 shows that the
scores after the differing treatments with analogies were practically identical.
The mean scores for the course's final exam were in fact identical for
the two groups.
|Table 1. Post treatment scores and differences|
|Written test: Percent||
|Written test: Mean||
|Written test: SD||
|Course Final Exam||
How important is the quality of analogies in the classroom? Ritualistic or procedural learning is broadened or enhanced by the development of a more complete conceptual base. However, higher-order thinking and problem solving skills require a level of understanding beyond simple ritual knowledge (Vockell & van Deusen, 1989). Classroom teachers need effective strategies for helping students achieve better understandings. Kintsch and Greeno (1985) support the relationship between understanding and problem solving ability with arithmetic word problems and the need for an internalized "conceptual representation upon which problem-solving processes can operate" (p. 110). The importance of sound conceptual models in problem-solving and higher-order thinking is further supported by Mayer (1989).
While it may be true that "Computing teachers need models for what analogies to use as well as for how and when to use them," (Bright, 1989, p. 5), maybe the most important focus should be on how analogies can be used most effectively rather than on building a repertoire of good analogies.
All data collected in this study, the written test scores, the course final exam and of course the journal essays, show that the two groups remained similar in spite of the dichotomized treatment of analogies in instruction. The students' achievement in the course, their conceptual development and their ability to use analogies (without being prompting to do so) and explain computing were all apparently affected equally by both good analogies as well as poor. These results assume of course that analogies are properly developed.
In trying to understand new ideas, students can benefit from an analysis of a concepts attribute's, necessary and coincidental, and in seeing how that concept relates to other similar ideas or situations. This examination of relationships, making comparisons and searching through information, is described as "the core of analytical reasoning," (Whimbey & Lochhead, 1984, p. 19). This approach is critical in the proper use of analogies.
There is evidence that analogies can be effective learning tools (Galloway, 1992, a.). However, as demonstrated in this study, the primary issue, contrary to the concerns of many educators, may not be whether the analogy is good or bad. This author has experienced many cases where educators expressed an immediate negative reaction to some analogies, but usually with little thought given to the details of the comparison. Such labeling may be inappropriate as analogies are selected for use in the classroom. The learning process seems to lie in analyzing relationships and attributes between the target and comparison domains. Analogies may be considered useful so long as they facilitate that process and it seems that so?called good and bad analogies can both serve that process equally ? if used properly in the classroom.
Future research needs to focus on classroom instructional implementation
of analogies. More work needs to be done to help teachers learn how
to use analogies in instruction with students in the classroom rather than
on discovering the best or on how to construct the best possible analogy.
Bright, G. W. (1989). Analogies in computer literacy and computer science textbooks. Paper presented at Southwest Educational Research Association Annual Meeting, Houston, TX.
Brown, D. E., & Clement, J. (1989, March). Overcoming misconceptions via analogical reasoning: Factors influencing understanding in a teaching experiment. Paper presented at the Annual Meeting of the American Educational Research Association, San Francisco, CA.
Curtis, R. V. (1988). When is a science analogy like a social studies analogy? A comparison of text analogies across two disciplines. Instructional Science, 17, 169-177.
Galloway, J. P. (1992, a.). Teaching educational computing with analogies: A strategy to enhance concept development. Journal of Research on Computers in Education, 24 (4), 499-512.
Galloway, J. P. (1992, b.). Analogies in educational computing. East Rockaway, NY: Cummings and Hathaway, Publishers.
Gentner, D. (1983). Structure?mapping: A theoretical framework for analogy. Cognitive Science, 7, 155-170.
Gick, M. L., & Holyoak, K. J. (1980). Analogical problem solving. Cognitive Psychology, 12 (3), 306-355.
Grandgenett, N., & Thompson, A. (1991). Effects of guided programming instruction on the transfer of analogical reasoning. Journal of Educational Computing Research, 7 (3), 293-308.
Halpern, D. F. (1987). Analogies as a critical thinking skill. In D. E. Berger, K. Pezdek, & W. P. Banks, (Eds.), Applications of cognitive psychology: Problem solving, education and computers. Hillsdale, NJ: Erlbaum.
Kintsch, W., & Greeno, J. G. (1985). Understanding and solving word arithmetic problems. Psychological Review, 92, 109-129.
Mayer, R. E. (1989). Models for understanding. Review of Educational Research, 59 (1), 43-64.
Sternberg, R. J. (1977). Component processes in analogical reasoning. Psychological Review, 84 (4), 353-378.
Vockell, E., & van Deusen, R. M. (1989). The computer and higher-order thinking skills. Watsonville, CA: Mitchell Publishing, Inc.
Whimbey, A., & Lochhead, J. (1984). Beyond problem
solving and comprehension. Hillsdale, NJ: Erlbaum.