ERIC ED438185: Student Expectations in Introductory Physics.

19 Pages·2000·0.38 MB·English

by  ERIC

DOCUMENT RESUME
SE 063 214
ED 438 185
Redish, Edward F.; Saul, Jeffery M.; Steinberg, Richard N.
AUTHOR
Student Expectations in Introductory Physics.
TITLE
2000-00-00
PUB DATE
19p.; Originally published in the American Journal of
NOTE
Physics, v66 p212-24, 1998.
Tests/Questionnaires (160)
Reports Research (143)
PUB TYPE
MF01/PC01 Plus Postage.
EDRS PRICE
*Classroom Environment; Higher Education; *Physics; Science
DESCRIPTORS
Education; Scientific Principles; *Student Attitudes;
Student Interests; *Student Reaction; Student Surveys;
*Teacher Expectations of Students; Teacher Student
Relationship
ABSTRACT
Students' understanding of what science is about and how it
is done and their expectations as to what goes on in a science course play a
powerful role in what they can get out of introductory college physics. This
is particularly true when there is a large gap between what the students
expect to do and what the instructor expects them to do. This paper describes
the Maryland Physics Expectations (MPEX) Survey, a 34-item Likert-scale
survey that probes student attitudes, beliefs, and assumptions about physics.
The results of pre- and post-instruction delivery of this survey to 1500
students in introductory calculus-based physics at 6 colleges and
universities are presented. Findings indicate a large gap between the
expectations of experts and novices and a tendency for student expectations
to deteriorate rather than improve as a result of a semester of introductory
physics. (Contains 36 references.)
(Author/WRM)
Reproductions supplied by EDRS are the best that can be made
from the original document.
Student Expectations in Introductory Physics
Edward F. Redish, Jeffery M. Saul, and Richard N. Steinberg
Department of Physics, University of Maryland, College Park, MD 20742
Students' understanding of what science is about and how it is done and their expectations as to what goes
on in a science course play a powerful role in what they can get out of introductory college physics. This is
particularly true when there is a large gap between what the students expect to do and what the instructor
expects them to do. In this paper, we describe the Maryland Physics Expectations (MPEX) Survey; a 34-
item Liken-scale (agree-disagree) survey that probes student attitudes, beliefs, and assumptions about phys-
ics. We report on the results of pre- and post-instruction delivery of this survey to 1500 students in introduc-
tory calculus-based physics at six colleges and universities. We note a large gap between the expectations of
experts and novices and observe a tendency for student expectations to deteriorate rather than improve as a
result of a semester of introductory physics.
Other questions, such as what happens over the longer
I. INTRODUCTION
term and how items of various clusters correlate with
each other, are left for future publications.
What students expect will happen in their introductory
We begin by reviewing previous work on the subject in
calculus-based (university) physics course plays a criti-
section II. The structure and validation of the survey is
It affects
cal role in what they learn during the course.
described in section III. Section IV contains the results
what they listen to and what they ignore in the firehose
of the survey for five calibration groups, ranging from
of information provided during a typical course by in-
novice to expert. The results of our survey with students
It af-
structor, teaching assistant, laboratory, and text.
are presented in section V, and section VI discusses the
fects which activities students select in constructing
implications of our work.
their own knowledge base and in building their own un-
derstanding of the course material.
II. BACKGROUND AND REVIEW OF
This paper explores student attitudes and beliefs about
PREVIOUS WORK
university physics and how those attitudes and beliefs
change as a result of physics instruction. In this paper,
Maryland Physics Expectations
(MPEX)
we present the
A. Recent Progress in Physics Education:
Survey,
a Likert-style (agree-disagree) questionnaire we
Teaching Concepts
have developed to probe some aspects of student expec-
In the past fifteen years, there has been a momentous
tations. We have used this survey to measure the distri-
change in what we know about teaching and learning in
bution of student attitudes at the beginning and end of
the introductory calculus-based physics course. In about
the first semester of calculus-based physics at six col-
1980, research began to show that the traditional class
leges and universities. Our survey is included as an ap-
leaves most students confused about the basic concepts
pendix.
of mechanics.I Subsequent work extended those obser-
Because so little is known about the distribution, role,
vations to other areas including optics, heat and thermo-
and evolution of student expectations in the university
In studying
dynamics, and electricity and magnetism.2
physics course, many questions can be asked. To limit
student understanding of the basic concepts of physics,
the scope of this paper, we restrict ourselves to three
much has been revealed about what students know and
questions.
how they learn. The crucial element is that students are
not "blank slates." Their experience of the world (and
Q1. How does the initial state of students in univer-
of school) leads them to develop many concepts of their
sity physics differ from the views of experts?
