A Survey of the Preference for MBA Teaching Method

method than younger students. Rationale: Prior research indicates that young people usage of technology is greater than their old counterparts. 3. Data and MethodologyA questionnaire was developed for the research, which is reported in Appendix 1. It comprised 13 items and was grouped into two major areas: preferences for teaching method across taught subjects and demographic information of respondents. The items were derived from intense discussion of group members and previous studies, and their appropriateness was checked through a pilot study. The respondents were first required to indicate the extent of preference between lecturer-based and technology-based teaching method. A short but focused explanation of technology-based teaching method is presented in introducing paragraph of questionnaire objective, facilitating the respondents’ understanding. In preference extent selection, a five-point labelled Likert-type scale was used. Specifically, respondents were required to give a rating between 1 = entirely technology-based teaching without lecturer presence and 5 = entirely lecturer-based teaching without technology assistance. Literature suggests that Likert-type scale is effective in measuring consumer attitudes and is easy to construct and manage. The choice of five-point scale is the trade off between data reliability and the use of general linear model techniques. Our Likert scale is balanced to employ middle point as equal split of technology-based and lecturer-based teaching method. The courses investigated are seven compulsory subjects in ABC MBA degree study, including full time and part time modules. An implicit assumption is that all respondents have a good understanding of seven courses delivered in ABC University. The actual survey was carried out with current ABC MBA students who attended lectures during June 2005. A pre-test conducted with some non-MBA students showed that this questionnaire posed no conceptual problems associated with technology-based teaching method. A total number of 99 MBA students filled questionnaires while only 89 respondents returned a complete evaluation of preferences across seven subjects. Among them, 46 respondents completed all questionnaire items. Given the limited time available for the survey , the response rate of preferences for teaching method is acceptable. Table 1 Statistics of VariablesThe statistical descriptions of data are presented in Table 1. There are two variables derived from original data: category of first degree and overall preference. The first variable was constructed by categorising first degrees into technical or non-technical subject. The second variable was mean of preference scores of all courses. We excluded the cases of respondents failing to provide all preference information for seven courses. The missing data was defined as 9 or 99 when we entered data into SPSS database. Before examining the hypotheses, we perform the spearman method to test significance of correlations between students’ demographics and overall preference of teaching method. The use of Spearman test is appropriate for our sample data, which has unknown distribution attributes due to small sample size for some subcategories of data. As shown in Table 2, there is no significant correlation between demographic variables and overall preference despite programme variable with significance level 0.022. Table 2 Correlation AnalysisWe then explore the impacts of diversity of students and course subjects on preferences for teaching method by one-way ANOVA test. Because our sample data consists of more than two groups, one-way ANOVA is appropriate to test group differences. Our data are further examined with factor analysis to identify the presence of ‘factors’ in taught subjects. The factor analysis is helpful to reduce the dimension of courses, which may be the fundamental sources of preference diversity. This operation simplifies the description of the survey results and offers a better economic interpretation of teaching method preference in our sample data. 4. Results and DiscussionsResults of one-way ANOVA in terms of students’ demographics are reported in Table 3-A through Table 3-D. Corresponding hypotheses in section 2 are examined based on the statistic results. Table 3 ANOVA Analysis3-A Hypothesis 13-B Hypothesis 23-C Hypothesis 33-D Hypothesis 4H1: Full time students have higher preferences for lecturer-based teaching method than part time students. (See Table 3-A) There were a total of 89 valid responses across seven course subjects. The mean of references for teaching method in terms of seven taught subjects is computed for each response, representing the overall preference of teaching method for each respondent. We compare the difference between mean of overall preferences of full time students and that of part time students by using one-way ANOVA. Our results show that both full time and part time students prefer lecturer-based teaching method to technology-based teaching method. However, the preferences for lecturer-based teaching of full time students are higher than those of part time students. The mean difference is significant at 5% significance level. In other words, we cannot reject the hypothesis based on our data sample. A further exploration of mean difference of preference for various courses shows that full time students have significantly higher preferences for lecturer-based teaching method than part time students when accounting or quantitative methods is taught. The results are significant even at 1% significance level. By contrast, there were no significant preference differences between full time and part time students in terms of the remaining courses. Overall, the results confirm our hypothesis and shed light on that teaching method preference may differ across courses. H2: Overseas students have higher preferences for lecturer-based teaching method than UK-based students. (See Table 3-B) Findings from our study