See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/241531834

Tutorials in Introductory Physics: The Pain and the Gain

Article  in  The Physics Teacher · October 2010

DOI: 10.1119/1.3488188

CITATIONS READS

7 3,313

5 authors, including:

Émerson Cruz W. Schaffenberger

University of São Paulo N.A.
3 PUBLICATIONS   10 CITATIONS    29 PUBLICATIONS   403 CITATIONS   

SEE PROFILE SEE PROFILE

Steven F Wolf Gerd Kortemeyer

East Carolina University Michigan State University
18 PUBLICATIONS   68 CITATIONS    91 PUBLICATIONS   900 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Applied Bayesian Networks View project

All content following this page was uploaded by Steven F Wolf on 01 October 2014.

The user has requested enhancement of the downloaded file.

Tutorials in Introductory Physics: The
Pain and the Gain
Émerson Cruz, Brian O’Shea, Werner Schaffenberger, Steven Wolf, and Gerd Kortemeyer, Michigan State
University, East Lansing, MI

I
n an introductory physics sequence with a large enroll- majors, though most simply took our non-majors course in
ment of premedical students, traditional recitation ses- previous years and received high grades.
sions were replaced by Tutorials in Introductory Physics, Instead of large-scale midterm exams, the first semester of
developed by the Physics Education Group at the University the sequence had six tests, typically covering two topics (e.g.,
of Washington. Initially, summative test scores (as well as FCI energy and momentum conservation). The second semester
scores) dramatically increased, but so did student complaints had three traditional midterm exams. At the end of both se-
and workload. Both effects decreased over time. The paper dis- mesters, a traditional cumulative final exam was given. Due
cusses issues that instructors should consider when contem- to the large size of the course and limited grader availability,
plating implementation of the tutorials. all tests and exams were multiple choice, with roughly half of
The series is composed of a workbook with in-class exercis- the questions being numerical and the other half conceptual
es,1 a book with homework assignments,2 and an instructor’s (“what if?” questions, ranking, graph interpretation, etc). The
guide3 that includes pre-tests and suggested exam problems (it grade is based on 40% midterms, 20% final, 20% homework,
is not immediately obvious that this guide is available; it needs 15% recitation, and 5% in-class “clicker” questions. A co-req-
to be ordered directly from the publisher). Some of the in-class uisite laboratory course accompanies the lecture course.
exercises use simple equipment, such as cardboard boxes, balls,
mirrors, and light bulbs. The in-class work is not graded, but is The students
designed to begin the students’ Socratic learning process. The The typical student in our course is a sophomore or junior
tutorial homework is very much in the style of the in-class tu- life science major who plans to go on to medical, veterinary,
torials, but meant to be solved individually. or dental school. The students’ background knowledge in
Tutorials are meant to be a supplement to traditional alge- physics is low, and typical Force Concept Inventory (FCI)10
bra- or calculus-based courses, not a full curriculum or text- pre-test scores are 35%. Also, their percentage of favor-
book. A unique feature of the tutorials is that they are based able answers on the Maryland Physics Expectations survey
on extensive research regarding students’ understanding of (MPEX)11 is low (48%), with a particularly low score in
physics concepts (e.g., Refs. 4 and 5). Yet, at the same time, the the independence cluster (38%), which indicates that their
tutorials are not “out there”—the presented physics is solid and approach to physics is very different from that of expert
exactly what most physics instructors desire their students to physicists. The independence cluster answers show that, on
master. average, students value information given by authority (in-
structor and materials) over learning independently. Typical
The course statements that the students react strongly to from this cluster
Rather than trying to make general statements about the are: “In this course, I do not expect to understand equations in
tutorials, we describe our experience in a particular course an intuitive sense; they must just be taken as givens,” and “My
with a particular student population and staff structure. Our grade in this course is primarily determined by how familiar
250-student introductory calculus-based physics sequence is I am with the material. Insight or creativity has little to do
run on a tight weekly schedule, with the standard mechanics with it.” This pattern has been consistent over at least the past
topics plus sound and thermodynamics in the first semester, three years.12 Experiences with other student populations and
and electricity and magnetism plus selected topics in relativity class settings have been reported in Refs. 13-15 and can vary
and quantum mechanics in the second semester. greatly.
The course has three lecture hours plus one recitation hour
per week. Course materials and homework are made available Previous recitation format
online using LON-CAPA,6 and students are expected to read Previously, recitation was composed of instructor-selected
the material before the first lecture on a given topic. Through exercises related to the topic of the week as well as small quiz-
embedded content-level questions and attached discussion zes. The sessions were led by LAs, and students turned in their
boards, Just-In-Time Teaching is implemented.7 During lec- quizzes and exercises immediately at the end of each session.
ture, peer teaching and “clickers” are used.8 After each lecture This format was largely unsuccessful: when putting a reci-
topic, homework questions are due online.9 The course offers tation problem on a test, success rates were frequently low,
several help room hours over the course of the week, staffed even though the recitation grades were high. In one case, only
by undergraduate learning assistants (LAs) and one or two 20% of the students successfully solved a recitation problem
graduate teaching assistants (TAs). Some of the LAs are physics when it appeared on a test, even though the vast majority had

