Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999

Is Computer-Based Learning Right for Everyone? Professor Jane H. Leuthold Department of Economics, University of Illinois at Urbana-Champaign [email protected] Abstract This study tests the hypothesis that a person's underlying learning style is a useful predictor of their attitude toward computer-based instruction and learning. Students in my undergraduate economics class participated in a learning style assessment based on the Gregorc Learning Style Delineator to determine their basic learning style: concrete or abstract, sequential or random. Students were also surveyed as to their attitudes toward the computer-based aspects of the class. Finally, correlation coefficients were computed to see whether or not certain learning styles were associated with positive attitudes toward computer instruction. According to the results, students with sequential learning styles use computerbased instructional techniques more frequently and prefer them to traditional instructional techniques when compared with students whose learning styles are random.

1. Introduction Computer-based instruction places particular demands on learners. Some learners flourish when a class goes online, find computer-enhanced learning fun, select classes that feature computer activities, and are comfortable in the world of mice and wizards. Other learners feel intimidated by computers, feel that computers reduce opportunities for face-to-face interaction, find computers boring, and generally prefer the traditional classroom format. Obviously, computer-based learning is not right for everyone, but is there any way we can predict which students will benefit from computer-assisted class formats and which will do better in traditional classrooms? Can studying learning styles within the context of the method of instruction assist us in improving instructional delivery? This study sought answers for these questions by correlating the learning styles of a class of freshman economics students at the University of Illinois at Urbana-Champaign with their preferences for computerbased instruction. Several earlier studies attempted to correlate learning styles and computer-related activities, however none to my knowledge has identified a strong relationship between learning styles and computer preferences. For example, Hart [6] found student attitudes to hypertext to be independent of learning style, Larsen [8] found that the effectiveness and acceptance of interactive video instruction are independent of students' learning style preferences, and Wesley et al. [11] found no relationship between locus of control (a cognitive style) and either learning or attitude of students in computer science classes. The results of the present study are based on a small sample and are therefore only

suggestive. However, it appears that for my freshmen economics students, learning style and computer preferences are indeed related.

2. Sample The class surveyed was a small (40 student) microeconomics principles. The class was taught in two sections, roughly 20 students each, and was part of the Discovery Program, a program designed to match freshmen with a professor in a small class setting. Students self-selected into this class, knowing from the class description that the class would involve computerbased elements of instruction. The class relied heavily on asynchronous learning techniques (on-line asynchronous chat sessions, weekly on-line quizzes, lecture slides and other course materials posted to the class website). Instead of a traditional term paper, students were asked to prepare a "webpaper" in HTML suitable for viewing on the net and containing links to other resources on the Internet. In other words, the class was much more computer oriented than the traditional introductory economics class. Students were invited, but not required, to participate in the learning style and computer preference assessments.

3. Assessing Learning Styles The Gregorc Learning Style Delineator was used to ascertain the learning style or styles of each student [4]. Gregorc's model is one of several models developed to improve understanding of the way we learn. Kolb [7] developed another well-known model of learning styles.

0-7695-0001-3/99 $10.00 (c) 1999 IEEE

1

Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999

Chart 1 shows the proportion of the class that scored in the dominant (27-40 points), intermediate (16-26 points), and low (10-15 points) ranges for each of the four learning styles. So, for example, CS was a dominant learning style for 53% of the students, an intermediate learning style for 47% of the students, and a low style for none of the students. CR was an intermediate learning style for the vast majority of the class (68%), while the AR style attracted the highest percentage of scores in the low range (16%). Although it is not shown in the chart, 95% of the students were dominant in one or more of the learning styles (only 5% were not dominant in any style) and over a third of the class was dominant in two learning styles. This is confirmed in other studies of learning style. According to Butler [1, p. 11], Gregorc's research showed

Chart 1. Learning Style Frequencies

80% 70% 60% 50% 40% 30% 20% 10% 0%

Dominant (27-40) Intermediate (1626) Low (10-15)

CS

AS

AR

CR

Learning Styles

Students were asked to plot their learning style scores on a graph. A representative learning style profile is shown in Chart 2. Each axis represents the person's score on one of the four learning-style characteristics. Connecting the scores gives a box that is distorted in the direction of the preferred learning styles. A person with scores as illustrated in Chart 2 favors abstract sequential and concrete sequential learning styles.