exhibit higher preferences for lecturer-based teaching method for either overseas students or UK-based students. The preference difference is not statistically significant at 30% significance level. Therefore, our data results reject the hypothesis. Running t test for each taught course gives insignificant preference differences between overseas and British students for different subjects except quantitative methods course. The exceptional case suggests there may exist asymmetry in quantitative backgrounds between overseas and domestic students. Overall, there was no difference of preferences for teaching method between overseas and home students in our sample. H3: Self-financed students have higher preferences for lecturer-based teaching method than sponsored students. (See Table 3-C) In total, 59 respondents provide information about their funding status and preferences for seven taught courses. Both self-financed and sponsored students were in favour of lecturer-based teaching method. The overall preferences for lecturer-based teaching of self-financed students were slightly higher than those of sponsored students. Yet the difference is insignificant even at 96% significance level. A further examination of preferences with respect to each taught subject provides insignificant results of differences. This suggests that funding status had little impact on students’ preferences for teaching method in this case. Hence the hypothesis is not supported by our sample data. H4: Preferences for lecturer-based teaching method of students with a non-technical background are higher than those of students with technical background. (See Table 3-D)We used first degree of respondents as proxy of students’ technical background. This measure has merits of measurement objectivity and collection convenience as well as disadvantage of inadequacy to capture technical background. For example, people may acquire rich technical experience in their jobs even their first degrees have nothing to do with technology. The first-degree data was coded into two categories: technical and non-technical background. All group members participated coding and final results reflected our consensus. Despite our agreement of categorisation the coding procedure may have our own perception bias. Therefore, our interpretation of results is viewed with extreme caution. There were totally 72 students stating first-degree background and teaching method preferences with all subjects. Initial analysis confirms that both two kinds of students in out sample favoured lecturer-based teaching method over its technology counterpart. The overall preference difference between two groups is insignificant as significance value is 0.269. Students with non-technical background exhibited higher preferences for lecturer-based teaching method than their technical counterparts and the difference is significant at even 1% (sig. = 0.008). Likewise, this significant result may reveal the unmatched quantitative foundation between sample students with technical background and those with non-technical background. Consequently, our sample data rejected the hypothesis but with cautionary note about our procedure of variable approximation. H5: Female students have higher preference for lecturer-based teaching method than their male students. In sum, 96 respondents offered information about gender status and preferences for teaching method in respect to seven courses. However, there were only 14 female students compared to 82 male students in our sample. The significant asymmetric distribution of data will deliver misleading results in further statistical analysis. Hence our data sample is not sufficient for examination of gender issue in preferences for teaching method. H6: Older students have higher preference for lecturer-based teaching method than younger students. The judgement of old generation and young generation is subjective and controversial. In our correlation analysis, non-parametric test gave insignificant result of correlation between age and overall preference. To put it differently, the age factor did not impact students’ preferences for teaching method in our sample. The one-way ANOVA analysis would deliver the same conclusion if age threshold were decided. Hence our data rejected the null hypothesis that age factor matters in preferences for teaching method. We further perform principal component analysis to extract factor in taught subjects. The factor analysis aims to examine the presence of factors in courses taught at ABC MBA module, which may justify respondents’ preferences for teaching method. Table 4 Factor AnalysisOur exploratory factor analysis used principle component method in extraction and varimax method in rotation. The results identify two significant factors existing in seven course subjects, as shown in Table 4. The conclusion is justified with the scree plot, which demonstrates that factor higher than 2 falls below eigenvalue of 1. Additionally, the first two factors explained 59.35% of total variance. With component plot in rotation space, we find that one group consists of marketing, organization behaviour, and presentation and communication skills while the rest courses go to the other group. Based on our discussion, we conclude that the first factor captures the nature of human-related courses while the second one reflects the nature of numeracy-related subjects. The reduction of seven variables (courses) to two factors obviously would simplify the interpretation of our results. Yet the reduction procedure needs further examination because the new variable created by summing individual variables may not be a reliable measure. Hence we run reliable scale test in SPSS to measure the internal consistency of scale. The reliability test gives alpha values of 0.6742 and 0.6720 against two factors, respectively. A rule of thumb for scale reliability requires alpha to be 0.70 or above. Since our results are approximately 0.70, we believe that the created factors fairly