Where are we after 30 years of physics education research?
Andrew Kovanen
This paper was completed and submitted in partial fulfillment of the Master Teacher Program, a 2-year faculty
professional development program conducted by the Center for Teaching Excellence, United States Military
Academy, West Point, NY, 2011.
Teaching the sciences, especially physics, to undergraduate students is a formidable task
even for the most seasoned instructor. Before even stepping foot inside the classroom, many
students believe physics is difficult and beyond their capabilities to comprehend. Their
epistemological beliefs lead them to view science as a jumble of unrelated facts and formulas to
memorize (Qian, Alvermann, 2000). Even worse, many students enrolled in non-physics degree
programs that require the successful completion of a physics course dread physics and may even
resent the fact they have to take a course they believe has little relevance in their lives. The
continuing challenge of mathematical and scientific competencies of public school high school
graduates in the United States (OECD, 2010) only exacerbates the misconceptions and negative
attitudes toward physics, not to mention the problems these declining competencies create once
the physics course actually starts. Caught in this self-reinforcing loop, these students often finish
their physics experience with a worse attitude than when they started. Unfortunately, the
teaching methods of many physics instructors encourage this outcome.
Physics education research (PER) started in earnest around 1980 after the publications of
numerous studies that showed the traditional lecture class was generally an ineffective method of
instruction (Redish, Saul, and Steinberg, 1998). Since then, a tremendous amount of effort has
been put into PER. Some of the fruits of this research are well known throughout the physics
education community such as Peer Instruction (Mazur, 1997), Just-in-Time Teaching (Novak,
et.al., 1999), and Interactive Lecture Demonstrations (Sokoloff, Thornton, 1997). An
appropriate question to ask after 30 years of PER is whether or not the research is effective at
improving students’ understanding of physics. This is a bit of a loaded question and needs to be
answered in two parts: the efficacy of instructional techniques and the actual use of these
techniques by physics instructors.
Ample evidence exists that instructional techniques based on PER are superior to traditional
lecture classes. Both the Peer Instruction and Interactive Lecture Demonstration techniques were
the subjects of ten-year studies (Crouch, Mazur, 2001) (Sharma, et.al., 2010) with both
instructional techniques resulting in significantly better student understanding of the material as
opposed to traditional lecture classes. With a vast (and growing) repository of evidence showing
how instruction methods based on PER result in better student understanding compared to the
traditional lecture class, one might predict that after 30 years of PER, the traditional lecture class
would be on the path to extinction. Unfortunately, this prediction is far from being realized.
Physics instructors, for the most part, still stick to the traditional lecture method of instruction
despite knowing about the benefits of alternative instruction methods based on PER. Dancy and
Henderson (2010) found in a survey of 722 physics faculty members from degree-granting
institutions across the United States that 87% of them have heard of at least one PER-based
instructional method and that 70% of them were interested in implementing PER-based
instructional methods. It can be reasonably concluded that not only is there good dissemination
of information resulting from PER, but that physics faculty members are motivated to act on that
information. However, the survey also found that physics faculty members overwhelmingly still
use the traditional lecture as their primary method of instruction. Additionally, physics faculty
members who claimed to use PER-based instructional methods often modified the methods,
sometimes to such a degree that critical components of the methods were excluded, bringing into
question their efficacy. Instructor implementation is the problem. Fortunately, significant time
and resources are not needed to begin fixing the problem.
Physics faculties have a number of options at their disposal to help improve physics
education. At the most base level, new instructors should receive some sort of physics education
training before they step into the classroom. Ideally, this training would include information
about PER-based instructional methods although any training would be better than the complete
lack of training received by new instructors at most institutions. Physics faculty members could
attend workshops such as the Workshop for New Physics and Astronomy Faculty. Physics
instructors completing this particular workshop tend to change their instructional methods
resulting in better student learning in their classrooms (Henderson, 2007). Maybe some brutally
honest feedback of instructors’ lectures from a department head would spur some action. Or a
more positive approach could be some words of encouragement and recognition of instructors
attempting to implement PER-based instructional methods. There are plenty of options
available, but in the end, it is the individual instructor who must act.
