Learning styles, sociodemographics and level of participation in a discussion forum

Dr Zlatko J Kovacic, School of Information and Social Sciences, The Open Polytechnic of New Zealand, [HREF1] , Private Bag 31914, Lower Hutt, New Zealand. Zlatko.Kovacic@openpolytechnic.ac.nz

Abstract

This paper reports initial research results on the relationship between student participation in an online discussion forum, their learning style profiles and various sociodemographics in a distance education computing course with Internet-based student support. The learning styles of students in a computer concepts class were evaluated and classified according to the Felder-Soloman Learning Style Index. The level of participation was measured using two indicators: the number of postings and lurking level. Using ANOVA and regression analysis we have identified statistically significant differences in the level of participation and academic performances between learner types, i.e. groups of students with different sociodemographic characteristics and learning preferences. We have shown that age and ethnicity are the most dominant factors with significant impact on the number of postings, lurking level and the course marks. Other factors such as gender, occupation and learning styles also played an important role in explaining the variation in the level of participation.

Introduction

This paper presents the results of an exploratory analysis of the relationship between level of participation in an online discussion forum, sociodemographics and learning styles of the first year students on a computer concepts course at The Open Polytechnic of New Zealand in the last two semesters in 2003. Computer mediated communication tools and online discussion forums in particular create an ideal learning environment to support collaborative learning model, which is based on social learning theory. One of the prominent protagonists of this school of thought is Russian psychologist Vygotsky (1978). According to him knowledge is constructed through social interaction and collaboration. When the knowledge is constructed it is absorbed individually. In other words knowledge does not exist before social interaction has taken place. Ho (2002 [HREF4]) suggested checking whether the students place similar value on collaboration and interaction in the forums as pedagogical theory suggests. A questionnaire routinely delivered to students at the end of the computer concepts course shows that our students regarded interaction in discussion forums as the third preferred way of communication on the course (the first two were weekly bulk emails and direct contact with the lecturers). In other words, interaction and collaboration in the forum was not the most preferred way of communication for these computing concepts students.

There are many factors/issues which influence effective participation in an online discussion forum in addition to sociodemographic characteristics and learning styles. Some of them, such as cooperation amongst students and between the educator and students and whether the forum is structured or not are discussed in an excellent overview of research articles on students' participation in online discussions written by Ho (2002 [HREF4]). A detailed discussion of all of these factors is beyond the scope of the present paper, which will concentrate only on identification of the relationship between the level of participation in discussion forums, sociodemographic characteristics, learning style profile and academic performance. More specifically the data gathered during two semesters in 2003 was used to address the following three questions:

• do the sociodemographic characteristics of the student affect their participation in an online discussion forum and academic performance on the course?
• is someone with a particular learning style more likely to participate in the forum than any other?
• is someone with a particular learning style more likely to succeed on the course than any other?

There have been numerous studies on the relationship between preferred learning styles, sociodemographic characteristics and academic performance. Some studies have focussed on specific characteristics or area, such as gender and learning styles (Blum, 1999; McLean & Morrison, 2000; Shaw & Marlow, 1999), ethnicity and learning styles (Auyeung & Sands, 1996; Jaju, Kwak & Zinkham, 2002), academic performance and learning styles in both IT and non-IT subject areas and in distance and contact courses (Aragon, Johnson & Shaik, 2002; Fowler, Allen, Armarego & Mackenzie, 2000 [HREF3]; McKenzie, Rose & Head, 1999; Neuhauser, 2002; Papp, 2001; van Zwanenberg & Wilkinson, 1999; Zywno & Waalen, 2002 [HREF8]). Based on these articles evidence of relationship between learning styles, sociodemographic characteristics and academic performance remains inconclusive. For example for first year programming courses, Thomas, Ratcliffe, Woodbury & Jarman (2002) suggest that there is a relationship between student learning style and academic performance, whilst Byrne & Lyons (2001) suggest that no such relationship exists.