own about how the world functions. These concepts are
often not easily matched with those that are being taught
Q2. To what extent does the initial state of a class
in physics courses, and students' previous conceptions
vary from institution to institution?
may make it difficult for them to build the conclusions
the teacher desires. However, it has been demonstrated
Q3. How are the expectations of a class changed as
that if this situation is taken into account, it is often pos-
the result of one semester of instruction in vari-
sible to provide activities that induce most of the stu-
ous learning environments?
dents to develop a good functional understanding of
many of the basic concepts.3
PERMISSION TO REPRODUCE AND
U.S. DEPARTMENT OF EDUCATION
DISSEMINATE THIS MATERIAL HAS
Office of Educational Research and Improvement
BEEN GRANTED BY
EDUCATIONAL RESOURCES INFORMATION
1
CENTER (ERIC)
midis document has been reproduced as
received from the person or organization
BESTCOPYAVAILABLE
originating it.
Minor changes have been made to
improve reproduction quality.
TO THE EDUCATIONAL RESOURCES
2
INFORMATION CENTER (ERIC)
Points of view or opinions stated in this
1
document do not necessarily represent
official OERI position or policy.
Success in finding ways to teach concepts is an excellent
memorizing long lists of uninterpreted facts or perform-
start (even though the successful methods are not yet
ing algorithmic solutions to large numbers of problems
without giving them any thought or trying to make sense
widespread), but it does not solve all of our teaching
of them. Although some students consider this efficient,
problems with physics. We want our students to develop
it is only efficient in the short term. The knowledge
a robust knowledge structure, a complex of mutually
supporting skills and attitudes, not just a patchwork of thus gained is superficial, situation dependent, and
ideas (even if correct). We want them to develop a quickly forgotten. Our survey is one attempt to cast
light on the hidden curriculum and on how student ex-
strong understanding of what science is and how to do
it. We want them to develop the skills and confidence
pectations are affected by instruction.
needed to do science themselves.
Previous Research on Cognitive Expectations
C.
Student Expectations
B.
There are a number of studies of student expectations in
science in the pre-college classroom that show that stu-
It is not only physics concepts that a student brings into
the physics classroom. Each student, based on his or her
dent attitudes towards their classroom activities and
own experiences, brings to the physics class a set of atti- their beliefs about the nature of science and knowledge
affect their learning. Studies by Carey6, Linn, and oth-
tudes, beliefs, and assumptions about what sorts of
things they will learn, what skills will be required, and
ers have demonstrated that many pre-college students
what they will be expected to do. In addition, their view
have misconceptions both about science and about what
they should be doing in a science class. Other studies at
of the nature of scientific information affects how they
interpret what they hear. In this paper, we will use the
the pre-college level indicate some of the critical items
phrase expectations to cover this rich set of understand- that make up the relevant elemerits of a student's system
ings. We focus on what we might call students' cogni- of expectations and beliefs. For example, Songer and
expectations about their under- Linn studied students in middle schools and found that
tive expectations
standing of the process of learning physics and the they could already categorize students as having beliefs
structure of physics knowledge rather than about the about science that were either dynamic (science is un-
derstandable, interpretive, and integrated) or static (sci-
content of physics itself.
ence knowledge is memorization-intensive, fixed, and
Our model of learning4 is a growth model rather than a
not relevant to their everyday lives)." Alan Schoenfeld
knowledge-transfer model. It concentrates on what
has described some very nice studies of the assumptions
happens in the student, rather than what the teacher is
high schools students make about learning mathemat-
doing. We therefore have chosen to focus our study on
ics.9 He concludes that "Student's beliefs shape their
cognitive attitudes that have an effect on what it is stu-
behavior in ways that have extraordinarily powerful
dents choose to do, such as whether they expect physics
(and often negative) consequences."
to be coherent or a loose collection of facts. The specific
Two important large scale studies that concern the gen-
issues our survey covers are discussed in detail in the
next section. Other issues, such as students' motivation,
eral cognitive expectations of adult learners are those of
Perry") and Belenky et al. (BGCT)" Perry tracked the
preferences, feelings about science and/or scientists, etc.