PER continues to investigate instructional methods and insight into how students think and
work. An instructional method called Science-Technology-Society (STS) aims to get students
more actively involved in their physics education by posing a real societal problem and using
science and technology to address the problem. Visualizations are being used to help students
bridge the gap between the often microscopic “theoretical” world of physics and the macroscopic
interactions they experience every day, making physics more relevant to them. The positive
benefits of expressing physics with language continue to be revealed by PER. Even seemingly
minute details are being researched to help physics instructors understand how students think.
For example, James (2006) showed an important effect on student discourse during Peer
Instruction caused by the grading of the Peer Instruction questions. Smith, Mestre, and Ross
(2010) went as far as to study the eye-gaze patterns of students to get a better idea of how much
time they spend studying worked-out example problems. (It’s higher than you think!) The
annotated readings contain these and other important new PER from the past six years. PER,
though, does not occur in a vacuum as whims of researchers needing to publish. PER is the
response to longstanding problems in physics education and the utility of PER ultimately comes
down to individual instructors who must make the choice of staying in their comfort zones or
taking a little risk and trying something new.
The answer to the original question posed is mixed. PER-based instructional methods clearly
result in better student mastery of physics, but the methods aren’t being implemented. Certainly,
factors exogenous to physics instructors’ spheres of influence contribute to poor student
performance and negative attitudes toward physics. However, before any significant change can
take place, physics instructors need to be willing to accept the fact that they themselves may be
their students’ (and their own) worst enemy.
References
Mazur, E. (1997) Peer Instruction: A User’s Manual. Upper Saddle River, NJ: Prentice Hall.
Novack, G., Patterson, E., Gavrin, A., & Christian, W. (1999) Just-in-Time Teaching. Upper
Saddle River, NJ: Prentice Hall.
OECD (2010) PISA 2009 at a Glance. OECD Publishing. Accessed at
http://dx.doi.org/10.1787/9789264095298-en.
Qian, G., & Alvermann, D. (2000) Relationship between epistemological beliefs and conceptual
change learning. Reading and Writing Quarterly: Overcoming Learning Difficulties, 16, 59-74.
Redish, E., Saul, J., & Steinberg, R. (1998) Student expectations in introductory physics.
American Journal of Physics, 66, 212-224.
Sokoloff, D., & Thornton, R. (1997) Using ILD’s to create an active learning environment. The
Physics Teacher, 35, 340-347.
Annotated Readings:
Casperson, J., & Linn, M. (2006) Using visualizations to teach electrostatics. American
Journal of Physics, 74, 316-323.
Students often have unconnected ideas and experiences about the same scientific phenomenon
and have trouble bridging the gap between everyday macroscopic experiences and the
microscopic world. This study used the Web-based Inquiry Science Environment (WISE) during
a five-day electrostatics project to provide visuals specifically targeted to bridge the gap between
the macroscopic and microscopic worlds. The authors found that both students with high prior
knowledge of electrostatics and students with low prior knowledge benefited significantly from
the WISE visuals although students with low prior knowledge benefited more than the high prior
knowledge students. The results suggest that being able to relate microscopic phenomenon to
everyday events helps students better remember and understand physics concepts.
Crouch, C., & Mazur, E. (2001) Peer Instruction: Ten years of experience and results.
American Journal of Physics, 69, 970-977.
The authors studied the effects of Peer Instruction over 10 years in both calculus-based and
algebra-based introductory physics courses at Harvard University. The study showed that Peer
Instruction significantly increased student scores on the Force Concept Inventory (FCI) and the
Mechanics Baseline Test and that some refinements to the Peer Instruction technique produced
even further gains. There was only one control group for each group: a calculus-based class in
1990 and an algebra-based class in 1999, so it would be a stretch to call the research conclusive.
However, the normalized gains on the FCI for calculus-based classes using Peer Instruction
ranged between two and three times greater than the control course. The normalized gains for
the algebra-based course using Peer Instruction was a little over 1.5 times greater than the control
course. Interestingly, the students in the calculus-based course were more accepting of the Peer
Instruction technique than students in the algebra-based course.
Dancy, M. & Henderson, C. (2010) Pedagogical practices and instructional change of physics
faculty. American Journal of Physics, 78, 1056-1063.
