VOLUMEN 33, NÚMERO 2 | Número especial | PP. 529-536

ISSN: 2250-6101

www.revistas.unc.edu.ar/index.php/revistaEF

REVISTA DE ENSEÑANZA DE LA FÍSICA, Vol. 33, no. 2 (2021) 529

La evaluación del presente artículo estuvo a cargo de la organización de la XIV Conferencia Interamericana de Educación en Física

Conceptual development through

computer simulations: a case study

in physics

Desarrollo conceptual a través de simulaciones

computacionales: un estudio de caso en física

Juan José Velasco

*, Laura Buteler

1,2

, Enrique Andrés Coleoni

1,2

Facultad de Matemática, Astronomía y Física, Universidad Nacional de Córdoba, Medina Allende y Haya de la Torre.

Ciudad Universitaria, CP 5000, Córdoba, Argentina.

Instituto de Física Enrique Gaviola, CONICET, Medina Allende y Haya de la Torre. Ciudad Universitaria, CP 5000, Cór-

doba, Argentina.

*E-mail: juan.velasco@unc.edu.ar

Recibido el 15 de junio de 2021 | Aceptado el 1 de septiembre de 2021

Abstract

This paper investigates how students engage with a simulation, during problem solving, and learn. It is a case study with three groups

of university students solving a thermodynamics problem (Carnot cycle) assisted by a computational simulation specifically designed

for that circumstance. Coordination Class Theory is used to interpret the results. These reveal that there are three distinct types of

interaction between students and simulation that promote the conceptual development of the participating groups.

Keywords: Computer simulation; Conceptual change; Coordination Class Theory; Thermodynamics.

Resumen

Este trabajo indaga cómo los estudiantes se involucran con una simulación, durante la resolución de un problema, y aprenden. Es un

estudio de caso con tres grupos de estudiantes universitarios que resuelven un problema de termodinámica (ciclo de Carnot) asistidos

por una simulación computacional específicamente diseñada para esa circunstancia. Se utiliza la Teoría de Clases de Coordinación para

interpretar los resultados. Estos revelan que existen tres tipos distintos de interacción entre los estudiantes y la simulación que pro-

mueven el desarrollo conceptual de los grupos participantes.

Palabras clave: Simulaciones computacionales; Cambio conceptual; Teoría de clases de coordinación; Termodinámica.

I. INTRODUCTION

Computer simulations are clearly within the vast set of technological devices that intervene not only in the advance of

scientific ideas, but also in the teaching of physics in classrooms. It is thus not surprising that the Physics Education

Research (PER) community has dedicated attention to the relation between learning and computer simulations. Nu-

merous studies report positive effects of simulations on students’ learning (Smetana & Bell, 2012). While a large num-

ber of studies report that beneficial impact, much less research has been focused on the details of how those tools

interact with the learning process of students in-depth (Velasco & Buteler, 2017). Among these less frequent examples,

Krajcik and Mun (2014) showed that computer simulations allow for deeper conceptual learning. Simulations function

as a bridge between theory and practice (Ronen and Eliahu, 2000) with dynamic animations allowing students to arrive

Conceptual development through computer simulations

www.revistas.unc.edu.ar/index.php/revistaEF

REVISTA DE ENSEÑANZA DE LA FÍSICA, Vol. 33, no. 2 (2021) 530

at an integrated conceptual understanding (Lowe, 2004). Some studies probed into the ways simulations can partici-

pate in conceptual development at a finer grain level, using Coordination Class Theory (diSessa & Sherin, 1998) as a

theoretical lens (Kluge, 2019; Sengupta et al., 2015 and Parnafes, 2007) Within this view of learning, these authors

were able to reveal how the representations provided by these devices participate in certain mechanisms of concep-

tual change.

Kluge (2019) showed, for the case of a situation involving a heat pump, that the simulation is a meeting point

between theory, existing knowledge and experience. It enables students to connect previous knowledge with physical

principles. The study also shows that simulation caters for exploring by incorporation and displacement, disregarding

some aspects and focusing on others. These issues are crucial for the process of conceptual learning.

