de aniversario: “argumentación en el salón de clase”

Cool Argument: Engineering Students’

Written Arguments about Thermodynamics
in the Context of the Peltier Effect in Refrigeration
Sibel Erduran1 and Rosa Villamanan2

Abstract
In this paper, we investigate university engineering students’ written arguments in the context
of the thermodynamics principles involved in refrigeration. The students were given writing
frames to complete reports following investigations on thermoelectric coolers, sometimes
called “thermoelectric module” or “Peltier cooler”. The device is a semiconductor-based
electronic component that functions as a small heat pump. By applying a low voltage DC
power source to a cooler, heat moves through the module from one side to the other. One
module face, therefore, will be cooled while the opposite face is simultaneously heated. The
task immerses students in the context of providing evidence and justifications for temperature
change using general principles of thermodynamics. The context of the study provides
examples on the nature of arguments in the applied field of engineering where appeals to
scientific principles are made to justify the design of an industrial product. The writing frame
engaged the students in the recording, analysis and interpretation of experimental
data including data from a simulation program. Several lines of analyses have been conducted
including the epistemic levels of students’ arguments. This paper will report on one aspect of
analysis focusing on the quality of students’ written arguments. A simplified version of
Toulmin’s Argument Pattern (1958) has been used as a guiding model to investigate the
written arguments. Results indicate that very few students were able to accurately interpret the
experimental data and only 35% of the arguments from 31 reports were valid. Most texts used
conclusions that were not derived from the evidence used. We propose some rubrics to further
support the writing and evaluation of arguments.

Keywords: argumentation, Peltier effect, epistemic levels, Toluminn’s Argument Pattern

Introduction Erduran, Simon and Osborne, 2004; Jimenez-Aléixandre,

In recent years, the learning and teaching of argumentation Bugallo-Rodríguez and Duschl, 2000; Kelly and Takao, 2002;
i.e., the coordination of evidence and theory to support or Zohar and Nemet, 2002).
refute an explanatory conclusion, model or prediction (Sup- Writing has been advocated as a tool in the learning of sci-
pe, 1998) has emerged as a significant educational goal. Of ence particularly in coordinating modes of inquiry and acqui-
growing importance in science education is the need to edu- sition of scientific ways of thinking (Keys, 1999) such as argu-
cate students about how we know and why we believe in mentation. Numerous researchers (e.g. Hand, Prain, Lawrence
certain claims (Erduran and Jimenez-Aleixandre, 2008). The and Yore, 1999; Kelly and Takao, 2002) have been investigat-
shift from what-we-know to how-we-know requires a re- ing students’ science writing. The evidence suggests that stu-
newed focus on how science education can promote students’ dents are not able to explain how knowledge claims are es-
skills in justifying claims with evidence. The case made is that tablished nor how ‘writing could act as an epistemological
argumentation is a critically important discourse process in tool’ (Hand, Prain, Lawrence and Yore, 1999, p. 160). In argu-
science (Toulmin, 1958) and that it should be taught and mentation studies, Toulmin’s Argument Pattern (1958) has
learned in the science classroom. Considerable research has been adapted as a model to support the writing of argument.
been carried out in argumentation in science classrooms (e.g. For example, Bell and Linn (2000) adapted the Toulmin
model for the design of the tool called SenseMaker. In a simi-
lar vein, researchers at Northwestern University developed a
software programme called BGuILE (Biology Guided Inquiry
1
University of Bristol, United Kingdom. Learning Environments) which supports the writing and jus-
Correo electrónico: [email protected] tification of causal claims in science (e.g. Sandoval and Reiser,
2
University of Valladolid, Spain. 2004).

abril de 2009  •  educación química de aniversario 119

 The purpose of this paper is to present an investigation Laboratory instruments used for data collection
conducted on a subset of data collected with 130 university Over 4 weeks, each student group had to study the Peltier
engineering students’ written arguments in the context of the effect, a principle that they would use in their professional
thermodynamics principles involved in refrigeration. Given future designing devices to either heating or refrigeration ap-
argumentation is a key aspect of scientific inquiry, its place in pliances in the industry context. The students were given the
tertiary science and engineering curriculum is critical though writing frame from Table 1 to complete reports following in-
few research studies have focussed on tertiary education in ar- vestigations on thermoelectric coolers. The device is a semi-
gumentation studies (e.g. Kelly, Chen and Prothero, 2000). In conductor-based electronic component that functions as a
this study, we aimed to better understand how argumentation small heat pump. By applying a low voltage DC power source
can be promoted in tertiary education, particularly in the to a cooler, heat will be moved through the module from one
context of engineering education. Investigations into argu- side to the other. One module face, therefore, will be cooled
mentation in a particular science content domain not only while the opposite face simultaneously is heated. One of the
illustrates how the nature of science operates but also could drawbacks of experimental work is the amount of time that
provide indicators for how science education be better aligned should spend, due to the complexity of the setups and the
with domain specific features of knowledge (Erduran, 2007). waiting time to achieve equilibrium conditions. Computer
The students were given writing frames to complete reports simulation is a complementary powerful tool in laboratory
following investigations on thermoelectric coolers, sometimes work and it allows improving the effectiveness of an interac-
called thermoelectric module or Peltier cooler. The device is tive learning and teaching, since this allows obtaining many
a semiconductor-based electronic component that functions data in different conditions. The effects studied are made
as a small heat pump. By applying a low voltage DC power through laboratory equipment and a simulation program dis-
source to a cooler, heat will be moved through the module played in Figure 1.
from one side to the other. One module face, therefore, will
be cooled while the opposite face simultaneously is heated. • The laboratory equipment is R533, P.A.Hilton Ltd,
The task context immersed students in the context of provid- U.K. This is a heat thermoelectric pump that consists of a
ing evidence and justifications for temperature change using Peltier module. This pump can be used for heating or re-
general principles in thermodynamics. Engineering students frigerating.
carried out investigations including the use of a simulation • The simulation is programmed in Visual Basic 6.0 and the
program (Chamorro, Segovia, Villamañán, Martín and Villa- screen interface is friendly, easy to understand and similar
mañán, 2004). to the equipment from the laboratory to easily identify the
entire element involved. This program had been develop-
Methods ing by the university team in engineering (Chamorro,
2004).
Participants
This study took place in the Faculty of Engineering in a Euro- Argumentation Scaffolds
pean university. Participants were 130 students of 3rd course A writing frame was produced including a set of questions to
in “Technical Thermodynamics II” subject of the academic guide the use, analysis, interpretation and presentation of data
curriculum. Students’ practical work involved collaboration in and conclusions. The questions proposed are convergent and
groups of three or four. The final written reports were pro- divergent with different difficulty and with an increasing ab-
duced by each group with a total of 31 reports for the straction degree. These questions have different purposes: (a)
course. convergent inquiries try to promote the ability to construct