attitudes of Harvard and Radcliffe students throughout
are important but have been probed extensively else-
their college career. Belenky et al. tracked the views of
where.5
women in a variety of social and economic circum-
Although we don't often articulate them, most physics
stances. Both studies found evolution in the expectations
instructors have expectation-related goals for their stu-
of their subjects, especially in their attitudes about
dents. In our university physics course for engineers
knowledge.I2 Both studies frequently found their young
and other scientists, we try to get students to make con-
adult subjects starting in a "binary" or "received knowl-
nections, understand the limitations and conditions on
edge" stage in which they expected everything to be true
the applicability of equations, build their physical intui-
or false, good or evil, etc., and in which they expected to
tion, bring their personal experience to bear on their
learn "the truth" from authorities. Both studies ob-
problem solving, and see connections between classroom
served their subjects moving through a "relativist" or
physics and the real world. We refer to this kind of
"subjective" stage (nothing is true or good, every view
a goal not listed in the course's syllabus
learning goal
has equal value) to a "consciously constructivist" stage.
or the textbook's table of contents
as part of the
In this last, most sophisticated stage, the subjects ac-
course's "hidden curriculum." We are frustrated by the
cepted that nothing can be perfectly known, and ac-
tendency many students have to seek "efficiency"
to
cepted their own personal role in deciding what views
achieve a satisfactory grade with the least possible effort
were most likely to be productive and useful for them.
often with a severe unnoticed penalty on how much
they learn. They may spend a large amount of time
Redish, Steinberg, and Saul
Student Expectations
2
3
mester of their university physics course, tracking their
Although these studies both focused on areas other than
progress through detailed problem-solving interviews.16
science," most professional scientists who teach at both
Each student was interviewed for approximately 10
the undergraduate and graduate levels will recognize a
hours. The interviews were taped and transcribed, and
binary stage, in which students just want to be told the
students were classified according to their statements
"right" answers, and a constructivist stage in which stu-
and how they approached the problems.I7 However,
dents take charge of building their own understanding.
conducting interviews with large numbers of students
Consciously constructivist students carry out their own
would be prohibitively expensive, and they are unlikely
evaluation of an approach, equation, or result, and un-
to be repeated at many institutions. Interviews therefore
derstand both the conditions of validity and the relation
cannot yield information about the distribution of stu-
to fundamental physical principles. Students who want
dent expectations in a large population. In order to
to become creative scientists will have to move from the
study larger populations, a reliable survey is needed
binary to the constructivist stage. This is the transition
which can be completed by a student in less than half an
that we want to explore.
hour and analyzed by a computer. We developed the
An excellent introduction to the cognitive issues in-
Maryland Physics Expectations (MPEX) survey to meet
volved is given by Reif and Larkin" who compare the
this need.
spontaneous cognitive activities that occur naturally in
everyday life with those required for learning science.
B. The Development of the MPEX Survey
They pinpoint differences and show how application of
We began to develop the MPEX survey in the Autumn
everyday cognitive expectations in a science class causes
difficulties. Another excellent introduction to the cogni-
of 1992 at the University of Washington. Students in
the introductory calculus-based physics class were given
tive literature on the difference between everyday and
in-school cognitive expectations is the paper by Brown,
a variety of statements about the nature of physics, the
Collins, and Duguid, who stress the artificiality of much study of physics, and their relation to it. They rated
these statements on a five point scale from strongly dis-
typical school activity and discuss the value of cognitive
agree (1) to strongly agree (5). Items for the survey
apprenticeships."
were chosen as a result of a detailed literature review,
All the above-cited works stress the importance of ex-
discussions with physics faculty, and our combined 35
pectations in how teens and young adults make sense of
years of teaching experience. The items were then vali-
their world and their learning. If inappropriate expecta-
dated in a number of ways: by discussion with other fac-
tions play a role in the difficulties our students com-
ulty and physics education experts, through student in-
monly have with introductory calculus-based physics, we
terviews, by giving the survey to a variety of "experts",
need to find a way to track and document them.
and through repeated delivery of the survey to the same
group of students.
III. Constructing the Survey
The MPEX survey has been iteratively refined and im-
plemented through testing in more than 15 universities
A. Why a Survey?
and colleges during the last four years. The final ver-
sion of the survey presented here has 34 items and typi-
Our interactions with students in the classroom and in
cally takes twenty to thirty minutes to complete. We re-
informal settings have provided us with preliminary in-
port here on the results of the MPEX survey given at six
sights into student expectations. As is usual in physics
colleges and universities to more than 1500 students. A
education research, repeated, detailed, taped and tran-
list of the institutions that have participated is shown in
scribed interviews with individual students are clearly
Table 1. All students were asked to complete the survey
the best way of confirming or correcting informal obser-
during the first week of the term'" (semester or quarter)
vations and finding out what a student really thinks.
and at the end of the term.
The education literature contains particularly valuable
transcripts of student interviews, especially in the work
In the rest of this section, we describe how we chose the
of David Hammer. In his Ph.D. thesis at Berkeley,
items of the survey and how we validated it.
