A case study of student

reasoning about
refraction and image-
object positioning
UPPSALA UNIVERSITET STUDENT REASONING ABOUT
David Nygren REFRACTION AND IMAGE-
OBJECT POSITIONING
2014-05-05

Abstract
This exploratory case study was undertaken to obtain a greater understanding of the difficulties that
physics students face when solving image-object projections in optics problems. This was carried out
by studying the students’ reasoning when facing new kinds of problem settings using the refraction
of light and the position of the virtual image and the real object as the frame for the research. The
results show that there is more than one reasoning possibility that is feasible for students to use when
dealing with the same problem. The results also illustrate how several different ways of reasoning
may be simultaneously needed to solve a refraction problem. The different kinds of reasoning have
been referred to as reasoning categories in this study. The analysis illustrates how the categories
complement each other, and the use of many reasoning categories is shown to be fruitful. However,
the vast majority of the participants made contradicting answer selections when solving similar
problems by using contradicting reasoning approaches. This lack of consistency in the participants’
reasoning could indicate that they have a fragmentary understanding of optics in general. Both the
capability to link reasoning approaches together, as well as the affordances that different modes of
representations offer, are needed for the construction of a better conceptual understanding. Only
mastering a few ways of reasoning and a few modes of representation could lead to fragmented
knowledge, which, in turn leads to making problem solving really challenging.
One purpose of this study was to find out if reasoning categories and modes of representations are
essentially linked. If so, then the reasoning categories would be determined by the representation of
the problem. The analysis shows that there is a connection, but that there are also other factors at
play.
The study is based on a questionnaire with same-problem, different representation Problem Sets
about refraction of light with respect to predicting image-object positions, and 20 follow-up
interviews with participants who had completed the questionnaire (in total 1200 international
university physics students completed it). After analyzing the interviews, 15 qualitatively different
ways of reasoning were found. These reasoning categories illustrate the reasoning processes behind
the answers, and also go towards explaining some of the overall results of the questionnaire. A
greater understanding of student reasoning is important to understand the challenges in learning and
teaching problem solving and this knowledge should be used to improve physics education in
general. A greater awareness of students’ reasoning categories and their fragmentary understanding
needs to be a resource to introduce changes in the approach to teaching physics in the area of optics.
Furthermore, the results provide a framework of reasoning categories in physics that could be used
for a more detailed mapping of reasoning processes in the area of refraction of light.

2
Acknowledgments
First and foremost I want to thank my supervisor Professor Cedric Linder at the Department of
Physics and Materials Science, Uppsala University for his countless hours of work making this
report understandable both in language and content. I am very thankful for everything he has taught
me about his field of research and in the work of writing a report. I also want to thank Anne Linder
and Jonas Forsman at Uppsala University for reading through and commenting on the report.
A big thank you goes out to Arvid Pohl, Department of Physics and Electrical Engineering, Linnaeus
University as well for his help with the study and valuable comments on this report.

3
Table of contents
Abstract ____________________________________________________________________ 2

Acknowledgments ___________________________________________________________ 3

Introduction_________________________________________________________________ 6
Cognitive studies ___________________________________________________ 6
The structure of the report ____________________________________________ 7
Background _______________________________________________________ 7

Methodology ________________________________________________________________ 9
Theory ________________________________________________________________ 9
Physics Education Research: case studies and interviews ___________________ 9
Disciplinary discourse ______________________________________________ 10
Approach _____________________________________________________________ 10
Ethical considerations ______________________________________________ 10
The questionnaire _________________________________________________ 11
Learning the needed scientific skills ___________________________________ 11
Analyzing interviews _______________________________________________ 11
Validation ________________________________________________________ 12

Results____________________________________________________________________ 14
Categories ____________________________________________________________ 16
Criteria for Category Formation _______________________________________ 17
Intuition__________________________________________________________ 18
Recalling ________________________________________________________ 19
Language recall ______________________________________________ 19
Visual recall _________________________________________________ 19
Principle _________________________________________________________ 20
Contemplating ____________________________________________________ 20
Extreme Case Reasoning ______________________________________ 20
Relation Reasoning ___________________________________________ 21
Analogical Reasoning _________________________________________ 21
Scientific Modelling ___________________________________________ 21
Logical Conclusion____________________________________________ 22
Procedure________________________________________________________ 23
Mathematics ________________________________________________ 23
Ray Tracing _________________________________________________ 23
Comparing __________________________________________________ 23
Using Snell’s Law on Object ____________________________________ 24

4
 Other ___________________________________________________________ 24
Alternative theory _____________________________________________ 24
Test-wiseness _______________________________________________ 24
Reasoning process _____________________________________________________ 25
Problem 2.a.1_____________________________________________________ 27
Problem Set 3b -- figure 1 ___________________________________________ 28
Problem Set 3c ___________________________________________________ 30
Consistency ______________________________________________________ 32

Discussion ________________________________________________________________ 34
Explaining the results of the questionnaire ______________________________ 36
Implications for physics education _____________________________________ 38
Future research ___________________________________________________ 40

References ________________________________________________________________ 41

Appendix __________________________________________________________________ 43
Appendix I: Questionnaire ________________________________________________ 43
Appendix II: Identifications of reasoning approaches ___________________________ 49
Question 2 _______________________________________________________ 49
Problem Set 1 ____________________________________________________ 50
Problem Set 1.1) _____________________________________________ 50
Problem Set 1.2), 1.3) and 1.4) __________________________________ 51
Problem Set 2 ____________________________________________________ 51
Problem Set 2.a.1 and 2.a.2 ____________________________________ 51
Problem Set 2.a.3 and 2.a.4 ____________________________________ 52
Problem Set 2b ______________________________________________ 53
Problem Set 3 ____________________________________________________ 54
Problem Set 3a ______________________________________________ 54
Problem Set 3b ______________________________________________ 54
Problem Set 3c ______________________________________________ 55
Appendix III: Processed transcripts _________________________________________ 56

5
Introduction
Physics is widely considered to be a challenging subject in school and at the university level. There
could be many reasons for this; one being that it is a subject that many students do not find
interesting or appealing. But it may be something within physics itself that makes it challenging.
Even after having studied physics for many years, it is still more challenging for me than many of
my other intellectual activities, even though I have a deep knowledge and interest in the subject.
Students interested in physics face many challenges when studying physics, and there is significant
failure and drop-out from the bachelor’s programme in physics at Uppsala University. Forsman
(2011) investigated the poor retention of physics students at Uppsala and presented and generated a
model of student retention using complexity thinking to show the many different influencing factors.
In this study report I will reveal some of the challenges in teaching and learning physics related to
problem solving in the area of refraction. And even though some of the students have “cracked the
code” and are now PhD students and professors in physics, the challenges faced by the vast majority
could lead to a shortage of engineers and scientists with sufficient knowledge in physics and
holdback further progress in technology and science.
Much research has been undertaken to find out why people find learning physic challenging
(McDermott 1991). In a survey given out by Ornek, Robinson and Haugan (2008:33) introductory
physics students were asked about some factors affecting the nature of physics. A majority of the
students thought that physics was a very difficult subject and 77% answered that this was due to
physics being cumulative and that if you miss one concept you cannot grasp another. A survey of
this kind can give valuable data, but to get a deeper understanding of the challenges in learning and
teaching physics, more extensive research is needed.

Cognitive studies

One thrust of this study aims at gaining insight into the kinds of thinking that students are evoking
when working in physics. Cognitive studies have, in recent years, provided valuable educational
research outcomes, showing that it is not as simple as either knowing or not knowing the answer, or
even on getting the right or wrong answer. At the heart of this is the attainment of a sound
conceptual understanding. There are different names for the cognitive processes at play; mental
models, patterns, schemas etc., but they all describe the complex cognitive processes of problem
solving, which are based on the tendency of the human brain to form patterns. As yet there is no
widely accepted theory describing the characteristics and representations of these cognitive
processes. However, Redish (1994: 796) proposes the following common properties for what he
calls mental models:

“Mental models have the following properties:

1. They consist of propositions, images, rules of procedure, and statements as to when and
how they are to be used.
2. They may contain contradictory elements.
3. They may be incomplete.
4. People may not know how to "run" the procedures present in their mental models.
5. Elements of a mental model don't have firm boundaries. Similar elements may get
confused.
6. Mental models tend to minimize expenditure of mental energy. People will often do extra
physical activities -- sometimes very time-consuming and difficult -- in order to avoid
a little bit of serious thinking.”

I will return to these properties when I explain my results in the Discussion chapter. In my analysis I
identified a set of qualitatively different mental models, called reasoning processes. The elements of
the reasoning processes are called reasoning approaches (see Figure 1).

6
The guiding research question for my study is: What reasoning approaches do university students
use, and how could these be categorized to generate teaching insight about the reasoning processes
and the challenges that image/object problem solving generate across different representations of
refraction.

The structure of the report

In the Methodology chapter, while there is an emphasis on describing the method, the chapter also
contains details of the relevant theories that I used to situate my study. The Results chapter is divided
in two parts; the reasoning categories found and the reasoning processes of three specific Problem
Sets. These are discussed by proposing possible explanations and suggesting a set of educational
implications.
To provide insight into the problem setting and subsequent analysis, a brief background story is
needed. It all started with an interest in representations and how optical refraction is being illustrated
in physics textbooks.

Background

Airey and Linder (2009) presented a sociocultural perspective on understanding university science
based on modes of representations, tools and activities of a discipline. In physics such
representations may be diagrams, images (schematic, pictorial, photorealistic), mathematics, written
words etc. Together with the tools and activities used in physics these constitute the disciplinary
discourse of physics.
Hüttebräuker (2010) investigated the different representations of optical refraction used in textbooks
and developed a questionnaire with Problem Sets all on optical refraction but using different
representations. This questionnaire was later modified and given to 1200 physics students from
Europe, North America, South Africa and Australia (Linder, 2013a; Linder & Priemer 2013). A
summary of the results they obtained showed that the majority of the participants had difficulties
with the Problem Sets in the questionnaire. However, links between sets of incorrect answers were
indecisive. Fredlund et al. (2013) then started the task of exploring this phenomenon further by
interviewing 15 students about the thinking that led them to select the questionnaire answers. In the
autumn of 2013 I travelled down to Kalmar to meet with Arvid Pohl to get insight into how analysis
of such interview data could be made. Using what I had learned from this experience I carried out
my own set of five interviews, transcribed them and then used the complete set of transcriptions (15
+ 5) as my data set.
Van Heuvelen (1991) showed how important the use of different representations is when solving
qualitative problems in physics. Students test scores on qualitative multiple choice problems
improved significantly for students who learned to use different representations to solve a problem,
e.g. drawing a picture of the situation, drawing a free-body diagram, and then associating
mathematical formulas. This is the background of why Airey and Linders’ (2009) work on
disciplinary discourse has positively influenced physics education. Their work will be further
explained in the methodology chapter.
Research on reasoning approaches has been shown to be useful in studies of learning. Clement
(2008) found a strong parallel between how scientists reason and how students could learn scientific
understanding (instead of rote knowledge based on memorization). In a case study of students in a
classroom situation, Stephens and Clement (2010) investigated the reasoning of the students, finding
three student-generated non-formal reasoning categories: “analogy”, “extreme case “and
“Gedanken”. They also identified combinations of these reasoning categories. Their work is similar
to the work done in this study, but with a slightly different terminology. In the terminology of this
study, what are called “reasoning categories” can be seen to be related to what Clement and
Stephens have referred to as “reasoning processes” and what I am calling “reasoning processes” in
this study can be seen to be related to what Redish (1994) refers to as “mental models”.

7
Optics is the part of physics that involves the behavior and properties of light. It is a crucial part of
physics education because of it being a keystone of some of the fundamental theories of physics and
the practical implications it has in everyday life. The properties and behavior of light are essential
for both macroscopic and microscopic physics.
The history of geometrical optics can be traced back thousands of years, where the refraction of light
was studied by the ancient Greeks, the great physicists of the medieval Islamic culture, and by
Willebrod Snell who in 1621 re-discovered the law of refraction and called it “Snells law” (Hecht
2002:1-2). In modern times, basic knowledge in geometrical optics still plays a great part in
technological advancements such as microscopes revolutionizing medicine, and fiber optics making
ultra-fast communications possible. Realizations of the wave properties of light that manifest
themselves by their patterns of interference have lead to great achievements in modern physics,
particularly in the field of quantum mechanics. The constant speed of light is the second postulate of
another great field of modern physics, Einstein’s special theory of relativity. All in all, the behavior
and properties of light are a critical part of physics knowledge, which explains why optics is an
important part of physics education all over the world.
In the context of optics, Harrison and Treagust (1993) did a case study of how students use analogies
to solve problems. They found that using analogies in the teaching of refraction could be effective if
the analogy is familiar to many of the students and if the shared and not shared attributes of the
analogy are presented. Jesper Haglund (2012) has done a similar study in thermodynamics, using a
theoretical framework of analogical reasoning, scientific modeling, are essentially the same thing
and semantics that will be revisited in the Discussion chapter.

8
Methodology
Theory

Physics Education Research: case studies and interviews

In Sweden education is a hot political topic! For example, the leading article of a major newspaper
recently proclaimed that “education is not nuclear physics”. I would agree, but not in the way the
statement was intended to dramatize; rather nuclear physics, like education, is a highly complex
affair (Davis & Sumara 2008), perhaps even more complex in many ways. For example, for physics
education research both quantitative data, usually from experiments and surveys, and qualitative data
is often called for. One reason for this is that the research aim is not only to explain how education
and learning takes place, but also to understand why students and teachers behave and think the way
they do. In this sense, education research uses sophisticated methodology. The human element in
education means that educational research rarely can be done in controlled laboratory environments
in the same way as in nuclear physics research. The complexity in education research often calls for
case studies to identify important focus points for more extensive studies in the area. Robson
(1993:5) describes this well:

“Case study is a strategy for doing research which involves an empirical

investigation of a particular contemporary phenomenon within its real life context
using multiple sources of evidence”.

To conduct case studies, observations or interviews are typically used. In my research, the data
comes from interviews, and the primary analyzing tool was me, the researcher, interpreting these
interviews. This is not entirely uncontroversial, many might argue that scientific research needs to be
objective, and both the data source and the interpretation in this study are in some way subjective.
Kvale (1996:64) points out that there are dozens of meanings of objectivity, often related to different
dichotomies; objective/subjective, unbiased/biased, public/private, intersubjective/personal and so
on:

“With the variety of conceptions of objectivity, the qualitative interview cannot be

objectively characterized either as an objective or subjective method”

There are no fixed set of rules that determines if a method is scientific or not. In modern philosophy
science thinkers like Michael Polanyi, Thomas Kuhn and Paul Feyerabend all agree that what
scientists do cannot be reduced to a set of rules to follow. Martin and Brouwer (1992) defend the
subjective side of science and argue that science is more an art form that needs curiosity and
imagination and not only a set of skills that can be logically explained. That being said, most
scientists agree on intersubjectivity being an important principle, and that the data should not depend
on the individual scientist. Kvale (1996:65) has two different definitions of intersubjectivity;
arithmetic and dialogical intersubjectivity. Arithmetic intersubjectivity refers to the same
measurement between independent observers, which is hard to apply to qualitative interviews in
general. Dialogical intersubjectivity is more valid for interviews, and interpretative research:

“Dialogical intersubjectivity refers to agreement through rational discourse and

reciprocal critique among those identifying and interpreting a phenomenon. This
may take the form of communicative validation among researchers as well as
between researchers and their subjects”

In my interpretative research Arvid Pohl, at Linneus University, and I used the same source material
and although we have made no attempt to claim arithmetic intersubjectivity, in our discussions we
agree that dialogical intersubjectivity has been achieved.

9
Disciplinary discourse

An analogy to the disciplinary discourse of a university science field is that of a language, and to be
familiar with all parts of a certain discipline can be seen as being analogous to being fluent in a
language. Airey and Linder (2009:1) describe a discipline as a shared way of knowing:

“By shared ways of knowing we mean here the coherent system of concepts, ideas,
theories, etc. that have been created to account for observed and theoretical
phenomena.”

To participate in this way of knowing one must master the semiotic resources of the discipline, much
like the semiotic resources in everyday language are needed to participate in the knowledge of
everyday life. Airey and Linder (2009) have described the system of semiotic resources of a
discipline in terms of “disciplinary discourse”. This disciplinary discourse consists of tools,
representations and activities. Linder (2013b:1) illustrates how physics uses sets of different
representations to share ways of knowing and doing physics:

“In semiotic terms, representations are constructed from collections of signs. In a

discipline such as physics these signs are multimodal representations, examples of
which are written and spoken language, gestures, symbols, diagrams, sketches,
pictures, simulation and imagery, mathematical formalism and so on.”

