Problem-based learning (PBL) is a teaching and learning method in which students engage a problem without preparatory study and with knowledge insufficient to solve the problem, requiring that they extend existing knowledge and understanding and apply this enhanced understanding to generating a solution.

Enthusiasm for PBL is widespread, yet there exists little rigorous experimental evidence of its effectiveness, especially in K-12 populations. Many studies are conducted in artificial environments, far removed from the realities of education. At the other extreme, studies taking place in the classroom often lack experimental control, due to non-random assignment of students to control and experimental conditions; students self-selecting to take part in curricula; variations in time and exposure to treatment; different instructors; and other extraneous variables impacting treatment.

Reported here is a tightly controlled experimental study of PBL in a middle-school population. Between- and within-subject comparisons are made of students in three equivalent classes learning the same two topics, randomly assigned to three instructional conditions: traditional lecture/discussion, characteristic small-group PBL, and solitary PBL. Class 1 learned topic 1 (groupthink) via PBL-individual and topic 2 (memory and learning) via PBL-team; Class 2 learned topic 1 via PBL-team and topic 2 via lecture/discussion; Class 3 learned topic 1 via lecture/discussion and topic 2 via PBL-individual. The instructor, curriculum objectives, content, and instructional time and schedule were equivalent and standardized across groups.

Students were given two assessments nine weeks after instruction. One assessment measured students' integration and application of the concepts to a new context by examining whether they spontaneously applied the concepts learned to a novel situation. Another assessment measured students' comprehension of the concepts. Both assessments of application and comprehension showed superior mastery in both PBL conditions, relative to the lecture condition, and equivalent performance in the two PBL conditions. This indicates that engagement with the problem is the critical feature of PBL's effectiveness, not the social component.

These findings indicate that PBL may be useful in K-12 education because it appears well-suited to helping novice students apply new learning broadly and gain deep and lasting understanding.