This project is a validation study of the specialized knowledge that characterizes effective physics teaching. There is a growing consensus that teachers need to develop content knowledge for teaching (CKT) that goes beyond simply knowing the content that their students are expected to learn. Teachers also need to understand the pedagogical content that is called for in carrying out the range of tasks and practices that make up the work or tasks of teaching a school subject.  Examples of tasks of teaching include: anticipating student thinking, analyzing student errors, and designing learning experiences. Each task requires specialized content knowledge that is often only used in the work of teaching.

We studied CKT in the domain of teaching and learning energy (CKT-E) in the mechanics part of the high school physics course. This work makes the case not only for the importance of CKT in teaching but also for effective ways of measuring teaching quality generally and in physics, specifically. Using a common theoretical framework we developed a set of measures to explore what teachers understand about CKT-E, how they teach a unit on energy, and what students learn with respect to energy.

Specific accomplishments of the work include:

1. Development of a theoretical framework of CKT-E that included tasks of teaching (ToTs) and student energy targets. ToTs describe the core practices of teaching that are common across subjects and grades, e.g., Anticipating student thinking around science ideas. Energy targets describe the specific learning goals of the energy domain for introductory physics, e.g., The student understands that the energy of a system is always conserved but might not be constant.

2.  Design, administration, and scoring of a CKT-E assessment given to 329 high school physics teachers. The assessment included items that were all framed within teaching scenarios; some items called for only foundational physics knowledge while others also queried knowledge only used in teaching. Items included both selected-response and constructed-response formats (see example). Findings include: 1) Scores on the assessment were sufficiently reliable to distinguish teachers in the sample; 2) Individual items varied in both their average difficulty and their ability to discriminate among individuals; 3) Test performance was differentially related to a number of teacher background characteristics. Teachers who were physics and engineering majors score more highly than do those who were biology majors. We found little evidence of multi-dimensionality.

3.  Administration of the assessment to undergraduate physics majors, a known group with similar content knowledge as physics teachers but no physics teaching experience. This study was conducted to explore whether CKT represented a construct that could empirically be differentiated from content knowledge (CK). We found that a far greater proportion of experienced teachers, compared to physics majors, were identified as having both CK and CKT. Considering possible relationships between performance on items that draw primarily on content knowledge and performance on items that draw on the unique types of physics knowledge used in teaching, the correlation between items is stronger for teachers than for non-teachers. We concluded that teachers with strong CK are able to develop their CKT in the context of teaching, yet those with weaker CK are limited in their ability to develop CKT. On the other hand, individuals with no teaching experience are less likely to develop CKT given the lack of instructional opportunities.

4.  Conducting a validation study of CKT-E with 32 high school physics teachers by examining the relationship of CKT-E assessment scores with measures of classroom interactions and student learning. All evidence was collected during the teaching of the unit on mechanical energy. For each teacher, all video lessons were recorded, and classroom assignments and assessments were collected. Students in each class were given a pre-test of science understanding and a post-test on energy concepts.

An observation protocol was developed and used to score seven lessons distributed across the unit. The protocol’s domains include: Lesson Purpose, Lesson Coherence, Errors, Content, Discourse, and Classroom Management. Scores were reliable and broadly distributed. Observation and CKT scores exhibited a stronger correlation (r=.43) than has been typically observed.

An artifact protocol was developed to evaluate the quality of assessments and assignments. The protocol domains included: Analysis, Math, and Representations. Scores were reliable and broadly distributed, and there was a substantial relationship between artifact scores and CKT-E (r=0.37).

Finally, we fit a multi-level model to examine the relationship of student learning and CKT-E performance. Estimates indicate that a teacher performing one SD above average on the CKT-E assessment could be expected to have students performing 0.2 SD above average on the post-test after controlling for student pre-test scores, a stronger association than has been detected in other studies examining teacher knowledge and student outcomes.

Taken together, the project has demonstrated the importance and relationship of CKT to teachers’ practice and student learning in physics classrooms.

Last Modified: 11/30/2017
Modified by: Michael C Wittmann