Because the nature of physics is multimodal, one may need different modes of representation to fully
understand a phenomenon, or rather see all facets of the phenomenon. A lack of fluency in the
disciplinary discourse may than lead to an incomplete understanding of a phenomenon in physics. If
physics education fails to give students the affordances that the different modes of representations
offer, the concepts, ideas and theories of the discipline may not be accessible as a coherent system,
but rather presented as fragmented bits and pieces.
Fredlund, Airey and Linder (2012) investigated the use of different representations in interactive
engagement between students when discussing refraction of light. Their results show that a number
of different representations were used when explaining refraction of light and that the different
representations were both complementary and supplementary.

Approach

Ethical considerations

Every research study involving people has to follow a certain set of ethical guidelines. The Swedish
Research Council (Vetenskapsrådet, 2002) has published four conditions for ethical research.
Openness and honesty: The researcher should in the best way possible inform the participants about
the research, their involvement and what will be expected of them, so that they can decide if they
want to participate or not.
Consent: The researcher needs to get consent from the participants to include them in the study and
they have the right to withdraw this consent whenever they want.
Confidentiality: The researcher is required to ensure confidentiality of the participant's identity and
the data collected in the best way possible.
Fair use: The researcher may only use the data from the participants to conduct research. The
research should not be commercially used or used to justify actions regarding the participants.

The participants in this study were informed about the purpose of this research and gave their
informed consent to be video recorded. To protect the participants’ identities their faces were not

10
filmed, only their hands and the questionnaire were visible. The questionnaires and recordings have
been stored properly without access by anyone but me and they are also numbered without any
names linked to them. I informed the participants that the data from the study was only to be shared
for research purposes, with their names remaining confidential.

The questionnaire

The questionnaire was developed by Hüttebräucker (2010) based on three modes of representation
used in physics textbooks to describe a refraction of light situation: ray diagram, schematic and
pictorial representations, and the additional representation of photorealistic pictures. Furthermore,
some of the questionnaire problems were of the kind that are commonly incorrectly explained in
textbooks. Nassar (1994) states that most textbooks seem to imply that the virtual image of an object
always is at a single unique position, regardless of the observer, and some of them describe a
principle, apparent depth, as the virtual image being straight above the real object. This is not the
case, Nassar (1994) shows that the position of the virtual image is dependant on the observer, and
that a general principle of the position would be that it is up and towards the observer.
The questionnaire was translated into Swedish, and in so doing some slight changes to enhance
clarification. The version used consisted of one question of the concept of apparent depth to be
answered with text and/or a drawing (Question 2 in Appendix I), nine multiple choice Problem Sets,
and five complete-the-picture Problem Sets.

Learning the needed scientific skills

This kind of questionnaire using multiple choice questions and open-ended problems is typical of
how data is often collected in research in the area of physics education research. There is a strong
theoretical basis for using these methods along with interviewing participants about their answer
selections. I needed to become competent with handling interview data for this study. I started out
with travelling down to Linneaus University where Arvid Pohl introduced me to doing analysis from
the interview data he had collected. It started with me doing the questionnaire myself and then
discussing the questionnaire and how interviews could help explore the nature of students sense-
making. I did the analysis of one of Arvid’s recorded interviews under his supervision to familiarize
myself with the skills needed before I undertook interviews on my own. Interviews were used to
understand how and why students think the way they do, as explained in the Theory section above.
The advantage of using recordings of the interviews instead of only memories and notes are that a
transcription makes it possible to carefully deconstruct every interview into parts, making
similarities, differences and patterns among the interviews become more apparent. As part of the
skill acquisition I also studied the methods of interpretive research (Clark, 2013) and a classic
introduction to the field of scientific interviewing for qualitative, interpretative studies (Kvale,
1996). I also studied an article (Fredlund, Airey, Linder, 2012) and a PhD thesis (Haglund, 2012) to
get an introduction on how to conduct qualitatively and interpretative physics education research.

Analyzing interviews
The data consisted of 15 interviews that were video recorded by Arvid Pohl (Linneaus University)
and 5 interviews that I recorded. As described earlier, I followed the procedure learnt as part of my
introduction to the type of research skills that I needed. The data setting for both sets of interviews
was identical, in both cases the participants were asked to explain their reasoning behind their
answer selections for the refraction questionnaire.
I started with the interviews that Arvid had made (each approximately 10 minutes long) and
transcribed them. After some initial analysis of these interviews and reading about using interviews
as a way to obtain research data (as described above) I felt ready to undertake my own interviews
and completed another five (each approximately 15 minutes) in a similar way to Arvid but a bit
longer with some more in-depth questions on some selected parts of the questionnaire.

11
From the initial round of analysis using Arvid’s data, I made a selection of Problem Sets to structure
the focus on. This decision was based upon the most interesting variation in reasoning descriptions
obtained. From the interviews, a focus point transcription of the interviews was made, in other
words, one that used the most pertinent descriptions of thoughts and methods by the participants. By
comparing reasoning for each set of Problem Sets across the interviews, themes in reasoning
approaches were identified.
An overview of all the interviews were made by arranging a Microsoft Word document with one
interview per page written in as a processed transcription, see Appendix III. This overview was used
to identify similar reasoning approaches for each part of the questionnaire, and finally to identify
similar reasoning approaches for all interviews on all parts of the questionnaires. Starting with
having created a coding for all reasoning approaches that I was able to identify the refined work of
constructing the categories and their characterizations began. This was done based on the theoretical
framing obtained from Airey and Linders (2009), which posits a model for disciplinary discourse as
a way to extend understand physics teaching and learning. The kind of interpretative research that I
undertook is filtered through my expectations and pre-understanding. In this way my experiences of
both learning and teaching physics played a critical role in the analysis. I started by formulating
categories with my preliminary coding and proceeded with refinement building using hermeneutic
iteration cycles through my coded data. As refined patterns emerge from this data analysis process
the category forms and characterizations became increasingly distinct. The process continued until a
point of saturation was reached (see Crotty, 1998: no further changes emerged from the refining
iterations).
My interviews were filmed using a video camera focused on the questionnaire at hand. The
interviews was guided by a semi-structured protocol that called for me to pose the same question for
all problems: how did you reason when you came up with this answer? This protocol left open to
what follow-up questions were needed to clarify the answers I received; each case had to be judged
on its own merits. All discussion about what the correct answers were was saved until after the data
collection discussions. A few of the participants did come up with new answers and ways of
reasoning during the interviews; these were not included in the study. An ipad recorded the sound
and any drawings that were used to supplement explanations. This was in addition to what was being
recorded by the video camera. I used these interviews to find complementary reasoning approaches
and also as well as investigating three Problem Sets in more detail. The combinations of reasoning
categories used for these three particular Problem Sets, 2.a.1, 3b and 3c, were noted for all 20
participants to show the nature of their reasoning processes. I also looked at the consistency of
answers and explanations given across these Problem Sets.

Validation
In order to validate the results of the study it had to achieve dialogical intersubjectivity (see theory
section above). Cedric Linder, Professor of physics education research at Uppsala University, and
Arvid Pohl, Department of Physics and Electrical Engineering, Linnaeus University, looked at the
analysis in terms of the work that they had done independently in the area, and after examination,
concluded that the analysis was in line with their own analytic outcomes. The usual concepts of
validity and reliability used with quantitative data are not appropriate for use with qualitative
interpretative studies (Lincoln & Guba 1985). Lincoln and Guba re-formulated these concepts. In so
doing, they argued for establishing trustworthiness, this trait of trustworthiness being made up of the
following components:
• Credibility - confidence in the 'truth' of the findings
• Transferability - showing that the findings have applicablity in other contexts
• Dependability - showing that the findings are consistent and could be repeated
• Confirmability - a degree of neutrality or the extent to which the findings of a study are shaped by
the respondents and not researcher bias, motivation, or interest.

12
Based on the methods used in this study and the feedback from the two independent researchers, I
claim the results of the study to have all the traits of trustworthiness:
Credibility: The interviews were engaging the participants in a serious discussion on their reasoning
for answering the different problems on the questionnaire. The participants were speaking freely
without any pressure to answer a certain way or being asked leading questions. There were no
indications of any incitements for the participants not to tell the truth or answer a certain way.
Transferability: No significant differences could be detected between the data set that I was
collecting on my own and the ones coming from Arvid, which were collected at a different
university, which I considered to manifest as a quite a different educational context.
Dependability: Arvid has looked at my results and concluded: "David's descriptions of the
individuals' reasoning match my own rather well" (Private correspondence with Cedric Linder).
The findings are consistent as the same kind of reasoning was found in different participants without
any irregularities.
Confirmability: Arvid and my analysis were done independently of one another. He did not share
any results of his analysis until mine were finished and vice versa. Sharing was then done in order to
ensure dependability and confirmability of the analysis, not to compare or use each other’s results.
The research was done without any preconception of what to find, in this way it was entirely
exploratory, and I had no motivation to get a certain result.

13
Results
The results show the reasoning categories found in the analysis of the interviews and how these
could be used to describe the nature of the reasoning process behind the answer selections of the
questionnaire. A category represents certain reasoning approaches but is not the same as a reasoning
process; a reasoning process is all the thoughts that lead to an answer and could, and often do, exist
of many reasoning categories. Further explained in {Figure 1}.

Figure 1. Illustrating the reasoning characterizations used in this study: reasoning approach,
reasoning category, reasoning group, and reasoning process.

A reasoning approach is what was first identified as a similar way of solving problems and
sometimes the same reasoning approach was found in several different problems and by several
participants (see Appendix II). When analyzing these reasoning approaches, they were categorized
into reasoning categories based on certain criteria. The reasoning categories are each made up of sets
of reasoning approaches. These are shown in {Table 1}. In the analysis of the interviews, 15
different categories of reasoning was found, 13 of them were sorted into five different groups of
reasoning categories: Intuition, Recalling, Principle, Contemplating and Procedure (see Figure 2).
All groups represent unique features of reasoning. The remaining two categories were left out of the
results because they dealt with only generalized reasoning that was not specific to the refraction of
light research setting of my project.

14
Figure 2. The framework of reasoning. Fifteen reasoning categories organized into six
reasoning groups. As shown in Figure 1, the reasoning categories are made up of a collection
of reasoning approaches (not shown in Figure 2).

The framework shown in {Figure 2} is the final result. The identification of reasoning approaches
from every part of the questionnaire was done prior to this framework (given in Figure 2) but will be
presented in the framework of categories in Appendix II. These reasoning categories are: Intuition,
Language recalling, Visual recalling, Principle, Extreme case reasoning, Relation reasoning,
Analogical reasoning, Scientific modelling, Logic conclusion, Mathematics, Ray tracing,
Comparing, and Using Snells law on object. The categories were based on the 15 interviews and
further investigated and complemented by the five more in-depth interviews. Alternative theories
and test-wiseness will also be mentioned but not further explained as they were not a part of the
study.

15
Categories

Intuition Visual Language Principle Extreme Relation Analogical Scientific Logic

recall recall case model

1.1-1 Q2-1 1.1-2 1.1-2 1.2-4 2b-4 1.2-5 1.1-4 1.1-5

2.a.1-1 1.1-3 1.2-1 1.2-1 2.a.1-5 3a-4 2b-5 2b-5 2.a.3-3

2b-1 2.a.1-2 1.2-2 3a-6

3c-1 2.a.3-1 1.2-3

2b-2 2.a.1-3

3a-1 2.a.3-2

3c-2 3b-3

3b-4

Mathematics Ray tracing Comparing Using Snells law Alternative Test-wiseness

on object theories

1.1-6 Q2-2 Q2-3 2.a.1-7 Q2-4 Q2-5

1.2-7 2.a.1-6 1.1-3 3c-6 2.a.3-4 3a-7

2b-6 1.2-8 2b-7

3a-3 1.2-9

3b-5 3a-5

3b-6

3c-5

Table 1. Overview of categories and identifications found, a summary of Appendix II.

The number-letter-combinations in {Table 1} refer to the identification of reasoning approaches

found in Appendix II and is in the form “question 2 (Q2) or question-set number (see Appendix I)”-
“approach number”/”identity number”. These are referred to within “[]” in the text. {Table 1} is only
an overview of the findings, where the results are purely qualitatively, and does not show the
frequency of the different categories.
Firstly, the findings do not necessarily correspond to the actual use of the reasoning category; the
participants tended to be more aware of certain specific kinds of their reasoning. The findings
correspond to what the participants expressed in the interviews. Secondly, as the categories are
qualitatively different, the columns represent different things. The Principle, Alternative theories,
and Test-wiseness list the different kinds of Principles/Alternative theories/Test-wiseness. The other
ones represent the use of the category in a certain problem (as well as the different kinds in some
cases). Thirdly, neither the amount of participants nor the amount of time per participant is being
presented in {Table 1}.
In most interviews Visual recall, Principle, and Ray tracing were identified but no category is found
in every transcript. However, every interview contains several categories. Each category will be
described and exemplified below and sorted by their group names as shown in {Figure 2}.
I need to clarify here that a reasoning category is not the same as the participants reasoning process
on a certain problem, see {Figure 1}. A category is a certain kind of reasoning approach. A
reasoning process is the complete cognitive processing of a problem for one person, and might
consist of a multitude of reasoning approaches of different categories. These reasoning approaches
have been identified and categorized into 15 reasoning categories in {Table 1} according to certain
criteria shown in {Table 2}.

16
Criteria for Category Formation

The categories were formed using a constant comparison method (Dye et al., 2000) that involved
refining the identifications of the reasoning approaches and the formulations of the categories
through repeated hermeneutic iterations. This process continued until no more refinement was
possible and then 15 distinct categories remained. . As seen in {Table 1} at least two identifications
for different parts of the questionnaire constitutes a category. The criteria for these categories (see
{Table 3}) evolved from the process of comparing and finding similarities in the reasoning
approaches. The decisions for the criteria of the categories were made so that distinctive categories
could be formulated. At the same time the process ensured that every participant fitted into one (or
more) of the categories. The criteria were also formulated in a way that would make it plausible for
later findings of reasoning approaches to fit into a given category. This was tested using my five
interviews where some new reasoning approaches were identified. These new reasoning approaches
easily fitted into the existing reasoning category scheme.

Reasoning category Criteria

Intuition Intuitive feeling of how things should be. Described as a feeling of

what is correct or not without any clear memories of a similar
situation.
Language recalling Recalling a memory of a spoken or written wording.
Visual recalling Recalling a memory of a visual image.
Principle A statement that is being applied as a rule.
Extreme case Using the extreme case situations of a phenomenon, tracing back
reasoning by logic reasoning to the situation of the problem.
Relation reasoning Using the relations coming from comparing situations close to each
other, tracing forward by logic reasoning to the situation of the
problem.
Analogical reasoning Using an analogy to map relational properties
Scientific modeling Constructing a physical model based on some physical
phenomenon
Logical conclusion Use of general logic principles
Mathematics Use of mathematics; calculations, formulas, or geometry
Ray tracing Using tools; pen, paper, hand gesture, or vision, to do a step-by-
step procedure of ray tracing
Comparing Compare different images or problems and use the
similarities/previous results
Using Snells law on Treating an object as a ray of light and apply Snells law on it
object
Alternative theory A theory that explains things based on an alternative conceptual
idea of certain phenomenon not established in physics
Test-wiseness Making logic reasoning on parts related to the test but not the
actual problems, like what the meaning of the test seems to be

Table 2. The criteria of the reasoning categories. Further explained on pages 18-24

17
All research data is filtered through the perspective of the researcher. I have a lot of experience
working with problem solving in physics myself as well as of observing other solving problems. I
have reflected over my own and other peoples reasoning during my own education of becoming a
physicist and teacher and in my practical teaching work in the Swedish high school environment. My
influences come from studying philosophy (especially a Masters course at Uppsala University called
philosophy for physicists), mathematics, physics and pedagogy and from my teaching experience.
Some of the abilities listed in the new curriculum of mathematics (Skolverket, 2012) have given me
inspiration, like the ability of procedure and mathematical modeling. Furthermore a presentation at a
teaching mathematics conference of the correlation of working memory and mathematical abilities
raised an interest in working memory and cognitive studies. My studies in philosophy of science and
philosophy of logic have influenced my categorization, especially the findings of general logic
principles of the Logical conclusion category. The reason that Alternative theory and Test-wiseness
was found and left out of the study was prior knowledge of their existence and complexity.
In the iteration phase of analysis which involved examining the relationships of the categories, some
of the categories were found to have partial similarities. This led me to re-grouping these as a
function of formulating the reasoning groups (see {Table 3} for criteria). These reasoning groups
show some fundamental differences in the reasoning approaches, which is further explained in the
Discussion chapter.

