• Tiada Hasil Ditemukan

CHAPTER 1: INTRODUCTION

2.3 Effec tiveness and outcomes of PBL

The research trend to compare PBL with conventional, lecture-based curricula has led to some interesting debate in literature. In their review article, Norman and Schmidt (1992) concluded that there were “small or negative differences between the overall knowledge or competence of students trained by PBL and by conventional curricular”. However, there was initial evidence pointing to better retention of knowledge and learning skills for PBL students. Review by Albanese & Mitchell (1993) showed that PBL students performed as well and sometimes better on clinical examinations and faculty evaluations compared to traditional medical students. However, their review also concluded there were potentially important gaps in PBL graduates’ knowledge base and expert reasoning skills. Vernon and Blake (1993), in their review of evaluative research published from 1970 through 1992, drew the following conclusions: (1) There was no significant difference between PBL and traditional approach on tests of factual and clinical knowledge, (2) Students from the traditional approach performed significantly better than their PBL counterparts on the National Board of Medical Examinations Part I Examination (NBMEI), (3) On outcomes that were less frequently researched (e.g. faculty attitudes, student mood, class attendance,

21

academic process variables), results generally supported the superiority of PBL approach over more traditional methods. Berkson (1993) in his review of literature published before 1992, had examined the effectiveness of PBL curricular over traditional methods in areas such as students’ motivation, development of problem solving skills and acquisition of content knowledge, drew the conclusion that ‘graduates of PBL are not distinguishable from their traditional counterparts’. Additionally, they pointed out that the practice of PBL can be stressful for students and faculty and that the cost of implementing PBL was unrealistically high. The review by Smits et al. (2002a) indicated that there was no consistent evidence to show that PBL is more superior to other educational strategies in improving doctors’

knowledge and performance.

More recent reviews on the effectiveness of PBL also yielded mixed or inconclusive results.

Colliver (2000), in his review of eight studies aimed at examining the effectiveness of PBL over traditional curricula, concluded that the literature either did not provide convincing evidence that PBL improved the content knowledge or clinical skills of students, or did not show significant improvement to justify the considerable resources needed to run a PBL course. Newman (2003), working with a review group in a pilot systematic review and meta-analysis using strict inclusion criteria, found that the results were mixed. In assessing the accumulation of knowledge, it was reported that out of the 39 effects studied, 16 favors the PBL intervention and 23 the control group. In measuring the improvement in practice (e.g.

attitude toward practice), study by Moore at al. (1994) showed that the result favors PBL group. Of the seven effects reported by Lewis and Tamblyn (1987), two favor the PBL group.

Of the nine effects reported by Grol et al. (1989) only one favors the PBL intervention group.

Mamede et al. (2006) drew a similar conclusion in their studies to compare the effectiveness

22

of PBL as an educational intervention with conventional curricula in that the studies have yielded inconclusive results.

However, there are other studies which support the positive effects of PBL. For example, research by Distlehorst & Robbs (1998) showed that PBL students performed better in clinical assessment compared to students who were from traditional curriculum. Study by Birgegard & Lindquist (1998) demonstrated that PBL fostered critical thinking and students showed improved in attitudes. An interesting meta-analysis was conducted by Walker &

Leary (2009) in which 47 PBL comparative outcomes outside the fields of medical education and allied health were used. These disciplines included teacher education, social science, business, science and engineering. This represents the first attempt to synthesize the results across various disciplines other than those in medical and related fields. The analysis showed that across almost all of these disciplines, PBL students did either as well or better than their lecture-based counterparts. PBL students from social science and teacher education in particular performed significantly better than those from the conventional curricular.

These conflicting and inconclusive results on the effectiveness of PBL as an educational strategy have prompted some researchers to suggest that the methods used for the evaluation were in most cases inappropriate and not congruent with the broader aim of PBL. For example, Van der Vleuten et al. (1996) pointed out the use of the regular multiple choice questions (MCQ) as an instrument for assessing the achievement of PBL students is inconsistent with PBL principles as most MCQ used tended to measure the lower taxonomic levels of knowledge of the students. In addition to that, they observed that the traditional MCQ tests caused the students to study to the tests. This had a negative effect on student

23

learning as they tended to adopt rote memorization to prepare for the test rather than to fully immersed in the PBL learning cycle. Hence, they suggested that new assessment tools or methods must be specifically designed for the PBL context. In a similar fashion, Gijselaers and Schmidt (1990) argued that if one intends to discover if an innovative educational method such as PBL works, it becomes necessary to evaluate the program based on its objectives.

They pointed out that the classical approaches for measuring students achievement are concerned mainly with the achievement of content knowledge and tend to ignore the influence of the learning process and context in shaping the final outcomes of the learning experience. Following the ideas of Cooley and Lohnes (1976), they posited that learning process in the classroom is complex and there are multiple variables and the interplay of these variables has a significant influence on the outcome of the students’ learning process. They developed and designed a causal model of evaluation study that assessed the causal relationships between the relevant variables and the outcomes in the PBL learning process.

The complex interplay of multiple variables was an idea that resonated with several other PBL researchers and some research in that direction was carried out to explore the relationships between the outcomes of the learning process and the various relevant variables in the PBL environment. Accordingly, some researchers argued that the trials of curriculum level interventions are quite pointless as PBL is a complex and multi-factorial learning environment and should not be treated as a uniform or pure intervention (Norman & Schmidt, 2000). In a complex and multiple variable environment such as PBL its effects on the outcomes of learning are affected by the complex interplay of these diverse variables, as a result of which it is impossible to attribute success or failure solely on the intervention. As rightly asserted by Faidley et al. (2000), PBL is “a sophisticated design that requires attention

24

to learner and to teacher, to content and to context.”

Schmidt and Gijselaers (1990) studied the various elements in the PBL environment and came out with a theory of PBL that was based on the complex interplay of three distinct categories of variables, namely, the input variables, the process and the outcome variables.

The input variables were the students’ prior knowledge, the block-book (i.e. the PBL problem) and tutor behavior. It was argued that the input variables would impact the process variables of study time and group functioning and this would in turn influence the outcome variables of achievement and interest in subject matter. They developed a causal model of PBL and evaluated the influence of various variables on each other using a complex path analysis method. This was an important attempt to frame the operational theoretical constructs in PBL. In line with these constructs, various studies had been carried out to investigate how these variables affected each other in the PBL environment. For instance, Gijselaers & Schmidt (1990) demonstrated that the quality of the problem (an input variable) presented to the group influenced the functioning of the group (a process variable). Using a similar causal model, Van den Hurk et al. (2001) showed that the quality of the learning issues identified and the depth of reporting had a positive impact on the students’

achievement. Schmidt & Moust (2000) discovered that group functioning affected the learning outcomes and intrinsic motivation of the students in PBL. What is clear from the above literature is that PBL environment is a complex and multi-factorial environment in which there are strong and complex interplay between various variables. As a result, a research direction to better understand how these variables interacted with each other and influenced the outcomes of the PBL process was set. The following sections outline such research trend.

25

2.4 Factors influencing the process of learning in PBL