Hammer followed six students throughout the first se-
Student Expectations
Redish, Steinberg, and Saul
3
4
whether it is useful to think about them together;
Choosing the Items of the MPEX Survey
The cognitive structures that we have referred to as
beliefs about the role of mathematics
Math Link
5.
"student expectations" clearly are complex and contain
in learning physics whether the mathematical
many facets. We decided to focus on six issues or di-
formalism is used as a way of representing informa-
mensions along which we might categorize student atti-
tion about physical phenomena or mathematics is
tudes towards the appropriate way to do physics. Three
just used to calculate numbers;
of these are taken from Hammer's study and we have
added three of our own.
Effort beliefs about the kind of activities and
6.
Building on the work of Perry and Songer and Linn
work necessary to make sense out of physics
cited earlier, Hammer proposed three dimensions along
whether they expect to think carefully and evaluate
which to classify student beliefs about the nature of
what they are doing based on available materials
learning physics:I9
and feedback or not.
Independence
beliefs about learning physics
1.
The extreme views associated with each of these vari-
whether it means receiving information or involves
ables are given in Table 2. We refer to the extreme view
an active process of reconstructing one's own un-
that agrees with that of most mature scientists as the ex-
derstanding;
pert or favorable view, and the view that agrees with
that of most beginning students as the novice or unfa-
beliefs about the structure of physics
Coherence
2.
vorable view. The survey items that have been selected
as a collection of isolated pieces or as
knowledge
to probe the six attitudes are given in the right hand col-
a single coherent system;
umn of the table. We refer to the collection of survey
items designed to probe a particular dimension as a clus-
Concepts beliefs about the content of physics
3.
ter. Note that there is some overlap, as these dimen-
as formulas or as concepts that
knowledge
sions are not independent variables.20
underlie the formulas.
Although we believe the attitudes that we have defined
as expert correspond to those attitudes needed by most
In the MPEX survey, we seek to probe three additional
creative, intuitive, and successful scientists, we note that
dimensions:
they are not always predictors of success in introductory
beliefs about the connection be-
Reality Link
4.
physics classes. In an earlier study, Hammer studied
whether physics is un-
tween physics and reality
two students in the algebra-based physics course at
related to experiences outside the classroom or
2t One student possessed
many novice charac-
Berkeley
teristics but was doing well in the course. The other stu-
dent possessed many of the characteristics preferred by
experts but was having trouble. The second student's
Instructional
Institution
N
desire to make sense of the physics for herself was not
Characteristics
supported and she did not begin to succeed until she
switched her approach to memorization and pattern
445
traditional lectures, some
University of Maryland,
College Park (UMCP)
classes with group-learning
matching. In this case the course supported an attitude
tutorial instead of recitation,
and an approach to learning that most physics instruc-
no lab
tors would not endorse and one which certainly would
467
traditional lectures, with
University of Minnesota,
cause her trouble if she were to try to take more ad-
Minneapolis (UMN) group-learning research de-
vanced science courses.22
signed problem-solving and
labs
One can imagine exploring a wide variety of character-
traditional lectures, group-
445
Ohio State University,
istics ranging from whether the students like physics to
Columbus (OSU) learning research designed
whether they are intimidated by physics to whether they
problem-solving and labs
think they should take notes in lecture. In creating the
Dickinson College (DC) Workshop Physics
115
MPEX survey, we have chosen to focus on issues that
have an effect on how students interpret and process the
a small public liberal arts Workshop Physics
12
university (PLA)
physics in the class. We have not considered the stu-
dent's feelings about physics, its value or its importance.
44
a medium sized public
traditional
two-year college (TYC)
Table I: Institutions from which first semester or quarter pre-
and post-instruction survey data was collected. All data is
matched, i.e., all students included in the reported data completed
both the pre- and post-instruction surveys.
Student Expectations
Redish, Steinberg, and Saul
4
BEST COPY AVAILABLE
5
perception of the learning characteristics described in
Validating the Survey: Interviews
Table 2 differs from the way that student actually func-
We conducted more than 100 hours of videotaped stu-
tions, the self-perception has a strong tendency to be
dent interviews in order to validate that our interpreta-
closer to the side chosen by experts. We therefore feel
tion of the survey items matched the way they were read
that while survey results for an individual student may
and interpreted by students. We asked students (either
be misleading, survey results of an entire classroom
individually or in groups of two or three) to describe
might understate unfavorable student characteristics.
their interpretations of the statements and to indicate
why they responded in the way that they did. In addi-
tion, students were asked to give specific examples from
IV. EXPERT EXPECTATIONS:
class to justify their responses.
