The impact of epistemology on learning: A case study from introductory

physics
Laura Lising* and Andrew Elby
Department of Physics, University of Maryland, College Park, Maryland 20742
We discuss a case study of the influence of epistemology on learning for a student in an
introductory college physics course. An analysis of videotaped class work, written work,
and interviews indicates that many of the students difficulties were epistemological in
nature. Our primary goal is to show instructors and curriculum developers that a students
epistemological stance her ideas about knowledge and learning can have a direct,
causal influence on her learning of physics. This influence exists even when research-
based curriculum materials provide implicit epistemological support. For this reason,
curriculum materials and teaching techniques could become more effective by explicitly
attending to students epistemologies.

I. INTRODUCTION attempted to describe not just whether, but how learning is

affected by epistemology and related factors. An excellent
In the past 15 years, physics education researchers example is Hogans thorough study on eighth-graders in
have identified student difficulties in learning a broad range which she observed relationships between students
of physics concepts. Curricula targeting these difficulties personal frameworks for science learning and their social
have produced dramatically improved conceptual and cognitive engagement patterns during group learning.12
understanding.1 In recent years, the physics education Ryder and Leachs study found some correlations between
research community also has begun to look at student college students ideas about the nature of scientific
attitudes, expectations, and epistemologies (ideas about knowledge and their self-reported activities during
knowledge and learning).2,3,4 For instance, students may investigative project work.13 Millar et al. observed that,
think of physics knowledge as disconnected facts and among 9 to 14-year-olds, students interpretations of
formulas, or as interconnected concepts (often expressible as classroom inquiry tasks varied according to their perceptions
formulas). Students may think of learning physics as of the aims of scientific investigation.14 Taylor-Robertson
absorbing information from authority or as building up their found differences in cognitive strategies used by college
own ideas.5 This discipline-specific epistemology research students according to their expectations of the
builds on extensive research on more generalized meaningfulness of laboratory work,15 and Edmondson found
epistemology.6 correlations between students reported learning strategies
The recent focus on epistemology in physics education and their epistemological stances as derived from
stems in part from two motivating ideas: (i) Students interviews.16 Dwecks work with students of varied ages
epistemologies may affect their science learning. In that showed some dramatic differences in learning behavior in
case, attending to epistemology may help us explain the the classroom which depended on students ideas about the
variations in student learning outcomes with research-based nature of intelligence.17 And Hammers study on college
curricula, create more effective curricula, and become better students described how students ideas about knowledge and
physics instructors. (ii) Fostering productive attitudes and learning in physics affected how they solved physics
epistemologies is in itself an important instructional homework problems during think-aloud interviews.2 Taken
outcome that could serve the students well beyond the together, these studies suggest a causal link between
course in question. epistemology and learning and also raise new questions and
Our study addresses the first of these ideas and builds issues. One issue is the distinction between personal and
on previous research on college and pre-college learners. public epistemologies. Public epistemology encompasses a
Most previous research has looked at correlations between students ideas about the nature of knowledge and learning
epistemological measures and learning outcomes, finding for society as a whole - or for a disciplinary community.
that specific clusters of epistemological beliefs correlate Personal epistemology concerns a students ideas about her
with academic outcomes such as grade point average7 and own knowledge and learning. A students public and
mathematical text comprehension.8 In the physical sciences, personal epistemologies can differ significantly. For
one study found that certain epistemological beliefs correlate instance, a student may doubt the possibility of coherence
with integrated conceptual understanding in middle school,9 in her own knowledge (personal epistemology), but may
while another found a correlation with ninth-graders ability expect scientists to seek and find coherence (public
to reason on applied tasks.10 In college physics, May and epistemology). Some of the previous correlation studies
Etkina found correlations between students gains on have looked at only one of these aspects of epistemology,
standard conceptual measures and their epistemologies as while others have not made this distinction. Of the three
inferred from weekly written reflections on their own personal frameworks for science learning in Hogans
learning.11 research, one aligns fairly closely with personal
A few studies have gone beyond these correlations epistemology while another aligns with public
to look at the causal influence of epistemology on students epistemology. She found that personal epistemology was
learning behavior. These studies, generally carried out by linked strongly to the students behavior, while public
observing students in the process of learning, have epistemology showed almost no effect. Thus her results
1 Submitted to the American Journal of Physics
point toward personal epistemology as being much more found that epistemologies assessed in interviews differed
relevant to learning. For that reason, we focus on the from the approaches students took in class. One might
personal epistemology of our student subject. This paper expect this difference between students tacit ideas and their
builds more on the work of Hogan,12 Hammer,2 and May explicitly articulated ones, but Hogans interview methods
and Etkina,11 which focused on personal epistemology, than included elicitation of tacit ideas through scenario-posing.12
on the other studies mentioned above,3,4,7-10,13-16 which Thus it has become clear that taking context-sensitivity into
looked at public epistemology or a combination of personal account when designing studies and analyzing data is crucial
and public epistemology and other attitude-related variables. in understanding epistemology and learning.
To build on this line of research, we have done an In our study, we look at a student, Jan and study
in-depth and naturalistic case study of a single student to both her personal epistemology and her learning and
distill and carefully describe the likely causal mechanisms. describe how one affects the other. By analyzing both
Of course, a case study cannot produce definitive, epistemology and learning from the same set of classroom
generalizable results about causality. But it can add depth data, we avoid many context-related interpretive challenges
and detail to the perceptive toolkit of the instructor and and also provide a description that is immediately relevant
curriculum developer by exploring specific causal to classroom learning and instruction. We use a separate set
mechanisms that might explain the correlations, and it can of data from interviews for a supplementary analysis,
generate specific hypotheses about causal mechanisms for carefully accounting for context-driven differences and
later testing in controlled-intervention studies. The factors that point to public epistemology, expectations, and
following hypothetical example illustrates this point. other influences. From this analysis, we are able to describe
Suppose a correlation is found between how quickly people direct, causal links that are likely to exist between Jans
learn rock climbing skills and how many safe exposures to epistemology and her learning in the classroom. Due to the
heights they experienced as children. A possible causal difficulty of making and describing such an in-depth
mechanism underlying this correlation might be that lack of argument about causality, we will do so for only one facet
safe exposure to heights as children leads to a fear of of Jans epistemology, although Jan certainly possesses a
heights, which then leads to some learners making more wide array of ideas about knowledge and learning. We will
cautious movements. Case studies of a few slow-learning focus only on how Jan selected and used conceptual
novice rock climbers might shed light on this hypothesis. resources in her physics learning, and not on other facets of
As they first attempt new moves, do they give clues to their her epistemology such as whether she treated knowledge as
fear of heights verbally or physiologically? Can we rule out static or evolving.
