Investigating the Effectiveness of Problem-Based Learning on Knowledge Acquisition and Retention: Empirical Evidence From a Controlled Experiment

• In Event: Arming Achilles’ Heel: Instructional Strategies and Approaches for Promoting Knowledge Acquisition in Problem-Based Learning

Abstract

Objectives & Theoretical Framework
One of the goals of problem-based learning (PBL) is the development of a flexible and rich knowledge base. Prior knowledge activation plays a prominent role in this respect and takes place when students discuss the problem at hand, relying on their prior knowledge. This is believed to foster knowledge acquisition and retention.
However, empirical studies on PBL’s ability to foster knowledge acquisition and retention compared to lecture-based curricula showed mixed results (McParland, Noble, & Livingston, 2004; Prince et al., 2003). Possibly, the design of these studies could account for the mixed results, since usually no randomization takes place. Therefore, the present study investigated the effects of PBL on knowledge acquisition and retention in a controlled way.

Methods
88 first-year Psychology students were randomly assigned to one of three conditions: PBL, lecture, and a self-study (control) condition. Table 1 illustrates the procedure for all conditions. The study topic was constructivist learning and all participants received a text on this topic to study it.

Table 1. Procedure of the present study
PBL condition (n = 28)
• Pre-discussion of the problem (15 minutes)
• Self-study of text on constructivist learning (30 minutes)
• Discussion (30 minutes)
• Test (15 minutes)

Lecture condition (n = 30)
• Lecture (45 minutes)
• Self- study of text on constructivist learning (30 minutes)
• Test (15 minutes)

Self-study condition (n = 30)
• Self-study of text on constructivist learning (75 minutes)
• Test (15 minutes)

After instruction or self-study, participants received a 30-item multiple-choice test (i.e., immediate test). After one week, participants received this test again (i.e., delayed test).

Results
A Repeated-Measure ANOVA was conducted with test performance as dependent variable, instructional type as between-subject factor, and test period (i.e., immediate and delayed) as within-subject factor. A significant main effect of time period was found [F(1, 85) = 20.16, p < .001, partial ² = .19], meaning that participants performed better on the immediate test compared to the delayed test. Also, a significant main effect of instructional type was present [F(2, 85) = 5.79, p < .005, partial ² = .12]. Multiple comparisons showed that the PBL condition significantly outperformed the lecture condition on both tests [t(85) = 3.40, p < .005, r = .35]. No differences were found between self-study and PBL and between self-study and lecture. No interaction effect was found [F(2, 85) = .23, p = .795, partial ² = .05].

Discussion & Significance
Results give evidence for the superiority of PBL in terms of knowledge acquisition compared to lectures. Although students in the lecture group also had the chance to read the study text, they performed less well. However, no differences were present between the self-study and PBL condition. This study shows that in terms of instruction, small group discussions of a problem are to be preferred over lectures.

Authors

• Marit Wijnen, Erasmus Rotterdam University

• Sofie Loyens, University College Roosevelt

• Lydia Schaap, Erasmus University