DOI: 10.1119/1.3488188 The Physics Teacher ◆ Vol. 48, October 2010 453
full points on the original recitation quiz. Students frequently complained to course faculty that “their peers” (i.e., other
described recitations as “useless” or “a joke” in course evalua- undergraduate students) intentionally withheld necessary
tions. information.
The tutorial materials do an excellent job of providing op-
Introducing the tutorials portunities for students to check their thinking for internal
Tutorials were introduced to replace the previous recitation consistency. Often the same scenario is approached from
materials. The nine tutorial sessions, spread over the course of two different angles, and possible inconsistencies and logi-
the day with up to 28 students each, were led by undergradu- cal errors are carefully exposed. Unfortunately, many groups
ate LAs. Institutional policies do not allow LAs to lecture; this simply work linearly through the material and hardly ever go
coincides with the philosophy of the tutorials that assistants in back and check their previous statements.
the room are there simply to facilitate and support collabora- Grading is work-intensive: some of the homework as-
tive work. All course staff participated in weekly preparation signments took an average of eight minutes to grade. In a
meetings, led by the TA. 250-student course, that is more than 30 hours of grading,
Students were required to work through the tutorials in and we needed to hire additional grading staff within the first
groups of three or four, get a receipt for their attendance in two weeks of the first semester (getting “bailed out” by our
the session, and turn in the associated tutorial homework for dean). It is also rather unfortunate that the homework assign-
grading at a later time. A key was prepared by a TA, applied ments are of varying lengths, ranging from 2 to 7 pages, which
to the homework by the LAs, and the graded homework was means very different grading workload between weeks. Some
returned the following week. of the assignments are simply too long, for both graders and
The instructor’s guide recommends using problems and students. This was moderated in the second semester by as-
scenarios from the tutorials on tests and exams so the students signing a subset of problems on larger assignments, though
perceive them as relevant. It has always been our practice to this decreased the ability of students to check their answers for
include recitation scenarios on tests, and we continued to have consistency. Overall, the increase in workload appears to be
a few tutorial-based (albeit multiple-choice) questions on ev- typical.15
ery exam. Recitation grades went down considerably: the average
Purchasing and building the hands-on equipment intro- recitation score went from 92% in the previous year to 71%
duces some cost in both time and funds, with the latter being this year. Naturally, students complained, and some stated
approximately $1200 initially for this size of course for the tu- that all the tutorials would do is lower their GPA; in previous
torials we chose. We anticipate that there will be annual costs years, recitations in this course had the reputation of “dili-
of a few hundred dollars a year to replace lost or damaged gence points.”
equipment and consumables. Rarely can a new curriculum be Our student comments regarding the tutorials were dif-
implemented at such a low cost for a course of this size. ferent from the favorable evaluations reported in Ref. 14, the
mixed reactions reported in Ref. 15, and even worse than the
The pain unfavorable evaluations in Ref. 13: the most positive com-
Aligning the tutorials with the lecture topics was challeng- ments were that the tutorials are “maybe ultimately” or “very,
ing. We go through basic kinematics and dynamics within the very slightly” helpful. In the evaluations, population and class
first four weeks of the course, yet eight out of the 20 tutorials standing effects seem to be very pronounced. The majority of
deal with these introductory topics. In the second semester, the complaints correspond to the students’ epistemology, as
topics such as RLC circuits and relativity are absent. indicated by the MPEX; in fact, the expert-like tutorial phi-
Using undergraduate non-physics major LAs to lead the losophy apparently goes against the grain of what our students
sessions presented a significant challenge, as it soon became expect regarding teaching and learning. Several students
apparent that many of them were overwhelmed by the tutori- stated that tutorials are not helpful because the correct answer
als. We had to schedule additional weekly preparation time is never given to them, and that their fellow students are just
and also scheduled volunteer graduate students, post-docs, as “clueless” as they are, and thus of no help during the discus-
and faculty to participate in recitations as additional helpers. sions. Students claim that, as they got things wrong during the
In the second semester, we modified teaching schedules so in-class sessions, their mistakes propagated into their home-
that we could have two LAs per session, which improved LA work. They also asked for a key with the correct solutions after
confidence and alleviated the need for the volunteers. The ad- they turn in their homework, to help them prepare for the
ditional staff requirements could have been anticipated had tests; this is not uncommon,15 and once again shows that stu-
we read Finkelstein and Pollock, for example.13 dents are not taking advantage of the built-in “checkpoints,”
Very different from the lively discussions reported in Ref. but instead strongly rely on authority. Other epistemological
14, the sessions were often depressing, as many groups just challenges become apparent in complaints that recitations are
sat quietly and waited for staff to come to their tables and help “pointless,” because “we are not taught any equations and log-
them out; little active learning occurred. As LAs followed ic.” Informal student interviews show that the tutorials have
directions and did not simply give away solutions, students not changed the way that students think—the tutorials are