Sengupta et al. (2015) were able to identify that conceptually integrated video games could favor and support

conceptual change by helping students bootstrap their intuitive reasoning about the physical world. This work was

carried out on the concept of force and the authors report that specific traits of video games can help students make

contact with the physical world and therefore to activate additional productive intuitive resources.

Parnafes (2007) showed that multiple representations make conceptual inconsistencies explicit, and that, in addi-

tion, the interactive dynamics of the simulation builds bridges between the real world and other representations.

In spite of the valuable contributions of these authors to the understanding of how simulations can participate in

different mechanisms of conceptual change, there are good reasons to go further in this direction. The existing studies

were carried out in particular contexts and given the context dependence of the phenomenon studied, it is important

to find out whether the same dynamics are replicated or if new ones can be unveiled. Also, while previous research

has reported how some specific features of the simulations participate in conceptual development, it is key this inter-

action between simulation and students' reasonings unfolds as the problem-solving task occurs, and what different

learning opportunities arise in each moment.

The goal of our research is to unveil the potentialities of computer simulations for specific stages of conceptual

learning during problem solving as described by Coordination Class Theory. In particular, we analyze the case of a

problem situation involving the analysis of a Carnot cycle, addressed by undergraduate students.

II. THEORETICAL FRAMEWORK

Coordination Class Theory was born as a proposal to add precision to the description of the conceptual change process.

The main purpose was to clearly define the meaning of the term “concept” and to focus on the process of conceptual

change (diSessa & Sherin, 1998). Within this perspective, knowing a concept consists of being able to get relevant

information from the world, across varied situations (diSessa, 2002).

A Coordination Class is a model for particular kinds of concepts, among which are physics concepts. The main

function of a Coordination Class is to allow people to read a particular kind of information out of situations in the

world. This reading takes place through specific processes and strategies. Many of the difficulties people have are

related to the context and circumstances in which they carry out those particular strategies and processes.

The architecture of a coordination class includes two elements: extraction and inferential net (diSessa, Sherin &

Levin, 2016). Extractions allow people to focus their attention on certain information of the phenomenon at hand. The

inferential net is the total set of inferences people make to turn those information read-outs into the required relevant

information.

According to this theory, “using” a concept in different contexts may well imply retrieving different pieces of

knowledge and/or articulating them in different ways. The particular knowledge and the particular way it is coordi-

nated in specific applications of the concept is called a concept projection. When projecting a class, students bring in

different elementary pieces of knowledge (incorporations), they establish links between those elements (connections),

create elements of the inferential net (inferences) or disregard some of them, (displacements)

Typically, students exhibit two characteristic difficulties in creating a new coordination class: the problem of span,

and the problem of alignment. Span refers to the ability (or lack thereof) to recruit and coordinate the elements of the

class in a sufficiently large set of contexts in which the concept is relevant. Alignment refers to the possibility of ob-

taining the same relevant information by means of different projections of the concept.

The theory also establishes a stronger form of alignment: articulate alignment, or articulation. Articulation happens

when students are not only able to determine the relevant information in different circumstances, but can also explic-

itly relate those different projections, noting differences and similarities between them. This stronger form of align-

ment is a metaconceptual process which is a natural extension of the theory in its original form. figure 1 shows a

schematic diagram of these elements (Buteler & Coleoni, 2016). For more details, we suggest addressing diSessa and

Wagner (2005).

Conceptual development through computer simulations

www.revistas.unc.edu.ar/index.php/revistaEF

REVISTA DE ENSEÑANZA DE LA FÍSICA, Vol. 33, no. 2 (2021) 531

FIGURE 1: Schematic representation of theoretical elements that constitute a Coordination Class.

III. RESEARCH CONTEXT AND METHODS

The present work intends to address the following research question: How do computer simulations assist conceptual

development throughout a problem-solving task on thermodynamics?

Our analysis will focus on a small-group discussion in a collaborative environment as students interact with a com-

puter simulation. The students interviewed volunteered to participate in the study. They had passed a thermal physics

course (second year of a career in physics) three months before the interview, with similar grades.