Finally, only few articles discussed participation in an online discussion forum in relation to sociodemographic characteristics and academic performance (McLean & Morrison, 2000 [HREF6]; Rowe & Vitartas, 2003 [HREF7]). In both articles a quantitative approach was adopted, the same approach we have used in this paper. However, we have added learning style profiles to the factors that may have impact on the level of participation in an online discussion forum. It is worth to note that McLean & Morrison (2000 [HREF6]) have based their quantitative study on only 30 learners which makes their conclusion limited to the sample they have used. Beside usage statistics for analysis of participation in online discussion we have considered an alternative approach - content analysis (used by McKenzie & Murphy, 2000 [HREF5]). However, we found this analysis to be too time-consuming in a class with 180+ students and over 1000 messages posted in a single semester. In other words, we tend to agree with Ho (2002 [HREF4]) when she said that the content analysis "as performed by teaching staff to be an unfeasible assessment procedure for larger class numbers (e.g. one hundred or more) because of the increased workload it represents."

Initially a brief overview of the learning environment is provided. Then in the second section the Kolb and Felder-Silverman learning styles models and the learning styles instrument are described. In the third section data and methodology are discussed and our findings along with critical comments are presented. Finally we make recommendations for further research.

Learning environment

This paper presents the results of research with the first year students on a Computer Concepts course at The Open Polytechnic of New Zealand in 2003. These students have no previous experience with discussion forums. The course emphasises the use of computers and information systems in a business scenario. The course has a large practical component of 60% covering applications, the Internet and the operating system. The course uses discussion forums and bulk email as the main means of learning support in addition to telephone and mail. Support is provided both by the lecturer and students via forum based peer support. Participation in the forums is an integral part of the assessment. At least 3 messages were expected to be posted to the forum, as it was explained in the "Data and results" section below. No other rewards were offered to students for participation. However, as one of the referees suggested, perceived rewards may also be contribution factor to higher level of participation. With a semester length of 17 weeks the course attracts upwards of 180 students per semester.

For this course we have set up two types of forums: an announcement forum for one-way communication from lecturer to student and a discussion forum as a two-way communication amongst students and between lecturers and students. In the discussion forum at the beginning of each semester we created a set of topics/folders which matches the learning modules in the course. There is also an area for informal social interaction amongst students. An initial message was posted to the each topic which explains what the purpose of the topic is and how to participate in the discussion.

Learning style models and instrument

The most dominant learning cycles models used to identify learners' preferences are Kolb's experiential learning cycles model (Kolb, 1984) and the Felder-Silverman model (1988). The Kolb model uses the learner's experience as a starting point in the learning process. Initially they are experimenting with the topic accumulating enough concrete experience to be able to reflect in the second stage on the observation gathered in the concrete experimenting stage. As a result of reflective activities learners derive abstract concepts and make generalisations in the third, abstract conceptualisation stage. Finally, new concepts are subject to testing to see if they provide a solid explanation in new situations. In other words, learners begin a new learning cycle, gathering new evidence and concrete experience. Though learners are moving through each stage they tend to use a specific learning mode. Therefore we can describe them as a learner with a preference for a particular mode.

The Felder-Silverman learning style model was developed as a four dimensional model, with the following dimensions: Perception, Input, Processing and Understanding. Both the Kolb and Felder-Silverman models belong to the information processing category of learning style models and are based on the same educational philosophy of John Dewey, which emphasises the nature of experience as of fundamental importance in education. There is a close relationship between the Felder-Silverman model and the Kolb learning style model. The Processing dimension, with active and reflective poles in the Felder-Silverman model matches the same Processing dimension in Kolb's model. Also, the Perception dimension in the Felder-Silverman model with sensing and intuitive poles matches the same Perception dimension in Kolb's model.

The Felder-Solomon Index of Learning Styles (2003 [HREF2]) questionnaire is used to determine learning styles in Felder-Silverman model. We have chosen this particular learning style instrument for the following reasons:

1. It covers all four learning styles dimensions and matches Felder-Silverman theoretical model.
2. The instrument has been widely tested and used successfully in helping to guide the design, development and use of effective learning environments.
3. This instrument is simple to use and the results obtained from this study are easy to interpret and can be applied easily.
4. This instrument has good validation results, which makes this instrument reliable in detecting preferred learning styles among students.

The Felder-Solomon Index of Learning Styles is constructed as a bi-polar instrument across four dimensions: Processing (with poles: active/reflective), Perception (sensing/intuitive), Input (visual/verbal) and Understanding (sequential/global). The dichotomous learning style dimensions of this model are continuous not discrete categories. This means that the learner's preference on a given scale does not necessarily belong to one of the poles. It may be strong, moderate, or almost non-existent.