454 The Physics Teacher ◆ Vol. 48, October 2010

 work and FCI scores, though, is surprisingly weak (R2 = 0.15),
which may be due to a high degree of close collaboration and
copying on the homework (Ref. 13 also reports low correla-
tion of tutorial homework scores with other performance
measures).
The increase in average test scores on the first three tests
was even higher, from 60% last year to 72% this year (p <
10-19, n = 256 and 250, respectively); Fig. 1 shows both distri-
butions. The correlation between tutorial homework and test
scores (R2 = 0.32) is much stronger than the correlation with
FCI scores. Disappointingly, later exam scores approached
previous years’ averages.
The course is graded on an absolute scale, without a curve.
Due to the relative weighing of the recitation scores (which
went down) and test scores (which went up), slightly higher
Fig. 1. Score distribution on the first three tests in a previous year overall grades were observed in the mechanics semester than
and this year. in the previous year. Grades in the second semester were com-
parable to previous years.
written by experts and attempt to mold students into an ex- Additional improvements in student learning, primarily
pert mode of thinking (understanding the underlying physics in the area of reasoning, may have gone unnoticed, as neither
before the mathematics is added). As students got hung up on the FCI nor multiple-choice exams would reveal them, and
surface features of the presented scenarios, they often failed to they might have gotten lost in the general and always-present
see the connection and progression between them. decrease of MPEX “scores.” Detailed student interviews would
The number of students who dropped the course between be needed to appropriately assess these.
the first day and the last day where it could be dropped with- A somewhat mixed blessing is how grading the tutorial
out academic penalty (albeit with loss of tuition) more than homework reveals the persistence of physics “myths,” such
doubled between last year (six students) and this year (14 as the infamous “force of the velocity” that shows up in free-
students). Several students stated that they would switch to an body diagrams. It is painful to the instructors to see that in
alternative physics course between semesters, and cited tutori- spite of what they believe to be their best efforts, they were un-
als as the reason for leaving the course. Indeed, 18 students able to debunk such notions.
who completed the first semester did not enroll in the second A final benefit to the implementation of the tutorials in this
semester of our course, as opposed to six the previous year (a course sequence is that, as also reported in Ref. 15, the course
7.8% intra-semester attrition rate as opposed to 2.4% in the personnel, including instructors, deepened their understand-
previous year). In a subsequent year, less than 220 students ing of basic physics during the preparation sessions. This
enrolled in our course, which typically enrolls 250. More than was especially evident in the LA training sessions, where we
30 students apparently chose an alternative, highly traditional regularly had lively discussions and an active learning process
physics course. —even at 9:00 in the morning!

