[:en]Can project-based learning adequately prepare students for AP-type tests?[:]

Issue 03 ・ December 10, 2015

Project-based learning (PBL), with its focus on solving real-world problems, is frequently advocated in opposition to more test-focused curricula such as Advanced Placement (AP). Teachers in traditional AP courses typically use direct instruction to cover the wide breadth of content students need for the exam. By contrast, teachers in PBL courses use projects that emphasize depth over breadth, which means that certain content is covered more deeply, but at the cost of covering less content. Given this breadth-for-depth tradeoff, PBL skeptics argue that a project-based methodology cannot adequately prepare students for the rigors of college-preparatory testing and coursework. Are there any well-researched examples of rigorous PBL that answer this charge?

The Knowledge in Action (KIA) Initiative1

The Knowledge in Action (KIA) initiative, sponsored by the George Lucas Educational Foundation (GLEF), promotes the design and delivery of rigorous PBL for college-preparatory courses in combination with research on learning outcomes. Five design principles define the KIA approach to PBL:

  1. Rigorous projects as the spine of courses (“the main dish, not the dessert”);
  2. Iterative, interleaved project cycles where each project builds on the other;
  3. Engagement that creates a “need to know”;
  4. Teachers as co-designers and collaborators;
  5. Courses that can be scaled and used by others.

This line of research received the 2013 Exemplary Research in Social Studies Award from the National Council for the Social Studies (NCSS).2

Summary of One KIA Experiment3

Research Question: Can project-based learning be used in a high school AP U.S. Government and Politics course to improve AP test scores?

Traditionally taught AP and PBL AP courses were compared across four different class scenarios across four schools: 289 students in total (control group: 114 students received traditional AP courses; experimental group: 175 received PBL courses). Students were compared both in terms of AP score results and in terms of a scenario exam that challenged them to apply the course learning to new areas.


AP Exam Scores: Students in the PBL course scored significantly higher on the AP test than those in the traditionally taught courses. Since many more students in the PBL course took the AP exam than in the traditional course (98% vs. 73%), the possible score difference may be even larger than indicated by the results.

Scenario Exam Scores: No significant difference in performance was found, a result attributed by the authors to a floor effect (i.e., the exam was so difficult and the time given to take it so short that everybody did poorly).

The authors explain the performance difference on the AP exam by suggesting that AP courses traditionally emphasize “fast, superficial learning at the expense of meaningful learning,” which results in low memory retention and low transferability. KIA-PBL, by contrast, emphasizes “deeper conceptual learning and capacity for adaptive transfer,” which results in high memory retention and high transferability.


The results of the KIA initiative prove that rigorous PBL is possible and that such approaches can effectively prepare students for the rigors of college-preparatory testing and coursework. Accordingly, the KIA methodology and five design principles should be considered when designing project-based courses.


  1. See http://www.edutopia.org/knowledge-in-action-PBL-research.
  2. “Rethinking Advanced High School Coursework: Tackling the Depth/Breadth Tension in the AP US Government and Politics Course” by W. Parker, S. Mosborg, J. Bransford, N. Vye, J. Wilkerson, and R. Abbott, J. Curriculum Studies, 2011, Vol. 43(4), 533–559.
  3. “Beyond Breadth-Speed-Test: Toward Deeper Knowing and Engagement in an Advanced Placement Course” by W. Parker, J. Lo, A. Jude Yeo, S. Valencia, D. Nguyen, R. Abbott, S. Nolen, J. Bransford, N. Vye, American Educational Research Journal, December 2013, Vol. 50(6), 1424–1459.

Special thanks to Matt Scott for his review and suggestions.[:]