According to Felder & Silverman (1988) active learners are described as those who learn by actively trying things and collaborating with others. Reflective learners' preferences are for thinking rather than trying things and they prefer working alone. Sensing learners prefer learning facts and procedures while intuitive learners are innovative and oriented more toward concepts, theories and meanings. Visual learners prefer visual representations such as pictures, diagrams and charts while verbal learners prefer written or spoken explanations. Sequential learners are linear, orderly and learn in small incremental steps while global learners are holistic thinkers who learn in large leaps.

The instrument consists of 11 questions for measuring each of the four dimensions, and thus a total of 44 questions. Each question along a dimension is designed to determine if a respondent tends to belong to one category or another on that dimension. It does so by asking the respondent to choose only one of two options where each option represents each category. Since there are 11 questions for each dimension, a respondent is always classifiable along each dimension. The range of data for each dimension is from 0 to 11. Obviously there are 16 possible combinations, or types of learner in this model.

Data and results

Data for this paper were gathered during the last two semesters in 2003. The ILS was self-administered by students who replied to the online questionnaire at the North Carolina State University web site, host of the ILS. Of 245 students in two semesters 156 of them (64% response rate) completed the online questionnaire and sent in their learning style profiles. Sociodemographic data came from the enrolment form - the demographic data is a requirement from the Ministry of Education who fund The Open Polytechnic. The data from the enrolment form was in some cases, such as ethnicity, education level, and occupation, further grouped to keep the number in each category relatively even and to ensure adequate numbers in each sample to make the statistics more reliable. For example, the occupation category was split to identify those who were studying whilst working (labelled in this paper as wage and salary workers) and those doing the courses while unemployed (labelled as the others).

Academic performance will be measured in terms of the final grade achieved rather than perceived learning as the author prefer to measure learning outcomes rather than perceived learning. Perceived learning is likely to be subjective and does not allow a comparison of one student with another since self perception will differ from one individual to another and may be more likely to indicate overall confidence levels than learning outcomes.

Finally, a complete discussion forum log provided relevant data/information about the level of participation in the forum. To measure a level of participation we have used two indicators: the number of messages posted by each student and the lurking level. Usually lurking is defined as no posting (Nonnecke & Preece, 2000). However, we have defined lurking as a minimal number of postings, which was set to 3. The rationale for this decision was as follows. In the first course activity we ask students to post an introductory message to the forum introducing themselves to the class. Also, as an integral part of the in-course assessment they have to post two messages to the forum. Therefore we consider 3 postings as involuntary participation in the forum and those students posting 3 or less messages to the forum are therefore described as lurkers.

The following tables (Tables 1 to 7) present the students' participation in the discussion forums and academic performance for each sociodemographic characteristic and learning styles. Column labelled "Number of Postings" contains the average number of messages posted by the learners which belong to a particular class. "Lurking level" is simply the proportion of all students in any classification that belong to a particular class and have posted less than 4 messages. Finally, the column labelled "Course Mark" contains the average overall course mark (composed of the in-course assessment and the final exam marks) scored by the learners that belong to a particular class.

These tables also contain the value of the F-ratio and P-value from a one-way ANOVA. The F-ratio is used to test the hypothesis of no differences in discussion forums participation among students with different sociodemographic characteristics. The P-value indicates the likelihood of obtaining a difference between mean values as large as that observed if it occurred simply from randomness in the data. A low P-value implies that we would probably not observe such a large difference from purely random data and the difference must be the result of some systematic effect. By convention, we usually label any difference with a P-value of 0.05 or less as meaningful, that is, statistically significant.

Before presenting results for each sociodemographic characteristics we have calculated Pearson correlation coefficients between level of participation and the final course mark. The correlation between the number of postings and the course mark is 0.28 and between lurking level and the course mark is -0.34 (both coefficients are significant at less than 1% level). This result gives an additional motive for students to contribute to online discussion forum (the more messages they post the higher marks they can expect).

Gender

In our previous paper (Kovacic, Green & Eves, 2004) we found that the female and male students on the Computer Concepts course were learning differently. Statistically significant differences were found in the way genders were adopting and inputting information (Understanding and Inputting dimensions in the Felder-Soloman model). In this study we have also detected difference in level of participation between genders (results are presented in Table 1). Female students tend to be more active in the forum posting on average 66% more messages than their male counterparts. Also, the lurking level is significantly lower among female students.