A. The problem-solving task

During the first minutes of the problem-solving session, students worked on their own. Once conflicting ideas or

doubts were detected, they were offered the simulation. It was students who decided when and how to make use of

it. They were completely free to explore, analyze, execute and control the simulation’s parameters in whatever way

they chose to.

FIGURE 2. Problem task.

B. The simulation

Three different sets of previous results informed the design of the simulation. i) Simulations that offer greater oppor-

tunities for the user to change parameters and manipulate the model are potentially more useful to foster conceptual

advancement (Adams, Reid, LeMaster, McKagan, Perkins, Dubson & Wieman, 2008). For this reason, the simulation

was designed to allow adjustment of reservoir temperatures and masses (figure 3, upper left). ii) Realistic schemes

favor the connection between models and phenomena (Martinez, Naranjo, Perez, Suero & Pardo, 2011); Also, anima-

tions offer opportunities for students to activate the most intuitive reasonings about the phenomenon (Lowe, 2004).

Conceptual development through computer simulations

www.revistas.unc.edu.ar/index.php/revistaEF

REVISTA DE ENSEÑANZA DE LA FÍSICA, Vol. 33, no. 2 (2021) 532

Thus, an animation of a device operating as a Carnot machine was included (figure 3, bottom). It consists of a cylinder,

with a mobile piston on one of its ends. Contact with either hot or cold reservoirs are represented by colored edges

on the cylinder and dashed lines indicate instances where heat flow stops and the process becomes adiabatic. This

simple animation offers an explicit depiction of heat flowing to and from the gas and work being done on and by the

gas. iii) Simulations that represent temporal events by means of spatial representations have a better chance of fos-

tering users’ conceptual understanding (Parnafes, 2007). Following this idea, two x-y plots of bath Temperature and

Entropy vs cycle were included (figure 3, upper right).

The simulation was designed with the Easy Java Simulation platform in html language.

FIGURE 3. Screenshot of the simulation designed.

C. Data Collection

Problem solving sessions were video-recorded. During the interview students discussed their decisions, relations,

changes of mind, etc. The interviews lasted a bit over 120 minutes. They were conducted by a researcher who was

not the students' instructor. Elements proposed by Halldén (2007) were considered: the interviewer’s mission was to

follow students’ ideas and to enable them to fulfill their project, as opposed to guiding their reasoning. Interviewer’s

interventions were oriented at asking for deeper explanations, checking understanding, or highlighting differences

between students’ reasonings.

The analysis was carried out on the audio-video data obtained during the interviews. It involved two distinct in-

stances. A first stage consisted of an individual (one single researcher) revision of the videos as they were transcribed.

In a second stage, these were reviewed by a research team as proposed by Jordan and Henderson (1998). This collab-

orative viewing is powerful for neutralizing preconceived notions of individual researchers and discourages the ten-

dency to see in the interaction what one is conditioned to see or even wants to see.

IV. RESULTS AND ANALYSIS

We made a specific purpose to identify the different ways in which students coordinate the elements of the class and

how the simulation intervenes in those dynamics. Through the analysis of the interviews, three different types of

interactions were observed between students and the simulation.

A. Interaction type 1: Extractive interaction

This fragment corresponds to the initial step of one of the groups. These students explain the process of a Carnot cycle

on a PV diagram, but they cannot associate this diagram with a concrete physical process. Moreover, despite pointing

out the cycle in the diagram, students are unable to recognize what happens to the gas. The same is true for reservoirs.

They know some things about their idealization but little about their nature. Therefore, they cannot choose any of the

answers proposed. After that, the interviewer offers them the simulation:

Conceptual development through computer simulations

www.revistas.unc.edu.ar/index.php/revistaEF

REVISTA DE ENSEÑANZA DE LA FÍSICA, Vol. 33, no. 2 (2021) 533

1. Int: I get the feeling that they have to see a concrete case of what a machine that performs a Carnot cycle does, right? to

understand what happens with these two reservoirs that you drew here... Maybe this simulation can help.