Reasoning group Criteria

Intuition Same as category
Recalling Browsing the long-term memory for specific representations
Principle Same as category
Contemplating Using certain features of a physical phenomenon and logic
deduction
Procedure The use of certain technical steps to get a result

Other None of the above (outside further study, Alternative theory and
Test-wiseness)
Table 3. The criteria of the reasoning groups. Further explained on pages 19-24

Intuition

When the participants described part of their reasoning as an intuitive feeling of what is correct or
plausible was categorized as intuition. In some cases the participants had created an internal image
(or drawn an external image) and assessed if it felt real or not. Even if this is similar to a visual
recalling of a memory, when the participants did not have a clear memory but only a “feeling of
what is real” to compare with, this has been categorized as intuition as well.
These cases were especially common in Problem Sets 2.a.1-2.a.4 [2.a.1-1/a,b,c] (refers to
identification in Appendix II). The other kinds of reasoning approaches tended to lead to multiple
possible answers, and the intuitive feeling of the image as being “shorter and bent upwards he
surface”, “bent in some way” or (misleadingly) “farther away from observer than object”resulted in
a singled out answer.
This categorization was not always based on the use of the word “intuitive”, but also on what I
discerned to be an intuitive response after my interviewing experience..
It could be argued that intuition is an easy explanation that hides the real reasoning behind an
answer. On the other hand the instant gut feeling all humans experience cannot be ignored, and
might in fact be a big part of all reasoning processes. The identification of intuitive reasoning has

18
only been done when this is an explicit part of an answer. The participants used estimations all the
time, especially estimation by the eyes (“ögonmått” in Swedish) and while these could also be
classified as intuitive they were left out of the data analysis. Trying to either define intuition further
or dissect it into other categories is beyond my study.
This category is based on the everyday semantics of what people mean by intuition. To even label
intuition as a “real world experience based”-reasoning should not be done without careful
consideration of the implications of such a label.

Recalling

Most of the reasoning starts with some kind of recalling of some aspect of physics knowledge,
however, a significant difference in recalling in general versus language and visual recalling, which
are reasoning approaches used by “browsing” the memory for specific representations. It is the
memories of specific words, not the meaning of the words, and specific memories of an image, not
the meaning of the image that is being recalled. The categories used in this reasoning group indicate
long term memory access, and are therefore relevant for my study.

Language recall

When the participants said they recalled some kind of spoken or written language, this has been
categorized as “language recalling”. Typically this was the language used to recall a certain
principle. This category draws a line between recalling the meaning of a principle directly and
recalling it via a language memory. The participants had a memory of specific wording, often
remembering it from elementary or secondary school, and formed this into a principle. This category
does not stands on its own, but is preliminary to other reasoning approaches; however it was used to
categorize the participants’ answers because it reveals something distinct about the nature of the
reasoning used.
In Problem Set 1 many participants recalled spoken or written language used to describe a principle
of refraction. There were four kinds of principles of refraction recalled, each one described in similar
way, sometimes so similar the description sounded almost like a chant. Some participants explained
that this was something they had repeated over and over again since elementary school in order to
get it memorized.

Visual recall

This category describes situations when the participants used a visual memory related to the
problems. If recalled, then these visual memories were used to make metaphorical leaps, or to
compare with another image. The Visual Recall category characterizes the recalling of a visual
memory. Creating an internal image and recalling a visual memory was not taken to be the same
thing. The category makes it possible to make a trajectory of the reasoning process, showing if an
internal image comes from a memory or not. To not recall any visual memory of a certain situation
could also be used in the selection of an answer. An example of this was used in Problem Set 2b [2b-
2] where participants had not previously noticed objects disappearing when adding water to a
container containing an object. Such was then thought to be unlikely to arise.
The most commonly used visual memory was that of an oar of a boat partially in water, or a pencil
put into a glass of water; this was used in the explanation of “apparent depth” in the second
questionnaire question and to select answers for Problem Set 2 and 3c. The visual memory could
either be used to confirm an answer, or to obtain an answer.

19
Principle

A Principle is an answer provided in the form of a rule-statement and being applied as such. Many
different principles were found in the interviews, and the same principle could have several different
origins. It could have been recalled directly, recalled from a language memory, derived or even
presumed. A distinction between a fundamental principle, rule of memory, reference, and
presumption could be made, but the categorization formulated does not take such origins into
account for its characterization. How the participants perceived one and the same principle could
differ, and it was often not evident, why a distinction had not been done. There were some more
obvious cases, but I still felt that such distinctions would not contribute to the formulation of a new
distinct category, and thus they have not been explored further here..
In the Principle reasoning group a fundamental Principle was the supplication of principles as facts
or fundamental axioms of physics. Examples of such fundamental Principles in the study are “lazer
is light” [3b-3], “the light follows the same path no matter the direction” [1.2-2] and “the change of
coordinates does not affect the physics” (Lorentz invariance) [1.2-3].
A rule of memory is a principle originated from language recall, which was typical for many of the
principles of refraction provided by the interviewees. The reasoning categorization then used was
Language Recall + Principle.
At times a Principle could be seen to be derived from other Reasoning categories. A typical example
is the use of “apparent depth” from Question 2 as the Principle “the image appearing closer to the
surface than the object” [2.a.3-2, 2b-3, 3a-2, 3b-1].
The reasoning would then be: Question 2 Reasoning (e.g. ray tracing) + Principle.
A Presumption principle was used when a participant thought that there must be some kind of
principle that they should be using to solve a problem. For example, this approach was used on
Problem Set 2:

“I was wondering if the location from where you looked did matter at first. I
thought that it did not matter in the end.”

The participant figured there would either be a correlation or none-correlation between observer and
image, and used the none-correlation as a presumed principle [2.a.1-3]. The interviewed does not
indicate why they decided on the none-correlation; maybe there were no other reasoning than
intuition, the reasoning categories would then be: Intuition + Principle. Maybe the dichotomy of
correlation/none-correlation comes from a logic conclusion (one must be true), to describe this with
the reasoning categories would then be: Logical Conclusion + Intuition + Principle.
The interview data do not provided sufficient additional precision to better describe the reasoning
processes, but such parts of the reasoning processes were captured by the combinations of reasoning
categories as illustrated.

Contemplating

These categories of reasoning are formulated on logical reasoning about physics phenomena.

Extreme Case Reasoning

This is the reasoning used when giving consideration to extreme cases of a phenomenon. It was used
when knowing or figuring out the situations at the extreme boundaries (maximum and/or minimum)
and then tracing back to the situation of the given problem. For example, this is how a participant
used this in Problem Set 2.a.3; they looked at extreme cases of the refraction indexes [same as 2.a.1-
5]:

20
 “The extreme case for an extremely high refraction index gives a near vertical
light path in the medium and an image close to surface. And if the refraction
index is the same as air the image is located at the real position, the image must
then be in between these extreme cases.”

Relation Reasoning

In this reasoning profile one considers at what would happen if small changes were made, and used
this thinking to project an answer for the given question. Finding relations between what is already
known in terms of small changes is seen to provide help with figuring out the answer for a new kind
of situation.
For example, a participant used this with Problem Set 2b [2b-4] as follows:

”The higher up you fill, the lower down the eye can be and still see the object,
because of light being refracted higher and higher up.”

Analogical Reasoning

This category characterizes the reasoning that calls on analogy to establish relational properties. A
participant did this as follows to answer Problem Set 1.4 [same as 1.2-5]:

“Thought of a trick from high school. If you have a small car water is a denser
medium and makes the car slower. The right wheel will come out in the lighter
material first and the car will turn because of the left side being slower than the
right side.”

Analogies formed part of another categories of reasoning as well, for example, Scientific Modelling.
The dividing line between Analogical Reasoning and using an Analogy Model is not made distinct
by my data set. Analogical Reasoning is reason that takes on the form of: “A doing B is like C doing
D” and an Analogy Model is reasoning where certain elements are being replaced by an analogy, for
example, the Bohr model of the atom using an astronomical-like orbital system for the movement of
electrons. . This is relevant for the interpretation of one particular reasoning approach [2b-5]:

“Thought that ”p1” and ”p2” (the other eyes/observers) could still see because
the ball is sending out light in all directions, some rays should then hit them.”

The most reasonable interpretation of this interview is that the participants were attempting
Scientific Modelling using an Analogy Model of a ball reflecting sunlight as a point source emitting
light in all directions. However, an Analogical Reasoning with light propagating out from a source in
space in all (infinite) directions could be based upon a substance analogy.
Although differences in criteria for Analogical Reasoning and Scientific Modeling are small because
analogical reasoning also forms part of the characterization of Scientific Modeling.

Scientific Modelling

Scientific Modelling emerged as a distinct category late in the iterative process described earlier. It
became needed to account for descriptions of correct solution for some of the problems.
This Reasoning Approaches uses physics modeling of physical phenomenon to solve problems. The
following correct solution to the 2.a.x Problem Sets illustrates this approach:
The way the eye sees a point at a certain distance is by focusing diverging light rays from the light
source with the lens of the eye into the retina (Hecht, 2002:201). The brain thinks that the diverging
light rays come from a point where these rays trace back to, not taking the refraction from water into
air into account. To model this we need two light rays, with only one we cannot see where all the
light rays intersect which gives us infinite points along a line, corresponding to 3 possible answer
selections.

21
Figure 3. Solid lines from point of pen to eye represent two light rays hitting the eye, the blue
rays behind the eye show how the rays get refracted by the lens of the eye into the retina. The
dashed lines shos the the brain extrapolates where the light rays whould intersect, and thus
where the object is.

The Scientific Modelling category is the reasoning that structures a physics mode. For the correct
modeling example given above mathematics and ray tracing procedures are used. The reasoning
categorization for this example would be Scientific Modelling + Ray Tracing + Mathematics.
Scientific Modelling got also used in a lesser way when the participants used only a single ray for
their ray tracing. The Ray Tracing category characterizes the procedure of ray tracing, whereas
Scientific Modelling was needed to formulate the ray tracing by connecting it to the principle or
mathematics in order to give the Ray Tracing result a physical meaning. Scientific Modelling could
be given a broader definition since it provides the physical interpretation and representation of the
other reasoning categories. However, the data was unable to support this in a sustainable way.
In one of the later interviews, a participant used Scientific Modelling in Problem Set 1 [1.1-4 and
1.2-6] by looking at the ray as indicating the direction of propagation in terms of a line that ran
perpendicularly off the front of the light wave, generating a physical model close to Huygen’s
principle model.

Logical Conclusion

This category characterizes reasoning approaches that are based solely on logic. The conclusion is
based on some general logic principle or principles. While logic is part of all the contemplating
categories, Logical Conclusion does not subsume all of these categories. In Problem Set 2.a.3-2.a.4
one participant argue that they could not see any reason for the image to bend left or right due to
symmetry.
This was misleading because the symmetry is in fact broken by the eye looking from left. He could
have been arguing the same with Problem Set 1.1 as well, stating that the solution was logical, with
greater success. The reasoning approach of finding no reason to bend right or left means that it does
not bend right or left could be based on Leibniz “Principle of sufficient reason”, stating that
everything has a reason or cause (Melamed & Lin, 2013). Although the participant almost certainly
is not aware of using a general logic principle, one could be found.
Stating for Problem Set 3a that the alternatives seeing and not seeing a coin both cannot be correct
when observed from the same position [3a-6] the participant is using the “law of contradiction”
(Horn, 2014) to find a logic conclusion.

22
Procedure

The Procedural categories of reasoning follow an almost mechanical format. Problem solving tools
and technical steps are used to create a reasoning recipe to solve a problem.

Mathematics

As an illustration of this category characterization a participant used the Snell’s law equation to
solve Problem Set 1.1 [1.1-6] by stating that if the angle of incidence is 90 degree then the angle of
refraction must also be 90 to get both sides of the equation equal to zero. The participant used the
same approach with Problem Set 1.2-1.4 [1.2-7], thinking that refraction index multiplied by the
angle is a proportional equality before and after refraction took place. They did not use any
computation but this category includes the use of Mathematics, Computations and Mathematics
Representations.

Ray Tracing

The Ray Tracing Procedure category is characterized by step-by-step ray tracing to solve a
refraction problem. Such ray tracing can be done using different representational forms, for example,
using a pen and paper ray diagram, or by hand gestures. The full process of Ray Tracing sometimes
included Scientific Modelling reasoning to establish a why the ray tracing was bing used and a how
to interpret the result, and associated physics Principles that helped guide their Ray Tracing. This
category is characterized by the actual ray tracing.
For Problem Set 2.a.1 Ray Tracing was a very common reasoning category Three useful and three
misleading ray tracing methods were identified [2.a.1-6/a-2.a.1-6/f], even though using only one ray,
resulting in more than one possible answer to select. In some cases the method of drawing a ray from
the eye-to-surface-to-object was used although the participants were aware that the photon really
comes from the seen object. Thus its origin could be based on the incorrect emission theory common
in Ancient Greek that involved light shooting out from the eyes to create images, a theory
originating from Plato (Langley, 2006).
The useful ray tracing procedures combined with some presumption of the image being at the same
depth as the object lead to the misconception of the image being further away from the observer than
the object (see {Table 5}).

Comparing

The Comparing procedure involves either comparing different images with each other, or by refering
to the solution of another problem by getting to see the two problems as being similar.
In Problem Set 1 a ray diagram provided an example of how the light path from air to water was
presented in the questionnaire. Some of the participants compared this diagram with the figures in
Problem Set 1.2-1.4. They used the principles “the light follows the same path no matter the
direction” [1.2-2] and “the change of coordinates does not affect the physics” (Lorentz invariance)
[1.2-3] to create an internal image of the diagrams needed top solve Problem Set 1.2-1.4. The same
positioning of the water and direction of light to compare this easily with the figure and complete the
Problem Set figures is illustrated in {Figure 4}.

23
Figure 4. An illustration of the reasoning process of using an image and principles to create an
internal image to compare with as the processinvolved in getting or checking an answer. The
reasoning categories used in this reasoning process are Principles + Comparing.

Using Snell’s Law on Object

Here the object’s position was taken to be a function of a light ray following an application of
Snell’s law. This involved procedurally following certain steps, without any indication that
understanding was being called for.

Other

Two categories were left unexplored because of the limitations of my study. These categories
represent well known tracks of reasoning, which can be expected to be found in all kinds of problem
solving. There are much to be found in the literature about this, and many studies have examined
these phenomenon in better ways than I could do for this study.

Alternative theory

An alternative theory characterizes sense making that does not conform to how physics understands
the given context (cf. Driver & Erickson, 1983). How these come about and to what degree the
participants believed in them are very interesting and it deserves to be separately studied. Alternative
theories are relates to misconceptions and conceptual change in the literarature.

Test-wiseness

Test-wiseness is when the participants try to figure out what the questionnaire and interviewer are
looking for in the answer set without actially solving the problem (cf. Wahlstrom & Boersma, 1968).
This could fall into the Logic Conclusion category, but on a meta-level it is not directly related to the
actual questionnaire problems.

24
Reasoning process
Rarely was only the use of just one reasoning category used to obtain an answer, particularly when
answering Problem Set 2.a.1-2.a.4.
As explained before, a reasoning process is the cognitive processing of a problem and the reasoning
process may be made up by different reasoning approaches of different reasoning categories. It
needs to be clarified that although the reasoning approaches and their categories describes this
reasoning process, they still do not completely describe it. To fully describe a reasoning process a
trajectory of the person’s thinking would be needed, showing how the person engages with different
tasks in real time, which is not possible with the data source that was possible for me to obtain for
my study. However, based upon the extensive descriptions that I did obtain I would argue that my
categories do substantially capture the reasoning used. Here a limitation is that some of the
participants may not have been fully aware of all aspects of their reasoning themselves and thus were
unable to describe it completely.
As such, the reasoning categories used on a certain problem reveal significant aspects about the
nature of the reasoning process and at the same time brought out some interesting patterns, all of
which are able to inform educational practices in this area of introductory physics.
The participants were not, in general, consistent in their reasoning processes, the way of thinking
changed between Problem Sets and sometimes they had different reasoning categories in mind at the
same time. This is postulated to be a consequence of the different forms of representation used. More
discussion on this issue is provided in the section dealing with the (lack of) answer consistency.
In Problem Set 2.a.3 a participant had a Visual Memory of a person’s legs looking shorter and fatter
in water [2.a.3-1] in mind when answering the questions. This perception suggests that the refracted
image will be shifted both upwards and towards an observer (such an image would look fatter). But a
simultaneous Logic Conclusion [2.a.3-3] reasoning would could easily rules this out and lead to a
participant ignoring their Visual Memory.
Trying to track the reasoning process, maybe going back and forth between different reasoning
categories at very high speed, is not easily done and cannot be done efficiently by retrospective
interviews. The data in this study is not detailed enough to map any trajectories of thinking, all it can
show is the combinations of reasoning categories used in a process and what kind of reasoning
processes these patterns of reasoning categories would imply.
The combinations of reasoning categories sometimes suggested that there are some reasoning
approaches preexisted one another. In Problem Set 1 a participants recalled a wording or an image
they thought would help them splve the problem correctly, but for this to work this must precede the
use of memory to manifest some particular chain of reasoning, for example, Language Recall ->
Principle and Visual Recall -> Comparing. But in most cases there is no evidence of a standard chain
of thought emerging. The use of another reasoning category with a Principle could both be to reject
or to confirm the Principle where the trajectory of reasoning may be jumping back and forth between
different Reasoning Approaches.
For example, for Problem Set 1.2 two different principles [se 1.2-2/a,c in Appendix II] (different
wording but same implications) were preceded/confirmed in three different ways, resulting in six
possible combinations:
Language recall + principle [1.2-2/a or 1.2-2/c]
Comparing Figure + principle [1.2-2/a or 1.2-2/c]
Extreme case reasoning + principle [1.2-2/a or 1.2-2/c]
A scheme of possible reasoning processes used for Problem Set 1 combinations are illustrated in
{Figure 5}, where Visual Recall + Comparing used fundamental Principles to Compare get added to
the reasoning scenario.