The gain Discussion

The introduction of the tutorials was the only major It is rare to make any changes to an introductory physics
change from the previous year implemented in the structure course for premedical students that make a major differ-
of the course sequence; most other aspects were the same, ence in student learning, and almost unheard of that a single
though a new faculty member taught the second semester. change between two years of the same course would have such
Due to logistics and time constraints, no FCI pre-test could a significant impact on student performance. In this course,
be given for the mechanics semester. A comparison of the the very same change also resulted in an astonishing increase
score distribution of the FCI post-test between the previous in student complaints, apparent frustration, additional mate-
year and this year showed a statistically strongly significant rials and staff cost, attrition, and increased workload on in-
increase from 48 percent to 55% (p < 10-5, n = 244 and 252, re- structors and volunteers, only to see that the effect eventually
spectively). Also, the post-test “failure” percentage (less than 7 decreased. This leaves us with a number of open questions:
out of 30 points) strongly decreased from 6.3% to 1.2% (while
we had increased attrition, drops happened after the data we • What was the reason for the initially strong per-
are considering here was gathered). This effect on the FCI formance increase? It could easily be understood if
appears to be comparable to Ref. 13 and much stronger than the recitation sessions were as active as our preparation
reported in Ref. 16. The correlation between tutorial home- sessions; alas, they were mostly far from it: many students