Chart 2. Example of a Learning Style Profile CS CONCRETE SEQUENTIAL

CR CONCRETE RANDOM

4. Learning Style Findings

that most people have strengths and preferences in one or two styles.

40 35 30 25 20 15

40 35 30 25 20 15

15 20 25 30 35 40

15 20 25 30 35 40

AS ABSTRACT SEQUENTIAL

The Gregorc model is a cognitive model designed to reveal two types of abilities, perception and ordering. Perceptual abilities, the means through which information is grasped, translate into two qualities: abstractness and concreteness. Ordering abilities are the ways the learner organizes information, either sequentially (linearly) or randomly (non-linearly). Gregorc couples these qualities to form four learning categories: concrete/sequential (CS), abstract/sequential (AS), abstract/random (AR), and concrete/random (CR). Although everyone has all four qualities, most people are predisposed toward one or two of them. The Gregorc learning style assessment, which was password protected and administered with Dr. Gregorc's permission, asked students to rank ten sets of words, four words in each set, according to which word provides the best and most powerful description of themselves. Based on their responses, indices were computed for each student indicating their scores for each of the four qualities, CS, AS, AR, and CR. The scores were normalized to sum to 100. Before the Gregorc Delineator was administered, students were told that everyone is different and there is no one best way to learn. They were also told that understanding their learning style can help to improve the way they learn and help them to interact with others who learn differently from themselves. Subjects were instructed to use themselves as a reference point and not to respond according to the way they think they should. If they were not familiar with the exact definition of one of the words, they were instructed to do their best, to work rapidly, and to react with their first impression. The survey was administered on-line. Students were given immediate feedback as to their scores in each of the four learning style categories: AR, AS, CR, and CS. Students were instructed to bring their scores to class where the scores would be explained.

AR ABSTRACT RANDOM

5. Learning Styles and Teaching Styles Certain types of classroom activities seem to work best with each particular learning style. According to Gregorc [5], students with a concrete-sequential dominant learningstyle tend to prefer programmed instruction, workbooks, lab manuals, field trips, and applications. Students with an abstract-sequential dominant learning-style tend to prefer

0-7695-0001-3/99 $10.00 (c) 1999 IEEE

2

Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999

lectures, books, syllabi, and guided individual study. Students with a concrete-random learning style prefer independent study, games, simulations, and problem solving, while students with an abstract-random learning style prefer television, movies, assignments with reflection time, and group discussions. Since computer-based instruction can be a complex combination of many of these activities, a focus of this study is to identify what type of learning style is most closely associated with a preference for the computer-based mode of instruction. To do this, a computer preference questionnaire was administered to the students.

6. Surveying Computer Preferences The computer-preference questionnaire was divided into three sections: frequency of use, satisfaction with computer learning, and interaction and motivation. The questions were patterned after those used in a classroom survey developed and administered at the University of Illinois at Urbana-Champagne by the Sloan Center for Asynchronous Learning Environments (SCALE). The questionnaire was delivered on-line with the results for each student computer-matched with the student's learning style scores. Results are divided into three categories: frequency of use, satisfaction with computer-based instruction, and interaction and motivation.

7. Frequency of Use Table 1 reports the mean, median and standard deviation of responses to the questions on frequency of computer use. Scores range from 0 to 5, where a score of 5 indicates daily use. Most students visited the class homepage at least once a week and participated in the weekly on-line quiz, which was required. Other homepage activities were optional but nevertheless attracted significant student usage. In order of greatest frequency of use, the nonrequired activities were exploring sites/links, reviewing lectures on-line, participating in the on-line chat, and consulting the class news.