THE CALIBRATION GROUPS
From these interviews, we have found that students are
In order to test whether the survey correctly represents
not always consistent with their responses and ap-
elements of the hidden curriculum, we gave it to a vari-
proaches to what appear to us to be similar questions
ety of students and physics instructors. We defined as
and situations. We feel that this does not represent a
"expert" the response that was given by a majority of
failure of the survey, but properly matches these stu-
experienced physics instructors who have a high con-
dents' ill-defined understanding of the nature of physics.
cern for educational issues and a high sensitivity to stu-
One reason for this was described by Hammer. He ob-
dents. We conjectured that experts, when asked what
served that some students in his study believed that pro-
answers they would want their students to give, would
fessional physicists used the favorable conditions, but
respond consistently.
that it sufficed for them to behave in the unfavorable
fashion for the purposes of the course. He referred to
A. The Calibration Groups
this by adding the marker "apparent" to the characteris-
We tested the response of a wide rangeof respondents
tic. This is only one aspect of the complex nature of
by comparing five groups:
human cognition. We must also be careful not to as-
sume that a student exists in one extreme state or an-
Group 1: engineering students entering the calcu-
1.
other. A student's attitude may be modified by an addi-
lus-based physics sequence at the U. of Maryland,
tional attitude, as in Hammer's observations, or even ex-
ist simultaneously in both extremes, depending on the
Group 2: members of the US International Physics
2.
situation that triggers the response.23 One must there-
Olympics Team
fore use considerable care in applying the results of a
limited probe such as our survey to a single student.
Group 3: high school teachers attending the two-
3.
week Dickinson College Summer Seminar on new
We are also aware that students' self-reported percep-
approaches in physics education
tions may not match the way they actually behave.24
However, the interviews suggest that if a student's self-
MPEX Items
Favorable
Unfavorable
independence
takes responsibility for constructing takes what is given by authorities
1, 8, 13,
(teacher, text) without evaluation
own understanding
14, 17, 27
believes physics can be treated as
12, 15, 16,
coherence
believes physics needs to be consid-
21, 29
ered as a connected, consistent
unrelated facts or "pieces"
framework
4, 19, 26,
stresses understanding of the under-
focuses on memorizing and using
concepts
27, 32
lying ideas and concepts
formulas
reality link
believes ideas learned in physics believes ideas learned in physics
10, 18,
are relevant and useful in a wide has little relation to experiences
22, 25
variety of real contexts outside the classroom
math link
considers mathematics as a conven- views the physics and the math as 2, 6, 8,
independent with little relationship
ient way of representing physical
16, 20
between them
phenomena
effort
does not attempt to use available in- 3, 6, 7,
makes the effort to use information
available and tries to make sense of
formation effectively
24, 31
it
Table 2: Dimensions c f student expectations
Redish, Steinberg, and Saul Student Expectations
5
6
sity teachers attending the two-week summer seminar
D D D D
28
19
10
(N=56), and group 5 college and university teachers
A
D
20 D D
29
2
11
implementing Workshop Physics in their classroom
D D
A
(N=19). The teachers in group 5 were committed to im-
30
3
21
12
plementing an interactive engagement model of teach-
D D D A
4 22
31
13
i
ing in their classroom. We asked the three groups of in-
A
D
D A
23 32
5 14
structors to respond with
the answer they would prefer
We expected these five groups to
their students to give.
D
24 1
A
D D
6
33
15
show an increasing level of agreement with answers we
(A) (A)
D
A
25 34
7
16
preferred.
=
D
A
D 26
8
17
B. The Responses of the Calibration Groups
A
(D)
9
27
18
1
I
The group we expected to be the most sophisticated, the
Table 3: Prevalent responses of our expert group. Where the
group 5 instructors, agreed strongly as to what were the
respondents did not agree at the >80% level, the item is
responses they would like to hear from their students.
shown in parentheses and the majority response is shown.
On all but three items, 80% or more of this group
The response "A" indicates agree or strongly agree. The re-
agreed with a particular position . Three items, num-
sponse "D" indicates disagree or strongly disagree.
bers 7, 9, and 34, had a strong plurality of agreement,
Group 4: university and college teachers attending
4.
but between X and X of the respondents chose neu-
the two-week Dickinson College Summer Seminar
tral. We define the preferred response of group 5 as the
on new approaches in physics education
We define a response in agreement
expert response.
with the expert response as and a response in
favorable
Group 5: college faculty who are part of a multi-
5.
disagreement with the expert response as
unfavorable.
university FIPSE-sponsored project to implement
For the analysis in this paper, the agree and strongly
Workshop Physics at their home institutions.
agree responses (4 and 5) are combined, and the dis-
agree and strongly disagree responses (1 and 2) are
The University of Maryland students are a fairly typical
combined. A list of the favorable responses to the sur-
diverse group of engineering students at a large research
vey items is presented in Table 3.
university. The entering class average on the FCI is
around 50%, comparable to the average for introductory
To display our results in a concise and easily interpret-
The number of students in
university physics classes.2
able manner, we introduce an
(A-D)
agree-disagree
the sample is N=445.