\\ Students interact with the simulation for a couple of minutes. They focus specifically on the animation, where they look

at the gas process and the heat flows:

2. N: Look what's going on there. I think I can see the process now. There the temperature increases but there the tempera-

ture decreases, doesn't it?

3. F: I don't know if that's showing you temperature... Here in the animation, you can see the heat flows...

4. N: Well... there the heat flows out... there the temperature increases... That would be like the two reservoirs.

From CCT, it is possible to identify extractions from the new representations offered by the simulation. As can be

inferred from the transcription, the program offers new representations that were not present in the statement of

the problem, and the students are mainly oriented to extracting this new information: what is the process about (turn

2) and the specific moment of heat exchange and its direction (turns 3 and 4). These extractions let them start rea-

soning about temperature. They connect specific aspects of the animation with what they have first represented in a

PV diagram. In this way, they have different opportunities to address their difficulties such as what process the gas

undergoes, what the reservoirs do, among others. Until that moment the students were stuck in their understanding

of the phenomenon since they could not connect their knowledge with the problem and the questions posed.

We define this type of interaction as Extractive: the simulation contributes new representations that become a

source of new extractions (as defined by CCT). These new extractions are inputs for the development of new projec-

tions. Students recognize new information that allows them to acknowledge new aspects of the phenomenon.

B. Interaction type 2: Inferential interaction

Throughout this fragment, students begin to interact differently with the simulation. Using the representations pro-

vided by the program, they develop new inferences. In the following excerpt we show how students use the simula-

tion to infer what happens with reservoirs’ temperature.

1. F: Then the reservoirs are going to maintain their temperature...the idea is that they maintain their temperature so that

they continue fulfilling the cycle...

2. N: Of course, that's what I think because if temperature decreases...

3. F: It would not complete the cycle...

4. F: Of course, if we have the reservoirs and this [for the T

] has to make the cylinder go and return, hasn't it? If when I come

back here the temperature is lesser this cylinder is not going to go there ... it is going to

get here [pointing to the piston of the simulation]. Because that temperature you lost

here [pointing to the gas in the simulation] is going to make the gas chamber not so hot

and it won't get to the end now. Then the next cycle is going to come here for the other

reservoir and when I get back there [pointing to the piston in the simulation] ... because

it was not at the same temperature as before.

5. N: Well, I think the initial cycle has this graph (see figure). Suppose you don't have constant

[he points reservoir]. Start with a T1, and suppose you don't add heat to the reservoir. So,

when a cycle ends you have a lower temperature [he points the gas], you don't get to T1

but brake earlier. In other words, if you do nothing, the temperature of reservoir 1 will

decrease and the other will increase the temperature because you are going to be delivering heat.

In this excerpt the students develop inferences using the representation provided by the simulation. They identify

heat flows and start to conjecture what happens with the piston if the reservoir's temperature goes down (turn 4).

They infer that if heat is flowing out of the hot reservoir, the piston will expand its volume and, in consequence, the

gas will reach temperature not as low as on the first cycle. The same happens with compression and the high temper-

ature of the isothermal curve (turn 5). So, the temperature of the reservoirs is modified in each cycle (the hot one

cools down and the cold one heats up) and this can even be represented on the PV diagram of the Carnot cycle for

reservoirs that modify their temperature (turn 5).

We define this type of interaction as Integrated. The simulation does not just provide extractions (as in extractive

interactions): it becomes part of students' explanations. In this sense, elements of these representations begin to

function as support for students to elaborate new reasonings, ideas and speculations. This includes the use of graphics,

animations and/or representations provided by the simulation that allow them to develop inferences. It is character-

istic of this type of interaction that students explicitly involve the simulation in their explanations.

Conceptual development through computer simulations

www.revistas.unc.edu.ar/index.php/revistaEF

REVISTA DE ENSEÑANZA DE LA FÍSICA, Vol. 33, no. 2 (2021) 534

C. Interaction type 3: Projective Interaction

In the following excerpt, another group of students are invited by the interviewer to estimate an equilibrium temper-

ature for the reservoirs. In the first instance, students use the heat-balance equation to propose the average temper-

ature of the two reservoirs as their final equilibrium temperature. By contrasting with the simulation, they identify

that work extraction causes the equilibrium temperature to be lower than the mean.

