University of North Dakota, Instructional Design & Technology, Education Building Room 218, 231 Centennial Drive STOP
7189, Grand Forks, ND 58202-7189, United States
a b s t r a c t
The design of problems is crucial for the effectiveness of problem-based learning (PBL).
Research has shown that PBL problems have not always been effective. Ineffective PBL
problems could affect whether students acquire sufficient domain knowledge, activate
appropriate prior knowledge, and properly direct their own learning. This paper builds
on the 3C3R problem design model, which is a systematic conceptual framework for guiding
the design of effective and reliable problems for PBL. To help practitioners apply the
3C3R model, this paper introduces a 9-step problem design process. The initial steps guide
an instructional designer through analyses on learning goal, content, and context to help
select problems. Later steps ensure that the problem appropriately affords the specifications
identified in the analyses. The last two steps incorporate a reflection component, as
well as ensure the integrity of the 3C3R components in the problem.
Problem-based learning (PBL) is perhaps the most innovative instructional method conceived and implemented in education. It aims to enhance students’ application of knowledge, problem solving skills, higher-order thinking, and self-directed learning skills. Its implementation began in medical education (Barrows & Tamblyn, 1980; Schmidt, 1983), and then gradually spread to various disciplines in higher education and K-12 education settings (Barrows, 2000; Dochy, Segers, van den Bossche, & Gijbels, 2003; Gallagher, Stepien, & Rosenthal, 1992; Hmelo-Silver, 2004; Hmelo, Holton, & Kolodner, 2000; Torp & Sage, 2002; Williams & Hmelo, 1998).
Fifty years after PBL was first implemented, its effectiveness as an instructional method has been a question open to debate. Advocates of PBL maintain that it is more effective than traditional methods in alleviating students’ problem of inert knowledge as well as enhancing students’ problem solving and self-directed learning skills (Barrows, 1996; Dods, 1997; Dolmans & Schmidt, 1994; Dombrowski, 1997; Kamin, O’Sullivan, Younger, & Deterding, 2001; Norman & Schmidt, 1992), while the skeptics argue that PBL is costly and ineffective because it requires more of students and instructor’s time to obtain similar learning outcomes (see Farnsworth, 1994). Even recently, Kirschner, Sweller, and Clark (2006) have continued to argue that PBL is less effective than traditional methods because its approach of providing minimum guidance is not compatible with human cognitive architecture. In addition to these theoretical research disagreements about the effectiveness of PBL, the empirical research also echoes these disagreements with contradicting results. Several meta-analyses on empirical studies in PBL were conducted in the past decade (see, for example, Albanese &Mitchell, 1993; Berkson, 1993; Colliver, 2000; Dochy
et al., 2003; Gijbels, Dochy, van den Bossche, & Segers, 2005; Vernon & Blake, 1993). The findings of these meta-analyses were mixed and inconclusive. The inconclusive or even contradictory results from these meta-analyses did not cease the battle between the advocates and opponents of PBL.
The effectiveness of an instructional method is a result of complex inter-causal relationships of numerous known and unknown variables involved in the instruction/learning processes. It may be a distant hope that we obtain a conclusive answer to the question of whether an instructional method is effective solely based on the end results of an implementation. There is little doubt that numerous factors contributed to the mixed results. Identifying these confounding variables and taking them into account could help elucidate a better picture of how an instructional method works; how effective it is; in what way and why. In attempting to do so, Gijbels et al. (2005) incorporated the format of assessment as an independent variable in their meta-analysis of the PBL studies. They found that PBLwasmost effective when the assessment of the studies focused on the students’ understanding of connection between principles and concepts.
Another potential variable that could also affect the effectiveness of PBL is the design of PBL problems. Four studies showed that the correspondence rates between instructors’ objectives and students’ generating learning issues were only about 62% (Coulson & Osborne, 1984; Dolmans, Gijselaers, Schmidt, & van der Meer, 1993; O’Neill, 2000; van Gessel, Nendaz, Vermeulen, Junod, & Vu, 2003). These low correspondence rates signal that the design of problems might have contributed to some ineffective PBL implementations in the past. Problems are at the heart of PBL. It is a reasonable conjecture that the design of PBL problems may play a critical role in affecting the effectiveness of the PBL curriculum. This paper proposes a 9-step PBL problem design process based on the 3C3R PBL problem design model (Hung, 2006) for guiding PBL educators and instructional designers to systematically design effective PBL problems. In the following, I will briefly discuss the nature of PBL and its development and research, followed by a discussion of some of the PBL implementation issues that pertain to the design of problems. Then I will discuss the 3C3R PBL problem design model and its application—the 9-step design process in detail.