Table 1: Gender and level of participation

Although there are statistically significant differences in the learning styles results and level of participation in the forum for female and male students their course performances are not significantly different. These results partly match the results of Rowe & Vitartas (2003 [HREF7]). They found that there were no differences between genders with comparisons between the overall contributions to the forum and the final examination mark in an Accounting Theory course. Similarly, McLean & Morrison (2000 [HREF6]) found that gender had no impact on participation.

Age

We tested the hypothesis that there are no differences in the level of participation among students from different age groups. If this hypothesis were true then we would expect that the average number of postings and lurking level for each age group would be equal. F-ratios and corresponding P-values in Table 2 would suggest rejection of the above hypothesis. Also, learners which belong to different age groups scored significantly different course marks.

Table 2: Age and level of participation

We also tested the hypothesis that the age of students has no impact on the level of participation and course performance. If this hypothesis is true then we would expect that the correlations between age and level of participation and age and course performance are equal to zero. The Pearson correlation coefficients between age and number of postings and age and course marks are 0.3 and 0.24 respectively (P-values are less than 1%), which means that the older students are more active in the forum and achieve better marks. Again, this result contradicts the result of McLean & Morrison (2000 [HREF6]) that age has no impact on participation.

Ethnicity

We found that NZ European learners are the most active participants in the discussion forum (measured by the average number of postings and lurking level) when compared with other two ethnic groups. From Table 3 we can see that they posted almost twice as many messages as Asian students. One reason for this might be the fact that the Asian learners are extremely visual (Kovacic, Green & Eves, 2004). Since discussion forums are written discourse they might give advantage to verbal learners. Also, for most Asian students English is their second language which could be a barrier for them to participate in discussion at the same level as other students. The last reason could be also used to explain the lowest course mark achieved by Asian students. While there are no significant differences between Asian and students from other ethnic backgrounds when compared with the in-course assessment, on the final exam their academic performance is relatively poorer.

Table 3: Ethnicity and level of participation

Education level

Students who have completed some tertiary study before are slightly more active in the forum and scored higher marks. However, these differences in level of participation and academic performance between those without tertiary education and those with some tertiary education are not statistically significant (P-values are above 5% in Table 4). This result contradicts McLean & Morrison (2000 [HREF6]) who claimed that the "learners with university degrees sent nearly three times the number of messages as did learners without degrees".

Table 4: Education and level of participation

Occupation

Results are presented in Table 5. Very significant differences were found in the lurking level. Wage and salary workers have lower lurking level than unemployed people. They also achieve higher course mark (marginally significant at 7% level). However, there are no significant differences in the average number of postings between these two groups of students.

Table 5: Occupation and level of participation

Learning styles

Active, sensor, verbal and sequential learners are posting more on average than their opposites (reflective, intuitor, visual and global). Active, sensor, verbal and sequential learners also scored a better course mark than their opposites (Table 6). However, none of these differences between their marks are statistically significant. In other words, learning styles can’t be used for predicting academic success on this course. Significant differences in the level of participation were identified in the Perception dimension only. For the number of postings this difference was significant at 8% level and for lurking level highly significant at 1% level. At the 10% level of significance differences in the number of postings were detected between active and reflective learners and differences in the lurking level between visual and verbal learners.

Table 6: Learning styles and level of participation

There is a positive and significant relationship between the active and sequential scores and the number of messages student posted to the forum measured by Pearson correlation coefficients (0.28 and 0.15 respectively, significant at 1% and 7% respectively). This result suggests that an active and sequential learner tends to post more messages than polar opposite (reflective and global learner). Also, there is a negative, statistically significant correlation between the Perception dimension (sensing scores) and the lurking level (-0.19 which is significant at 2% level).

In the next stage of analysis we take a more holistic approach to learners. As we know, with 4 dimensions and two poles on each dimension in the Felder-Silverman model we have 16 distinctive types of learners. However in this course 83% of the class belonged to one of only six types of learners listed in Table 7. Our hypothesis that these six types of learners participate equally in the forum was accepted for the both indicators: number of postings and lurking level. However, the hypothesis that they perform equally well was rejected for the overall course mark at the 1% level.