reliable in scale quality. The scale liability test results are reported in Table 5. Table 5 Scale Reliability of Factor Analysis5-A Factor 15-B Factor 2 With the created factors available, we further explore the relation between factors and preferences for teaching method. One-way ANOVA test was employed to examine the mean difference of preferences for teaching method with respect to students’ demographics. The results are reported in Table 6-A through Table 6-D. Findings from ANOVA tests reject our Hypotheses 2 and 3. Concerning Hypothesis 1, we find full time students had higher preferences for lecturer-based teaching method than their part time counterparts when taught courses belonged to numeracy-related subject. The result is significant at 5% level (sig. = 0.013). There was insignificant difference of preferences for teaching methods with regard to two groups of respondents when courses were human-related. As for Hypothesis 4, our data suggests that the case held when taught courses were of numeracy-related subject despite weak evidence (sig. = 0.064). The hypothesis was not supported by our data when courses were taught in the context of human-related subject. Table 6 Created Factors in ANOVA Analysis6-A Factors * Programme6-B Factors * Nationality6-C Factors * Funding Status 6-D Factors * Category of First DegreeDue to the size limit of our data, we did not examine hypotheses within the context of sub-segmentation, such as the intersection of full time students and respondents with technical background. However, our factor analyses shed light on the impact of underlying factors of course subjects on students’ teaching-method preference. Although the general conclusion is that ABC MBA students preferred lecturer-based teaching method to technology-based teaching method, the preference extent differed with the diversity of students and course factors. Our results suggest that students taking full time module or owning non-technical background demonstrated higher preferences for lecturer-based teaching method when taught subjects were numeracy-related. A more detailed and focused study about teaching-method preference within sub-segmentation would be helpful to identify the implication of technology in teaching method. 5. ConclusionsDespite our meaningful results of factor analysis for teaching method preference, our research project is subject to a number of shortcomings. This section gives some recommendations for future research based on insufficiencies of our study, including survey design, data processing, and data analysis. First, our survey design was not adequate to materialise our research intention. On the one hand, it did not fully measure respondents’ preferences for technology-based teaching method. For example, we did not differentiate between people’s preferences to technology-based teaching and the need to stay at home. Some potential influences on preferences include work commitment, family responsibility, and previous experience of course materials (e.g., quantitative methods for students with technical background). It would be helpful for identifying potential influences by asking respondents to state their rationales to prefer particular kind of teaching method or not. On the second hand, the variable approximation of technical background was questionable. The initial suggestion was to require respondents stating their technical experience. Due to obvious biases from self-evaluations, we adopted first degree to approximate technical background variable, which was an objective but inadequate measure. A possible solution is to use several variables including objective and subjective measures to describe respondents’ technical background. Consistent test results against these measures will confirm or reject the hypothesis. Second, our data processing needs more careful consideration of reliability. As less than half of sample respondents provided complete questionnaires, the procedure of missing data was very important. In our group presentation results, the missing data was taken as system missing. Overall preferences for teaching method were computed by total preference scores over the number of valid scores. For instance, if a respondent did not provide preference score for economics and gave five for the remaining courses, the overall preference of this respondent was five (5+5+5+5+5+5/6=5). However, such procedure was open to question, as we would bias the results by assuming the missing data was mean value of valid data. Our procedure was changed to drop missing data out of sample to keep results objective and reliable. The comparison of two procedures shows some differences of results. The first procedure did not lend support to all hypotheses. By contrast, the second procedure offered some evidence for Hypothesis 1 although other hypotheses were rejected. Our recommendation is to use a larger data sample and more sufficient collection time period to reduce missing data in order to improve analysis reliability. Third, our data analysis was generally sound despite some uncertainty of scale reliability. The sample data did not give sufficiently high scale reliability to create factors. The improvement of reliability could be achieved by employing large data sample and rich course index. Overall, we explore the teaching method preferences of ABC MBA students in this project. A number of hypotheses concerning students’ demographics and course subjects were raised and tested. In general, our sample students prefer lecturer-based teaching method to technology-based teaching method. Further factor analysis identifies two fundamental factors of course subjects that impact students’ preferences for teaching method. However, a number of shortcomings in our research project give cautionary notes about our conclusions from sample data. Future research should take more appropriate methods to tackle these problems and give more convincing results of students’ preferences for teaching method. Appendix 1 Survey Questionnaire