In this plot, the percentage of respondents in each
plot.
group answering favorably are plotted against the per-
The US International Physics Olympics Team (USIPOT)
centage of respondents in each group answering unfa-
is a group of high school students selected from appli-
vorably. Since the fraction of students agreeing and dis-
cants throughout the USA. After a two week training
agreeing must add up to less than or equal to 100%, all
session, five are chosen to represent the US in the Inter-
points must lie in the triangle bounded by the corners
national Physics Olympics. In 1995 and 1996, this
(0,0), (0,100), (100,0). The distance from the diagonal
group trained at the University of Maryland in College
line is a measure of the number of respondents who an-
Park and we took the opportunity to have them complete
swered neutral or chose not to answer. The closer a
survey forms. The total number of respondents in this
point is to the upper left corner of the allowed region,
group is N=56. Although they are not teachers, they
the better the group's agreement with the expert re-
have been selected by experts as some of the best high
26
sponse.
school physics students in the nation. Our hypothesis
was that they would prove to be more expert than the
The results on the overall survey are shown in Fig. 1.
In
average university physics student, but not as expert as
this plot, the percentages are averaged over all of the
our groups of experienced instructors.
items of the survey, using the preferred responses of
calibration group 5 as favorable. The groups' responses
The physics instructors who served as our test groups
are distributed from less to more favorable in the pre-
were all visiting Dickinson College. Attendees came
dicted fashion.27
from a wide variety of institutions. Many have had con-
siderable experience in teaching, and all of them were
Although the overall results support our contention that
sufficiently interested in educational development to at-
our survey correlates well with an overall sophistication
tend a workshop. We separated them into three groups:
of attitudes towards doing physics, the cluster results
high school teachers attending a two-week
group 3
show some interesting deviations from the monotonic
ordering. These deviations are quite sensible and sup-
summer seminar (N=26), group 4
college and univer-
Redish, Steinberg, and Saul Student Expectations
6
7
port our use of clusters as well as overall results. In or-
pect their students to have.
der to save space and simplify the interpretation of re-
sults, we present the data in Table 4. Displayed in this
V. STUDENT EXPECTATIONS:
table are the percentages of each group's favorable and
DISTRIBUTION AND EVOLUTION
unfavorable responses (in the form favorable / unfavor-
able). The percentage of neutrals and not answering can
In this section, we discuss the results obtained from giv-
be obtained by subtracting the sum of the favorable and
ing the MPEX survey at the beginning and end of the
unfavorable responses from 100.
first term of introductory calculus-based physics at six
different institutions. In each case, the subject covered
From the table we see that most of the fraction of re-
was Newtonian mechanics. The schools involved in-
spondents agreeing with the favorable response tends to
clude the flagship research institutions of three large
decrease monotonically from group 1-5 with a few inter-
state universities: the University of Maryland (UMCP),
esting exceptions. The high school teachers (group 3)
Ohio State (OSU), and Minnesota (UMN); plus three
are farther than their average from the favorable corner
smaller schools: Dickinson College (DC), a small public
in the coherence and math clusters, while the Physics
liberal arts college (PLA), and a public two year college
Olympics team is closer to the favorable corner in those
(TYC). At the named colleges, we have data from mul-
categories than their average. These results are plausi-
tiple instructors. In the case of the last two institutions,
ble if we assume that high school teachers are less con-
data was only collected from a small number of instruc-
cerned with their students forming a coherent and a
tors and students. These are included in order to dem-
mathematically sophisticated view of physics than are
onstrate how the MPEX survey can be used as a diag-
university teachers. The results also agree with our per-
nostic tool, but are kept anonymous to protect the iden-
sonal observations that the members of the USIPOT are
tity of the instructors and institutions involved.
unusually coherent in their views of physics and excep-
tionally strong in their mathematical skills.