other possible causes by watching their behavior in detail? If After discussing our methods in Sec. II, we present in
so, the next step toward establishing causal mechanism Sec. III two examples of Jans classroom behavior in group
might be a controlled-intervention study, safely exposing work. In Sec. IV we use these examples to argue that a
children to heights, enrolling them in a rock climbing class component of Jans epistemology, her perception of a wall
15 years later, and comparing their learning speed to a between formal reasoning and everyday/intuitive reasoning,
control group who received a different intervention as contributes to her troubles learning the material. We then
children (for example, reading about rock climbing). Our use an independent data set from clinical interviews to argue
goal is to develop a plausible existence argument and that Jans epistemology does include this wall. Section IV
descriptive analysis for one particular causal mechanism also addresses alternative, non-epistemological explanations
between epistemology and learning, a mechanism that we of Jans classroom behavior. Although some of those factors
hope will be tested in future controlled-intervention contribute to Jans actions, we argue that no combination of
experiments. them adequately accounts for her behavior, unless our
The various previous studies we have cited also epistemological explanation is included. This strengthens
vary in the extent to which they disentangled students our case for a causal link between Jans epistemology and
personal epistemologies from their expectations about her learning. In Sec. V we summarize this argument and
whats rewarded in a particular course. It can be difficult to discuss implications for instruction and research.
distinguish between what a student thinks is productive for
her learning and what she perceives is required by the II. METHODS
teacher or the curriculum. Yet these can be quite disparate at
IIA. Selection of our case study subject and collection of
times. Hammers work with one student illustrates an
data
example where a student ruefully and self-consciously
abandoned her productive learning strategies to survive in a The subject of this case study, Jan, was a third-year
memorization-focused physics course.18 A 1999 study by student in the second semester of an algebra-based
Elby gave some insight into the magnitude of the introductory physics course at the University of Maryland.
epistemology/expectations gap.19 The course, taken by about 100 students and taught by a
Yet another issue arising in previous studies is the physics education researcher, consisted of 3 hours per week
context-sensitivity of students epistemologies. Survey- of interactive lectures (including interactive lecture
based research on students epistemologies has established demonstrations20 and other physics education research
differences in approaches according to discipline, motivating inspired elements), one hour of tutorial (worksheet-led
research that is discipline-specific (such as Ref. 3). conceptual group work21), and two hours of traditional-style
However, studies that involved observations of learning laboratories. Jan had taken this courses prerequisite in the
behaviors and studies with multiple epistemological previous semester in a large lecture, purely traditional
assessments also uncovered a sensitivity of epistemology to format from a different professor. Although we will
context within a given discipline. Hogan, for example, highlight some of Jans difficulties, overall she was a
2 Submitted to the American Journal of Physics
capable student. She has excellent mathematical skills, did single, stationary source charge, Q, on several test charges,
well on the more traditional homework problems, and put q, of differing magnitudes placed at the same point. They
in considerable effort, seeking help from peers. Some then work out the ratio of the force to the test charge, define
concepts she learned quite deeply while others she did not. the field, E, as the ratio, and then continue to explore which
Jan was in one of the two groups of students we factors affect the field and which do not. The main point of
videotaped working in tutorials and laboratories over the this part of tutorial is that E expresses the influence of the
course of the semester. For Jans group, we had two usable source charge in a way that doesnt depend on the test
hours of videotape. The other videotapes of her group were charge used to measure that influence. The full transcript
unusable because they were inaudible or because the appears in the Appendix22 of the electronically archived
discussion focused primarily on logistics rather than physics version of this paper.
concepts. From among the students in her group, we chose Jan participates quite a bit, as do Veronica and Nancy.
to study Jan because she was neither a top nor a low- During the first part of the hour, Jan answers the tutorial
performing student, and because we believed that we were worksheet questions using mathematical reasoning.
seeing epistemological indications in her behavior that we Specifically, she reasons using the functional dependences
could explore with further analysis. (Again, since we are between force, charge, and field in the relations F = kQq/r2
trying to make an existence argument and a descriptive and E = F/q. While doing so, she makes a series of errors
analysis for a certain mechanism, rather than a generalizable that her group members and the teaching assistants catch
conclusion, a random representative sample isnt necessary.) and help her correct. After the students look at the forces,
The following semester, she agreed to undergo six the worksheet asks them to describe the dependence of the
interviews about student reasoning with one of us (AE), electric field on the magnitude of the test charge. Veronica
whom she had not met previously. Over the following year figures out that the field is independent of the test charge
the interviews were audiotaped and transcribed. Jan received and Jan agrees with her. However, a few minutes later, Jan
$10 per interview. The first four interviews consisted claims she doesnt get it, and explains her math reasoning.
primarily of Jan reasoning aloud in response to physics Veronica helps her and Jan eventually seems to understand.
questions about real-world objects and phenomena. The Veronica: Its the same ratio, cause the higher the test
final two interviews consisted of more formal, quantitative charge the bigger the force.
problems and of increasingly direct probes of Jans Jan: Right, so theyre proportional.
epistemology. [Veronica and Nancy digress for a while and then
Veronica explains the ratio idea to Nancy.]
IIB. Analysis of the data and interpretation of the Jan: I dont really get that, though. Cause like you
results know how you were saying that E = F/q. Cause like
they're saying that that's -
We reviewed the two usable hours of videotaped Veronica: Its force per test charge. So if you have a
classroom data and looked for instances in which big test charge its
epistemology seemed to affect Jans approach to learning Jan: I thought that meant that the electric field i s
and doing physics. From this review we developed a gonna get, if you have a small one, then the E-field i s
hypothesis about Jans epistemology and its causal gonna be big. But then if you have, cause you know,
relationship to her learning. To test this hypothesis, we cause like my understanding is that it says like
describes the ratio of the force felt by the test charge
attempted to explain her classroom behavior in non- and the strength of the test charge, right?
epistemological terms, by focusing on expectations (her Veronica: Yeah the q changes, but that makes the force
perceptions of what is rewarded in the course), confidence, different.
skills and habits, and the social dynamics in her tutorial Jan: So its not the E-field that changes but the force
group and in the interviews. We also used Jans written that changes.
homework to test predictions of the hypothesis we generated When asked to consider the E-field at a different
from the classroom and interview data. To quantify patterns distance, Jan claims it cannot change. This time Nancy
in Jans reasoning during the interviews, we developed and corrects Jans math error.
applied a coding scheme designed to pinpoint when she Jan: No, but what I am saying is E is equal to F over q,
used formal, classroom-taught reasoning versus everyday right? That doesnt include radius in it.
and intuitive informal reasoning; when she was sense- Nancy: But F includes, um, includes r. Jan: Because
making versus just trying to remember or throwing out further away is smaller.
ideas with little thought; and when she attempted to When asked to consider how the field changes when
reconcile different lines of reasoning. We describe this the source charge changes, again Jan returns to E = F/q, this
scheme more fully in Sec. IV. time reasoning as if q represented the source charge rather
than the test charge. Nancy tries to make a non-
III. CLASSROOM DATA mathematical argument, but Jan ignores her.
IIIA. Episode 1: The electric field tutorial Jan: If q is on the bottom.
Nancy: The point charge becoming smaller is the same
The earliest usable video segment shows the students thing as the distance becoming greater. It affects the
working for an hour on a tutorial developed at the outcome of this yet, so its the same thing.
University of Maryland to address student difficulties with Jan: If this [q] becomes smaller then that [F] becomes
the concept of the electric field. At this time in the bigger. Thats all it is, right?
semester, the four students (Jan, Veronica, Carl, and Nancy) Later, when the TA asserts that the field doesnt
have been working together for just a few weeks. Early in depend on the test charge, Jan protests that it does and
the tutorial, students find the electric force, F, exerted by a Veronica agrees with Jan, using the erroneous math
reasoning that Jan has been persistently using (reasoning
3 Submitted to the American Journal of Physics
with the formula E = F/q while ignoring the functional Nancy: Oh, thats a good point. I dont know.
dependence of F on q), and Jan verifies that this is what she Veronica: No, theres no mirror for it to reflect off.
is thinking. Veronica immediately catches her own error, Jan: But its not direct, right? Cause if it were direct,
but Jan does not comment, continuing to appear confused. then wouldnt it move up when we move the thing up?