The authors of this paper conclude that the dissemination of information gleaned from physics
education research (PER) has been largely effective and that physics instructors are generally
willing to implement changes based on PER. However, for the most part, physics instructional
techniques have not changed and for those instructors who have attempted to implement a
research-based instructional technique, critical portions of the technique are modified or not
performed at all. Most instructors surveyed (52.7%) cited a lack of time as the reason why they
did not implement more research-based instructional techniques followed by lack of knowledge
or access to research-based instructional techniques (25.5%).
Duda, G., & Garrett, K. (2008) Blogging in the physics classroom: A research-based approach
to shaping students’ attitudes toward physics. American Journal of Physics, 76, 1054-1065.
This article focuses on the attitudes of students taking an introductory physics course at
Creighton University. In the study, instructors maintained course blogs in which they would post
interesting topics or articles about physical phenomenon and students could comment and ask
questions. The authors discovered the course blogs helped students maintain a positive attitude
about physics whereas the students in control groups without course blogs saw a deterioration of
attitudes toward physics. Despite the better attitudes, the students in courses with blogs earned
about the same grades as students in the courses without blogs. Although the blogs did not seem
to aid in conceptual understanding, the better attitudes of students resulting from the blogs may
have encouraged more students to consider physics as a major rather than be totally turned off by
it.
Henderson, C. (2007) Promoting instructional change in new faculty: An evaluation of the
physics and astronomy new faculty workshop. American Journal of Physics, 76, 179-187.
The author’s research indicates the Workshop for New Physics and Astronomy Faculty is
meeting its goals of introducing new physics and astronomy faculty to the latest developments in
pedagogy and having participants integrate the Workshop’s ideas into their classrooms. The
Workshop takes physics and astronomy faculty who typically have one or two years of teaching
experience and exposes them to new instructional techniques based on physics and astronomy
education research (PAER). The responses to survey questions by department chairs of
Workshop attendees indicate that Workshop attendees do change their instructional style
(72.4%), the students of Workshop attendees are learning better (72.6%), and that Workshop
attendees have had influence on instructors who have not attended the Workshop (51.0%).
However, the survey of Workshop attendees also showed that many attendees modified the
research-based instructional techniques (sometimes heavily) the Workshop exposed them to,
bringing into question the effectiveness of the reinvented techniques.
James, M., & Willoughby, S. (2010) Listening to student conversations during clicker
questions: What you have not heard may surprise you! American Journal of Physics, 79, 123132.
This paper discusses the types of conversations students engage in during peer instruction using
an electronic response system (“clickers”) in a freshman-level astronomy course. In this study,
38% of student conversations were standard conversations an instructor would expect from
students while 62% were nonstandard. The authors conclude that clicker statistics can provide
misleading feedback to instructors due in large part to the effects these nonstandard
conversations have on student responses. The authors make numerous recommendations to
improve the efficacy of clicker questions including the addition of a none-of-the-above response,
having students rate their confidence in their responses, and making clicker questions a low-stake
event. The authors also present a list of recommended conversation behaviors instructors could
share with students so as to keep their discussions more meaningful.
James, M. (2006) The effect of grading incentive on student discourse in Peer Instruction.
American Journal of Physics, 74, 689-691.
This paper discusses the impact of scoring clicker responses during peer instruction. The author
compared student conversations during peer instruction for two freshman-level astronomy
courses taught by different instructors. One instructor awarded three times as much credit for
correct clicker responses as incorrect responses (high stakes) while the other instructor awarded
equal credit for correct and incorrect clicker responses (low stakes). The peer instruction
conversations in the high stakes class showed significant discourse bias where one student (often
the more knowledgeable) out of a pair would dominate the conversation. The discourse bias in
the low stakes class was about half that of the high stakes class. The author concludes that using
high stakes clicker questions during peer instruction results in response statistics that exaggerate
the level of understanding in a class making it more difficult for instructors to tailor their
instruction to fit their class’s needs.
Kortemeyer, G., Kashy, E., Benenson, W., & Bauer, W. (2007) Experiences using the opensource learning content management and assessment system LON-CAPA in introductory physics
courses. American Journal of Physics, 76, 438-444.