Table 1. Frequency of Use How often this semester did you use the web to*: Mean Med. St. Dev. Visit the class homepage 3.28 4 1.23 Take on-line quizzes 3.00 3 0.91 Review the lectures on-line 2.33 2 1.03 Explore sites/links 3.00 3 1.50 Consult the class news 2.06 2 1.21 Participate in the on-line chat 2.28 2 1.13 *Scale: 0 = no opinion, 1 = not at all, 2 = a few times, 3 = Once a week, 4 = a few times a week, 5 = every day

8. Satisfaction Students' responses to the questions on satisfaction with computer learning were on average very positive (Table 2). Most students agreed or strongly agreed that

Table 2. Satisfaction with Computer Learning Please answer these questions exploring your general satisfaction with the use of computer learning in this class*: Mean Med St Dev 4 0.76 I consider the combination 4.1 of lecture sessions and computer sessions relevant to my success in this course. 4 0.81 In general, being able to 3.8 utilize the computer web site has actually made a difference in helping me understand the concepts, which were described in the lecture. The electronic on-line chat 2.8 3 1.22 increased my interest in the course material. 4 1.14 The on-line lecture slides 3.7 helped me better understand the course materials. 4 1.32 I prefer using the computer 3.9 lab sessions in combination with lectures instead of what is done in traditional class sessions. I would prefer to have more 3.4 4 1.54 courses that used the world wide web. When examining a new web 3.1 3 0.80 page, I read the page fully before exploring the links. Having the lecture slides 1.9 2 0.76 available on-line discouraged me from attending class. I found the web site easy to 3.9 4 1.37 navigate and use. I was generally able to find 3.8 4 1.35 a terminal and access the web site without problem. I would recommend this 4.2 4 0.86 course to other students. *Scale: 5 = strongly agree, 4 = agree, 3 = neutral, 2 = disagree, 1 = strongly disagree, 0 = no opinion computer-based instruction contributed to their learning and understanding of the course material; although they

0-7695-0001-3/99 $10.00 (c) 1999 IEEE

3

Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999

were slightly less enthusiastic about the on-line chat than about other on-line aspects of the course. Most students said they would prefer to have more courses that used the web, they liked the computer lab sessions, and they had little trouble navigating and accessing the website. Nonetheless, the high standard deviation on these items suggests there were a small number of dissatisfied students. One area where there seemed to be strong agreement was that having lecture slides available on-line did not discourage their class attendance.

coefficient, which is the ratio of the covariance to the product of the variances, ranges between -1 and +1 and is unaffected by the scaling of the variables [3, p.65]. Positive correlation coefficients indicate that those with higher scores with respect to a particular learning style also have stronger preferences for a particular computerbased learning technique; conversely for negative correlation coefficients.

9. Interaction and Motivation

Table 4, showing correlation coefficients between frequency of computer use and learning style scores, indicates that students with high abstract sequential (AS) scores most frequently used the computer-based features of the course and students with high concrete random (CR) scores least frequently used of these features. On average, the correlation between items indicating frequency of class computer usage and the strength of the student's abstract sequential score was .36, compared with an average correlation of -.43 for those with high concrete random scores.

Students agreed that web-based instruction increased interaction among students with each other and with the instructor in both quantity and quality, although more so with the instructor (Table 3). Most students agreed or strongly agreed that computer-based instruction increased their motivation to learn and most students felt that the course had increased their familiarity with computers, the Internet, and HTML programming. High standard deviations on the latter items may be due to the fact that many students in the class came from engineering backgrounds and were already familiar with computers and the Internet.

11. Learning Styles and Frequency of Use

Table 4. Correlations between Learning Styles and Frequency of Computer Use

Table 3. Interaction and Motivation Please answer these questions about how the use of the computer in this class affected*: Mean Median St Dev The amount of your interaction 4.2 4 0.62 with the instructor? The quality of your interaction 4.0 4 0.77 with the instructor? The amount of your interaction 3.3 3 0.83 with other students? The quality of your interaction 3.3 3 1.02 with other students? Your motivation to learn? 4.1 4 0.87 Your familiarity with 4.2 5 1.26 computers and the web? Knowledge of HTML 4.4 5 1.29 programming? *Scale: 5 = strongly agree, 4 = agree, 3 = neutral, 2 = disagree, 1 = strongly disagree, 0 = no opinion

10. Learning Styles and Preferences Nineteen students out of a total of forty completed both the learning style and computer-preference questionnaires. Correlation coefficients between the learning style scores and the computer-preference indices were computed to test the direction and strength of the relationship between computer preferences and learning styles. The correlation