Veronica: No. Because its going like this. When you
It is important to notice that Jan is using mathematical move it up. Its going through a hole. Well, I mean, I
reasoning that is sophisticated for this population. guess it is reflected light. Cause look, heres the hole.
Confronted with an equation, she does not try to plug and Its down here. Its going to go through the hole like
chug. Instead, she tries to extract information from the that. You know what I mean? Because its its position
functional dependencies of different quantities; shes good at relative to that hole.
attending to proportionalities and inverse proportionalities. Jan: But I mean, if it, if if it was direct, right, then the
However, despite her facility with mathematics, she makes light wouldnt come through if it wasnt aligned.
several math errors. Although she is corrected each time and Veronica: If it was direct, then it would go like this.
acknowledges her mistake, four times in a row, Jan makes [gesturing horizontally with hand from light bulb,
the same math error, repeatedly reasoning with the equation bangs into board above aperture to show that light
would not pass through]
E = F/q without considering the fact that the force F Jan: Right. Thats what I mean.
depends on distance and on the test charge, the very Veronica: It would hit, it would just hit the board.
quantities she is being asked to vary. Well, the light goes out, like that, like that. So, its
Jans problem here seems to be her failure to check her going to go, whatever path of that light its going t o
mathematical reasoning against her common sense hit right through the circle, its going to keep going
reasoning. Specifically, she does not link her math to a straight that way.
sense of physical mechanism in the way that Veronica does. Jan: Right.
For instance, it is highly unlikely that Jan would find Veronica seems confused about the reflecting and
sensible her prediction that the field does not change as the direct issues, but may realize that she and Jan might mean
distance changes, were she to think it through intuitively. It something different by direct. So Veronica then goes on
is unlikely that she would continue to ignore the functional to explain in everyday language what she thinks is
dependencies of F in the relation E = F/q were she happening. Jan indicates that she understands. The dialogue
attending to more than just mathematical accuracy each time continues with Veronica helping the group to understand
a group member corrects her. In the interviews, Jan how the model explains their observations.
considers it obvious that greater distance leads to weaker Nancy: How is it possible for the things to, like when
fields. She momentarily acknowledges her understanding of we have the two bulbs, for one little circle to create the
this common sense idea in class by emphasizing Because two . . .
Veronica: Because they are two different directions.
further away is smaller in response to Nancys comment. Ones going in like that and one is going in like that.
But then she ignores this idea, returning to reasoning with Jan: So you are saying that . . .
E = F/q in isolation, as if the common sense ideas were Nancy: But whats the normal direction of the light?
irrelevant. Jan seems almost not to even hear the next Cause thats what Im asking.
attempt by Nancy to make a common sense argument, and Veronica: It, it spans out, and whatever part goes
then says about the mathematical reasoning, Thats all it through that circle is the part were going to see.
is, right? Jan: [drawing as she talks] So the light is like that and
We think this behavior is both a window into Jans these are the rays, and the vector points that way will
epistemology and evidence of how it is affecting her go through the hole.
learning. In the following discussion, we propose that Jans Nancy: Okay, so then if you move it up, then its
going to be?
epistemology is causing her to act as if a wall separates Carl: So if here is the hole and the light is down here,
formal reasoning from informal, common-sense reasoning the light is going to go in the direction
and that this wall accounts for her lack of checking her Jan: Right, so like it has
mathematical answers against her informal understanding. [Nancy, Jan, and Carl talking, unintelligible]
Veronica: Really, its just normal.
IIB. The light and shadow tutorial Jan: All the rays are going like this. So, its kind of
Eight weeks later, the same four students are working like polarized.
on the Light and Shadow Tutorial .21 The group has several Veronica: Mmm, not really.
light bulbs, a board with several small apertures, and a large Jans behavior here is puzzling. She is engaged and
screen. They manipulate the bulbs and apertures and observe indicates her understanding of others explanations, but she
the changes in the pattern of light on the screen. As the is using more technical and mathematical language: rays,
students work through the activities, the worksheet guides vector, and polarized. Her use of this last term is
them to build a model of light to explain their observations. particularly striking. What is Jan doing in this exchange? Is
Early in the hour, the students observe that moving the bulb she, like the others, trying to make sense of her
to the right causes the bright spot on the screen to move to observations and her groups explanations, or is something
the left, the opposite direction. else going on? Veronica takes on this issue.
While attempting to answer the worksheet question, Veronica: Its just, well, its just, guys youre making
it, youre trying to make it more difficult. Its just, the
What do your observations suggest about the path taken by light goes out. It only goes through that one circle.
light from the bulb to the screen?, Jan initiates a So, obviously, if it is down here, and Im looking
discussion of physical mechanism. through that circle. Look, youre sitting down here.
Jan: So does that mean that the path is not a straight Youre looking at this big cardboard. Youre looking
line? . . . Does that mean its reflecting?

4 Submitted to the American Journal of Physics

 through that little circle. All youre going to see i s affected by her epistemology. In particular, Jans
whats up there. Its a direct line. epistemology divides the reasoning that can be used to
In accusing the group of trying to make it more understand physical phenomena into two disparate
difficult than it really is, Veronica suggests that something categories: formal, technical reasoning and everyday,
other than simple sense-making is going on. Why would intuitive reasoning. Between these two types of reasoning is
the group and Jan in particular do this? Fortunately, Jan a barrier (a wall, metaphorically) that keeps Jan from
tells us what she is up to. looking for connections between ideas from the different
Jan: Look, I see what youre saying, alright. But, Im sides.
just trying to make it like physics- physics-oriented. Jan is quite adept at some types of formal reasoning.
[laughs]
An open question for us at this point is how adept is she at
Veronica: It is, it is physics-oriented. Thats just the
way it is. informal reasoning. However, we claim that even if Jan were
Jan: Alright. to use informal reasoning or accept that of her peers, her
Jans behavior during this episode seems puzzling at tendency not to link the formal and informal would
first, but Jan is quite explicit in describing her motives. continue to cause her difficulties.
Im just trying to make it like, physics- physics-oriented. In the following, we evaluate and expand our
Her words and her behavior reveal her epistemology and its hypothesis in two ways: by analyzing additional interview
impact on her choices. Although she desists when data, and by exploring several counter-hypotheses (alternate
challenged by Veronica, Jan strongly implies that she is not interpretations that dont involve this epistemological
looking for an informal, common-sense explanation. mechanism).
We should notice that, in searching for the more IVB. Results from the interviews: Jan uses everyday
formal explanation and rejecting the common-sense one, Jan reasoning more often but still shows evidence of the
still claims to understand what the group has been epistemological barrier between formal and everyday
discussing. (I see what youre saying.) It may be that Jan reasoning.
is considering the intuitive explanation but searching for
more technical language for the worksheet. Alternatively, it Given Jans behavior in the tutorials, what stands out
may be that Jans rejection includes passing up an most in the interviews is her willingness to approach
opportunity to understand intuitively. In isolation, her problems using the kind of everyday knowledge and
behavior here cannot distinguish these two possibilities. intuitive reasoning that we see her rejecting in tutorial.
