This research compares differences between standardized test scores in problem-based learning (PBL) classrooms and a traditional classroom for 6 ^th grade students using a mixed-method, quasi-experimental and qualitative design. The research shows that problem-based learning is as effective as traditional teaching methods on standardized tests. The most significant finding from this study is that the use of problem-based learning can increase standardized test scores at least as much as traditional teaching methods. This is true for initially low-scoring students and disadvantaged groups and when increasing higher-thinking skills is a classroom goal.

The mixed-method research design demonstrated differences in scores both between the experimental groups as an aggregate and within the groups on the pre- and posttests. The aggregate test score improvement over traditional teaching methods for the PBL group was very low. Improvement is so small it can be dismissed as unimportant. However, an additional regression analysis of test question levels of difficulty, defined by CTB (2002), provides more important results. When incorrect answers and corrected answers are compared, this analysis shows more students chose correct answers at higher levels of difficulty for the PBL classrooms while the control classroom chose more correct answers at the lower levels of difficulty.

Overall, the research increases the body of knowledge about PBL because it compared PBL and traditional teaching relationships to achievement scores on standardized tests. Previous PBL research was also supported. The earlier research focused on qualitative studies that examine teacher classroom observations, opinions, and emphasize non-standardized assessments. The test score focus provides a unique and needed starting point for new research.