At Maryland, Ohio State, and Minnesota, classes were
presented in the traditional lecture-lab-recitation frame-
Note also that the Olympics team results are very far
work with some modifications. At Maryland, there is no
from the favorable corner in the effort cluster. The main
laboratory in the first semester and some of the recita-
discrepancies are in items 3 and 7. We suggest that the
tion sections were done with University of Washington
reader peruse the survey items of that cluster (3, 6, 7,
style tutorials.28 At Minnesota, the laboratory and reci-
24, 31). These items represent highly traditional meas-
tations involve carefully designed group work.29 At
ures of effort (reading the textbook, going over one's lec-
Ohio State, lectures are traditional but are enhanced by
ture notes) which we conjecture are not yet part of the
use of various interactive elements, while recitation and
normal repertoire of the best and brightest high school
laboratory are done in a group problem-solving format
physics students before they enter college. We also con-
similar to that developed at Minnesota. At Dickinson
jecture that most of them will have to learn to make
College and at the public liberal arts institution, the
these kinds of efforts as they progress to increasingly so-
classes surveyed were done in the Workshop Physics en-
phisticated materials and the level of challenge rises.
vironment which replaces lectures with a combined lab
This analysis of both the overall responses of the cali-
and class discussion.3d The two-year college used a
bration groups and the variations in the ordering con-
purely traditional lecture-recitation framework. Like
firms that the MPEX survey provides a quantitative
Maryland, they have no lab in the first semester. The
measure of characteristics which experts hope and ex-
schools involved, the structure of their courses, and the
number of students in our sample are summarized in
Table 1.
100%
00%
In order to eliminate the confounding factor of differen-
tial drop-out rates, we only include students who com-
80%
+Experts
pleted the survey both at the beginning and at the end of
+
70%
College 789050,6
the term. We say that the data is matched. Our results
60%
X
show some differences among different classes at the
HS Taw hers
50%
same institution, but the variation is statistically consis-
+US1POT
40%
tent with the sample size. Therefore, we have combined
30%
X UsACP pre
results for similar classes at a given institution.
20%
The overall survey results for the six schools are pre-
10% -
sented in an A-D plot in Fig. 2. In order to simplify the
0%
1
reading of the graphs, we have displayed the results
0% 80%
30% 40% 60% 90%
20% 70%
100%
50%
10%
from the three large research universities in one part of
Fig. I: A-D plot for Calibration Groups.
Average of all items
Redish, Steinberg, and Saul
Student Expectations
7
8
an
MIR
(gAtuage.
Dm Gra
I depen-
Overall
Reality
lkita3
Experts
87/6
93/3
92/4
89/6
85/12
93/3
85/4
80/10
College
80/8
80/12
80/8
94/4 82/6
84/9
Teachers
HS Teachers
75/16
62/26
95/2
73/15 71/18
68/13
67/21
USPOT
68/18
64/20
81/12
79/8
50/34
73/13
85/8
I
UMCP pre
61/14
53/24
54/23 42/35 67/17
54/25
67/13
1
UMCP post
48/27
44/32
49/25 49/27
59/20
58/18 48/27
UMn pre
45/27
59/19
72/9
59/18 57/20
i
72/11
72/11
UMn post
69/10
58/20
57/20
61/17 46/28
72/12
63/16
-
,
OSU pre
51/24
65/10
53/23
37/36 66/16
65/13
52/21
OSU post
46/28 46/26
44/30
45/28
54/17
55/20
35/35
j
DC pre
62/14
76/4
58/17 47/23
70/10
61/15
75/7
DC post
60/19
67/14
72/9
71/12
57/26
66/18 58/23
PLA pre
57/27 57/26
57/23
38/46 71/13
72/8
74/11
-----4/31
PLA post
45/34
48/30
-----2722 47/33 52/25
54/19
!
TYC pre
41/29
69/16
55/22
30/42 58/17
50/21
80/8
TY----C--P-o-si----- 49/26 -----ii/3i-T- 48/29
58/17
58/18
35/41
65/21
Table 4: Perm rages of students giving favorable/unfavorable responses on overall and clusters of the MPEX
survey at the beginning (pre) and end (post) of the first unit of university physics.
the figure (Fig. 2a) and those from the smaller schools
age deterioration rather than an improvement
in another (Fig. 2b). The pre-course results are shown
of student expectations.
by filled markers and the post-course results by open
markers. The results of the expert group are shown by a
The overall survey includes items that represent a vari-
cross.
ety of characteristics, as displayed in Table 2. In order
to better understand what is happening in the classes ob-
We make two observations.
served, let us consider the initial state and the change of
The initial state of the students at all the
student expectations in the various clusters. The results
1.
schools tested differs substantially from the ex- are presented in Table 4.
pert results. The expert group was consistent,
A. The Independence Cluster
agreeing on which survey responses were desir-
able 87% of the time. Beginning students only
One characteristic of the binary thinker, as reported by
agreed with the favorable (expert) responses
Perry and BGCT, is the view that answers come from an
about 40-60% of the time, a substantial dis-
authoritative source, such as an instructor or a text, and
crepancy. What is perhaps more distressing is
it is the responsibility of that authority to convey this
that students explicitly supported unfavorable
knowledge to the student. The more mature students
positions about 20-30% of the time.
understand that developing knowledge is a participatory
process. Hammer classifies these two extreme views as
In all cases, the result of instruction on the
2.