CS

AS

AR

CR

CS+ CR+ AS AR

Visit the class 0.05 0.51 -0.10 -0.55 0.36 -0.36 homepage Take on-line 0.25 0.47 -0.27 -0.46 0.45 -0.45 quizzes Review the -0.02 0.38 -0.03 -0.42 0.25 -0.25 lectures on-line Explore 0.36 0.33 -0.22 -0.45 0.41 -0.41 sites/links Consult the class 0.26 0.26 -0.10 -0.43 0.31 -0.31 news Participate in the -0.05 0.20 0.06 -0.29 0.11 -0.11 on-line chat Average 0.14 0.36 -0.11 -0.43 0.31 -0.31 Note: CR = concrete random, AR = abstract random, AS = abstract sequential, CS = concrete sequential Correlating the sum of the sequential scores (CS + AS) and the random scores (CR + AR) with frequency of course computer use produced an interesting result. Students with high sequential scores (those who like to organize information linearly) have a positive correlation (.31) while students with high random scores have a negative correlation (-.31) with frequency of class computer use. The correlation coefficients are largest in absolute value for frequency of taking on-line quizzes and exploring sites/links.

0-7695-0001-3/99 $10.00 (c) 1999 IEEE

4

Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999

Note the correlation coefficients in the (CS + AS) column of Table 4 are equal but opposite in sign to the correlation coefficients in the (CR + AR) column. This arises because the sum of (CS + AS) and (CR + AR) learning style scores is one-hundred; when two variables sum to a constant, the sum of their correlation coefficients with another variable are equal and opposite in sign. The last column is included in the table (and in subsequent tables) simply to reinforce the finding that sequential and random learners differ in their preferences for computerbased learning.

12. Learning Styles and Preferences for Computer-Based Instruction Table 5 reports correlation coefficients between satisfaction with computer-based elements of instruction and learning styles. The table shows sequential learners prefer computer-based instruction while random learners tend to prefer traditional instructional techniques. The correlation coefficient between computer preferences and learning style score is .23 for sequential learners (those with high CS + AS scores) and -.23 for random learners (those with high CR + AR scores). It is interesting to note that random learners have more problems navigating and using the website and are more bothered by access problems than are the sequential learners.

Table 5. Correlations between Learning Styles and Satisfaction with Computer Learning CS 0.14 I consider the combination of lecture sessions and computer sessions relevant to my success in this course. In general, being 0.14 able to utilize the computer web site has actually made a difference in helping me understand the concepts, which were described in the lecture. The electronic on- 0.17 line chat increased my interest in the course material.

AS

AR

CR

CS+ CR+ AS AR 0.37 -0.28 -0.21 0.32 -0.32

The on-line lecture 0.22 0.25 -0.23 -0.21 0.28 -0.28 slides helped me better understand the course materials. I prefer using the 0.23 0.15 -0.27 -0.03 0.22 -0.22 computer lab sessions in combination with lectures instead of what is done in traditional class sessions. I would prefer to -0.08 0.32 -0.04 -0.25 0.17 -0.17 have more courses that used the world wide web. When examining a 0.26 0.12 -0.36 0.11 0.21 -0.21 new web page, I read the page fully before exploring the links. Having the lecture -0.08 0.45 0.01 -0.50 0.26 -0.26 slides available online discouraged me from attending class. 0.13 0.36 -0.17 -0.34 0.31 -0.31 I found the web site easy to navigate and use. -0.02 0.17 0.01 -0.20 0.10 -0.10 I was generally able to find a terminal and access the web site without problem. 0.07 0.13 -0.22 0.07 0.13 -0.13 I would recommend this course to other students. Average 0.11 0.26 -0.18 -0.18 0.23 -0.23 Note: CR = concrete random, AR = abstract random, AS = abstract sequential, CS = concrete sequential

0.31 -0.20 -0.24 0.28 -0.28 The question, "When examining a new web page, I read the page fully before exploring the links," was included to see if sequential learners with their preference for outlines would also approach web pages linearly. A correlation coefficient of .21 for sequential learners and -.21 for random learners suggest sequential learners do approach the web differently than do random learners. Is it possible that random learners may have a tendency to go exploring (and get lost?) before reaching the end of a web page? 0.23 -0.21 -0.16 0.24 -0.24

0-7695-0001-3/99 $10.00 (c) 1999 IEEE

5

Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999

13. Learning Styles and Interaction and Motivation Interaction and motivation to learn also showed higher correlations for sequential than for random learners (Table 6). On average, the correlation coefficient between items related to interaction and motivation with learning style was .31 for sequential learners and -.31 for random learners. Sequential learners' reported feeling strongly that interaction, particularly with the instructor, is improved with computer-based learning. It may be that sequential learners appreciated opportunities for on-line interaction because they allowed the students to have their questions answered in a more timely fashion than waiting for the instructor's office hours or class. Sequential learners also experienced increased motivation to learn and their familiarity and knowledge of computers and the Internet increased with computer-based instruction.