The Physics Teacher ◆ Vol. 48, October 2010 455

 did not engage in active collaborative learning. More often to point out the deeper physics that may otherwise never
than not, they appeared to completely “miss the point.” be recognized by the students.
Still, the performance increase shows that somehow tu-
In summary, within our introductory course sequence
torials work in spite of many students not really taking
the tutorials led to a significant, albeit temporary, increase in
advantage of them. Some side effects are increased time-
student performance, but at a high cost. They cannot simply
on-task (some students reported spending three hours per
be used as a drop-in replacement for ineffective recitation
week on tutorial homework) and presumably increased
materials, but require a restructuring of several other course
studying due to insecurity (or lack of a false sense of secu-
components. Since (at least with premedical students), the
rity).
introduction of tutorials leads to increases in student com-
• How can the initial success be sustained? plaints and attrition, the department chair or dean should be
Increased familiarity (and false comfort level) as well involved in their introduction, especially if the instructor is
as copying of solutions may have subverted the active pre-tenure. Is the gain worth the pain? We decided to con-
learning process. As one of the TAs put it, the students tinue using tutorial-style sessions, as we cannot, in good con-
eventually “fixed the problem of tutorials” by main- science, go back to an ineffective recitation format.
streaming them into their “efficient” way of surviving the
course—the tutorials had “worn out.” It is clear that the Reference
initial shock effect of the Tutorials is unsustainable, and 1. Lillian C. McDermott and Peter S. Schaffer, Tutorials in Intro-
that long-term success depends on a broad-based change ductory Physics (Prentice Hall, Upper Saddle River, NJ, 2001).
in staff and student attitude toward acceptance of active 2. Lillian C. McDermott and Peter S. Schaffer, Tutorials in Intro-
learning. ductory Physics Homework (Prentice Hall, Upper Saddle River,
• How can we foster active learning? In most cases, NJ, 2001).
non-physics major undergraduate LAs should not be 3. Lillian C. McDermott and Peter S. Schaffer, Tutorials in Intro-
the facilitators of these sessions in the first place, as their ductory Physics Instructors’ Guide (Prentice Hall, Upper Saddle
grasp on physics is generally not firm enough for them River, NJ, 2001), available from the publisher on request.
to be effective Socratic teachers. More time needs to be 4. Lillian C. McDermott and Peter S. Schaffer, “Research as a
guide for curriculum development: An example from introduc-
spent training staff to raise their comfort level with both
tory electricity. Part 1: Investigation of student understand-
the material and the teaching method. Furthermore, all ing,” Am. J. Phys. 60, 994–1003 (Nov. 1992); "Part 2: Design
staff need to understand the point of the tutorials and take of Instructional Strategies,” Am. J. Phys. 60, 1003–1013 (Nov.
them seriously—otherwise, this will sabotage the learn- 1992), and “Erratum, ‘Research as a guide for curriculum devel-
ing process. In the future, we plan to have more intense opment: An example from introductory electricity. Part I: In-
initial training for our LAs on how to effectively facilitate a vestigation of student understanding,” Am. J. Phys. 61, 81 (Jan.
tutorial session, and to periodically revisit this during our 1993).
weekly meetings. 5. Lillian C. McDermott, Peter S. Shaffer, and M. D. Somers, “Re-
• How can we get the students to “buy into” the search as a guide for teaching introductory mechanics: An il-
lustration in the context of Atwood’s machine,” Am. J. Phys. 62,
active learning process? It may be difficult, if not
46–55 (Jan. 1994).
impossible, to completely change the students’ epistemol-
6. www.lon-capa.org/.
ogy within our current course structure and limitations;
7. G. M. Novak, E. T. Patterson, and A. D. Gavrin, “Just-in-Time
however, we certainly could have improved the imple- Teaching: Blending active learning with web technology,”
mentation of the tutorials. It is critical (particularly for (Prentice Hall, Upper Saddle River, NJ, 1999).
premedical students) to demonstrate that doing the tutori- 8. C. H. Crouch and E. Mazur, “Peer Instruction: Ten years of ex-
als increases their understanding of the material and, as perience and results,” Am. J. Phys. 69, 970–977 (Sept. 2001).
a result, their exam scores—we did this only once at the 9. Gerd Kortemeyer, Edwin Kashy, Walter Benenson, and Wolf-
beginning of the second semester. gang Bauer, “Experiences using the open-source learning
• How can this be achieved without ignoring stu- content management and assessment system LON-CAPA in
dent concerns and worries? The pre-tests and in- introductory physics courses,” Am. J. Phys. 76, 438–444 (April/
class introductions to the tutorials are an avenue that we May 2008).
should explore in the future, even at the expense of “click- 10. D. Hestenes, M. Wells, and G. Swackhamer, “Force Concept
Inventory,” Phys. Teach. 30, 141–158 (March 1992).
er” questions and lecture content. The call for eventually
11. E. F. Redish, R. N. Steinberg, and J. M. Saul, “Student expecta-
giving out the “correct answers,” though, cannot be ac-
tions in introductory physics,” Am. J. Phys. 66, 212–224 (March
commodated: if the students expect to get them at the end 1998).
of the session, they will not engage in any discussion with 12. Gerd Kortemeyer, “The challenge of teaching introductory
each other, and homework keys would be handed down physics to premedical students,” Phys. Teach. 45, 552–557 (Dec.
to next year’s students. To partly accommodate this desire, 2007).
we started producing “Tutorial Guides,” in which we tried 13. N. D. Finkelstein and S. J. Pollock, “Replicating and under-

456 The Physics Teacher ◆ Vol. 48, October 2010

 standing successful innovations: Implementing tutorials in
introductory physics,” Phys. Rev. ST-PER 1, 010101 (2005).
14. R. Steinberg and K. Donnelly, “PER-based reform at a multicul-
tural institution,” Phys. Teach. 40, 4–10 (Jan. 2002).
15. Heidi V. Mauk and Dan Hingley, “Student understanding of in-
duced current: Using tutorials in introductory physics to teach
electricity and magnetism,” Am. J. Phys. 73, 1164-1171 (Dec.
2005).
16. Mel S. Sabella, “Implementing Tutorials in Introductory Phys-
ics at an Inner-City University in Chicago,” Physics Education
Research Conference Proceedings (PERC), 79-82 (2002).

Gerd Kortemeyer is an associate professor of physics education and the

director of the LON‑CAPA Project at Michigan State University.
Lyman‑Briggs College, Michigan State University, East Lansing, MI
48825; [email protected], www.lite.msu.edu/kortemeyer

Brian O'Shea is an assistant professor of physics and astronomy at

Michigan State University.

Émerson Cruz and Steven Wolf are graduate students of physics educa‑
tion; Werner Schaffenberger (when he is not helping out with tutorials) is
a postdoctoral researcher in solar dynamics.

The Physics Teacher ◆ Vol. 48, October 2010 457

View publication stats