1. P: I think it is the average

2. E: The average?... Well, I think that the heat flows with a constant rate...

3. P: It remains the same …

4. E: It is the same, so the average makes sense...

5. P: If you still doubt, compute it… You know the masses, suppose a coefficient...

\\They take the pencil and they write down the heat equation...

P: Well as you can see, it is the average

6. E: Yes. You’re right

7. Int: Do you agree that it’s the average?

8. E-P: Yes, we do…

9. Int: Ok, so according to your input the average temperature must be...?

10. P: 250 K…

\\They run the simulation

12. E: The result is 245K … Well, stop… The problem is that we are not considering dissipations, isn’t? The temperatures are

going to be the same but there is a loss of energy.

13. Int: Is the simulation considering dissipations?

14. P: No, I don’t think so...

\\They run the simulation but with different temperatures.

15. E: It doesn’t fit again. Wait...We are not considering [during their computation] that the gas is doing work. That energy

is going out, so the temperatures are not going to converge at the average. They must fit to a value less than the

average.

16. P: Yes, it is not because of the dissipation. It is because of work.

Students initially complete a projection basing their inferences on the formalism of the heat equation. This leads

them to conclude that the equilibrium temperature of both reservoirs will be the mean temperatures (turn 7). How-

ever, when contrasting it with the result of the simulation they doubt what they obtained. Simulation offers the op-

portunity to manipulate parameters, and even to simulate the phenomenon under different conditions. This let

students convince themselves that something was wrong with their prediction and made them look for another ex-

planation.

From the dissonance with the result of the simulation, the students begin a process of articulation. They have two

projections that are not aligned so they check their inferences again. After some executions of the simulation, they

identify that the presence of an amount of work done by the machine implies that the equilibrium temperature must

be lower than the average (turn 16 and 17). It is important to highlight that simulation fosters this process of articula-

tion. The multiple representations offer the opportunity to make explicit conceptual inconsistencies and make stu-

dents work on aligning their own projection with simulation results. This was previously reported by Parnafes (2007).

In this fragment, the simulation presents a different role in the conceptual development of the students. It is used

to compare with the projection constructed by the discussion group. We define this type of interaction as Projective.

V.DISCUSSION

In this work we focus on the study of conceptual development assisted by simulations. From Coordination Class The-

ory, we focus on analyzing the mechanisms of interaction that occur during the problem solving involving this re-

source. We address the following research question: How do computer simulations assist conceptual development

during a problem-solving task on thermodynamics?

Students’ conceptual development showed different interaction types with the simulation. In the first instance,

students presented great difficulties in dealing with the problem situation. Faced with the difficulty, they begin to

interact with the simulation. The animation provides them with new representations that serve as a bridge between

the model and the phenomenon (link that was weak), as well as allowing them to focus on important aspects of the

phenomenon. It is important to remember that similar aspects were reported by Segnupta (2015), who finds that

Conceptual development through computer simulations

www.revistas.unc.edu.ar/index.php/revistaEF

REVISTA DE ENSEÑANZA DE LA FÍSICA, Vol. 33, no. 2 (2021) 535

these representations function as bootstrap for students’ reasoning. However, the study allowed us to look further.

Simulation during this stage worked as a provider of new extractions. Students here recognized new information. We

call these interactive dynamics extractive.

In a second instance, they abandoned the passive posture of receiving new information and began to use that

information and make inferences with it. At this stage, representations began to be part of their explanations and

discussions. The analysis reveals that the students constructed new inferences from the representation, that is, they

built relationships that were neither in their minds nor in the simulation. We define this type of interaction as an

inferential type.

Ultimately, students use the simulation to check their predictions. We say this type of interaction is projective.

From CCT, we can identify that students compare their projections with simulation outcomes. When they find disa-

greements, they begin an articulation process, by means of which they identify differences and similarities between

projections. This is similar to that reported by Parnafes (2007) who mentions that different representations make

conceptual inconsistencies explicit. In this sense, we saw that students assign a value to simulation very similar to that

of experimentation, as if the program were actually the phenomenon.