25
Figure 5. Illustration of different possible reasoning processes used in Problem Set 1 (the
category in grey were not identified in Problem Set 1 but they were theoretical possible).
There are more theoretically possible combinations Principles of refraction derived/confirmed from
Snell’s law (Mathematics), the car model (analogical reasoning), or even with other principles like
Fermats and/or Huygens principles (which could be applied directly as well).
An adapted transcript of interview 12 and the interpretation shows the dynamics of the categories in
Problem Set 1 and 2 in a nice way in {Figure 6}.

Figure 6. An interpretation of selected parts of an interview, showing the use of the categories:
Language Recall, Extreme Case, Principle, Ray Tracing, Intuition and Comparing reasoning.

Three Problem Sets from the questionnaire, Problem Set 2.a.1, 3b figure 1 and 3c, were particular
interesting to compare with respect to the reasoning processes. These Problem Sets had a similar
conceptual meaning, but were represented in different ways and with slightly different framing of
the question. When compiling the reasoning processes below, some of the “help”-categories of
reasoning, being implicit categories, have been removed for an easier overview in {Table 4}, {Table
5} and {Table 6}. “Visual Memory” is the combinations of “Visual Recalling + Comparing”, while
“Comparing” in the tables only stands for the reasoning approaches of referring to results from other
Problem Sets. The participants that used the procedure of ray tracing did in some degree also use
scientific modeling to explain why they used it (although a unfinished model) and the category
Principles of Refraction to guide them how to do it. The reasoning category of procedure of ray
tracing is the doing, but in this section “Ray tracing” may refer to the combinations of reasoning
categories: “Scientific modeling + Ray tracing + Principle of refraction”.

26
Problem 2.a.1

The Problem Sets 2.a.1 and 2.a.2 (see {Figure 7} for 2.a.1 and Appendix I for both) were amongst
the most challenging for the 1200 international university students that completed this questionnaire,
54% and 52% selected incorrect answers [c] or [e] for the extended object to be “bending down”
(“[]” refers to the options of answer selections of the Problem Sets in this and the following
sections). My analysis shows that the participants in this study that got these two correct did so either
used Intuition or Visual Memory reasoning to solve the problem. None of them used Principle
reasoning associated with apparent depth, although they later used this approach with Problem Set
2.a.3, 3b and 3c. The effective use of Ray Tracing calls for at least two rays being used to obtain the
kind of correct answer provided in the Scientific Modeling characterization given earlier. If only one
ray is used then three possible options emerge, namely options [a], [c] and [d]. The Ray Tracing
procedure categorizations for my data all had single ray tracings and any elimination of options
came from using another reasoning approach. One exception was found in interview 13 where a
misleading ray tracing procedure provided the correct answer by coincidence. The eight other correct
answer selections [d] was not grounded in any alternative conceptions (conceptually misleading) as
in the case of interview 13, shown in {Table 4}.

Interview Answers Reasoning process (reasoning categories used)

1 c Using Snells law on object

2 c discarded because of principle Visual memory (oar) + Principle (2.a.1-3/b) + Ray tracing (2.a.1-6/b)

3 d Intuition (2.a.1-1/a) + Principle (2.a.1-3/a)

4 d Intuition + Ray tracing

5 d Visual memory (stick) + Ray tracing

6 c Ray tracing (2.a.1-6/b)

7 c Ray tracing (2.a.1-6/c) + Alternative theory (2.a.1-8)

8 c, e Ray tracing (2.a.1-6/d)

9 b,e Ray tracing (2.a.1-6/e)

10 a,c,d Analogical reasoning + Ray tracing

11 d Intuition + Visual memory (pencil)

12 d Intuition (2.a.1-1/b) + Extreme case reasoning + Ray tracing

13 d Ray tracing (2.a.1-6/f)

14 a Ray tracing

15 d Visual memory (pencil) + Ray tracing

*1 d, a Visual memory (pencil) + Ray tracing (2.a.1-6/f) + Alternative theory (2.a.1-9)

*2 d Visual memory (oar) + Principle (2.a.1-3/a)

*3 e Intuition + Using Snells law on object

*4 c Could not explain

Visual memory (looking flatter) + First thought later discarded Using Snells law on object + doubted own
*5 d
interpretation of visual memory a while Alternative theory (could look flat due to refraction)

5 Intuition, 7 Visual memory, 2 Principle (presumption of image being/not being related to observer),
Sum 9 / 20
1 Extreme case, 1 Analogical reasoning, 13 Ray tracing, 3 Using Snells law on object, 2 Alternative theory

Table 4. Shows the reasoning process behind the answers of Problem 2.a.1

27
Based on these results, the “bending down” options of [c] might be the result of a single Ray Tracing
Procedure reasoning and [e] to be the result of Procedure (Using Snell’s law on object), while the
correct [d] option got chosen when using Intuition or Visual Image reasoning. Why the [c] option
got chosen instead of [a] and [d] when using single-ray tracing is not explained by my data. The
geometry of the ray tracing procedure shown in {Figure 7} indicates that option [c] may intuitively
look more likely, because it involves the least bending of the light.

Figure 7. Problem 2.a.1: The use of a single ray tracing generates three options [a], [c] and [d]. I
suggest that the green single ray tracing approach would be the most likely single ray used
because it involved the least light bending..

Problem Set 3b -- figure 1

Using {Figure 7} I posit why participants selected the incorrect [c] option for Problem 2.a.1 yet the
correct option for Problem 3b-figure 1. All three ray tracing options suggest that the eye-to-image-
line intersects the surface further away than the eye-to-object line does. The single Ray Tracing
Procedure proved more fruitful in this problem and no Intuition or Visual Memory reasoning were
used (an additional explanation for this could be that they have not experienced as many fish-and-
spear-related situation as pencil-like object-in-water phenomenon). That being said, there was
generally a lack in consistency in the reasoning process used for Problem Set 2.a.1 and 3b figure 1.
{Table 5} illustrates this. Although not expanded upon for my study, it appears that some kind of
reasoning process link is made with the representation format of the question (cf. Linder & Priemer
2013).
The Problem 3b figure 1 gives a sketch of a fish in water and an arm holding a spear and asks how to
best aim the spear to catch the fish (see Appendix I). The correct answer [c] which called for aiming
below the fish was the most selected option in my data set (85 %). This contrasted with the 36%
selecting this in the 1200 international student survey (36 %). My analysis shows that the incorrect
options come from either “thinking backwards” Principle reasoning approach. Some of the correct
answers do arose from an alternative conception that had he image of the fish being further away
from the observer than the actual fish. The results of the survey based on the questionnaire answers
alone provide limited insight into conceptual understanding and it was beyond the scope of my study
to include a classification of the different forms of conceptual understandings.

28
Interview Answers Solid Reasoning process (reasoning categories used)
reasoning

1 c yes Principle (apparent depth)

2 c yes Ray tracing + Principle (apparent depth)

3 c yes Principle (apparent depth)

4 c yes Principle (apparent depth)

5 c yes Comparing (Problem Set 2b) + Ray tracing

6 c no Ray tracing notion of image further away from observer but at same depth

7 c no Ray tracing notion of image further away from observer but at same depth

8 c no Ray tracing notion of image further away from observer but at same depth

9 c no Ray tracing notion of image further away from observer but at same depth

10 c no Ray tracing notion of image further away from observer but at same depth

11 e yes Principle (apparent depth) thought “backwards” about the options

12 c yes Ray tracing

13 c yes Ray tracing

14 c yes Comparing (Problem Set 2b) + Ray tracing

15 c no Ray tracing notion of image further away from observer but at same depth

*1 c yes Principle (apparent depth) + Ray tracing

*2 c yes Principle (apparent depth) + Ray tracing

*3 b yes Thought of the picture as spear right above fish Comparing (Problem Set 1.1)

*4 c no Comparing (Problem Set 3a picture [ii.]) + Ray tracing

*5 e yes Thought of the picture as spear being thrown from the right making [e] the option closest to
thrower Ray tracing + Principle(apparent depth)

Sum 17/20 13/20 8 Principle (apparent depth), 15 Ray tracing, 3 Comparing

Table 5. Shows the reasoning process behind the answers of Problem 2.a.1

As shown in {Table 5}, no visual recalling or intuition was used by anyone; instead the principle of
apparent depth is being present. This principle is either referring to “apparent depth” in Question 2
or described as a tendency of an image being closer to surface. As described in the Principle section,
the same principle could be explicit or implicit, sometimes a fundamental principle and sometimes
derived. Principle of apparent depth have these different statuses, although most times the status in
unclear. The way the Problem Sets are ordered in the questionnaire could explain that a principle
was used in this but not in the similar Problem 2.a.1. The “visual recalling + comparing” and
“intuition” could have emerged to a principle naturally, making the reasoning processes very much
the same although the categories being different.

29
Problem Set 3c

In this Problem Set the participants were asked to draw how a pencil would look on a photo of a
situation similar to that of Problem 2.a.1, with the part of the pencil in water edited out. Being a 2-
dimensional image of a 3-dimensional situation, the task of evaluating the students’ answers is not
easily done; the same drawing could represent different solutions (the pen bending towards or away
from the observer looking the same for instance). In my analysis two things have been noted, how
“real looking” the image is in my point of view and how solid their reasoning have been when the
participants explained what their images really are suppose to show in the interviews. {Figure 8}
shows my classifications of real looking, partially real looking and not real looking and {Table 6}
gives a summary of the result.

Figure 8. Real looking, partially real looking and not real looking images of pen-in-water
refractions. Examples of the “Yes”, “a bit” and “No” assessments of Table 6.

The assessments of how “real looking” the drawings was rather subjective, and it did not reveal what
kind of reasoning approach was being used. The real looking drawing shown in {Figure 8} could be
interpreted as the pen either bending towards observer, or away from observer. If the drawings could
be interpreted as a real scenario, they have been labeled “real looking”. Fifteen of the answers
included looked least a bit real, but only 10 had a solid reasoning that went with that. Solid reasoning
means that it does not lead to alternative conceptions being formed. Examples of solid reasoning is
being found in using Visual Memory (Recalling + Comparing) and a Comparison (with solution of
Problem 2.a.1) reasoning categories. One might argue that these reasoning approaches sometimes
are being based on the (false) assumption that the image is not related to an observer. None of the
participants revealed an entirely correct conceptualization of objects-image positioning in water for
this problem set. None appeared to be aware of the image being slightly towards the observer and
above the object. In interview 12, the participant compared this problem with Problem 2.a.1 in order
to account for the tip of the pen being closer to the observer, but they did not reach a correct
conclusion that had the pen appearing lifted above the object in general without any shifting towards
the observer.

30
Interview Real Solid Reasoning process (reasoning categories used)
number looking reasoning

1 No no Using Snells law on object

2 Yes yes Visual memory (oar)

3 Yes yes Visual memory (oar, pencil)

4 Yes yes Comparing (Problem 2.a.1)

5 a bit yes Visual memory

6 a bit no Visual memory (leg in water getting higher up and (misleading) further away)

7 No no Visual memory + Using Snells law on object

8 No no Comparing (Problem 2.a.1)

9 No no Visual memory (spoons) found the multiple dimensions tricky while comparing images

10 Yes no Intuition (image felt real) + Ray tracing higher up and further away

11 Yes yes Intuition (image felt real)

12 Yes no Principle (apparent depth) + Visual memory (legs) + getting stretched out Comparing
(Problem 2.a.1)

13 a bit no Ray tracing higher up and further away + Visual memory (arms)

14 Yes yes Visual memory (pencil)

15 a bit yes Visual memory (pencil)

*1 Yes yes Visual memory (oar, stick)

*2 Yes yes Visual memory (oar) + Comparing

*3 No no Comparing (Problem 2.a.1)

*4 Yes no Ray tracing further away

*5 Yes yes Comparing (Problem 2.a.1)

Sum 15/20 10/20 2 Intuition, 12 Visual memory, 1 Principle (apparent depth), 3 Ray tracing, 5 Comparing, 2
Using Snells law on object

Table 6. This table shows my assessments of the drawings, if the reasoning behind the
solutions where found solid and what reasoning categories were used in the reasoning process
of Problem 3c.

31
Consistency

Interview Problem 2.a.1 Problem Set 3b Problem Set 3c Solid

number figure 1 reasoning
1 Using Snells law on object Principle Using Snell’s law on 1/3
object
2 Ray tracing + Visual Ray tracing + Visual memory 2/3
memory + Principle Principle
3 Principle + Intuition Principle Visual memory 3/3
4 Intuition + Ray tracing Principle Comparing 3/3
5 Visual memory + Ray Comparing + Ray Visual memory 3/3
tracing tracing
6 Ray tracing Ray tracing Visual memory 0/3
7 Ray tracing + Alternative Ray tracing Visual memory + Snells 0/3
theory law on object
8 Ray tracing Ray tracing Comparing 0/3
9 Ray tracing Ray tracing Visual memory 0/3
10 Analogical reasoning + Ray tracing Ray tracing + Intuition 1/3
Ray tracing
11 Intuition + Visual memory Principle Intuition 3/3
12 Intuition + Ray tracing + Ray tracing Principle + Comparing 2/3
Extreme case reasoning + Visual memory
13 Ray tracing Ray tracing Ray tracing + Visual 1/3
memory
14 Ray tracing Comparing + Ray Visual memory 2/3
tracing
15 Visual memory + Ray Ray tracing Visual memory 2/3
tracing
*1 Visual memory + Ray Principle + Ray Visual memory 2/3
tracing + Alternative tracing
theory
*2 Visual memory + Principle Principle + Ray Visual memory + 3/3
tracing Comparing
*3 Intuition + Using Snells Comparing Comparing 1/3
law on object
*4 - Comparing + Ray Ray tracing 0/3
tracing
*5 Using Snells law on object Ray tracing + Comparing 3/3
+ Visual memory + Principle
Alternative theory
∑Answers 9/20 17/20 15/20
∑Solid 8/20 13/20 11/20 32/60

Table 7. Consistency of the use of reasoning processes across three Problem Sets. Summary
of reasoning categories for the three Problem Sets. Consistency of the reasoning process for all
three are shown in bold. The number of Solid Reasoning refers to if the solution did not include
any alternative concepts; concepts not part of the general knowledge of modern physics.

32
{Table 7} shows that only four of the participants were consistent with the reasoning processes that
they drew on to solve the image-object problem that the Problem Sets presented. Only two of these
participants presented correct answers. Interestingly enough, the reasoning processes were unique
even for these two consistent participants. Some participant were consistent in their reasoning of
Problem 2.a.1 and either 3b (interview 6, 7, 12) or 3c (interview 1, 5, 11, 15, *1, *2, *5), but others
showed no consistency at all (even though the reasoning categories would suggest it, Ray Tracing,
for example, was often not used in consistent ways). Interview *5 is an example of the participant
trying to find consistency in the answers of 2.a.1 and 3a (coin in a cup) and 3b. This led to the
participant re-prioritizing the contradicting reasoning approaches of the reasoning process in
Problem 2.a.1 leading to the Visual Memory approach to be the basis of the answer instead of
Procedure “Using Snell’s law on object” or “Alternative theory” approaches. This led to the
contradictions that they saw disappear, but the difference in reasoning processes was too large for
me to be able to judge them to be coherent. The same thing goes for interviewee *2, which being
more coherent, is very close to interviewee 3. The borders between consistency and inconsistency
and its relation to coherency cannot, with my data set, be distinctly clarified.
An example of a lack of consistency: interview 2
2.a.1) The participant did have a visual recalling of an oar in water but decided not to use it because
of a principle (presumption) of image-position being related to observer. This is a correct principle,
but it is a small effect that does not affect the image-position much. Instead he use a procedure of ray
tracing rendering the wrong answer of [c]: the image of the tip of the pen at deeper altitude than real
tip.
3b figure 1) Ray tracing was used, but also a statement of the fish “looking to be higher up than it
really is”, contradicting the results of 2.a.1.
3c) Uses memory of an oar in water, although he did not use that same memory in 2.a.1. The
observer in the Problem Set is still not in the same place as the observer of the memory.