Table 7: Dominant types of learners and level of participation

The learners from the sixth group (REF-SEN-VRB-SEQ) make only 6% of the class, but they have the highest average number of postings, the second lowest lurking level and the highest course performance, scoring the highest course mark among all other learner’s types. This would suggest that the Computer Concepts course is designed and taught in a way that caters for this particular type of learner or that this type of learner is generally more perceptive or a better learner. On the other side, REF-SEN-VIS-GLO type of learner (14% of the class) has the poorest course performance in terms of having the lowest average course mark, the highest lurking level and the second lowest average number of postings among these six learner’s types. Comparing these two opposite types of learners we can say that the way they are inputting information (visually or verbally) is the most significant factor which influences their level of participation and their academic performance. To avoid any confusion when comparing the results in Table 6 (visual/verbal dimension) with what we just said, it is worthwhile to emphasis that the calculation in Table 6 was done separately for each learning style dimension. In Table 7 we are analysing each learner’s type described with 4 learning style dimensions simultaneously.

Distance education courses in general, by the very nature of the learning mode may give advantage to verbal learners. The discussion forums used for peer-to-peer and tutors support are dominantly written discourse, if we ignore the fact that we can insert emoticons and attached graphic/multimedia file to the message. With the opportunity for including multimedia components in the course material brought by the Internet technologies, an effort should be made to adjust the learning environment in a way that suits the majority of the learners (visual learners). Consequently the participation and course performance would be expected to improve as a result of these changes.

Regression analysis

Finally a regression analysis was utilised to assess the relationship between the level of participation measured by number of postings and lurking level and a set of sociodemographics and learning styles. Based on the results from the previous steps we have specified two regression models, one for each indicator of the level of participation in the discussion forum. Two regression models were estimated and the results are presented in Table 8.

Table 8: Regression models for level of participation

Based on the results from the first regression model we say that only 16.3% of variation in the number of postings was explained by the following variables: age, gender, ethnicity and whether the student is active learner or not. Among these variables it seems that age is the most dominant factor. Negative coefficient for ethnicity simply reflects the way we have coded three ethnic groups. Asian students are posting fewer messages in the forum than Maori & Pacific Islanders and their NZ European counterparts. Age is also the most dominant factor in the regression model for the lurking level. Together with ethnicity it explained 14% of variation in the lurking level. These two percentages (16.3% and 14%) simply mean that using this regression models to predict number of postings and lurking level results in prediction that are 16.3% and 14% more accurate than using the average number of postings (6.44) and average lurking level (0.46) for predicative purposes. The last two average figures were taken from Table 1 (row labelled Total).

Concluding remarks

This article explores the impact of sociodemographic variables and learning styles on the level of participation and academic performance in an online discussion forum. From five sociodemographic variables, three were significantly related to the number of postings, the lurking level and the course mark. These three variables are gender, age and ethnicity. Previous education level was not related to the level of participation or academic performance, while occupation has impact on the lurking level only. We have identified that the particular learner type (reflective, sensor, verbal and sequential) was the most active participant in the forum scoring the highest course mark.

These findings have practical implications for the instructional designers, student support group and lecturers. Those learners whose level of participation is below the class average should be subject of a special attention to encourage them to actively participate in the forum. Active participation of each student could bring benefit to the whole class and would probably help them to have better academic performance.

This study raises a large number of questions that require further investigation into the learning styles, level of participation and sociodemographic characteristics. Does the size of the class/forum have impact on the level of participation/lurking? How different learner’s types react to the alternative organisational form of discussion forum? What impact would a different forum moderation style have on the level of participation? These are some of the questions we are planning to address in our future research.

Acknowledgements

I would like to thank John Green, Senior Lecturer in the School of Information and Social Sciences at the Open Polytechnic of New Zealand who has made valuable contributions to the development of this paper. I also take this opportunity to thank two anonymous Conference reviewers for their time and valuable work which helped me to improve the final version of this paper. However, the author should be held responsible for any remaining errors.

References

Aragon, S. R., Johnson, S. D., & Shaik, N. (2002). The influence of learning style preferences on student success in online versus face-to-face environments. The American Journal of Distance Education, 16(4), 227-244.

Auyeung, P., & Sands, J. (1996). A cross cultural study of the learning style of accounting students. Accounting and Finance, 36, 261-274.

Blum, K. D. (1999). Gender differences in asynchronous learning in higher education: Learning styles, participation barriers and communication patterns. Journal of Asynchronous Learning Networks, 3(1).

Byrne, P., & Lyons, G. (2001). The effect of student attributes on success in programming. Proceedings of the 6th Annual Conference on Innovation and Technology in Computer Science Education, Canterbury, United Kingdom, 49-52.

Felder, R. M., & Silverman, L.K. (1988). Learning and teaching styles in engineering education. Engineering Education, 78(7), 674-681.