This paper explores the creation of LON-CAPA (an online content management and assessment
system) and its effects on grades of introductory physics students. The authors conclude that
online homework results in a noticeable (although not necessarily statistically significant)
increase in students’ final course grades and that the grade increase comes primarily from female
students. The authors offer as reasoning behind the increased grades students spending one to
two hours more per week doing physics when online homework is assigned. The authors also
analyzed the use of online discussion boards and found a negative correlation between the use of
unauthorized third-party discussion boards and examination grades. Conversely, the use of the
provided online discussion board had a positive correlation to examination grades.
Sadaghiani, H. (2011) Using multimedia learning modules in a hybrid-online course in
electricity and magnetism. Physical Review Special Topics – Physics Education Research, 7,
010102.
This study took one section of an introductory electricity and magnetism course and introduced
multimedia learning modules (MLMs) for the students to complete before select classes. This
section also had its lecture time reduced from 2, 75-minute lectures per week to 2, 50-minute
lectures per week. Despite less time in the classroom, students in the hybrid online/face-to-face
section had an 8% higher normalized gain on the Conceptual Survey of Electricity and
Magnetism than the control group.
Sharma, M., Johnston, I., Johnston, H., Varvell, K., Robertson, G., Hopkins, A., Stewart, C.,
Cooper, I., & Thornton, R. (2010) Use of interactive lecture demonstrations: A ten year study.
Physical Review Special Topics – Physics Education Research, 6, 020119.
The authors conducted a study of the effectiveness of interactive learning demonstrations (ILDs)
for introductory physics classes. The study was done over two periods of time: 1999-2001 and
2007-2009. During the 1999-2001 portion of the study, students exposed to ILDs showed
significantly higher scores on the Force and Motion Concept Evaluation (FMCE) test than
students who only had traditional lectures with normalized gains for the ILD classes being
between two and four times that of the non-ILD classes. The study was repeated from 20072009 but without a control group. The normalized gains for the ILD classes were similar to the
ILD classes in the 1999-2001 study. The study also found that ILDs were very labor intensive to
set up and operate and that having a technician do those tasks greatly aided in the delivery of the
ILD.
Smith, A., Mestre, J., & Ross, B. (2010) Eye-gaze patterns as students study worked-out
examples in mechanics. Physical Review Special Topics – Physics Education Research, 6,
020118.
This study analyzed the eye-gaze patterns of introductory physics students when studying
worked-out example problems with explanations including both textual and mathematical
information. When studying the example problems, students averaged about 40% of their gaze
time reading the material. However, despite the significant time spent reading the worked-out
example problems, the students performed poorly on conceptual questions posed to them
afterward. These results suggest the students either did not learn the conceptual information
while studying the example problems or did not retain it.
Stewart, J., & Ballard, S. (2010) Effect of written presentation on performance in introductory
physics. Physical Review Special Topics – Physics Education Research, 6, 020120.
The authors of this paper conducted a two-year study of how students in an electricity and
magnetism course at the University of Arkansas expressed solutions to physics problems. The
study showed that the total amount of language used and the fraction of writing (as opposed to
mathematics or drawing) used in solutions strongly correlated with topic understanding. This
conclusion may seem obvious, but it gives quantitative backing to making introductory physics
students use language to explain their answers in order to better understand the material.
Interestingly, the study also showed an inverse correlation between graphical means of
expression and topic understanding.
Yager, R., Choi, A., Yager, S., & Akcay, H. (2009) A comparison of student learning in STS vs
those in direct inquiry classes. Electronic Journal of Science Education, 13, 186-208.
The authors of this study conclude that using a model of Science-Technology-Society (STS), or
teaching science and technology in context of human experiences, is a more effective teaching
method than traditional direct inquiry methods. Of the five domains studied (concept, process,
applications, creativity, and attitude), students in STS classes performed better in all but the
concept domain, where they performed about the same as students in direct inquiry classes.
Considering the primary focus of the direct inquiry classes was concept mastery, students in STS
classes performing as well as those in direct inquiry classes demonstrated that the STS method of
teaching does not sacrifice understanding.