Competency based teaching of college physics: The philosophy and the practice

The practice of learning physics contributes to the development of many transdisciplinary skills learners are able to exercise independent of the physics discipline. However, the standard practices of physics instruction do not explicitly include the monitoring or evaluation of these skills. In a competency-based (CB) learning model, the skills (competencies) are clearly defined and evaluated. We envisioned that a CB approach, where the underlying competencies are highlighted within the instructional process, would be more suitable to teaching physics to learners with diversified disciplinary interests. A model CB course curriculum was developed and practiced at Purdue University to teach introductory college physics to learners who were majoring in the technology disciplines. The experiment took place from the spring semester in 2015 until the spring semester in 2017. The practice provided a means to monitor and evaluate a set of developmental transdisciplinary competencies that underlie the learning of force and motion concepts in classical physics. Additionally, the CB practice contributed to produce substantial physics learning outcomes among learners who were underprepared to learn physics in college.

Figure 1

A learner moving through the process of a content area test (CAT). A given CAT (CAT X) had two versions $\{\mathrm{CAT}\mathrm{X}.1,\mathrm{CAT}\mathrm{X}.2\}$. CAT X.1 could be taken up to 6 times, while CAT X.2 was available only for a single taking. The first four horizontal bars (solid) represent four successive attempts of the learner on CAT X.1. The learner starting from a weak base (23%) has progressed towards a high level of proficiency (97%) within the CAT X.1, earning the opportunity to take the CAT X.2. The fifth horizontal bar (bricks) represents the outcome of CAT X.2. In CAT X.2, the learner acquired 87% exercising the single available chance. A proficiency level of 80% was required for passing a CAT.

Figure 2

A background evaluation of the two classes (T and CB) at the beginning of the spring (top) and fall (bottom) semesters in 2016. The height of the vertical bars indicate the % learner population that chose one of the categories listed along the horizontal axis. The striped columns represent the T class and the solid columns represent the CB class. Both classes were similarly populated by learners in terms of their prior to college physics experience and self efficacy towards the college physics learning experience.

Figure 3

Distribution of learners of the two classes (T and CB) in terms of their diagnostic FCI scores at the beginning of spring (top) and fall (bottom) semesters in 2016. The height of the vertical bars indicate the percent learner population that fell into each percentile group listed along the horizontal axis. The striped columns represent the T class and the solid columns represent the CB class. In both classes, 65% to 70% of the learners remained in the below 40% range of the diagnostic FCI score.

Figure 4

Learners’ opinions about learning basic vector methods in the CB physics course. The height of the vertical bars are the percent learners who fell into each category listed along the horizontal axis. 31% of learners claimed it was their first time to learn about vectors, while 38% claimed that although they were not new to vectors, they felt as if they acquired the knowledge of vectors for the first time. 28% acknowledged it was a helpful review and 10% of learners thought it was unnecessary. The results indicate weak mathematical preparation overall to learning physics in college.

Figure 5

Distribution of learners of the two classes (T and CB) in terms of the post instructional FCI scores at the end of fall 2015 (left), spring 2016 (middle), and fall 2016 (right) semesters. The height of the vertical bars indicate the percent learner population that fell into each percentile group listed along the horizontal axis. The striped columns represent the T class and the solid columns represent the CB class. Unlike at the diagnostic survey (Fig. 3), the learners in the two classes demonstrate distinctly different distributions at the end of the learning process. In the T class, the learners are concentrated in the below 60% range and in the CB class the learners are populated in the above 60% range.

Figure 6

Comparison of changes to the conceptual understanding due to the instructional process, as measured using the normalized FCI gain ($⟨g⟩$ score). The height of the vertical bars indicate $⟨g⟩$ score for each semester listed along the horizontal axis. The striped columns represent the T class and the solid columns represent the CB class. The $⟨g⟩$ score of the learners who participated in the CB model instructions are $\sim 3$ times higher than that of the learners who participated in the traditional instructional process.