However, her homework sheds light on this issue by Often she immediately responds to a question using
revealing a lack of understanding on Jans part. The everyday/intuitive reasoning. Other times, if a formal line of
assignment asks students to apply the model for light they reasoning doesnt work for her, she switches to
just constructed. When asked to predict the shape and size everyday/intuitive reasoning. In the following example,
of a shadow, Jan draws straight lines indicating the rays of from the third interview (3:243, interview 3, line 243), Jan
light from the bulb. However, her rays reach only the is asked whether it matters if you choose a long wrench or a
blocking object and do not extend all the way to the screen. short wrench when loosening a stuck bolt.
She does not attempt to answer the question further and Jan: Well I think it matters, I definitely think i t
when asked to explain her prediction writes, I dont know matters. Because one of the things that we did i n
physics was torque, and al-, you know when you have
how else to think about it except for the rays from the light to draw like a lever arm? And um, I think it was T = l r,
bulb. This directly contradicts the understanding she is that what it was? I dont know.
claims to have had during the discussion. If she really Interviewer: Whats l and whats r?
understood what Veronica was saying, she would have Jan: Like, like, okay this is like the pivot point, you
another way of thinking about it, the common-sense way know like here, and so you would draw like a line, and
Veronica keeps describing. (Look, youre sitting down this is like the place from which you are going t o
here. Youre looking at this big cardboard All youre change it, you know, or like, you know
going to see ) Veronica, using this understanding, gives Interviewer: This is like where youre holding
complete and correct qualitative and quantitative responses Jan: And you draw like some line here. I cant
on the homework, whereas Jans formal labeling of the remember exactly. I shouldve learned physics better.
I should keep these things in my mind. So I think like
phenomena has not given her enough real understanding for the further you go out, you know, the easier, thats not
either task. to say you go all the way out, but as things, its better
When we combine this information from her if you have it here than if you have it here. I think
homework with the statements she makes, the what I can think about is like a door. Youve got like
epistemological implications become clearer. As in the first the hinge here and you know youve got like the
episode (with E = F/q), Jan behaves as if common sense swinging door. I think if you push here [closer to the
reasoning is a separate endeavor from formal hinge] is the door is going to feel more heavy that if
(mathematical or technical) reasoning, and that she you push it out here [farther from the hinge].
considers only the latter to be physics-oriented. Jan starts out here using formal reasoning, trying to
apply the concept of torque and the (incorrect) formula T =
IV. DISCUSSION Ir. However, she doesnt seem to understand the formula or
IVA. Preliminary hypothesis: Jans epistemology places remember how to use it. So then, after implying that she
a barrier between formal and everyday reasoning doesnt trust her formal physics reasoning in this case, she
switches gears and tries her common sense instead.
From these two episodes we make our preliminary Specifically, she comes up with an everyday experience
interpretation. Jans learning behavior seems to be strongly (pushing a swinging door) that helps her solve the problem.
5 Submitted to the American Journal of Physics
In contrast to her behavior in the tutorial, we find that Jan is force there, right? And if you just add force in the
far more likely to use everyday/intuitive reasoning to solve same direction, then its probably just going to add
physics problems in interviews. This does not mean that up
Jans preference for formal versus everyday/intuitive Interviewer: So its like your adding onto something
which is already there, as opposed to--?
reasoning stems entirely from non-epistemological origins. Jan: Trying to like oppose it, or to do something else.
Rather her epistemology might depend on context, a context Interviewer: Right, so by that reasoning the very
dependence we can account for in a resource-based (as worst thing that you could do is like hit it the
opposed to belief-based) cognitive framework.23 In the opposite way, going to, trying to beat it into going
interviews, Jan may (consciously or unconsciously) activate the other way.
an epistemological resource that guides her to use Jan: Right, right. I mean, this would probably like
everyday/intuitive reasoning, a resource that stays off in take a lot of energy out of you, but I think it would be
the classroom context. good. [Laughing]
This context-sensitivity of Jans epistemology is the Interviewer: Right, so each time you hit it, it would
subject of a separate paper.24 We believe that this context- start going a little bit higher, and to that effect. Cool,
OK, as long as were on this big bowling ball
sensitivity is good evidence against the notion of pendulum, so lets say that you are done whacking it,
epistemology as constructed of consistent, unitary, and now its swinging higher, and--
context-insensitive beliefs, but we will not argue that point Jan: Its also like when you, have a person on a swing,
in this paper. Our goal here is to make a detailed well actually but when the person is on the swing you
plausibility argument for the impact of epistemology on actually hit them when theyre up here, but you push
learning. That argument relies on the detailed refutation of them that way, so [Drawing a picture of a swing at
counterarguments. Thus we need to deeply analyze an its highest point.]
aspect of Jans epistemology that is fairly consistent across Interviewer: Huh, what do you make of that?
the two contexts from which our data is based. Although we [Laughing.]
do not take a belief approach, we will spend most of the Jan: I dont know [Laughing.] well, I mean its a little
different when a person is on a swing because its hard
remaining discussion describing an aspect of Jans to get underneath them when theyre like in this
epistemology that is somewhat belief-like in these two position.
contexts. Jan starts by reasoning that you should whack the ball
Our data show that Jans view of the separateness of when its going fastest (at the lowest point in the swing)
formal and everyday/intuitive reasoning is consistent across because you want to add force to force. We code this as
contexts. To get at this issue in the interviews, we look at intuitive rather than formal reasoning; although she says the
the frequency with which Jan checks and reconciles multiple word force, she seems to be using it in an informal,
lines of reasoning. In the wrench example, for instance, Jan colloquial way, as a term that expresses the energy or
had an opportunity to reconcile her idea that torque is motion of an object. We think she is reasoning,
relevant with her idea that the wrench problem resembles intuitively, that you want to add whatever youre going to
pushing a door. As teachers, we want Jan to ask herself, add whenever theres the most of the target stuff already
Does torque and the equation I thought of have anything to there.
do with the door example Ive given? But Jan does not do Jans second line of reasoning comes from her
that. We claim that her failure to do so stems in part from everyday experience. She thinks of pushing a person on a
the barrier she places between formal and everyday/intuitive swing as an everyday instantiation of the problem at hand.
reasoning; if those two kinds of reasoning arent connected, But then she notices a conflict. When you push a person on
it makes no sense to try to reconcile them. a swing, you push at the high point of the swing, not the
In contrast, Jan often reconciles two lines of reasoning low point she has decided would be best for the bowling
when theyre both formal or when theyre both ball. Notice that rather than ignore the conflict by
everyday/intuitive, For example, consider this exchange abandoning one or the other line of reasoning, Jan
about a bowling ball swinging on a chain (1:33). reconciles. She notes that, although we push people at the
Interviewer: Imagine that hanging from the ceiling b y
high point, this may not indicate the most efficient
a chain is a bowling ball, and somebody gets i t
swinging back and forth, like a pendulum. And approach; we cant push the swing at its low point because
youve got like a stick or a mallet and youre allowed its hard to get underneath them when theyre like in this
to whack as hard as you want, five times. And the position.
purpose of you whacking it is to get it swinging as Jans successful reconciliation in the bowling ball
high as possible, so my question is, how--what would example but lack of reconciliation in the wrench example is
your strategy be for the whacking? How, and where i n consistent with our epistemological hypothesis. She can and
its swing would you want to whack it? will reconcile when she doesnt have to overcome the barrier
Jan: I think Id probably want to whack it when its between formal and everyday/intuitive knowledge. This
kind of like on its way up, and whack it like from the pattern emerged robustly in the coding results.
side going up, do you know what I mean?