"by authority" and "independent." Survey items 1, 8,
overall survey was an increase in unfavorable
13, 14, 17, and 27 probe students' views along this di-
responses and a decrease in favorable re-
mension. On this cluster, students' initial views were
sponses.
Thus, instruction produced an aver-
Redish, Steinberg, and Saul
Student Expectations
8
9
16, 21, and 29 have been included in order to probe stu-
favorable in a range from 41% (TYC) to 62% (DC). All
dent views along this dimension.
groups showed essentially no significant change as a re-
sult of one term of instruction. For comparison, the
Our expert group was in agreement as to what responses
USIPOT showed favorable views on these items 81% of
were desirable on the elements of this cluster 85% of the
the time.
time. The initial views of students at our six schools
were only favorable between 50% and 58% of the time.
Survey items 1 and 14 are particularly illuminating and
Most classes showed a small deterioration on this clus-
show the largest gaps between experts and novices.
ter, except for UMN (slight improvement from 57% to
61% favorable responses) and DC (improvement of 58%
#1: All I need to do to understand most of the basic
to 66% favorable responses).
ideas in this course is just read the text, work most
of the problems, and/or pay close attention in class.
Two specific items in this cluster are worthy of an ex-
plicit discussion.
#14 Learning physics is a matter of acquiring new
knowledge that is specifically located in the laws,
#21: If I came up with two different approaches to a
principles, and equations given in the textbook and
problem and they gave different answers, I would
in class.
not worry about it; I would just choose the answer
that seemed most reasonable. (Assume the answer
The expert group was in 100% agreement that students
is not in the back of the book.)
should disagree with item 1 and in 84% agreement that
they should disagree with item 14. Disagreeing with
#29: A significant problem in this course is being able
these items represents a rather sophisticated view of
to memorize all the information I need to know.
learning; but favorable shifts on these items are exactly
the sort of changes that indicate the start of a transition
Item 21 is a touchstone. Coming up with two different
between-a binary and a more constructivist thinker. The
answers using two different methods indicates some-
interviews strongly support this view. Students who
thing is seriously wrong with at least one of your solu-
disagreed with these items were consistently the most
tions and perhaps with your understanding of the phys-
vigorous and active learners.
ics and how to apply it to problems. Our expert group
This cluster of items, and items 1 and 14 in particular,
appear to confirm that most students in university phys-
100%
ics enter with at least some characteristics of binary
90
learners, agreeing that learning physics is simply a mat-
80%
ter of receiving knowledge in contrast to constructing
+ Evens
70
one's own understanding. We would hope that if a uni-
.osu Pre
60%
X UMCP pre
versity education is to help students develop more so-
UMN pre
50%
phisticated views of their own learning, that the intro-
,,OSU peel
MUMCP post
40%
ductory semester of university physics would begin to
UMN post
30%
move students in the direction of more independence.
20% -
Unfortunately, this does not appear to have been the
10% -
case. In the touchstone items of 1 and 14, the only sig-
nificant improvement was DC on item 14 (26% to 53%),
0%
90%
70% 60%
30% 40%
100%
50% 60%
0% 20%
10%
and overall, only DC showed improvement.
100%
B. The Coherence Cluster
90
Most physics faculty feel strongly that students should
80 %-
see physics as a coherent, consistent structure. A major
+ Evens
70%
strength of the scientific worldview is its ability to de-
0
41 TYC po
ir
60%
DC we
scribe many complex phenomena with a few simple laws
%
W PLA pro
50
-
and principles. Students who emphasize science as a
()roc post
0 CC post
40%
collection of facts fail to see the integrity of the struc-
VPLA post
ture, an integrity that is both epistemologically convinc-
30% -
ing and useful. The lack of a coherent view can cause
20%-
students many problems, including a failure to notice er-
10%
rors in their reasoning and an inability to evaluate a re-
0%
90%
80%
50%
30% 40%
100%
60%
70%
0% 20%
10%
called item through cross-checks. Survey items 12, 15,
Fig. 2a: (top) A-D plot for large schools. Average of all items.
Fig. 2b: (bottom) A-D plot for small schools. Average of all items
Student Expectations
9
Redish, Steinberg, and Said