33
Discussion
My categorization of reasoning approaches can be seen to have connections with the framing that
Haglund (2012) describes in terms of three theoretical reasoning platforms: Analogical reasoning,
Scientific Modeling and Semantics. Semantics is spoken and written language, gestures,
mathematics and visual (pictorial, schematic and diagrammatic) representations. My categories of
Principle, Recalling and Comparing fit within Semantics and one might argue that the other
Procedure categories would fit into Scientific Modeling. What extends Haglund’s framing are my
Intuition, Extreme Case Reasoning, Relation Reasoning, Logic Conclusion, Alternative Theories and
Test-wiseness categories. My results can bee seen as a confirmation of Haglunds (2012) framework
(the termss Analogical Reasoning and Scientific Modeling were found in his PhD thesis after I had
completed my analysis). My results also suggest that Analogical Reasoning and Scientific Modeling
are not commonly used by students doing refraction problems.
The reasoning categories that I formulated have two basic explanatory properties, one of describing
the participants reasoning processes in general and one of analyzing the questionnaire results. Being
a case study, the value of reasoning categories still needs to be further explored. In my study, the
reasoning categories show the nature of the reasoning processes, but may not fully describe them.
Moreover, the reasoning process may not be exactly the same thing as what Redish (1994:796) calls
“Mental models”, in that my categories of the reasoning process characterize the reasoning being
used to explain how a problem was solved. Redish’s Mental Models refers to thinking that is
accessible. My reasoning categories do, however, fit in with Redish’s properties of “mental models”:

“Mental models have the following properties:

1. They consist of propositions, images, rules of procedure, and statements as to when and
how they are to be used.
2. They may contain contradictory elements.
3. They may be incomplete.
4. People may not know how to "run" the procedures present in their mental models.
5. Elements of a mental model don't have firm boundaries. Similar elements may get
confused.
6. Mental models tend to minimize expenditure of mental energy. People will often do extra
physical activities -- sometimes very time-consuming and difficult -- in order to avoid a
little bit of serious thinking.” (Redish 1994: 796)

Perhaps the only exception for my categories is Intuition. The reason why Intuition does not occur
on this list is that it is unlikely that intuition would be an approach that is purposefully “chosen.
My results show that the reasoning that the participants in the study used often were not consistent
(essentially for the same problem), Procedures were being used although without knowing how to
appropriately use them. The last property, a tendency of minimizing expenditure of ”mental energy”,
which I rather call “cognitive effort”, may explain why some of the reasoning categories in the study
were being used less frequently than others. The reasoning categories were divided into six major
reasoning groups: Intuition, Recalling, Principle, Contemplating, Procedure and Others, which
represent different ways of problem solving. The use of Contemplating (Extreme case reasoning,
Relation reasoning, Analogical reasoning, Scientific modeling, Logic conclusion) and Other
(Alternative theory and Test-wiseness) were the ones used less frequently; perhaps they are the ones
that call for the most “cognitive effort”.

34
Stamovlasis and Tsaparlis (2005:9-10) lists four cognitive variables in problem solving:
 Working-memory capacity – A persons limited capacity to store and process
data at one time (usually 7 ± 2 chunks of data)

 Cognitive Style/Disembedding Ability – Ability to distinguish relevant information from

irrelevant noise

 Developmental Level – Ability to use formal reasoning (logic reasoning)

 The Mobility–Fixity Dimension – How field-dependent/independent a person is

Given that these cognitive variables (measured in different kinds of tests) describe not only what
person are capable of but also the cognitive effort they put in, this could be used to estimate the
cognitive effort of a reasoning approach. The Contemplating categories and Alternative theory
category acquire more use of the working memory in general than the other categories, as there are
more things to keep in mind and process simultaneously. The limited working-memory capacity
might even make some reasoning approaches impossible for some participants (without using tools
like notes and calculators to reduce the amount of data to keep in mind). The Contemplating
categories and Test-wiseness include formal reasoning, acquires a higher developmental level than
the other categories. My own experience, using the everyday meaning of the term “cognitive effort”,
corresponds to these categories being more effort consuming as well. Recalling, being the browsing
of the long-term memory, could also require a great use of cognitive effort, at least for me, but have
the benefit of being a one-time “cognitive cost”, a memory found do not acquire much cognitive
effort to re-recall, being a low effort reasoning approach in the long run. This argument is being
illustrated in {Figure 9}:

Figure 9. A hypothetical graph of the reasoning groups related to the amount of cognitive effort
they acquire, on average, for solving a physics problem.

A use of analogical reasoning and scientific modeling is crucial to attaining conceptual

understanding of physics and to be able to solve new physics problems. As these categories are not
widely found in the analysis of the questionnaire, this could be an explanation for the many incorrect
answers and lack of consistency. Why are these reasoning categories not used by the majority of the
participants? Why is no one trying to understand the situation of Problem 2.a.1 by asking
themselves, how does the eye create an image of an object at a certain position?
There are a couple of possible explanations or factors for this, one being on the cognitive level,
another being the different representations and the framing of the problems, the survey situation in
itself should also be scrutinized.

35
Explaining the results of the questionnaire

Assuming that {Figure 9} shows a correct generalization of cognitive effort and the reasoning
categories, the simple conclusion would be that the participants in this study, and the students doing
the questionnaire, performed badly because of their unwillingness to expend cognitive effort. This
conclusion does not correspond to what was found in the interviews though. The participants did
give the impression of wanting to perform well, as well as not being sure about their answers, so it is
hard to see why they would not go the extra mile to satisfy themselves with a correct answer. Maybe
because the other categories were more “effort-efficient”, they would usually use those categories
for most problem solving, making them unfamiliar to the reasoning approaches of the other
reasoning categories. This may be caused by,for instance, that the teacher explains by analogical
reasoning but never gets the student to use that kind of reasoning themselves to solve problems
because of the principles or procedures derived from this more conceptual analogical reasoning
being more effective to use in the problem solving of the exams. Redish (1994:1) points out that
physics teachers fails to make the students understand instead of just memorizing bits and pieces:

“We have readjusted our testing so that the students can succeed and we have
then either fooled ourselves that we are teaching them successfully or
lowered our standards by eliminating understanding from our definition of
successful learning.”

Now, if students, like the participants in my study,experience never being asked to reason in a
certain way, a rational conclusion of this is that they would not reason this way, even if they have
been presented with this kind of reasoning in class. Maybe this kind of reasoning is too advanced to
ever be requiredfor an answer, in other words, there must be a way of solving a problem in the usual
way, otherwise it would not be presented to them, the person giving the problem would never be that
cruel. Another possibility is that reasoning approaches of this kind have too much uncertainty in
themselves; it could either be that they render easy errors or give too little information.
Going back to Problem 2.a.1, why are no participants trying to understand the problem by
Contemplating? Possible factors are:
 Unwillingness: It acquires too much cognitive effort
 Educational system: This kind of reasoning is never needed, the hidden curriculum is
teaching the students that there is always another way to reason that is better in some way.
I would suggest the following additional factors:
 Too hard a problem: The participants might have tried to understand the problem but failed
and felt too embarrassed to talk about it in the interviews.
 Representations: The reasoning categories of Contemplating might be associated with a
certain type of representation.
 Design of the problems: the framing of the problems might influence the participants
 Design of the survey: the framing of the questionnaire might influence the participants
The understanding of Problem Set 2.a.1 is not trivial. To understand what an apparent position or
virtual image really is, knowledge in physics alone is not enough, ontologically it is an immaterial
entity – nonphysical - which is not common in physic problems, scientific it is linked to how the eye
and brain works. Ontology and medicine is not a physics student’s greatest field of expertise. When I
myself did this questionnaire, prior to knowing about the answers, I did try to understand the
situation of Problem Set 2.a.1 by constructing a scientific model of two rays, but not in the correct
way. I could not figure out how the eye and brain perceived distances to objects, my model was
based on an alternative theory of this depth perception being based on the distance between the rays
of different points of the object hitting the eye at the same angle. I knew that I was treading on thin
ice and maybe some of the participants had a similar experience of reasoning and did not want to
account for it in the interviews. If some of them did try to use a Contemplating reasoning approach

36
but immediately found it useless or over their cognitive capacity, it is not hard to understand why
they either forgot about it or choose not to mention it.
Redish (2003:30-39) has selected five general principles to help understand how student thinking
works in the physics classroom:
 The constructivism principle: “Individuals build their knowledge by making connections to
existing; they use this knowledge by productively creating a response to the information
they receive.” (Redish 2003:30)
 The context principle: “What people construct depends on the context—including their
mental states.” (Redish, 2003:31)

 The change principle: “It is reasonably easy to learn something that matches or extends an
existing schema, but changing a well-established schema substantially is difficult.” (Redish,
2003:33)

 The individual principle: “Since each individual constructs his or her own mental structures,
different students have different mental responses and different approaches to learning. Any
population of students will show a significant variation in a large number of cognitive
variables.” (Redish, 2003:37)

 The social principle: “For most individuals, learning is most effectively carried out via
social interactions.” (Redish, 2003:39)
The constructivism principle could help explain the factors of unwillingness, the problems being too
hard and educational system described earlier. Ultimately this explains why the more associative
reasoning categories that connect to existing knowledge and give a response more directly like the
Recalling and Principle categories are more frequent. The Logic reasoning imbedded in the
Contemplating reasoning categories make the connections more far off and gives a delayed response.
Redish (2003:31) addresses standard testing methods as a problem because of the narrowed
situations, hint and clues that result in not testing the more complex patterns of associations needed
to really understand physics.
The context principle relates to the representations used and design of the problems as well as the
design of the survey. The “change principle” shows why misconceptions are hard to get rid of and
the existence of some of the Alternative theories. The “individual principle” seems to correspond to
my data, {Table 7} shows the big variety in reasoning, the reasoning processes were very individual.
The “social principle” could explain the poor results of the survey. The social aspect of learning that
was removed in the survey is especially important when facing new kinds of problems as these were
to the majority of the participants.
The question of how the representations of a physic problem affect the way the students try solve it
is what started this area of research in the first place. The results of Problem Set 2.a.1 (schematic
drawing), 3b (pictorial drawing) and 3c (photograph) show that the reasoning categories differ a lot
with the kind of representation being used: 2.a.1 being mostly Recalling and Procedure, 3b being
Principle and Procedure and 3c being Recalling. There are only two instances of Contemplating
categories (though Science modelling is implicit in some of the Ray tracing operations), both in
Problem Set 2.a.1, one of Extreme case reasoning and one of Analogical reasoning. As there were
only two instances, these three representations cannot really explain why the Contemplating
categories were not used more frequently. I do however believe that the representations do matter.
As shown in the results, no Principle of apparent depth was used in Problem Set 2.a.1 and no Visual
memory was used in 3b. Maybe an even more condensed representation similar to the ray diagrams
(could be called physical representation) in Problem Set 1 is needed for the participants to begin
reasoning in a more Contemplating way. What if the tip of the pen would have been represented by
points instead {see Figure 10}, the Procedure of “Using Snell’s law on object” might then be
reduced.

37
Figure 10. A copy of the image for Problem 2.a.1 with the object cut out and only the given
object and image positions being shown (as points)
If more “physical” detail were being added to this representation, such as n=1 and n=1,33 for the
respective refractive indices in the place of “air” and “water” and an actual lens image replacing the
eye, then I suggest that then the participants would have been more inclined to use Scientific
Modeling with the help of Ray tracing.
It could be that it is not all about the representation of the situation and the information given, but
also the design of the question that influence how students solve it. Problem Set 2 was designed as
five sub-Problem Sets on an A4-page with multiple choice options. The pictures were rather small, a
two-ray tracing method in Problem Set 2.a.1 would require a magnifying glass, along with a
protractor and ruler to help solve the problem. This design makes it harder to use the picture along
with other tools. In the framework of Airey and Linders (2009) disciplinary discourse the pictures is
not only a representation but also a tool, but the small size of it makes it less of a tool. The design of
the survey in general does not encourage the use of tools like the calculator, protractor or ruler. The
message the questionnaire sends out is that it should be done without using any calculations, notes or
drawings. This, the design of the survey, might be a big factor in constituting the reasoning
processes. The number of Problem Sets, the multiple choice answers, and the small pictures
indicates that every Problem Set could be done quickly and pain free (in the sense of low acquired
cognitive effort).
Question 2 is an exception to this. Here, the participants were being asked to describe “apparent
depth” in text and drawings. However, the phrase was unknown to most of them, and only a few
described it properly before doing Problem Set 2.a.1, some of them saved Question 2 until the end.
Still, no one used the principle of apparent depth to guide them, which may be because Problem Set
1 and Ray tracing were still fresh in their minds. Another reason for this may be that the participants
did not realize that a tip of the pen could be regarded as an individual object, maybe {Figure 10}
where the pen is not shown would render more Principle reasoning as well.

Implications for physics education

An awareness of the different kinds of categories of reasoning approaches would be very helpful for
teaching design and practice aimed at optimizing learning.
Students’ difficulties with physics and a possible reason for the drop-out rate from university physics
could be that to attain fluency in the disciplinary discourse of physics one does not only need to
master different modes of representation, but also the different modes of reasoning associated with
those representations. Such a double kind of fragmentation is disconcerting for many students and its
consequences need to be taken seriously in physics education at all levels.

Students need more guidance with their reasoning processes than they are currently receiving, and if
the different reasoning approaches are contradicting each other need guidance in their choices.

38
Are some reasoning categories better than others? That will depend on the problem at hand. To
construct a conceptually holistic understanding, my study raised the following issues for teachers of
university physics:
Intuition: Some physics is known to be counter-intuitive and generally the further away from our
macroscopic world the problem is, the less useful intuition typically is. Quantum mechanics and
cosmology in particular present are counter-intuitive learning environments in the everyday sense of
the word.
Recalling: The same applies for Visual Recalling where a real world image is arguably the most
useful if the problem is set at the macroscopic level. A Recalling of a Principle is generally not the
best way to tackle a problem unless it is back by a sound conceptual understanding.
Principle: Fundamental principles are essential aspects of physics learning (like the speed of light
being a postulate in Einstein’s relativity theory) but some principles are just applicable in certain
situations. Several of the principles used by the participants in my study derived those principles
from other reasoning categories. It would be more educationally useful if such derivation of the
principle is known. Otherwise, the conceptual understanding is being lost.
Contemplating: Is the best way of reasoning to acquire new knowledge in physics, but often needs
the help of Principles and Procedures to be effective. It seems as if the students require some training
in these kinds of reasoning for problem solving instead of, for example, simply watching teachers
doing scientific modeling and solving problems.
Procedure: An essential part of problem solving that needs a solid conceptual backing. . This means
that a solid conceptual understanding of the problem is needed before the procedure can be used
effectively.
Other: Alternative Theories lead to a alternative conceptions that are not appropriate for use in
physics problem solving scenarios. Teachers need deeper insight into these alternatives in order to be
able to address them effectively. Test-wiseness is a problem which occurs for all answer selection
tests and teachers should be aware of the possibility that their tests do not measure the knowledge
they are meant to.
The students in my study tended to use a Principle or Procedure to solve the given physic problems,.
I can see a rationale for this – it reduces cognitive effort. However, such an approach quickly
becomes problematic if one cannot link those reasoning approaches back to a more conceptual,
logical and coherent understanding of physics. There seems to be at least two kinds of skill-sets that
are needed to facilitate the attainment of a more comprehensive conceptual understanding; a
familiarization of different kinds of reasoning categories, and modes of representations. To only
master a few reasoning approaches and a few representations will lead to fragmentation of the kind
that turns problem solving into a really challenging experience.
To present and solve the same problem using different representations and using different reasoning
approaches would be a good didactic approach to avoid this fragmentation. The educational system
could then be strengthened by not reproducing problems presented and solved in one particular way
without any need for conceptual understanding. The results of the survey shows that the university
has partially failed to teach students how to understand refraction and object-image positioning and
to give the students the means to solve problem set in refraction.

39
Future research

To validate a more generalized theory of reasoning processes more studies like the one I carried out
need to be done across a broad spectrum of areas in physics. More of these studies could also
determine the relationships between the reasoning categories and different representations; this could
be done by giving the exact same conceptual problem in different representations to different people
with similar knowledge in physics.
Not only are the relationships between the reasoning categories and representations of pedagogical
importance and interest, but offer the knowledge to enable change in course and teaching design. To
find out how the framing of the questionnaire are influencing the reasoning approaches a couples of
different surveys with essentially the same Problem Sets with some variation on some parameters
could be send out to students. These changing parameters could be different sizes of the pictures,
Problem Sets in different orders or different ways to select answers. To find out if unwillingness to
make cognitive effort influence the reasoning approaches, the changing parameter could be the
motivations to solve a given problem correctly. One way to do this would be to reward the correct
answer with some sort of prize.
Moreover, with Clement’s (2008) non-formal reasoning and Airey and Linder’s (2009) disciplinary
discourse as a starting point, my reasoning categories could be used to find the most effective ways
of teaching refraction through an application of a greater theory of student reasoning on refraction.
The most worthwhile future research would be a mapping out of the reasoning processes by
trajectories of students’ thoughts while solving problems. If such studies could include exploring
how different reasoning categories interact and complement each other in detail, the pedagogical
value would be enormous. This is not easily done; some sort of observations combined with
interviews would be required, maybe even including neuroscience and brain scans. This would
facilitate studies involving student reasoning related to measuring of the cognitive variables
(Stamovlasis & Tsaparlis, 2005) giving another powerful synergy of different field of research.