Thirty-six problems from the first five interviews were
Interviewer: Can you draw me a little--?
Jan: Yeah, like the chain is this way, and so its on its coded. (Interview 6 had a different format.) Each line of
way up, right? [Drawing a picture of a pendulum with reasoning in each problem was coded as involving
the bowling ball at the lowest point.] So Id probably everyday/intuitive reasoning or formal reasoning. In the
whack it with the mallet right here. bowling ball example, for instance, we coded two lines of
Interviewer: I see, and whats the, what sort of everyday/intuitive reasoning: the adding force to force
triggered you to think that? argument and the analogy to person on a swing.
Jan: Well, its already on its way up, so theres already Altogether we coded 106 lines of reasoning. Jan used
6 Submitted to the American Journal of Physics
everyday/intuitive reasoning three times as often as formal Within-type Reconciliation Opportunities
reasoning, 71 versus 22. (We discuss the remaining 13
instances in the following.) Idea 1. Idea 2.
We then coded when Jan did or didnt reconcile given Everyday/Intuitive Everyday/
an opportunity (generally a conflict or lack of connection Reasoning Intuitive
between two lines of reasoning.) Reconciliation
opportunities involving two formal lines of reasoning or Idea 1. Formal Idea 2.
two everyday/intuitive lines of reasoning were coded as Reasoning Formal
within-type. For instance, in the swinging bowling ball
example, Jan reconciles within-type between adding force Results: 13/28 opportunities reconciled (46%)
to force and analogy to person on a swing.
Reconciliation opportunities involving formal versus
everyday/intuitive lines of reasoning were coded as Between-type Reconciliation Opportunities
between-type, as illustrated in the wrench example with 1. Everyday/ 2. Everyday/
the torque equation versus the analogy to pushing a Intuitive Intuitive
door. Of the 36 coded reconciliation opportunities, Jan
reconciles about 40% of the time (14 reconciles.) To test our 1. Formal 2. Formal
hypothesis about Jans epistemology, however, we must
look for differences between her tendency to reconcile within Results: 1/8 opportunities reconciled (13%)
type versus between types.
Within type (28 coded opportunities), Jan reconciled Figure 1. Reconciliation opportunities occur when the
about half the time (13 reconciles.)25 Most are unprompted, relationship between two ideas Jan is using is ripe for
and the others involve only mild prompting, as in the examining. For instance, in the pendulum problem described
above, the line of reasoning about adding force where the
bowling ball problem when the interviewer says, What do movement was greatest conflicted with the observation that
you make of that? By contrast, between types (8 codings), you push a person on a swing at the turning points. If we call
Jan reconciled only once.26 Although the number of codings the adding force idea Idea 1 and the swing idea Idea 2, we
involved does not give very reliable statistics, the dramatic see that in the diagram this is a reconciliation opportunity
difference between her within-type and between-type represented by the topmost arrow, since Idea 1 and Idea 2 both
reconciliation rates, 46% versus 13%, supports our use everyday/intuitive reasoning. Since Jan does reconcile
attribution of a barrier between formal and these two ideas, we count this in the results as a within-type
everyday/intuitive thinking in Jans epistemology. A reconcile. The differences in the relative rates of reconciliation
summary of our findings is included in Figure 1. in the within-type and between-type categories support our
At times Jans reasoning seems to be neither hypothesis of the wall Jans epistemology places between the
two reasoning types.
everyday/intuitive nor formal but rather a hybrid of the two.
Thirteen of our 106 lines of reasoning were coded as hybrid. Our argument is that, due to Jans epistemology, she
Hybrid reasoning is not a simple mix of the two other does not search on her own for the connections between the
types. Rather, it is a type of reasoning in which the two types of reasoning. As mentioned, in the interview
everyday/intuitive and formal are already integrated. (By coding, we found only one between-type reconciliation out
contrast, between-type reconciliation is an action to address of eight opportunities. This finding becomes more
conflicts between everyday/intuitive and formal ideas that compelling and interesting if we look in detail at this one
are not already integrated.) For the hybrid reasoning to instance. Jan is asked to explain the motion of a pencil that
occur, the ideas must have been integrated somehow at has been thrown into the air (3:274). She starts with a line
some point in the past, probably involving multiple of reasoning that we coded as hybrid, integrating her
between-type reconciliations. This type is rare in beginning everyday/intuitive ideas about a pencils motion with formal
physics students but is common in practicing scientists. concepts of velocity. Jan then discusses the influences of
The existence of a hybrid reasoning type seems to gravity and air resistance and the force from your hand in
contradict our hypothesis that Jan views formal and a hybrid fashion, but then gets confused about what happens
everyday/intuitive reasoning as unconnected. However, our to the force of the hand: Does it permanently die out at the
hypothesis claims not that it is impossible for these types peak or is it still influencing the motion when the pencil
of reasoning to be integrated for Jan, but rather that her comes back down? At this point the interviewer pushes her
epistemology generally prevents her from searching for these very strongly not to let that go, but to try to figure out what
connections on her own. We found that most of Jans happens to the force of the hand. Jan brings in her last,
integrated formal and intuitive reasoning can be explained in formal line of reasoning, discussing kinetic and potential
a way that is consistent with our epistemological energy, and uses it to explain how the force that you added
interpretation. The instances of hybrid reasoning we found would be energy that you gave it, which is being
were either instances where the connection between the interconverted in the system [between kinetic and
intuitive and formal were exceedingly transparent (almost potential], and hence, the force (kinetic energy) from the
unavoidable) or instances where the connections between the hand is still there as the ball falls even though it died away
intuitive and formal had been stressed strongly and temporarily at the peak (3:307). She is reconciling two lines
repeatedly in the course. We believe that this effort by the of reasoning. However, this between-type reconciliation is
course instructor on particular topics helped Jan scale the weak in the sense that it was strongly prompted and
epistemological wall that normally would have prevented involved a hybrid and a formal line of reasoning rather than
these connections. fully scaling the barrier between formal and
7 Submitted to the American Journal of Physics
everyday/intuitive. Given that Jans lone instance of a One could argue that Jan simply lacks skill with
between-type reconciliation is weak, her tendency to informal reasoning in physics, perhaps for lack of practice.
reconcile only within type becomes a more robust coding This might explain, for example, why she would fail to
result. catch her math errors were she actually checking her math
To test the reliability of our coding, we trained an against her informal reasoning. However, weak informal
education researcher not previously involved with the reasoning cannot explain why she rejects as not physics-
project and then had the coder apply our four coding steps oriented the common-sense reasoning of the rest of the
to a subset of our data, some randomly chosen problems, group. Furthermore, it is clearly not the case that Jan lacks
and others deliberately culled from cases we ourselves this skill. In interviews, Jan proficiently uses common-
considered particularly difficult. The external coder saw the sense, informal reasoning to work out physics problems.
same general trends as we did in the preponderance of
iii. Jans difficulties are not due to lack of skill at
everyday/intuitive over formal or hybrid reasoning and in
checking.
the greater relative rate of within-type versus between-type
reconciliation. To further validate this result, we then Perhaps Jan fails to check her interpretations of E =
described the specific reconciliation opportunities from our F/q, and fails to check how well she actually understands
original codings. For each case, the coder decided whether Veronicas model of light, because she has a general
Jan reconciled, compromised, checked, or didnt reconcile tendency not to check her reasoning, or because shes not
at all. The external coders codings matched our codings good at checking. Again, the interviews suggest otherwise.