Felder, R.M., & Soloman, B. A. (2003). Index of learning styles questionnaire. Available online [HREF2].

Fowler, L., Allen, M., Armarego, J., & Mackenzie, J. (2000, February). Learning styles and CASE tools in software engineering. Proceedings of the 9th Annual Teaching Learning Forum, Perth, Australia. Available online [HREF3].

Ho, S. (2002). Evaluating students' participation in on-line discussions. Proceedings of AUSWEB02 The Eighth Australian World Wide Web Conference, Twin Waters Resort, Sunshine Coast Queensland 6 - 10 July 2002. Available online [HREF4].

Jaju, A., Kwak, H., & Zinkham, G. M. (2002). Learning styles of undergraduate business students: A cross-cultural comparison between the US, India and Korea. Marketing Education Review 12(3), 49-60.

Kolb, D. (1984). Experiential learning: Experience as the source of learning and development. Prentice-Hall: Englewood Cliffs, NJ.

Kovacic, Z., Green, J., & Eves, C. (2004). Learning styles of Computer Concepts students in a distance tertiary institution. Manuscript submitted for conference.

McKenzie, W., & Murphy, D. (2000). "I hope this goes somewhere": Evaluation of an online discussion group. Australian Journal of Educational Technology, 16(3), 239-257. Available online [HREF5].

McKenzie, J., Rose, S., & Head, C. (1999). A study of the relationship between learning styles profiles and performance on an on-line team-based collaborative learning project', in : J. Hill, S. Armstrong, M. Graff, S. Rayner & E. Sadler-Smith (Eds.), Proceedings of the 4th Annual Conference of the European Learning Styles Information Network, 313-332. Preston: University of Central Lancashire.

McLean, S., & Morrison, D. (2000). Sociodemographic characteristics of learners and participation in computer conferencing. Journal of Distance Education 15(2). Available online [HREF6].

Neuhauser, C. (2002). Learning style and effectiveness of online and face-to-face instruction. The American Journal of Distance Education 16(2), 99-113.

Nonnecke, B., & Preece, J. (2000). Lurker demographics: Counting the silent. Proceedings of the ACM CHI 2000 - Conference on Human Factors in Computing Systems, The Hague.

Papp, R. (2001, December). Student learning styles and distance learning. Proceedings of the 16th Annual Conference of the International Academy for Information Management - International Conference on Informatics Education & Research, New Orleans, Louisiana, USA, 14-20.

Rowe, S., & Vitartas, P. (2003). An assessment of contribution to an on-line discussion forum in Accounting Theory, Proceedings of AUSWEB03 The Ninth Australian World Wide Web Conference, Hyatt Sanctuary Cove, Gold Coast Queensland 5 - 9 July 2003. Available online [HREF7].

Shaw, G., & Marlow, N. (1999). The role of student learning styles, gender, attitudes and perceptions on information and communication technology assisted learning. Computers & Education, 33, 223-234.

Thomas, L., Ratcliffe, M., Woodbury, J., & Jarman, E. (2002). Learning styles and performance in the introductory programming sequence. SIGCSE Bulletin, 34(1), 33-37.

van Zwanenberg, N., & Wilkinson, L. J. (1999). Felder and Silverman's Index of Learning Styles and Honey and Mumford's Learning Styles Questionnaire: How do they compare and do they predict academic performance? The European Learning Styles Information Network Conference, Lancashire, UK.

Vygotsky, L.S. (1978). Mind in society: The Development of Higher Psychological Processes. Cambridge, MA: Harvard University.

Zywno, M. S., & Waalen, J.K. (2002). The effect of individual learning styles on student outcomes in technology-enabled education. Global Journal of Engineering Education, 6(1), 35-44. Available online [HREF8].

Hypertext References

HREF1

http://www.openpolytechnic.ac.nz

HREF2

http://www.engr.ncsu.edu/learningstyles/ilsweb.html

HREF3

http://lsn.curtin.edu.au/tlf/tlf2000/fowler.html

HREF4

http://ausweb.scu.edu.au/aw02/papers/refereed/ho/paper.html

HREF5

http://www.ascilite.org.au/ajet/ajet16/mckenzie.html

HREF6

http://cade.athabascau.ca/vol15.2/mclean.html

HREF7

http://ausweb.scu.edu.au/aw03/papers/vitartas/paper.html

HREF8

http://www.eng.monash.edu.au/uicee/gjee/vol6no1/Zywno.pdf

Copyright