In addition to having a positive impact on learning, it has been shown that computational simulations correspond

to a very useful tool as a scaffolding for processes of conceptual change. This line of work has made it possible to

identify in greater detail how this tool intervenes in the learning process. We have been able to show that it does so

in different ways at different stages of conceptual development and according to the need for learning at each time.

The contextual factor is key to understanding this tool in action. This line of research calls for further development, so

that the community can not only acknowledge the potential of these tools but, above all, understand the best ways

to make use of them.

REFERENCES

Adams, W. K., Reid, S., LeMaster, R., McKagan, S., Perkins, K., Dubson, M., & Wieman, C. E. (2008). A study of educa-

tional simulations Part II–Interface Design. Journal of Interactive Learning Research, 19(4), 551-577.

Buteler, L., & Coleoni, E. (2016). Solving problems to learn concepts, how does it happen? A case for buoyancy. Physical

Review Physics Education Research, 12(2), 020144.

diSessa, A. A. (2002). Why “conceptual ecology” is a good idea. In M. Limon, & L. Mason (Eds.). Reconsidering concep-

tual change: Issues in theory and practice (28-60). Dordrecht, the Netherlands: Springer Netherlands.

diSessa, A. A., & Sherin, B. L. (1998). What changes in conceptual change? International journal of science education,

20(10), 1155-1191.

diSessa, A. A., Sherin, B., & Levin, M. (2016). Knowledge analysis: An introduction. In A. diSessa, M. Levin, & N. Brown

(Eds.), Knowledge and interaction: A synthetic agenda for the learning sciences, (30-71).

diSessa, A. A., & Wagner, J. F. (2005). What coordination has to say about transfer. Transfer of learning from a modern

multi-disciplinary perspective, 121-154.

Halldén, O., Haglund, L., & Strömdahl, H. (2007). Conceptions and contexts: On the interpretation of interview and

observational data. Educational Psychologist, 42(1), 25-40.

Jordan, B., & Henderson, A. (1995). Interaction analysis: Foundations and practice. The journal of the learning sciences,

4(1), 39-103.

Kluge, A. (2019). Learning science with an interactive simulator: negotiating the practice-theory barrier. International

Journal of Science Education, 41(8), 1071-1095.

Krajcik, J. S., & Mun, K. (2014). Promises and challenges of using learning technologies to promote student learning of

science. Handbook of research on science education, 2, 337-360.

Lowe, R. (2004). Interrogation of a dynamic visualization during learning. Learning and Instruction, 14(3), 257-274.

Conceptual development through computer simulations

www.revistas.unc.edu.ar/index.php/revistaEF

REVISTA DE ENSEÑANZA DE LA FÍSICA, Vol. 33, no. 2 (2021) 536

Martinez, G., Naranjo, F. L., Perez, A. L., Suero, M. I., & Pardo, P. J. (2011). Comparative study of the effectiveness of

three learning environments: Hyper-realistic virtual simulations, traditional schematic simulations and traditional la-

boratory. Physical Review Special Topics-Physics Education Research, 7(2), 020111.

Ronen, M., & Eliahu, M. (2000). Simulation—A bridge between theory and reality: The case of electric circuits. Journal

of computer assisted learning, 16(1), 14-26.

Parnafes, O. (2007). What does “fast” mean? Understanding the physical world through computational representa-

tions. The Journal of the Learning Sciences, 16(3), 415-450.

Sengupta, P., Krinks, K.D., Clark, D.B. (2015). Learning to Deflect: Conceptual Change in Physics during Digital Game

Play. Journal of the Learning Sciences, 24(4), 638-674.

Smetana, L. K., & Bell, R. L. (2012). Computer simulations to support science instruction and learning: A critical review

of the literature. International Journal of Science Education, 34(9), 1337-1370.

Velasco, J., & Buteler, L. (2017). Simulaciones computacionales en la enseñanza de la física: una revisión crítica de los

últimos años. Enseñanza de las ciencias: revista de investigación y experiencias didácticas, 35(2), 161-178.