80% of the time.27 In the 28 within-type reconciliation opportunities, there
In summary, in comparison to the classroom data, the were also 6 instances we coded either as "checking only,"
interviews yielded opposite results regarding Jans when Jan makes sure one line of reasoning agrees with the
inclination to use everyday/intuitive versus formal answer yielded by another line of reasoning (without
reasoning. The interviews also provide evidence that these reconciling or even comparing the two lines of reasoning),
two types of reasoning can be integrated by Jan under or as "compromise," when Jan glosses over rather than
certain circumstances. However, both sets of data are robust directly addresses a contradiction she notices between two
in showing Jans tendency not to look for connections lines of reasoning. Our point is that, in 19 of 28 instances
between these two types of reasoning. What the interviews (68%), Jan notices and in some way addresses a connection
cannot establish completely is that this wall Jan places between two lines of reasoning within type, proving that
between everyday/intuitive and formal reasoning accounts she often notices the potential tension between different
for some of Jans learning difficulties in the tutorials. To lines of reasoning and that she has the skills to address the
further that argument, we must refute alternative tension.
explanations. Jans tendency to check and reconcile isnt confined to
the interviews. For instance, later in the electric field
IVC. Refutation of alternative explanations: Jans
tutorial (after the portions presented previously), the group
difficulties cannot be accounted for without
considered a scenario in which a positive test charge is
epistemology.
pushed directly toward a positive source charge. Does that
We now discuss other possible contributors to Jans push force do positive, negative, or zero work? Veronica
behavior in the tutorials. Although some of these factors reasoned that because the potential energy becomes greater,
play a role in Jans actions, no combination of these effects, the change in work is going to be negative, because the
without epistemology, can account for all that we observe. The work in, like, an electric field, its the opposite, the
opposite of the change in potential energy. Jan
i. Jans difficulties are not due to lack of facility or immediately wants to check this conclusion using an
confidence with mathematics. analogy the professor pointed out between electrostatic
We might consider attributing Jans repeated troubles forces and gravity. She correctly notes that moving the
using E = F/q correctly in the electric field tutorial to a lack point charge is analogous to changing the height of a mass,
of facility or confidence with mathematics. However, showing that the potential energy changes. In this instance
evidence from that episode and from the interviews suggests in which she has two formal lines of reasoning (for
otherwise. We have already noted that Jan reasons in a example, reasoning about electric field and potential versus
sophisticated manner with the equation E = F/q, using the a professor-supplied analogy to gravitation), Jan wants to
functional dependencies between variables in this case reconcile (although the group goes in a different direction
proportionalities and inverse proportionalities to draw before Jan can make progress).
information from the equation. iv. Jans actions cannot be fully explained by an
Perhaps even more striking, Jan shows the ability to expectation that only formal reasoning will be rewarded
use mathematics intuitively. She refers to herself as a in the class.
proportions person (2:194) and can formulate her own
equations to express her intuitive reasoning (for example As previous studies show,19,20 a students expectations
1:128, 5:70), although she still differentiates these about what will be rewarded in a physics class need not
equations ontologically from the classroom equations. align with her epistemological views about what constitutes
Furthermore, she explicitly states her confidence in her use learning and understanding. In tutorial, perhaps Jan focuses
and learning of mathematics (for example 6:144). on formal reasoning to the exclusion of everyday/intuitive
reasoning because she thinks formal reasoning is what the
ii. Jans difficulties are not due to lack of skill with course requires. This counterargument also would explain
everyday reasoning in physics. why Jan is much more willing to use everyday/intuitive
8 Submitted to the American Journal of Physics
reasoning in the interviews, which were not part of the It turns out that Jans lack of confidence with
physics course. However, formal expectations cannot everyday/intuitive reasoning in physics does indeed have a
explain why Jan rejects everyday/intuitive reasoning that large impact on her behavior, as we have discussed. But it
could help her use E = F/q correctly. Even if exams reward is Jans epistemological stance that a barrier separates formal
formal reasoning only, they dont reward incorrect formal from everyday/intuitive reasoning that determines how she
reasoning. And after being corrected, Jan acknowledges that deals with her lack of confidence. She feels that her
she is using E = F/q incorrectly. Her classmates suggest perceptions are imperfect and not to be relied upon.
some common-sense reasoning that could help Jan apply E It always seems like, you know, theres like a trick that
= F/q correctly, but Jan acts as if that reasoning is Ive missed, you know, something that Ive overlooked
irrelevant. or something that I havent thought about. (5:138)
Similarly, in the light and shadow tutorial, formal She gives static friction as an example, explaining that
expectations could explain why Jan rejects her groups although you have ample experience with pulling things,
everyday/intuitive model of light. But once again, you might fail to observe the need to pull harder at first
expectations alone cant explain why Jan doesnt engage in (5:178). There are several ways a student might deal with
learning just enough of Veronica-style everyday/intuitive the unreliability of perception-based everyday/intuitive
reasoning to apply formal resources correctly. Formal reasoning. One response would be to use the formal and
expectations alone, unsupplemented by our epistemological everyday/intuitive in conjunction, incorporating the two to
interpretation, cannot explain Jans behavior in tutorial. make a more robust understanding, thereby avoiding the
pitfalls of relying solely on imperfect intuitions and
v. Jans confidence in her informal reasoning does perceptions. But Jan does not take this stance. Instead, she
have an impact, but it can only account for her generally keeps (unreliable) everyday/intuitive reasoning
behavior in concert with her epistemology. separate from (reliable) formal reasoning. Our point is that
Another counterargument is that Jan has low confidence Jans response to the unreliability of everyday/intuitive
in her ability to use everyday/intuitive reasoning in physics, reasoning is driven in part by epistemology. Instead of
and she therefore hides behind formal reasoning during seeing problematic everyday/intuitive reasoning as refinable
group work. This lack of confidence could explain not only and hence reconcilable with formal reasoning, she sees the
her nervous, perhaps self-deprecating laughter when she two kinds of reasoning as too separate to inform each other.
explains her quest for a more physics-oriented At one point, Jan laments that physics is unlike
explanation, but also her willingness to use common-sense chemistry, because chemistry is
kind of totally new, you know, like you kind of have
reasoning in the interviews, where she perhaps feels safer a fresh brain, I mean, youre talking about
away from her peers and away from grade pressure. molecules and things you cant really see, you know
We can quickly rule out one version of this so you have to kind of start fresh and I think so, so i t
counterargument, the idea that Jan hides behind formal makes it a little easier to think. (5:186).
reasoning during group work not because she lacks faith in Jan would rather start with a blank slate (fresh brain) than
her ability to use everyday/intuitive reasoning, but because try to refine and build on her own ideas. Jans epistemology
shes afraid of stepping out on a limb by expressing her and confidence are entangled and mutually reinforce one
ideas publicly. According to this argument, Jan feels safer another. Epistemology tells her the two domains are
sticking to more objective, formal reasoning. This version separable. Her confidence says to reject one, and then she
of the counterargument fails, however, because it cannot only experiences success in the one domain, which leads to
explain why Jan rejects other students everyday/intuitive reduced confidence in the other domain, and so on.
reasoning in the light and shadow tutorial, or why she has Although it is clear that confidence is playing a role, as an
such trouble using other students qualitative ideas to help alternative explanation it cannot stand alone, because it
her interpret E = F/q correctly. relies on epistemology to have the effects we observe.