Table 6. Correlations between Learning Styles and Interaction and Motivation CS

AS

AR

CR

CS+ CR+ AS AR The amount of your 0.11 0.23 -0.16 -0.17 0.21 -0.21 interaction with other students? The quality of your 0.03 0.24 -0.17 -0.09 0.17 -0.17 interaction with other students? The amount of your 0.51 0.38 -0.27 -0.61 0.53 -0.53 interaction with the instructor? The quality of your 0.33 0.28 -0.03 -0.63 0.36 -0.36 interaction with the instructor? Your motivation to 0.33 0.44 -0.42 -0.28 0.47 -0.47 learn? Your familiarity 0.26 0.04 -0.18 -0.06 0.17 -0.17 with computers and the web? Knowledge of 0.22 0.22 -0.31 -0.05 0.26 -0.26 HTML programming? Average 0.26 0.26 -0.22 -0.27 0.31 -0.31 Note: CR = concrete random, AR = abstract random, AS = abstract sequential, CS = concrete sequential These results can only be taken as suggestive for a number of reasons. First, the sample size is very small; only nineteen of the forty students in the class participated in both surveys, despite a lot of nagging. Second, students self-selected to register for this class, which was advertised as a class that would include computer-based activities. This could have biased the results in favor of

computer-based instruction. Finally, the survey was conducted two years ago, when computer-based instruction was not widely used. For this reason, I resurveyed the students (now juniors and seniors) to see if their preferences for computer-based instruction had changed.

14. Results of the Follow-up Survey Fourteen students responded to the follow-up survey on computer learning preferences. Questionnaires were emailed to the now juniors and seniors. Results of the follow-up survey are shown in Table 7. Interestingly, they split evenly on their preferences for traditional vs. computer-based instructional formats. Roughly half experienced better learning, higher motivation, and harder work in the traditional class than in the computer-based class. While most would select a computer-based class if given the choice, a large proportion (79%) felt that class attendance suffers in computer-based classes. As in the earlier survey, most reported their feeling that studentinstructor interaction increases in a computer-based class, interaction among students suffers. One thing that had not changed for these students -- most viewed themselves as heavy computer users. Some preliminary correlations between learning styles and computer preferences for the follow-up students suggest that sequential learners still show a preference for computer-based instruction, although the correlation coefficients are smaller than for the original group. Unfortunately, sample size was so small that not a lot of confidence could be placed in this result.

15. Implications for Instruction and Future Research I set out to test the hypothesis that learning style is a useful predictor of preference for computer-based instruction. I found some evidence that sequential learners make more frequent use of computers, prefer and benefit from computer-based instruction, and experience improved interaction and motivation as a result of webbased instructional techniques than do random learners. Most importantly, the results of this study confirmed the wonderful diversity of approaches to learning that characterize our students. In an attempt to determine whether or not there is a "best" learning style, I ran a simple regression of course grade on learning style scores. I found that no learning style score has significant predictive power with respect to course grade. Since my sample was so small, I was not able to consider possible influences on learning styles such as race and gender. However, an interesting study of undergraduate business majors by Gentry and Helgesen [2] found a significant

0-7695-0001-3/99 $10.00 (c) 1999 IEEE

6

Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999

Table 7. Responses to the Follow-up Survey Questions I learn best in classes that follow a traditional (0) or computer-based (1) format. I am more motivated to learn in classes that follow a traditional (0) or computer-based (1) format. I work harder in classes that follow a traditional (0) or computer-based (1) format. My classroom attendance is better in classes that follow a traditional (0) or computer-based (1) format. The workload is usually heavier in classes that follow a traditional (0) or computer-based (1) format. I would generally select the section that was traditional (0) or computerbased (1). I feel that computer based instruction increases (1), decreases (0), or doesn't affect (.5) interaction among students in the class. I feel that computer based instruction increases (1), decreases (0), or doesn't affect (.5) interaction between students and the instructor. I would rank myself as a heavy computer user (2), moderate computer user (1), or infrequent computer user (0).

