The relationship between learning style and learning environment Diana H J M Dolmans & Ineke H A P Wolfhagen Does problem-based learning (PBL) have an effect on learning style and student performance? This question is addressed in a study 1 published in this issue, which compares a traditional psychiatry attachment and a prob- lem-based attachment. Both attachments consisted of 12 days of formal teaching, comprising lectures about illnesses in the tra- ditional attachment and discussion around patient problems in the PBL attachment. The authors assumed that the PBL students (n ¼ 191) would use less surface learning, more deep and strategic learning and achieve better student performance scores than the stu- dents on the traditional attachment (n ¼ 188). The Study Process Questionnaire was used to measure students’ learning styles at the beginning and end of both attach- ments. Student performance was measured at the end of each attachment with 125 multiple choice questions (MCQs) and the presentation of the case of a patient seen during the attachment and answers given to questions on that case. The results demonstrated that the PBL attachment resulted in better student performance scores on both the MCQs and the case presentation. Contrary to expecta- tions, no differences in learning styles were found between the tra- ditional and PBL attachments and neither were any differences in learning style found between the beginning and end of both attach- ments. 1 The strength of this study relies on its authors’ investigation into the relationships between different vari- ables to examine causal links The strength of this study relies on the fact that its authors have tried to investigate what might cause PBL students to perform better. The theory is that a PBL attachment will result in less surface learning, more deep learning and more strategic learning, resulting in better student performance scores. A multiple regression analysis was conducted to examine causal links between type of attachment or treatment, student learning style and student performance. Future research should focus on the relationships between different vari- ables that might influence student learning in a PBL attachment instead of comparing a PBL attachment with a traditional one The weakness of this study lies in its use of a curriculum-level interven- tion design to investigate the rela- tionship between learning style and student performance. Curriculum- level intervention studies have some weaknesses that limit inter- pretation of the findings. 2,3 This is also true for the study reported in this issue. First of all, neither stu- dents nor teachers in the study were blind to the intervention and therefore the better student per- formance score on the PBL attach- ment is probably not due to the nature of the attachment alone. Secondly, one might question whe- ther the intervention was a real PBL intervention or rather a hybrid one. When facilitators in the PBL attachment gave a summary of Ôtake homeÕ messages after discussing patient problems, there seemed to be a high level of faculty direction. In other words, if both attachments were characterised by a high level of faculty direction, it is not so strange that no differences in learning styles were found. In con- clusion, it is difficult to interpret the findings of this curriculum-level intervention study. Therefore, we would recommend that in future the researchers focus their investi- gation on the relationships between different variables that might influence student learning in a PBL attachment instead of com- paring a PBL attachment with a traditional one. Making decisions about student performance based on the presenta- tion of one patient case is unreliable Another weakness of this study concerns the fact that one of the assessment instruments used was commentaries Correspondence: Diana H J M Dolmans PhD, University of Maastricht, Department of Educational Development and Research, PO Box 616, 6200 MD Maastricht, The Netherlands. Fax: 00 31 43 388 4041; E-mail: D.Dolmans@educ.unimaas.nl. doi: 10.1111/j.1365-2929.2004.01941.x Ó Blackwell Publishing Ltd MEDICAL EDUCATION 2004; 38: 800–804 800