Another version of the confidence counterargument has
more traction: Jan avoids engaging in everyday/intuitive IVD. Summary: The epistemological mechanism is
reasoning during group work largely because she lacks faith plausible
in her ability to learn and understand physics in those
terms. This could help to explain her resistance to using We believe that we have made a strong case that Jans
everyday/intuitive reasoning when interpreting E = F/q as learning difficulties in the two tutorial episodes stem in part
well as her quest for a more physics oriented model of from her epistemology, in particular from the barrier Jan
light. It would also explain why Jan seems so much bolder places between formal and informal reasoning. This barrier
and more confident with her group when pursuing formal prevents her from searching for connections between these
explanations such as an analogy between gravitational and two types of knowledge. By examining supplementary data
electric potential. The interviews further buttress this and evaluating alternative explanations,28 we have
counterargument. Perhaps because the interviewer repeatedly established a highly plausible argument for this part of our
emphasized that he studies student reasoning and doesnt hypothesis, that Jans epistemology has a direct, causal
care whether her answers are correct, Jan willingly uses effect on her learning.
everyday/informal reasoning. Even then she often hedges her In testing our hypothesis, we have also discovered
reasoning with qualifiers such as maybe and it could be, more detail and subtlety to Jans epistemology and its
and on several occasions she says her everyday way of impact on her learning. We found that Jans epistemology
thinking doesnt work reliably in physics. (Well give an does not prevent her from using everyday/intuitive
example in the following.) reasoning in some contexts (for example, interviews) and
that there is a deep entanglement between Jans

9 Submitted to the American Journal of Physics

epistemology and her distrust of her intuition and well, Kanim noted the extreme, iterative effort needed to
perceptions. Another intriguing issue is the existence in develop them; the bridging worksheets seem to keep
Jans reasoning of a hybrid form in which formal and exposing new student difficulties. For instance, even when
everyday/intuitive reasoning are integrated, generally when students could answer an intuitive qualitative question
explicitly taught and sanctioned by the professor. Clearly about resultant vectors, they had trouble applying that
this integration is not impossible for Jan. Yet her knowledge to formal vector addition.
epistemology leads her only rarely to strive for this In our view, an epistemological barrier between
integration on her own, and even to resist it in many everyday/intuitive and formal knowledge can help us
situations. understand why students have such trouble bridging those
One hallmark of an epistemological effect is when we two types of knowledge, even when strongly supported.30 If
see students failing to use skills or knowledge they clearly we are right, then Kanims bridging worksheets might
possess. Jans skills, her abilities, her store of ideas -- none become even more effective by explicitly addressing this
of these are the limiting reagent for her learning in these epistemological issue. The worksheets could include
episodes. She is capable and fluid with mathematical, activities and reflection questions designed to help students
technical, and everyday, common-sense reasoning. She is realize that their thought processes sometimes incorporate
capable of checking her understanding and reconciling such a barrier and that some of their ah hah moments of
inconsistencies. She is capable of working through difficult understanding occur when they scale or tear down the
problems for which she has very little relevant formal barrier. When students epistemologies become more
knowledge. Despite all these strengths, her epistemology aligned with the goals of the bridging worksheets, students
sometimes gets in the way of her learning. might become better at spotting and addressing new
difficulties they encounter. For instance, consider a student
IVE. Implications who has learned to expect that her intuitive mathematical
knowledge for instance, her common-sense ideas about a
Our case study has built on previous research into hiker who walks two miles north and then three miles east
epistemology and learning to show, in causal detail, how should mesh with formal mathematical tools. Especially
epistemology can have a profound effect on the learning- when supported, that student would probably use her hiker
relevant behavior of students. This is important for several knowledge when figuring out or making sense of formal
reasons. First, this type of analysis provides supporting vector addition rules. Consequently, she could resolve her
evidence that the effects of epistemology on learning vector difficulties more quickly than would otherwise be the
outcomes observed in correlational studies are in fact causal. case.
Learning also most likely plays a causal role in the We lack space to discuss the details of
development of a personal epistemology; but making a epistemologically-focused curricula and teaching practices
strong case for local causality in one direction is an our research group has used.31 Key components include
important first step in understanding the complex interplay listening for students epistemological strengths and
between the two. difficulties during office hours and recitation section; using
For the curriculum developer and the classroom tutorials and interactive lecture demonstrations designed to
teacher, understanding how epistemology affects learning tap into productive epistemological resources (for sense-
or just keeping in mind that epistemology affects learning making and consistency) and highlight learning strategies;
has broad implications. In designing curriculum, and using some reflective tutorial questions, homework
developers must attend to epistemological as well as questions, and class discussions to focus students attention
conceptual, social, or affective factors. For instance, on their approach to learning.
epistemology can help us understand why a piece of
curriculum optimized to address conceptual difficulties is V. CONCLUSION
ineffective for some students. Curriculum developers can
take up the challenge of helping students associate their Our biggest challenge as instructors is listening to our
productive epistemological resources with the activity, the students, responding to their difficulties, and facilitating
course, and the discipline. their use of productive cognitive resources they possess. In
There are also some implications that link specifically diagnosing student learning, we must consider their
to the epistemological resource of a barrier between formal strengths and difficulties of an epistemological nature.
and everyday reasoning. Numerous physics teachers and Specifically, we must learn to identify the epistemological
researchers have noticed that students rarely hook up their resources that students possess and to understand which
conceptual/intuitive knowledge to their formal knowledge resources they are using during the learning process, so that
and problem-solving techniques. An epistemological barrier we can help them to choose the more productive approaches
helps explain why this disconnect exists for many students, to learning. Our strong argument about the plausibility of a
and what can be done about it. causal mechanism by which epistemology can affect
For instance, Kanim found that even when students learning gives more reason than ever to believe that
gain a deeper, more intuitive conceptual understanding of a epistemological interventions could lead to better conceptual
topic (such as batteries and bulbs), they dont apply that learning.
knowledge to traditional quantitative problems (for
example, about circuits).29 To address this gap, Kanim ACKNOWLEDGEMENTS
started creating topic-by-topic bridging worksheets designed
We would like to thank Joe Redish, David Hammer, and
to help students connect their conceptual knowledge to their
the members of the Physics Education Research Group at
formal reasoning. Although many of these worksheets work
the University of Maryland for playing a large part in this
10 Submitted to the American Journal of Physics
 19
investigation, with special thanks to David May and D. Hammer, Two approaches to learning physics, Phys. Teach. 27,
664-671 (1989).