40
References
Airey, J., & Linder, C. (2009). A Disciplinary Discourse Perspective on University Science
Learning: Achieving Fluency in a Critical Constellation of Modes. Journal of Research in Science
Teaching, 46(1), page 27-49.
Clark, J., (2013) ‘Into the fields…’ Matters and methods in Interpretive research. Uppsala University
notes for a workshop on interpretative research.
Clement, J.J., (2008). Creative model construction in scientists and students: the role of imagery,
analogy, and mental simulation. Dordrecht, NL: Springer.
Crotty, M. 1998. The foundations of social research: Meaning and perspective in the research
process. Sydney: Allen & Unwin.
Davis B, & Sumara D., (2008) Complexity as a theory of education, Transnational Curriculum
Inquiry, 5(2), page 33-44.
Driver, R & Erickson, G., (1983) Theories in action: Some Theoretical and Empircal issues in the
Study of Students’ Conceptual Framework in Science Studies in Science Education 10, page 37-60.
Dye, J. F., Schatz, I. M., Rosenberg, B. A., & Coleman, S. T., (2000). Constant comparison method:
A kaleidoscope of data. The qualitative report, 4(1/2), page 1-9.
Forsman, J., (2011). Exploring using complexity thinking to extend the modelling of student
retention in higher education physics and engineering. Licentiate Thesis, Department of Physics and
Astronomy, Uppsala University, Sweden.
Fredlund, T., Linder C., Priemer B., Boczianowski, F. & Pohl, A. ( 2013). Perceptions and forms of
reasoning. Refraction of light: a study of virtual image predictions. Poster presented at the
Foundations and Frontiers of Physics Education Research Conference, Bar Harbor, Maine, 17-21
June.
Fredlund, T., Airey, J., & Linder, C. (2012). Exploring the role of physics representations: an
illustrative example from students sharing knowledge about refraction. European Journal of Physics,
33(3), page 657-666.
Haglund, J., (2012). Analogical reasoning in science education – connections to semantics and
scientific modelling in thermodynamics. Doctoral Thesis, Linköping, Sweden: The Department of
Social and Welfare Studies at Linköping university.
Harrison, A. G., & Treagust, D. F. (1993). Teaching with analogies: A case study in grade-10 optics.
Journal of Research in Science Teaching, 30(10), 1291-1307.
Hecht, E., (2002), Optics, 4e Edition, Addison-Wesley.
Horn, Laurence R., (2014), "Contradiction", The Stanford Encyclopedia of Philosophy , Spring 2014
Edition.
H ttebr ucker N., (2010) Empirische Untersuchung zum Verständnis ikonischer Repräsentationen
am Beispiel der optischen Brechung. Bochum, Germany: Der Fakult t f r Physik und Astronomie
der Ruhr-Universität Bochum.
Kvale, S., (1996) Interviews : an introduction to qualitative research interviewing. Thousand Oaks,
CA:Sage
Langley, J., (2006). Anatomizing the early-modern eye: a literary case-study. Renaissance Studies,
20(3), page 340-355.
Lincoln, Y.S. & Guba, E.G., (1985). Naturalistic Inquiry. Newbury Park, CA: Sage Publications
Linder, C. (2013a). An investigation of student ability to use different forms of representation in
applying basic principles of the refraction of light to make image predictions. Paper presented in the

41
Discipline Based Education Research - A View from Physics Invited Symposium at the ESERA
(European Science Education Research Association) Conference, University of Cyprus, Nicosia, 2-
7th Sept.
Linder, C. (2013b). Disciplinary discourse, representation, and appresentation in the teaching and
learning of science. European Journal of Science and Mathematics Education, 1(2), page 43-49.
Linder, C. & Priemer, B. (2013). A social semiotic understanding of representation linking and
affordance in physics: The refraction of light. Paper presented at the ESERA (European Science
Education Research Association) Conference, University of Cyprus, Nicosia, 2-7th Sept.
Martin, B. & Brouwer, W. (1992) The subjective side of science. Alberta Thy-13-92, page 1-21.
McDermott, L. C. (1991). Millikan lecture 1990: What we teach and what is learned—Closing the
gap. American Journal of Physics, 59(4), page 301–315.
Melamed, Y & Lin, M., (2013). "Principle of Sufficient Reason", The Stanford Encyclopedia of
Philosophy, Summer 2013 Edition.
Nassar, A. B., (1994) Apparent Depth. The Physics Teacher, 32, page 526-529.
Ornek, F., Robinson, W.R. and Haugan M.P., (2008) What makes Physics difficult? International
Journal of Environmental & Science Education, 3(1), page 30 – 34.
Redish. E.F., (1994). The Implications of Cognitive Studies for Teaching Physics. American Journal
of Physics , 62(6), page 796-803.
Redish. E.F., (2003). Teaching Physics with the Physics Suite, New York: John Wiley & Sons.
Robson, C., (1993). Real world research: a resource for social scientists and practitioner-researchers.
Oxford, UK: Blackwell
Skolverket. (2012) Mathematics
http://www.skolverket.se/polopoly_fs/1.174554!/Menu/article/attachment/Mathematics.pdf (2014-
03-11)
Stamovlasis, D & Tsaparlis,G., (2005) Cognitive variables in problem solving: a nonlinear
approach. International Journal of Science and Mathematics Education, 3(1), page 7-32.
Stephens, A. L., & Clement, J. J. (2010). Documenting the use of expert scientific reasoning
processes by high school physics students. Physical Review Special Topics - Physics Education
Research, 6(2)
Van Heuvelen, A., (1991). Overview, Case Study Physics. American Journal of Physics, 59(10),
page 898-907.
Vetenskapsrådet. (2002). Forskningsetiska principer inom humanistisk-samhällsvetenskaplig
forskning. Sweden, Stockholm: Vetenskapsrådet.
Wahlstrom, M & Boersma, F.J., (1968) The influence of test-wiseness upon achievement,
Educational and Psychological Measurement, 28(2), page 413-420.HE INFLUENCE OF TEST-
WISENESS
UPON ACHIEVEMENT

42
Appendix
Appendix I: Questionnaire
In Swedish, not in real size but a bit contracted to fit.
Question 2 is referring to the question on page 44 (Question 1 is the background info on page 43 and
not part of the study)
The numbering of the Problem Sets follows the one in the questionnaire.

43
44
45
46
47
48
Appendix II: Identifications of reasoning approaches

Question 2

In question 2 the following reasoning approaches were identified:

Visual recalling
The participants recalled a visual memory in order to explain and/or figure out what apparent depth
meant.
Q2-1/a) A visual memory of putting an oar or a pencil in water.
Q2-1/b) A visual memory of looking at the bottom of an aquarium.
Ray tracing
The participants used ray tracing to show themselves and/or the interviewer the difference in
apparent depth and real depth.
Q2-2) The ray do not travel in a straight line due to refraction but the eye thinks it does, resulting in
an apperent depth more shallow than the real depth.
Comparing

The participants saved this question to the end and used the result of Problem Set 2b) to explain
and/or figure out what apparent depth meant.
Q2-3) Comparing with Problem Set 2b)
Alternative theories
Q2-4/a) Light traveling at a lower speed in thicker medium, giving a longer virtual distance.
Q2-4/b) Effect of the light being absorbed and/or scattered.
Q2-4/c) Misinterpretation. Perspective in a painting.
Q2-4d) Misinterpretation. Thinking of ”siktdjup” (line of sight in water).
Testwiseness
Q2-5) Based on the following Problem Sets in the questionnaire, it has to do with refraction of light.

49
Problem Set 1

The *-sign marks that these findings comes from the 5 later interviews.

Problem Set 1.1)

*Intuition
The participants stated that they did solve it by intuition, although a further explanation of this
wasn’t always given.
1.1-1) *When looking on something straight ahead I get the intuitive feeling that it doesn’t bend in
either direction.
Language recalling and principle
The participants used a principle to get the solution; often they recalled a spoken or written wording
of this principle.
1.1-2) Principle of refraction: No refraction when a ray is perpendicular to the surface
Visual recalling and comparing
The participants had an experience of working with similar refraction problems when applying
Snells law and recalled a visual image of the ray path and compared with figure 1.1).
1.1-3) Experience from working with Snells law
* Scientific modelling
1.1-4) * Looking at the ray as the front of a plain wave, when reaching the border surface the wave
hits it perpendicular, not resulting in the same effect of the front turning as when the front hits the
surface at different times .
Logic conclusion
The participants stated that they found a straight trajectory of the ray logic. One of these participants
argued in a Problem Set 2a.3) that when there is no reason for the ray to go left and no for going
right the logical conclusion is that it does neither (similar to Leibniz principle of sufficient reason)
and it could very well be the same reasoning In this Problem Set. Other participants said that no
angle of incident would implicate no angle of refraction.
Maybe the same reasoning as in 2a.3-2) No reason for the image of the object to bend to left/right so
it doesn’t.
1.1-5) * No angle of incident would implicate no angle of refraction
Mathematics
The participants used the Snells law formula to solve the problem.
1.1-6) If one angle is 90 degree the other one must also be 90 to get both sides to 0 in the formula

50
Problem Set 1.2), 1.3) and 1.4)

Figure refers to the figure on the right, Problem Set figure refers
to the figures the participants was asked to complete.
Language recall and Principle
1.2-1/a) When light travels from a optical thicker into a optical
thinner medium, it refracts away from the normal of the border
surface
1.2-1/b) When light travels from a optical thinner into a optical thicker medium, it refracts towards
the normal of the border
1.2-1/c) The angle to the normal of the border surface is always lower in the thicker medium
Principle
1.2-2) The light going the same path no matter the direction
1.2-3) Change of coordinates doesn’t affect the physics (Lorentz invariance)
Extreme case reasoning
1.2-4) Derive the principles of refraction by looking at total internal reflection as an extreme case
Analogical reasoning
1.2-5) The light being modeled as a car, passing thru the border surface into a thicker or thinner
medium leading one of the wheels getting slower or faster and the car to turn.
* Scientific modelling
1.2-6) * Looking at the ray as the front of a plain wave, when reaching the border surface the wave
hits it at different times, making the wavefront be turned clockwise or counterclockwise depending
on the angle of incidence and if refraction index gets greater or smaller.
Mathematics
1.2-7) Using Snells law formula: refraction index times angle is an proportional equality after and
before refraction
Comparing
1.2-8) Use figure given in questionnaire combined with principles 1.2-3) and 1.2-4) to compare with
Problem Set figures
1.2-9) Use figure given in questionnaire to derive the principles of refraction

Problem Set 2

Problem Set 2.a.1 and 2.a.2

Intuition
2.a.1-1/a) Feeling that image should be shorter and bent upwards.
2.a.1-1/b) Feeling that image should be bend
2.a.1-1/c) Misleading. Feeling that image should be farther away from observer than object.
Visual recalling
2.a.1-2) same as Q2-1/a) A visual memory of putting an oar or a pencil in water.

Principle

51
2.a.1-3/a) Image being related to the position of observer
2.a.1-3/b) Image not being related to the position of observer
2.a.1-4) Same as 1.2-3) Change of coordinates doesn’t affect the physics (Lorentz invariance)
Extreme case reasoning
2.a.1-5) The extreme cases of (n being refraction index) n_1=n_2 and n_2 growing very big being,
the image should be somewhere in between (and stay near the object).
Ray tracing
The participants used ray tracing with or without help from Problem Set 1.x) reasoning on refraction.
Some of them helpful and some misleading.
2.a.1-6/a) Straight line eye-image then surface-object to check (with Snells law or experience) if
possible
2.a.1-6/b) Object-surface-eye then extend the eye-surface line to eye-image (or check with Snells
law or experience)
2.a.1-6/c) Eye-surface-object then extend the eye-surface line to eye-image (or check with Snells law
or experience)
2.a.1-6/d) Misleading. Straight line Object-eye then refracted line surface-position below eye.
Then draw parallel line eye-surface-image
2.a.1-6/e) Misleading. Image-surface-eye then extend eye-surface to eye-object
2.a.1-6/f) Misleading. Object-surface-eye, image is somethere on the object-surface line
Using Snells law on object (misleading)
2.a.1-7) Applying Snells law (by principle or formula) on the object by treating it as a light ray and
the refracted light ray as the image of the object.
Alternative theories (misleading)
2.a.1-8) same as Q2-4/a) Light traveling at a lower speed in thicker medium, giving a longer virtual
distance.
2.a.1-9) Even if there is images behind the image closest to observer ([d]), they ([a] or [f]) are
covered by this image, and can’t be seen.

Problem Set 2.a.3 and 2.a.4

Basically the same reasoning as in 2.a.1 and 2.a.2 (* interesting note is that 2.a.1-6/f) gives the right
answer but is misleading conceptualwise) with the following additions:

* Visual recalling
The participants recalled a visual memory in order to find the right alternative (mention that people
look fatter in water put do associate this with looking shorter but not with image getting closer to
observer and therefore looks slightly larger).
2.a.3-1) *A visual memory of the legs of people standing in a pool, looking shorter (and fatter).
Principle
The participants referred this principle to the “apparent depth” concept from Question 2.
2.a.3-2) Principle of apparent depth: The image appear closer to the surface than the object

Logic conclusion (misleading)

52
The participants didn’t see any reason for the image to be left or right of the object due to the
symmetry in the figure. However this is misleading, and the symmetry is in fact broken by the eye
looking from left.
2.a.3-3) No reason for the image of the object to bend to left/right so it doesn’t.
* Alternative theory
2.a.3-4) *Looking at an object from behind (pointing away from you) makes it look longer, and
looking at it in front of you (towards you) makes it look shorter.

Problem Set 2b

Intuition
2b-1) It feels plausible
Visual recall
2b-2) Havn’t seen any object disappearing when adding water
Principle
2b-3) same as 2.a.3-2) Principle of apparent depth: The image appear closer to the surface than the
object
Relation reasoning
The participants finds a relation for small changes and apply it for greater changes.
2b-4) Being that the angle of refraction is higher than the angle of incidence, adding water makes
observers closer and closer to the edge see the object
Analogical reasoning and *Scientific modelling
The participants looked at the object as a point sending light in every direction. At first this was
categorized as analogical reasoning but with the addition of Scientific modelling this could be both,
the interview didn’t show if this was based on an analogy (objects is like a point sending light in
every direction/light is a substance radiation from a point source) or Scientific modelling (it is really
sunlight reflected from the object, it could be modeled as a point light source for this purpose).
2b-5) An object (reflecting light) could be seen as light sent out from a point in all (infinite)
directions
Ray tracing
The participants used different ways of ray tracing, some of them helpful and some not. No new way
of tracing, same procedure as 2.a.1-6/a)-2.a.1.-6/f) but applied on a new case.
2b-6) same as 2.a.1-6/a)-2.a.1-6/f)
Alternative theory (misleading)
2b-7) Light bending too much, resulting in no one seeing the object.

53
Problem Set 3

Problem Set 3a

Visual recall
3a-1) same as 2b-2) Havn’t seen any object disappearing when adding water
Principle
3a-2) same as 2.a.3-2) Principle of apparent depth: The image appear closer to the surface than the
object
Ray tracing
3a-3) same as 2.a.1-6/a)-2.a.1-6/f)
Relation reasoning
3a-4/a) Being that the angle of refraction is higher than the angle of incidence, the probability to see
the coin/the possibility to see deeper down gets higher when adding water
3a-4/b) Being that the angle of refraction is higher than the angle of incidence, adding water moves
image from coin
Comparing
3a-5/a) With 2b) results
3a-5/b) With 2.a.x) results
Logic conclusion
3a-6) [ii] and [iii] can’t be true on the same time when observed from same position.
Testwiseness
3a-7/a) There is no point to consider these cases ([iii] and [iv]) for the purpose of the survey
3a-7/b) The text says “which one”, indicating only one answer, and [ii] is the most meaningful
3a-7/c) The checkboxes indicates more than one possible answer

Problem Set 3b

Principle
3b-1) same as 2.a.3-2) Principle of apparent depth: The image appear closer to the surface than the
object
3b-2) same as 1.2-2) The light going the same path no matter the direction
3b-3) Lazer is light
3b-4) Light have (almost) the same refraction no matter the wavelength
Ray tracing
The participants often used ray tracing in a misleading manner conceptualwise, along with some
unknown presumption (that image is on the same depth as object), the image would be further away
the observer than object.
3b-5) same as 2.a.1-6/a)-2.a.1-6/f)
Comparing
3b-6) Comparing with 2b) results

54
Problem Set 3c

Intuition
3c-1) Believing what I am seeing
Visual recall
3c-2/a) Oar/pencil/spoon in water
3c-2/b) Leg/arm in water when bathing
Principle
3c-3) same as 2.a.3-2) Principle of apparent depth: The image appear closer to the surface than the
object
3c-4) Misleading. Apparent distance: The image appear farther away from observer compared to the
object
Comparing
3c-5) Comparing with 2a.x) results
Using Snells law on object (misleading)
3c-6) same as 2.a.1-7) Applying Snells law on the object by treating it as a light ray and the refracted
light ray as the image of the object.