Mean Response 0.50

0.50

0.46

0.21

0.43

0.57

0.39

0.57

1.64

at particular aspects of computer-based instructional techniques need to be explored. With better information about which aspects of computer-based instruction appeal to which learners, we would be better able to enhance the appeal of our instructional approach for a wide range of learners. This will be best accomplished when academic practitioners and software developers join forces to work toward the development of effective computer-based learning technologies.

16. Summary and Conclusion This study asks the question: "Is computer-based learning right for everyone?" By correlating students' learning style scores and their preferences for computerbased instruction, we were able to identify learning technologies that appeal to particular learning types. Computer-based instruction, in its present form, is more preferred by students whose learning orientation is sequential. Random learners report less benefit from computer-based instruction, perhaps because our early attempts at computer-based instruction were simple extensions of old-style sequential learning technologies (posting our lecture outline on-line, for example). The next generation of computer-based instruction needs to broaden its appeal to reach all types of learners. But in our enthusiasm for computer-based instruction, we should be mindful of the very real benefit of the interactive atmosphere of the classroom that is hard or impossible to replicate on-line. REFERENCES

difference in learning styles among male and female students and among students of different races. There are two lessons to take from the present study. First, as instructors we need to expand our computer-based teaching technologies so that they are not merely an online extension of earlier sequential learning tools (lectures, outlines, tests/quizzes). There should be a place on our class webpages for games and simulations, group projects and on-line discussions, audio and video, problem-solving, and shorter lessons -- teaching techniques that appeal to random learners and stretch sequential learners to become better learners. For some suggestions for additions to your class websites, see Leuthold [9]. Second, we need to recognize that neither computerbased instruction nor traditional instruction is right or wrong for everyone; a combination of the two is required unless we want to leave half of our class behind. There is a place for both traditional classroom and computer-based instruction in our teaching and learning repertoire. Future research is needed to replicate this study using larger classes. Also, questions directed more specifically

[1] K.A. Butler, It's All in Your Mind: A Student's Guide to Learning Style, The Learner's Dimension, Columbia, CT.,1988. [2] J.A. Gentry and M.G. Helgesen, "Using Learning Style Information to Improve the Core Financial Management Course," Faculty Working Paper, University of Illinois at Urbana-Champaign, Spring, 1998. [3] W.H. Greene, Econometric Analysis, 2nd Edition, Macmillan Publishing Company, New York, 1993. [4] A.F. Gregorc, An Adult's Guide to Style. Maynard, MS: Gabrial Systems, Inc., 1982. [5] A.F. Gregorc, "Individual Differences: Teaching for Active Learning," Keynote Address, University of Illinois at UrbanaChampaign Faculty Retreat on College Teaching, June 19, 1996. [6] G. Hart, "Learning Styles and Hpertext: Exploring User Attitudes," paper available in pdf format at: http://ascilite95.unimelb.edu.au/SMTU/ASCILITE95/papers/hart .pdf, 1995.

0-7695-0001-3/99 $10.00 (c) 1999 IEEE

7

Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999 Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999

[7] D.A. Kolb, Experiential Learning: Experiences as a Source of Learning and Development. Englewood Cliffs, NJ: PrenticeHall, 1984. [8] R.E. Larsen, "Relationship of Learning Style to the Effectiveness and Acceptance of Interactive Video Instruction," Journal of Computer-Based Instruction, 19 (1), 1992, pp. 17-21.

[10] A. Miller, "Personality Types, Learning Styles and Educational Goals," Educational Psychology, 11(3-4), 1991, pp. 217-238. [11] B.E. Wesley, G.H. Krockover, and C.R. Hicks, "Locus of Control and the Acquisition of Computer Literacy," Journal of Computer-Based Instruction, 12 (1), 1985, pp. 12-16.

[9] J.H. Leuthold, "Building a Homepage for your Economics Class," Journal of Economic Education, 29 (3), 1998, pp. 247261.

0-7695-0001-3/99 $10.00 (c) 1999 IEEE

8