THE EFFECTIVENESS OF PROBLEM-BASED INSTRUCTION

Blumberg, P. (2000). Evaluating the evidence that problem-based learners are self-directed learners: A review of the Literature. In D. H. Evensen & C. E. Hmelo (Eds.), Problem-based learning: A research perspective on learning interactions (pp. 199-226). Mahwah, NJ: Lawrence Erlbaum. 

Introduction

            Problem-Based Learning (PBL) is an appealing instruction strategy. Rather than reading or hearing about the facts and concepts that define an academic field of study, students solve realistic (albeit simulated) problems that reflect the decisions and dilemmas people face every day. Many argue that PBL is a powerful and engaging learning strategy that leads to sustained and transferable. PBL, it is argued, fosters the development of self-directed learning strategies, and makes it easier for students to retain knowledge and apply knowledge and solution strategies to new and unfamiliar situations.  PBL deviates from more conventional instructional strategies by restructuring traditional teacher/student interactions toward active, self-directed learning by the student. In PBL, teachers coach students with suggestions for further study or inquiry but do not assign predetermined learning activities. Instead, students pursue their own problem solutions by clarifying a problem, posing necessary questions, researching these questions, and producing a product that displays their thinking. These activities are generally conducted in collaborative learning groups, and these groups often solve the same problem in different ways and arrive at different answers.

Purpose

To summarize, the following study tested three hypotheses:

There is no difference in achievement, as measured via pretest-posttest changes in macroeconomics knowledge, between students in PBL and traditional instructional environments.

There is no difference in achievement, as measured via pretest-posttest changes in macroeconomics knowledge, between students with different levels of verbal ability in PBL and traditional classes,

There is no difference in achievement, as measured via pretest-post changes in macroeconomics knowledge, among students with different levels of interest in learning economics, preference for groupwork, and problem-solving efficacy.

Population/Sample

A total of 346 twelfth-grade students in 11 classes completed one or more of the instruments used in the study. The data analysis was based on data collected from the 246 students who completed the pre- and post-macroeconomics knowledge instrument and the verbal ability measure described below. These students made up 71% of students enrolled in the classes. Some of the students did not complete one or more of the aptitude assessments. When this occurred, they substituted the population mean for the missing score.

Methods

Five veteran teachers at four different high schools participated in this study conducted during Spring semester of the 1999-2000 academic year. All of the high schools were located in a large metropolitan area in Northern California. Two of the high schools were suburban and two were urban. To control for teacher effects, all teachers taught the same macroeconomics content using a PBL approach with one or more classes and a traditional lecture/discussion approach with one class. Teachers were allowed to select which class they would instruct using a lecture/discussion approach, but this choice was made before the school year began, and before teachers had received their class lists. Consequently, teachers had no advance indication of the student composition of each class. PBL and traditional classes were distributed throughout the school day, with four of the five teachers teaching the PBL and traditional classes within 2 periods of each other. The remaining teacher’s PBL and traditional classes were within 3 periods of each other.

Findings

To determine if there was a statistically significant difference in the learning of macroeconomic concepts between students in the PBL and traditional classes they calculated independent-samples t tests on the pretest-posttest change on the macroeconomics tests. For PBL students, the average pretest-posttest change was +1.48 (SD=2.52); for students in the traditional classes it was +.82 (SD=2.81). This difference was statistically significant, t = 1.94, p = .05, and equivalent to an effect size of .59 for students in the PBL instructional approach and .29 for students in the traditional approach. When pretest-posttest changes in macroeconomics knowledge for students in the PBL and traditional classes are analyzed at the teacher level, PBL class gains were greater than traditional class gains for four of the five teachers, although this comparison only reached statistical significance for teachers A (p < .05) and D (p < .01). Students in teacher C’s traditional class gained more in macroeconomics knowledge than students in the PBL class, although this difference was not statistically significant at the .05 level. These data suggest that the first hypothesis was rejected, and that the PBL instructional approach was more effective than the traditional approach in helping students to learn basic macroeconomic concepts.

Implication

            I believe the rejection of their first hypothesis concerning the equivalent learning of macroeconomic concepts in PBL and traditional classes to be a compelling finding, since statistically significant differences at the .05 level (or lower) were found for the population and for two of the five teachers. Across all teachers, the average effect size difference.  I do not consider the size of this difference to be negligible. Most students would not consider the mean difference in pretest-posttest score between the PBL and traditional classes to be trivial. Across all teachers, PBL classes gained .66 more than the traditional classes. This is equivalent to a raw score difference of 4% – or the distance between a B and a B+ in a grading system based on a maximum score of 100%.  It is interesting to compare results for teachers A and C. Both teachers taught at the same high school with a population of generally upper income students, whose verbal ability and macroeconomic knowledge pre-test scores were higher than that of students in the other schools. One teacher (A) was more effective using the problem-based approach. The second teacher (C) was more effective with a traditional approach. Although a PBL instructional approach may not be suited for all teachers, our results should be encouraging to teachers, instructional designers, and researchers who seek alternatives to traditional “sage-on-the-stage” pedagogy. In this study, PBL not only “did no harm,” but it did some good. This may prompt educators to experiment with PBL to better understand the classroom conditions and social arrangements necessary to maximize its effectiveness.