55
Appendix III: Processed transcripts

Intervju 1 2b) Alla

Man kan tänka sig strålgångar då alla ser.
Exempel på Snells lag.
Uppgift 3:
1.1)
Här har vi ingen vinkel mot normalen så det kommer fortsätta att gå rakt a) [ii]
Enligt brytningslagen. Beror på djupet av koppen, kan tänka mig [iii] eller [iv]
1.2) men [i] iaf felaktigt.
Tätare->tunnare bryts från normalen
b)
1.3) - fig.1 -) c
Tvärtom bryts mot normalen Det här med skenbart djup, det ser ut som fisken är högre upp än vad den
egentligen är pga Snells lag.
1.4)
Samma som 1.2) - fig.2 -) 1
Tänkte att om man lyser med lasern i punkt 1 så skulle den åka ner till punkt 2 där
Uppgift 2: fisken egentligen är. Men är ju ljuset ifrån fisken som är punkt 2 kom jag på nu.

a.1) c c) 2
Pennan bryts enligt Snells lag mot normalen när det går ifrån tunnare till tätare Tänkte att skulle se ut som den bryts av mot normalen, såsom ljus gör.

a.2) c Fråga 2: (sist)

Spegelvänt
Ser ut som om det är närmare ytan om du tittar på nånting i optiskt tätare ifrån
a.3) optiskt tunnare material.
Förmodligen det h r med ”skenbart djup”, pennan blir kortare om man kollar ifrån N r fråga 2b) st lldes förstod jag vad ”skenbart djup” innebar
sidan. Flyttar du ögat närmare pennan går du mot dess riktiga längd. (Ledande fråga?) har själv egna erfarenheter ifrån att man ligger i ett bad.

a.4)
-||-

56
Intervju 2 2b) Alla
Antar ingen totalreflektion. Ljuset som går ifrån kulan kommer träffa även [P3]
1.1) eftersom ljusstrålen bryts från normalen.
Går längs normalen så bryts inte alls
Uppgift 3:
1.2)
När ljuset går ifrån optiskt tunnare till optiskt tätare material så bryts mot a) [ii], [iv]
normalen. Tvärtom när ifrån tätare till tunnare. Ljusstrålen ifrån botten bryts från normalen, mot ögat, så om jag ser myntet utan
vatten ser jag det även med. [i] och [iii] skulle kunna stämma om man tänker
1.3) -||- totalreflektion men räknade inte med det. Istället stämmer [ii].
1.4) -||-
b)
Uppgift 2: - fig.1 -) c
Spjutet fortsätter rakt på i bana, det gör inte ljuset fisk-öga som bryts vid ytan.
a.1) c Eftersom ljuset bryts ifrån normalen så måste jag sikta under fisken eftersom den
Utgick ifrån verklig pennspets, fundera på hur ljuset kommer gå till ögat. ser ut att befinna sig högre upp än den egentligen gör.
Ljusstrålen går genom [d] och bryts mot normalen till ögat. Ögat ser då pennan i - fig.2 -) 2
linjens förlängning så vi hamnar på [c]. Lasern följer samma väg som ljuset fisk-öga tillbaka.
Hade minnesbild av åra i vatten, men ögat felplacerat mot när man sitter i båten.
Hade ögat varit på andra sidan hade jag svarat [d]. c) 3
Antar ingen totalreflektion. Jämför med en åra som ser ut att bli krokig när man
a.2) c sticker ner i vatten. Kröker sig upp mot vattenytan.
a.3) c Fråga 2: sist
Strålgång. Vi ser förlängningen till [c], [g] kändes orimligt av någon anledning. Inte pratat om på optikkurserna. Antar baserat på uppgifterna att det är djupet
någonting ser ut att vara på i vatten då man tittar ifrån luft. Tänker på åran i vatten,
a.4) a ser ut att vara grundare och böjt sig.

57
Intervju 3 2b) Alla
Kom på att det handlar bara om brytningar i vatten och inte totalreflektion. När
1.1) man fyller på vatten kommer det se ut som att kulan åker upp litegrann och då
Eftersom ljusstrålen är vinkelrät emot gränssnittet mellan de två medierna så kanske P3 också kan se.
kommer den fortsätta rakt genom.
Uppgift 3:
1.2)
Tänkte att när det går ifrån optiskt tätare till tunnare bryts det bort ifrån normalen.
a) [ii], [iii]
1.3) Tänkte på 2b), objektet ser ut att vara närmare ytan. (i) helt fel och (iv) fel
När det går ifrån tunnare till tätare bryts det mot normalen. eftersom det inte ser ut att hända någonting med brytning av ljuset. (ii) tänkte som
på uppgift 2 b), om jag var P3 skulle jag kunna se objektet. Tänkte att (iii) också
1.4) Samma som 1.2) kan stämma för det är inte säkert man ser kronan, man vet ju inte hur djup koppen
är.
Uppgift 2:
Funderade först på om det spelade någon roll varifrån man tittade. Tänkte att det b)
inte borde spela någon roll. Tänkte först att det var ett koordinatsystem som visade var fiskens egentliga
position men insåg det konstiga i att sikta på en fisk man vet var den är men inte
a.1) d ser så ändrade till att det borde vara representationer för hur det ser ut (bild).
Tänkte att det borde se kortare ut på något sätt och uppvinklat ifrån den riktiga
pennan. - fig.1 -) c
Tänkte inte så mycket på strålgångar utan bara intuitivt vad som kändes rätt. Fiskens bild borde vara närmare ytan än fisken så då får man sikta under fisken för
att träffa.
a.2) d
Spegelvänt - fig.2 -) 2
Då är det bara att sikta rätt på fisken då ljuset tar samma väg fram och tillbaka.
a.3) b
Kom fram till att eftersom pennan r ”rakt ner” så borde det inte h nda så mycket, c) 4
den borde inte se skev ut. Den borde se kortare ut pga det här med skenbart djup Tänkte tillbaka på hur det kan se ut om man stoppar ner en penna eller åra i vatten.
(saker ser ut att vara närmare ytan). Då känns min figur (böjd upp) helt rätt.

a.4) b Fråga 2: Först

Samma sak som a.3) Borde vara vad djupet ser ut att vara när man t.ex. tittar ner i vatten. Tänkte på att
när man ser ner i ett akvarium så ser det grundare ut än vad det egentligen är.
Föresökte rita en skiss av det som blev lite konstig.

58
Intervju 4 2b) Alla
Jobbade mer teoretiskt, det är inte så ofta man hamnar i en sådan situation. Tänkte
Hade Snells lag i bakhuvudet. mig att om ytan är vid övre karkant så borde strålarna brytas av ifrån normalen och
då borde även [p3] kunna se kulan.
1.1)
Här är den parallell och kommer fortsätta vara parallell, rätt logiskt. Uppgift 3:

a) [ii]
1.2) Fortsättning på 2b). Men funderade även på om (iii) kunde stämma men det är inte
Snells lag säger att det måste brytas bort ifrån normalen när det går ifrån tätare till säkert. (iv) har inte ändrats någonting vilket är orimligt.
tunnare. Tänkte dels på teorin vi gått genom om Snells lag och memorerat och dels
på erfarenheter, hur åran ser ut när man är ute och ror. Är snarare erfarenheterna b) Tänkte på fråga 2 – skenbart djup.
av Snells lag än den matematiska formeln som jag har tillämpat.
- fig.1 -) c
1.3) - Fisken är egentligen djupare än den ser ut att vara så man får sikta under.

1.4) - - fig.2 -)
Om man vill träffa med laserstrålen måste man sikta på fisken då lasern följer
Uppgift 2: samma bana som fiskens ljus.

Lite mer hur skulle det se ut på riktigt men jag konfirmerade även med Snells lag c) 3
om det verkade rimligt. Fortsättning på tidigare uppgifter 2a.x). Tänkte dock att då såg vi pennan rakt
emot oss och nu ifrån sidan men jag tror det stämmer ändå.
a.1) d
Valde [d] och kollade strålgångar om det kunde stämma Fråga 2: Först

a.2) d Jag tänkte, med tanke på hur resten av uppgifterna såg ut - ögnade igenom uppgift
Samma fast spegelvänt 1 – så om vi har två medium så kommer ett objekt som befinner sig på verkligt
djup se ut att ligga på skenbart djup om man ser utifrån som t.ex. fiskar i vatten
a.3) b ifrån luft. Den heldragna linjen representerar det faktiska ljuset ifrån objektet
Finns inget givet håll som den borde brytas av mot så den borde bara bli kortare, medan den streckade kommer ifrån det virtuella objektet – där objektet ser ut att
och så konfirmerade jag med Snells lag. vara.
a.4) b
Samma sak som a.3)

59
Intervju 5 2b) Alla
Borde vara vinkelberoende. Egentligen kan man inte säkert säga om [p3] ser eller
Tänkte när man går ifrån icke-tätare till tätare medium bryts det av på visst sätt ej förrän man har avstånd och vinklar, de andra är jag säker ser. Tanken var
och vice versa. Jag använde figuren i uppgiften i huvudet och applicerade den på brytningsindex.
de här fallen.
Uppgift 3:
1.1) -
1.2) -||- a) [ii]
1.3) -||- Använder 2b). [i] försvann då det inte stämmer med brytningsindex, [ii] evetuellt
1.4) -||- rätt svar. [iii] skulle kunna stämma om tillräckligt djup kopp. [iv] ser det ut som att
myntet inte flyttat sig öht vilket måste vara fel.
Uppgift 2:
b)
a.1) d - fig.1 -) c
Intuitivt ifrån minnesbilder av verkligheten, när man sticker ner en pinne i vatten Erfarenhet att man lätt missar fisk. Använde 2b), ljusstrålen bryts ner mot mig
så ser den ut att brytas upp mot ytan. Bekräftade med strålgångar. ifrån mitt håll sett, så måste sikta nedanför.

a.2) d - fig.2 -) 2
-||- Lasern följer samma väg som ljuset fisk-öga.
a.3) b
-||- c) 3
a.4) b Lite erfarenhet och lite tänka efter, pennan bryts upp mot ytan.
-||-
Fråga 2: Sist
Utan vatten går ljuset rakt mot ögat. Med vatten så går det inte rakt men ögat
förväntar sig det vilket gör att vi ser en illusion som ögat skapat. (Text och figur
förklarar allt korrekt).

60
Intervju 6 2b) Alla
Ju högre upp man fyller, ju lägre ner kan ögat vara och fortfarande se då ljuset
1.1) bryts högre upp. Tänkte att [p1] och [p2] fortfarande ser eftersom kulan skickar ut
Rät vinkel mot ytan så då kommer det inte brytas. ljus i alla riktningar så borde någon stråle fortfarande träffa dem, känns inte rimligt
att ljuset skulle brytas mot [p3] men inte till dem andra längre.
1.2)
Vet att det bryts inåt för att det är ett tätare material, såg direkt. Uppgift 3:

1.3) - a) [ii]
Samma som 2b). [i] känns orimligt att man först skulle se och sedan inte se. [iiI]
1.4) borde vara liksom 2b) att om jag är [p3] så ser jag när det fylls upp hela vägen. [ii]
Som 1.2) fast upp-och-ner. kändes rätt då det var samma som 2b). [iv] verkar orimligt att man ska se den på
samma ställe efter att man fyllt på vatten.
Uppgift 2:
b)
a.1) c - fig.1 -) c
Pennspetsen kommer skicka strålar i alla riktningar, tittar vi på en som bryts utåt Man har en ytan med fisken under, ljuset bryts så att fisken ser ut att vara längre
in i ögat så kommer den se ut att ligga längre ner här i [c] när vi drar en rak linje. bort än den är så man måste sikta närmre än vad man ser.

a.2) c - fig.2 -) 2
Spegelvänt bara Lasern bryts likadant som ljuset fisk-öga.

a.3) g c) 3
Om ljusstrålen går upp och bryts av till ögat så kommer den se ut att ligga i [g] när Den kommer lite längre bort från betraktaren. Vet ifrån när jag badar och stoppar i
vi drar en rak linje. ett ben att det kommer längre bort och högre upp. Högre upp också så jag vill
ändra mitt tidigare svar a.3) till [c] som jag inte tänkte på förut, för ser konstigt ut
a.4) e med [g].

Fråga 2: Först

Fick minnesbild ifrån grundskolan. När någonting bryts så ser det ut att vara på en
annan plats och därför kommer det se ut att ligga på ett annat djup. Bilden i mitt
huvud var att det hamnar högre upp men kan inte motivera.

61
Intervju 7 a.4) e
Samma som a.3)
Figuren förklarade hur man ska göra resten av enkäten. Det är svårt att komma
ihåg vilket håll ljuset bryts i det tätare materialet, figuren visar mindre vinkel mot 2b) Alla
normalen. Då minns jag det. Resonemanget att eftersom ljuset böjs mot normalen så kan [p3] titta på vattenytan
i en rak linje och se kulan. [p1] och [p2] kan titta var som helst och se kulan.
1.1)
Rakt mot normalen, då går det rakt fram och bryts inte alls. Uppgift 3:

1.2) a) [ii]
Ifrån tätare till tunnare bryts det från normalen, vilket man kan se i figuren då man Samma som 2b). [i] inte möjlig då man ju kunde se mer innan vattnet. [ii] funkar
kan följa ljuset åt andra hållet likaväl. Man kan också se uppgifterna som att bäst. [iii] skulle kunna vara möjlig beroende på hur djup koppen är. [iv]
strålen kommer ifrån andra hållet och använda figuren direkt. Man kan också förkastades eftersom [ii] var bättre.
tänka att en tillräckligt stor vinkel i vatten ger totalreflektion, vilket är en
minnesregel för att se att det blir större vinkel i luft tills vinkeln blir för stor. b)
- fig.1 -) c
1.3) -||- Spjutet går rakt medan ljust böjs. Tänker på hur ljuset böjs. Strålen går rakt fram
1.4) -||- så då ser fisken ut att vara längre bort än den är eftersom ljuset bryts av nedåt i
vattnet.
Uppgift 2:
- fig.2 -) 2
a.1) c Lasern följer samma väg som fiskens ljus.
Jag prövade med raka linjer ifrån ögat till de skenbara pennspetsarna för att se
vilka strålgångar som kunde stämma och träffa verkliga pennspetsen när de bryts c) 1
av. Troligare att det är [c] än [d] om man tänker optisk väg, att det går Tänkte att pennan skulle böjas mot normalen när den kommer ner i vattnet. Man
långsammare för ljuset i vatten så kommer det se ut som att ljuset har färdats en minns ju att pennor böjs av i vatten sen gäller det att komma ihåg hur den böjs och
längre sträcka. Kom fram till [a] först men ändrade efter sista uppgiften 3c), (böjs hur mycket. Mindes inte vilket håll. Tänkte att den skulle böjas likadant som
på samma sätt som ljus?) gjorde mycket fram och tillbaka. ljuset.

a.2) c Fråga 2: Först

a.3) g Tänkte först siktdjup men kom sen på att det är djupet som man tror att det är.
Ett par olika svar utifrån samma metod som a.1), den kommer till [g] snarare än Förklarar i texten: ”Det djup som ljuset ser ut som det har f rdats j mfört med
[a] då den ska se ut att färdas en längre sträcka. Kom fram till [b] först men vakuum optisk väg. Ljuset färdas långsammare i optiskt tätare material, alltså ser
ändrade efter sista uppgiften 3c), gjorde mycket fram och tillbaka. det ut som det har färdats en längre sträcka. Denna sträcka kallas skenbart djup.”

62
Intervju 8 (Österrike/Frankrike) Spegelvänt. [c] skulle kunna vara ett alternativ men jag fick för mig att den borde
vara längre bort ifrån betraktaren.
Använde reglerna jag fick ut av figur och text samt minnen av Snells lag. 2b)
Missförstod frågan.
1.1) -
Uppgift 3:
1.2)
Använde att vinkeln när det går till ett tunnare material som luft blir större ifrån a) [ii]
normalen. Använde tidigare uppgifter 2.a.x). Utifrån dessa försvinner iaf inte objektets bild
på det sätt som [i] så den uteslöts. Visste inte om det var möjligt att kryssa fler
1.3) svar, valde bara en. [iv] orimlig att botten skulle vara så högt upp att myntet sågs
I det mindre täta materialet luft så ska vinkeln vara längre ifrån normalen än i det från början. [iii] vore möjlig med stor kopp men denna såg liten ut. Använde att
tätare materialet vatten. bilden flyttas bort ifrån betraktare som i 2.a.x) för att övertygas om [ii].

