Cultural differences in students’ learning habits

Susann Kowalski
FH Köln
susann.kowalski@fh-koeln.de

1. Introduction and concern

Evaluations of lectures show over and over again that one and the same lecture is valued very differently. These observations lead to the question, whether students from different countries learn differently. To prove this hypothesis we carried out a survey. In doing so we were interested in interaction forms, types of tasks, learning contents, group behavior etc. The survey also covered two questions concerning the learning styles aural, communicative, motor and visual. The purpose of this project is to provide a guide for teachers to make it easier for them to deal with the diversity of their students.

The available material allowed analyses of China, Germany, Mexico, Poland, Romania, Russia and Sweden. The results of a discriminant analysis show significant differences in the learning behavior of students from different countries. The analysis of the two questions concerning the learning styles show significant differences between students from several countries within each question, but the comparison of the results of these two questions show some contradictions.

After a short introduction into the structure of this survey and the used statistical methods, the results of a discriminant analysis and the results of the analyses of the two questions concerning the learning styles will be presented. An outlook on how the survey may be continued will complete this article.

2. Structure of the survey

The survey is in the first place based on data out of a questionnaire. The questionnaire contained 36 questions partly with sub questions, which were divided into four parts and asked for different aspects of learning behavior, requirements and of the personal environment:

Part A) – Self assessment of learning habits and motivation of learning

Part B) – Exposure to learning materials, results and types of tasks

Part C) – Need for communication

Part D) – Personal situation, environment, computer accessibility

The questionnaire was offered in German, English, Russian and Spanish. It was available via the internet from April 2005 till January 2006. After that it was also possible to fill in a paper version of the questionnaire and send it in. The basic population relating to the concrete analyses within this article consisted of all students in China, Germany, Mexico, Poland, Romania, Russia and Sweden, which were brought up and still live in the respective countries. Participation took place based on self-choice. All answers are based on a self-assessment of the participants. The results of the survey are descriptive.

3. Choice of statistical methods

Using a discriminant analysis (cf. Backhaus et al., 2003, chapter 3) it was determined, whether students from the specified countries differ regarding the prompted features.

To determine connections between the answers of the participants of different countries the Chi Square independence test was used. As the level of significance for the Chi Square independence test the value α=0,05 was chosen (cf. Schlittgen, 2003, chapter 19.2).

The comparison of proportion values of multiple-choice questions was carried out using the t-test (cf. Schlittgen, 2003, p. 337).

4. The sample

2888 students from 83 countries filled in the whole questionnaire. The majority of the participants came from China (66), Germany (986), Mexico (75), Poland (56), Romania (143), Russia (86) and Sweden (605). The amount of male and female participants from Germany and Mexico was relatively even. More women than men coming from the other countries took part in the survey.

5. Discriminant analysis

Do students from different countries learn differently? To answer this question a discriminant analysis was carried out.

37 variables were included into the discriminant analysis. All of them discriminated significantly.

Six discriminant functions have been constructed. The first and the second of them contribute a share of 78,6% (43,3% and 35,3%) in the explanation of the differences. Each of the other functions contributes a share less than 10%. All six discriminant functions separate the seven countries clearly. The level of significance of the quality criterion Wilk’s Lambda for the test of the function or the functions was p=0,000 in all cases. While the a priori probability of a correct classification is 14,3%, 67,9% of all cases may be correctly classified using the found discriminant functions. Chinese and Russian students are most frequently classified correctly with more than 70%.

Using the first and the second discriminant functions, which contribute the highest share in the explanation of the differences, the graphical representation shown in figure 1 was generated.

Figure 1: Differentiation of the groups based on the first and second discriminant functions

It is visible that the groups of participants from the several countries differ from each other concerning their learning behavior. German, Polish, Romanian, Russian and Swedish students differ clearly from each other. Chinese and Mexican students do not differ that clearly from each other and from the Polish, Romanian and Russian students, however clearly enough to distinguish them from the other students.

6. Learning types

Two multiple choice questions asked for learning types (aural, communicative, motor and visual). Participants should have marked the one or two most appropriate answers out of four possible answers. If they gave more than two or none answer at all, the answer of this question was declared invalid and not included into the analysis of this question. The answers are based on self-assessment of the participants and do not measure learning types.

Figure 2 shows the relative amount of answers (number of answers for this learning type divided by the total number of given answers for this question) per country for each learning type of question 1 and 2 respectively.

Figure 2: relative amount of answers to questions 1 and 2

Within each question clear differences are visible. Also the t-test to compare the proportion values shows significant results for several combinations of countries. But the comparison of the two questions exposes some contradiction. Thus, Mexican students indicated to learn aural more rarely in question 1 than in question 2. Russian students indicated to learn motor more often in question 1 than in question 2.

It is not possible to derive a clear statement concerning the learning types from these two questions. However, the results indicate that students from different countries differ in their self-assessment concerning their preferred styles of learning. A detailed survey, which on the one hand measures learning types in different countries and on the other hand compares these measures with the self-assessment of the students, could come up with interesting results.

7. Conclusions

As the discriminant analysis could show, Chinese, German, Mexican, Polish, Romanian, Russian and Swedish students differ within the prompted aspects concerning their learning behavior. The detailed analysis of the questions concerning the learning types lead to the presumption that students from different countries differ in their learning types. A special survey should find these differences.

Answering some basic questions this survey brought up new and continuative questions. On the one hand further statistical analyses will be conducted to determine, which aspects, except of the country, influence the learning behavior. Furthermore executing other statistical analyses should discover, whether there are single variables which could be combined to characteristic factors. Thus more substantiated statements can be given. Moreover additional countries should be included into the survey.

8. References

Backhaus, K., Erichson, B., Plinke, W., Weiber, R. (2003). Multivariate Analysemethoden. Heidelberg: Springer
Schlittgen, R. (2003). Einführung in die Statistik, 10. Auflage, Oldenbourg Verlag, München

9. Biography

Dr. Susann Kowalski has been professor of Business Administration at the department for Economics and Business Administration at Cologne University of Applied Sciences (CUAS) since July 1999. She specializes in Business Computer Sciences. Before joining CUAS, she worked as a research assistant at Dresden Technical University, being assigned to GMD (national research center for information technology) where she was involved in the development of an object-oriented operating system. This was also the subject of her PhD thesis. After that, she implemented a computerized call center at a large Re-insurance company where she was also responsible for development and implementation of a procedural model for client server software.

At CUAS her research interests are on e-learning, cultural influence on learning behavior and web applications.

International survey on cultural differences in student’s learning behavior needs your help. We are looking for universities from each country to improve our data base. What could you do: Let your students fill in the questionnaire and send them to us. We would need about 500 or 600 answers from a good mixture of women and men and of different subjects. What is your benefit: In turn you would get all data collected at your university and all data collected from German students in electronic form. Interested? Please contact us via susann.kowalski@fh-koeln.de