Abstract:

All students have their own preferred learning style which means that they use different techniques for studying and demand diverse modes of communication. E-ducation (abbr. for electronic transmitted education) entirely depends on the availability of diverse forms of media such as print, video, or podcasts (Keegan 1996). Blended learning combines e-ducation and face-to-face education (Young 2002; Graham 2005). This implies that students get ‘the best of both worlds’ (Graham 2005, p. 8) and an enlarged variety of media. These forms of media should be used continuously for practicing, because the learning curve suggests that with sufficient or substantial training all individuals can gain comparable performance levels (Ritter and Schooler 2002), or even turn average persons into experts (Staszewski 1988). Furthermore, literature indicates that multiple study sessions are more effective than practicing all tasks on a single occasion (Ebbinghaus 1985; Bahrick and Hall 2005).

Based on these streams of research it is hypothesized that different learning style patterns exist. Moreover, frequency of training and usage of hands-on-exercises [which have the same format than exam-questions] will have an effect on grades achieved.
To test these hypotheses several data sources are used: online-questionnaires, grading reports, and log files which record the actual usage-frequency of online-exercises as well as the success-rate while practicing (n=354).

The data is analyzed by applying the software package ‘Mixed Interaction Modeling’ (MIM) (Edwards 1987; Edwards 2000) for graphical modeling. This software is the only one that supports maximum likelihood estimations with categorical and continuous responses; rearrangements are done with multidimensional scaling (MDS).
The undirected model shows that students who attend the lecture also read the recomme
nded literature, and take notes. Thus, these students prefer face-to-face education. Further, it is shown that reading additional online-notes to the subject are associated with the ‘article of the week’ which are also published online as well as with the recommended book, and with note-taking. All media mentioned are preferred modes of visual learners, who learn most efficiently through reading and writing tasks. When attending lectures body language and facial expressions are important to these students (Peacock 2001). Literature suggests that auditory learners can best absorb information through lectures, discussions, and talking things through (Peacock 2001). This patter is empirically apparent in the present study due to the relationship between studying with colleagues and the usage of discussion boards. Online-exercises and online-sample-exams are also related, and represent the preferred learning style of kinesthetic learners who like hands-on-trainings. These learners also try to answer questions asked by peers.

Further insights are given by the directed graphical model. It indicates that the examination result depends on the amount of exercises executed, the percentage of these questions correctly solved while practicing, the usage of additional online-notes, and the frequency of studying. This is in accordance with Ebbinghaus (1985) and Bahrick (2005) who state that inter-session training is important. Indirect effects can be found for all other modes apart from studying in groups and whether students read ‘articles of the week’ or not.
These results allow for numerous implications.

Keywords:

Learning curve, learning style, learning media, graphical modeling.