1.4) - b)
- fig.1 -) c
Uppgift 2: Spjutet går rakt fram medan ljuset bryts. Jag ser fisken framför mig som betraktare
och antar att fiskens bild flyttat sig bort ifrån mig pga ljusbrytningen så då borde
a.1) c, e fisken vara i [c].
Testade att hitta strålgång ifrån verkliga pennspetsen som blev vettig men är inte
säker jag lyckades. Vinkeln ifrån yta-öga bör öka gentemot pennspets-yta. Jag - fig.2 -) 2
ritade en rak linje pennspets-öga och ritade den linje som borde bli pga brytningen Lasern går samma väg som ljuset fisk-öga fast andra hållet tillbaka. ”Mon Patrice
ifrån ytan till punkt under ögat. En idé var att rita en linje parallell med denna fast law”
ifrån ögat ner i vattnet där den bryts till mindre vinkel och se ungefär var den
hamnar. Då fann jag att den kunde hamna på [e] eller [c]. c) 1
Gjorde bara en strålgång fast du egentligen behöver två för att vara säker på att Borde flyttas bort ifrån betraktaren, enligt 2.a.x)
bilden är där.
Fråga 2:
a.2) c, e Skenbart djup är den effekt man får av perspektiv i en målning. Objekt ser ut att
Spegelvänt vara på ett visst avstånd beroende på vinklar fast det är målat på en platt yta, antar
att det finns alternativa betydelser som när man ser fisk i vatten som har en
a.3) b, c korrelation till denna betydelse.
Liknande metod som a.1).

a.4) a, b

63
Intervju 9 2b) ingen
Tänkte att det nog bröts ganska mycket och isåfall kommer ingen av dem kunna se
1.1) för ljuset har brutits för mycket.
Rät vinkel infallande så då går den rakt genom.
Uppgift 3:
1.2)
Ifrån tjockare till tunnare medium bryts det bort från normalen a) [iv]
Valde det som kändes mest verklighetstroget. Inte minnen direkt men det som
1.3) kändes mest verklighetstroget. [ii] [iii] kändes fel att det inte skulle synas innan
Ifrån tunnare till tjockare medium bryts det mot normalen. vattnet. [i] skulle kunna stämma om det var mer vatten.
Traglats sen högstadiet.

1.4) -||- b)
- fig.1 -) c
Uppgift 2: Ljuset från fisken bryts från normalen när det går upp ur vattnet så betraktaren
borde se det som att fisken ligger längre bort.
a.1) [b], [e]
Försökte rita strålgång (ifrån skenbara pennspetsar, vilket ger verkligt läge i - fig.2 -) 2
linjens förlängning). När ljuset kommer upp från vattnet bryts det från normalen, Lasern tar samma väg som ljuset fisk-öga.
så då borde det vara något läge närmare betraktaren för då får man mindre vinkel.
c) 2
a.2) [b], [e] Borde se ut så när jag tänker på minnesbilder av skedar. Hade svårt att tänka med
Spegelvänt alla dimensionerna.

a.3) [f] Fråga 2: Sist

Likadant som a.1). Mindre vinkel mot normalen i vatten vilket ger en bild längre Ingen term jag associerar med fysik utan vardagsspråk. Tänkte att det är hur djupt
ner. något verkar vara för att ljuset som går in i materialet sprids eller absorberas
istället för att reflekteras. Istället för att se botten där den verkligen är så ser man
a.4) [f] det högre upp eftersom ljuset spridit sig för mycket för att kunna se den längre ner.
-||-

64
Intervju 10 2b) Alla
Fyller man upp så kommer ljuset att brytas och alla att kunna se.
1.1)
Enkel, bara rakt fram. Uppgift 3:

1.2) a) [i]
Tänkte på ett knep ifrån gymnasiet. Om man har en liten bil så är ju vatten tätare Samma princip som 2b). Vattennivån gör att ljuset kan böjas, och gör att myntet
medium så trögare för bilen. ser ut att vara längre bort då man uppfattar ljuset att komma rakt framifrån.

1.3) -||- b)
- fig.1 -) c
1.4) Fisken ligger egentligen lite närmare oss, inte längre ner i vattnet utan närmare i
Bilmodellen: Högerhjulet kommer först ut i ett lättare material och bilen svänger horisontalled.
eftersom det är trögare på höger sida än vänster. Använde den modellen överallt,
men egentligen bara ett sätt att komma ihåg hur det bryts inte varför. - fig.2 -) 3
Vet inte hur jag tänkte, gjorde snabbt, ser att det ska vara [2] nu
Uppgift 2:
c) 4?
a.1) [a], [c], [d] Då ljuset ifrån fisken böjs efter det kommer upp ifrån vattnet emot betraktaren så
Använde bilmodellen för att se vilken väg ljuset tar till ögat. Sen ser ögat det som kommer den se ut att vara längre bort ifrån betraktaren än den är. (Går ej se på bild
att kommer rakt mot ögat, så måste vara något av [a], [c] eller [d]. vilket håll den är vinklad åt, och ohörbart/utan för bild i intervjun). Dessutom så
kände jag igen bilden jag ritat så det såg rimligt ut.
a.2) [a], [c], [d]
Samma fast tvärtom Fråga 2: först
Gissar på att det är hur långt ner i ett medium som man ser. Tänkte siktdjup.
a.3) [a], [g]
Samma grej med strålgången. Det kommer någonstans på linjen mellan [a] och
[g].

a.4) [c], [e]

Lika som a.3)

65
Intervju 11
2b) Alla
1.1) Eftersom vattnet har en mindre vinkel då det bryts till en större i luft så betyder det
Kommer strålen infallande 90 grader så finns det ingen brytning för då går den att när det fyllts på vatten så kommer vinkeln vid kanten brytas och folk längre ner
längs normalen alltid. kunna se. Kändes också rätt, men fick en konflikt mellan intuition och strålgången
i figuren jag försökte rita upp.
1.2)
Vände på bilden för det kändes naturligast intuitivt. Låtsades att strålen kom ifrån Uppgift 3:
andra hållet för att direkt kunna se hur det ska se ut eftersom strålen går likadant
oavsett riktning. Detta för att intuitivt veta hur jag skulle rita strålen, sedan ritade a) [ii], [iii]
jag normal och infalls- och brytningsvinkel. Använde tänket ska det vara större (iii) rimligt om det är litet mynt. [ii] möjlig då vatten ”lyfter” saker genom att man
eller mindre brytningsvinkel, vände på bilden i huvudet för att vara säker. tror det är närmare ytan. (iv) ej möjlig då den knappast är kvar på samma plats.

1.3) b)
För att veta vilken av infalls- och brytningsvinkel som är störst vände jag på bilden Magen säger en sak och huvudet en annan.
i huvudet och ritade ut det. Sen ritade jag in normal och vinklar. - fig.1 -) e
Om fisken ser ut att vara där, fast är i vatten, så måste den egentligen vara närmre
1.4) -||- ytan och då antog jag att den flyttades bakom, tolkade [e] som närmre ytan. Inser
nu att jag tänkte baklänges, fisken ser ut att vara närmre ytan.
Uppgift 2:
Såg först inte ögat så ritade först normalen där pennan skär vattenytan. - fig.2 -) 1
Tänkte att den verkliga fisken är närmre ytan, vände på det även här.
a.1) d
Ser inte ut att vara där det är pga hur ljuset bryts. Svarade utifrån min intuition, c)
den borde vara närmare ytan. Tänkte mig det rimligaste när man stoppar en penna Jag tänkte hur jag skulle rita så det ser verklighetstroget ut, så jag tror på det jag
i vatten är att den ser kortare ut och närmre ytan. Försökte rita upp i figur men ser.
visste inte hur jag skulle göra med normal, infalls- och brytningsvinkel.
Fråga 2: Först
a.2) d Antog att det handlade om brytning av ljus då enkäten handlade om detta.
-||- Det ser mindre djupt ut än vad det verkligen är då ljuset bryts men ögat uppfattar
det som att det kommer ifrån en rak linje. (Ritad figur med normal, infalls- och
a.3) b brytningsvinkel samt brytningsindex.)
Det jag intuitivt trodde, svårt att analysera.

a.4) b
-||-

66
Intervju 12
2b)
1.1) Misstänkte att p3 skulle se. Ritade strålgångar för att bekräfta, ljusstråle ifrån p3
Fortsätter rätt fram. bryts ner till kulan. p1 och p2 ser fortfarande fast i en liten annan vinkel. Finns en
gräns som vattnet ska höjas till vilket jag anser bör ha passerats då kort avstånd
1.2) kant och ljusstråle mot p3 före vattnet.
Bryts från normalen när den kommer in i luft som är tunnare.
Bekräftar med totalreflektion som tänkt extremfall. Uppgift 3:

1.3) a) [ii]
Bryts mot normalen Myntet kommer se ut att vara högre upp för betraktare. Bekräfta med
tankemodellen att strålen bryts till brantare vinkel när den når vattenytan. [iv]
1.4) stiger inte myntet så fel. [iii] kan vara så att bilden stiger men man ändå inte ser
Bryts från normalen. Strålningens riktning ingen betydelse. men [ii] känns mer rätt.

Uppgift 2: b)
- fig.1 -)c
a.1) d Tittar man på fisken så är det först inte så brant (vinkel) och sen brant så skulle
Ljuset kommer gå i en mindre brant vinkel öga-yta än yta-pennspets. Spetsen man sikta på fisken skulle man kasta över fisken, därför måste man sikta under.
kommer därför garanterat se ut att vara högre upp för ögat än den är. Givet en rak
linje genom ögat och verklig pennspets så kommer bilden vara ovanför den linjen, - fig.2 -) 2
så [a], [d] och [c] blev aktuella. Jag använde extremfall genom att tänka ett ökande Lasern bryts precis som synliga ljuset. Beror lite på frekvens men liten effekt då
brytningsindex ger större och större vinkel pennspets-yta så virtuella pennspetsen det är i synliga ljusspektrat.
kommer högre och högre upp (eller flyttar sig orimligt långt bakåt), vilket tar bort
[c]. [d] k ndes mer ”logiskt” då pennan bör se bruten ut. c) 3
I och med att ljuset bryts nedåt så kommer den synliga positionen längre fram (?).
a.2) d Och den kommer alltid att vara högre upp än den egentligen är. Minnesbild av
Spegelvänt benen på någon som står i vatten är att de alltid ser kortare ut. Så om man tar
riktiga pennan och trycker ihop den uppåt så blir det såhär.
a.3) b
Extremfall ger att extremt högt brytningsindex får en nästan lodrät strålgång och Fråga 2: Först
bild precis vid ytan och vid n_vatten=n_luft att den är på verklig position, så
någonstans mittemellan blir [b] Ljus sprids och försvagas i ursprungsriktningen när det passerar ett medium. Så
långt ljuset färdats tills det försvagats till viss grad.
a.4) b

67
Intervju 13 2b) Alla
Ljuset böjs från normalen då luft har lägre index än vatten så objektet kommer för
Tänkte på formeln sin theta_1 n_1 = sin theta_2 n_2 så förhållandet mellan mediet p3 se ut att vara i en förlängd linje ifrån ögat i strålgången, uppe till höger under
och vinkeln ska vara detsamma. kranen.

1.1) Uppgift 3:
Eftersom vinkeln mot normalen är 0 här så är det ingen vinkel man räknar på så
vinkeln är samma. a) [ii]
Samma som 2b). Om vattennivån höjs så kommer sannolikheten att se myntet öka.
1.2)
Eftersom luften har lägre index än vatten så kommer brytningsvinkel vara större b)
än infallsvinkel för att index*vinkel ska vara likhet. - fig.1 -) c
Skissade lite, om fisken är under ytan så bryts ljuset av fisken på vattenytan och vi
1.3) får större vinkel pga index. Eftersom ju inte spjutets bana bryts så måste man sikta
Vatten har högre index än luft så då kommer vinkeln vara lägre. under.

1.4) Samma som 1.2) - fig.2 -) 2

Lasern går samma väg som ljuset.
Uppgift 2:
c) 2
Baserat på formeln, större vinkel i luft. Drog linje pennspets-yta-öga. Tänkte dels hur det ser ut på badhus då armen böjer sig upp mot ytan. Också
strålgång i huvudet, att när ljuset böjer sig så ser vi att det som att det kommer rakt
a.1) Drog linje pennspets-yta-öga och eftersom den strålgången tangerar [d] så är emot oss vilket gör att det hamnar längre bak och högre upp.
det [d].
Fråga 2: Först
a.2) -
Samma som a.2) då bara spegelvänt

a.3) a
Drog linje pennspets-yta-öga och eftersom den strålgången tangerar [a] så är det
[a].

a.4) c
Samma som a.3)

68
Intervju 14 2b) Alla
I och med att ljuset bryts kan alla se.
1.1)
Går det rakt fram fortsätter det rakt fram. Uppgift 3:

1.2) a) [ii]
Ifrån tätare till tunnare bryts det från normalen. Visste det och såg det i figuren. När koppen är tom ser man in i koppkanten, men sen med vatten bryts ljuset så
man ser myntet
1.3) [ii] det som var relevant, uteslöt de andra pga [i] är orimlig och de andra
Ifrån tunnare till tätare bryts mot normalen. poänglösa.

1.4) b)
Samma som 1.2.) fast upp-och-ner - fig.1 -) c
Samma som i 3a). Ljuset bryts så då måste man sikta närmare sig.
Uppgift 2:
- fig.2 -) 2
a.1) a Lasern går precis samma väg som ljuset ifrån fisken.
Ljusstrålen som man ser bryts ju mot normalen så då tror man att ljusstrålen
kommer härifrån [a] då den borde kommit rakt framifrån. c) 3
Tänkte mest erfarenhetsmässigt att det brukar bli så.
a.2) a
Samma sak Fråga 2: Först
Vi uppfattar t.ex. en sten i vatten på ett annat djup än den verkligen är på pga
a.3) c ljusbrytning.
Visar strålgång ifrån objekt-yta-öga och fortsatt rakt fram till [c]

a.4) a
-

69
Intervju 15
2b) Alla
1.1) Ljuset böjs runt hörnet så [p3] borde kunna se. Ingen motivering till att [p2] och
När ljus har 90 grader infallsvinkel mot en yta så går den rakt genom utan att [p3] skulle se men kan inte minnas att jag någonsin sett ett fat fyllt med vatten som
brytas. skulle göra att jag inte längre ser objekt när det fylls på. Ritade strålgångar men
blev inte 100% övertygad av dem (Stålgången för [ p1] följer inte Fermats
1.2) princip/Snells lag riktigt).
Kommer ifrån ett medium med högre index så bryts från normalen, infallsvinkel
mindre än utfallsvinkel. Uppgift 3:

1.3) a) [ii], [iv]

Ifrån lägre index bryts det mot normalen. Utgick ifrån tidigare kunskaper i skallen, Tänkte på 2b).
ej figur. Aldrig sett saker försvinna när man fyller på vatten. Kan inte se framför mig att
det skulle hända i verkligheten. [ii] strålarna böjs runt hörn när vatten fylls på så
1.4) man kan se längre ner. [iii] beror på ögats position. Ser ut att vara taget ifrån
Likadant som 1.2) fast upp-och-ner. samma position (så paradox om [ii] och [iii] stämmer samtidigt). [iv] är samma
som för [p2] och [p3] i 2b), ser de innan bör de kunna se det efter.
Uppgift 2:
b)
a.1) d - fig.1 -) c
Strålgång ifrån ögat till verklig penna med brytning mot normalen, förlängd linje Strålgång fisk-yta-öga-bild som visar fiskens bild bakom verklig bild fast samma
till [d]. Bekräftar [d] genom minnesbild av att en penna i vatten böjs lite uppåt. djup, så fisken bör vara lägre placerad än bilden sett ifrån figuren.

a.2) d - fig.2 -) 2
Samma fast spegelvänt Även om ljuset går mot eller bort ifrån betraktaren så bryts det på samma sätt,
lasern bryts på samma sätt som fiskens ljus.
a.3) b
Ljusstrålen kommer att brytas från normalen när den går ifrån ett tätare till tunnare c)
medium. Ritade strålgång ifrån ögat till ytan och sedan med brytning mot Minnesbild av hur pennan bryts uppåt mot ytan.
normalen ner i vattnet, spelar ingen roll vilket håll strålgången ritas då den ser
likadan ut. Ögat tror sig se pinnen vid [b] så den ser kortare ut. Fråga 2: Först
-
a.4) b
Samma som a.3)

70