Rebecca Lippmann. This work was done in part with the 20
A. Elby, Another reason physics students learn by rote, Am. J. Phys.
support of NSF grant REC-0087519. We would also like 67 (7), S52-S57 (1999).
to thank advisory board members John Fredriksen and Paul 21
D. R. Sokoloff and R. K. Thornton, Using interactive lecture
Feltovich for useful suggestions early in the analysis. demonstrations to create an active learning environment, Phys. Teach.
Thanks are also due to several anonymous reviewers and 35, 340-347 (1997).
22
The tutorials used were a mixture of University of Washington tutorials
many members of the American Association of Physics [L. C. McDermott, P. S. Schaffer, and the Physics Education Group,
Teachers whose questions helped us refine our focus. Tutorials in Introductory Physics (Prentice Hall, Upper Saddle River, NJ,
2002)] and tutorials written by the University of Maryland Physics
Education Research Group. Tutorials lead students through a structured
*
Current address: Science Education Group, Department of Physics, series of questions and activities usually aimed at developing qualitative
Astronomy, and Geosciences, Towson University, Towson, MD 21252 conceptual understanding (rather than traditional problem solving) and
1
L. C. McDermott and E. F. Redish, Resource letter PER-1: Physics often focusing on common student difficulties such as the difference
education research, Am. J. Phys. 67, 755-767 (1999). between velocity and acceleration.
2 23
D. Hammer, Epistemological beliefs in introductory physics, Cog. We will refer to the archival version of the appendix here. The
Instruct. 12 (2), 151-183 (1994). appendix will contain transcripts and more detailed arguments.
3 24
E. F. Redish, J. M. Saul, and R. N. Steinberg, Student expectations in See D. Hammer, Student resources for learning introductory physics,
introductory physics, Am. J. Phys. 66, 212-224 (1998). Am. J. Phys. 68 (7), S52-S59 (2000), and D. Hammer and A. Elby, On
4
W.-M. Roth and A. Roychoudhury, Physics students epistemologies and the form of a personal epistemology, in Personal Epistemology: The
views about knowing and learning, J. Res. Sci. Teach. 31 (1), 5-30 Psychology of Beliefs about Knowledge and Knowing, edited by B. K.
(1994). Hofer and P. R. Pintrich (Lawrence Erlbaum, Mahwah, NJ, 2001), pp.
5
For clarity, we draw this distinction between metacognition and 169-190.
25
epistemology. Metacognition refers to awareness and manipulations of A paper about the context-sensitivity of Jans epistemology is currently
ones knowledge, such as monitoring ones understanding while reading in preparation.
a textbook and checking an answer for plausibility. Epistemology, in In addition to those 13 strong reconciles, Jan also displayed two
contrast, refers to ideas about the nature of knowledge and learning instances of checking-only and four instances of compromise. One
viewing textbook physics knowledge as something that can be example of checking-only is when Jan made sure one line of reasoning
understood or checking as a way to reconcile qualitative and agreed with the answer yielded by another line of reasoning, without
quantitative knowledge. A students epistemology can help to drive reconciling or even comparing the two lines of reasoning. Compromises
which metacognitive strategies she does and does not use. involve comparing two lines of reasoning but glossing over rather than
6
For a review, see B. K. Hofer and P. R. Pintrich, The development of addressing the contradictions. We mention these codings because they
epistemological theories: Beliefs about knowledge and knowing and involve at least some kind of contact between two kinds of reasoning, a
their relation to learning, Rev. Ed. Res. 67 (1), 88-140 (1997). contact noticeably absent in Jans reasoning in the electric field tutorial
7
M. Schommer, Epistemological development and academic episode.
26
performance among secondary students, J. Ed. Psych. 85 (3), 406-11 She also displayed two instances of between-type checking-only, a
(1993). coding category explained in Ref. 25. Even if were generous about
8
M. Schommer, A. Crouse, and N. Rhodes, Epistemological beliefs and what is counted as reconciliation and therefore include compromises
mathematical text comprehension: Believing it is simple does not make it and checking-only instances in our final tallies, Jans within-type
so, J. Ed. Psych. 84 (4), 435-443 (1992). reconciliation rate of 68% (19 of 28) is still much higher than her
9
N. B. Songer and M. C. Linn, How do students views of science between-type reconciliation rate of 33% (3 of 9), supporting our
influence knowledge integration?, J. Res. Sci. Teach. 28 (9), 761-84 hypothesis of a barrier in Jans epistemology between formal and
(1991). everyday/intuitive reasoning.
10 27
G. Qian and D. E. Alvermann, The role of epistemological beliefs and This paper and its appendices are available at
learned helplessness in secondary school students learning science http://www.arxiv.org/abs/physics/0411007.
28
from text, J. Ed. Psych. 82, 282-292 (1995). Readers may wonder if some of the behavior we see on the videotape is
11
D. B. May and E. Etkina, College physics students epistemological due to the students nervousness at being taped. Although the students
self-reflection and its relationship to conceptual learning, Am. J. Phys. agreed to tapings and the cameras are in plain view, we have evidence
70 (12), 1249-1258 (2002). that they seem to forget about them. For example, a few minutes after
12
K. Hogan, Relating students personal frameworks for science Jans physics-oriented comment, the group starts to discuss things
learning to their cognition in collaborative contexts, Sci. Ed. 83, 1-32 more personal than we believe they would discuss in front of an
(1999). instructor or researcher. (We did not watch or transcribe this portion of
13
J. Ryder and J. Leach, University science students experiences of the video.)
29
investigative project work and their images of science, Int. J. Sci. Ed. S. Kanim, Connecting concepts to problem-solving, Proceedings of
21 (9), 945-956 (1999). the 2001 Physics Education Research Conference: Research at the
14
R. Millar, F. Lubben, R. Gott, and S. Duggan, Investigating in the Interface (Rochester, NY, July 25-16, 2001), pp. 37-40.
30
school science laboratory: Conceptual and procedural knowledge and We dont dispute Kanims claim that new conceptual difficulties keep
their influence on performance, Res. Pap. Ed. 9 (2), 207-248 (1994). appearing. But we think thats only part of the problem. In many cases,
15
M. Taylor-Robertson, Use of videotape-stimulated recall interviews to the epistemological barrier between formal and everyday/intuitive
study the thoughts and feelings of students in an introductory biology knowledge may make it especially difficult for students to spot and
laboratory course, unpublished masters thesis, Cornell University, resolve what would otherwise be fairly easy conceptual problems.
32
Ithaca, NY, 1984. D. Hammer and A. Elby, Tapping epistemological resources for
16
K. M. Edmondson, The influence of students conceptions of scientific learning physics. J. Learn. Sci. 12 (1), 53-90 (2003); A. Elby, Helping
knowledge and their orientations to learning on their choices of learning physics students learn how to learn, Am. J. Phys. 69 (7), S54-S64
strategy in a college introductory level biology course, unpublished (2001); E. F. Redish, Developing student expectations in algebra-based
doctoral dissertation, 1 7Cornell University, Ithaca, NY, 1989. physics, presentation at the Conference on Integrating Science and
18
C. S. Dweck, Self-Theories: Their Role in Personality, Motivation and Math Education Research, Orono, ME, June, 2002,
Development (Psychology Press, Philadelphia, PA, 1999). <http://www.physics.umd.edu/perg/talks/redish/
Maine02/EFRMaineExpects.pdf>
.

11 Submitted to the American Journal of Physics