Figure 1. Laboratory equipment and simulation program.

120 de aniversario educación química  •  abril de 2009

arguments about the knowledge associated with mathematics Table 1. Questions guiding the writing of arguments.
and science. Such argumentation to converging questions are DATA MEASUREMENT
expected to hold truth value, that is, to be verifiable. For ex- Data from the laboratory equipment
Data from simulation program SimBCT
ample the question of “What is the temperature of the cool
side at different voltage flow?” promotes and records a con- DATA ANALYSIS
What is the temperature of the cool side at different voltage flow
vergent inquiry; (b) divergent inquiries take place in the con-
through the cell?
cept domain, where the argumentation does not have truth How did the temperature of the hot side change?
value, which is not necessarily verifiable. Such argumenta- How much electric power is consumed by the cell in each case
tions involve a higher level of abstraction and difficulty. For (situation)?
example, the question “Can you identify the logic for the be- DATA INTERPRETATION
haviour of what you observe?” invites the students to reason Can you identify the logic to the behaviour of what you observe?
by appealing to a range of principles about thermodynamics. What is the lowest temperature of the cell and what are the
Thus students’ written tasks were supported with a range of conditions for obtaining this temperature?
Do the data depend on the environmental (room) temperature or
questions so as to enable the manifestation of a range of argu- the temperature of the hot side?
ments from those that are readily derived from the use of al- Do you believe that the temperature of the cooler side should
gebraic equations to those that would require further abstrac- change if we do not put a heat sink with a fan to dissipating the
tion and reasoning with qualifiers from a range of theories heat from the hot side?
and principles. The questions on the writing frame are sum- DATA PRESENTATION & COMMUNICATION
marised in Table 1. Write a report analysing the behaviour of TE module using
information — including graphics and suitable variables — that you
consider necessary.
Data sources for students’ arguments
Data sources were 31 written reports produced by all groups.
Examining students’ answers across the inquiries of the prob-
lems can show the effect of practice on students’ use of data account the system variables together with their accuracy.
and may show differences regarding the nature of the data as Despite the emphasis in instruction to use the data in the
well as how well students make sense of data for variable right scientific-technological language, just only 19.4% of stu-
temperature change using general principles in thermody- dents found out the right expression of the observed phe-
namics. nomenon related to the thermoelectric effect.

Approaches to data analysis Formal validity of the argued text

A simplified version of Toulmin’s Argument Pattern (Toul- Formal validity refers to the presence of the data, warrant and
min, 1958) allowed us to carry out the analysis of the stu- claim as the key components of the argument. Without these
dent’s argued texts from the empirical data. Toulmin’s frame- three components as well as a logical and accurate link be-
work in more depth has been used by researchers to identify tween these components, the text is not considered a valid
the quality of argumentation in science classrooms (Erduran argument. Only 35% of argued texts are formally valid. Most
et al., 2004). We have emphasised the key claims made, the of the students used tautologies, propositions and ambiguous
data used to validate the claim and the further justifications sentences. Examples of students’ arguments are illustrated in
used to warrant the use of the data to make the claim. We the following paragraphs.
have produced a framework based on the quality of the
epistemic levels and argument quality in the written reports. Example 1: Peltier effect as a warrant
Here we will provide a brief overview with respect to broad The following is an excerpt from a student’s written report:
categories for the quality of the empirical data gathered by
the students, validity of the argued text and the inferences In the previous tables (tables I and II) we can observe how the
made between facts and observed phenomena. temperature changes similarly between the cold and the hot side
of the cell. One can realize that when we increase the applied
Results voltage, we increase the absolute value of both sides, this is, the
hot side increases its temperature and the cold side decreases its
temperature. According to this/therefore, the flow heat that can
Description of the empirical data gathered be profited increases (in this case the heat is dissipated by a suit-
by the students able heat sink attached to hot side). This is due to Peltier effect,
Data are the factual background to any argument. They trace according to this, the heat has been produced depending on the
the path towards a conclusion or a claim. An example of data voltage (or the intensity) and a coefficient.
from a group’s work is illustrated in Figure 2. During the pro-
posed problems, students should be able to take relevant in- In this example, we identified the main claim of the argument
formation (data, evidence) from measurements either from as that “the heat that has been produced depends on the voltage
experiments or simulation program. Students should take in or the intensity and a coefficient.” The Peltier effect is used as a

abril de 2009  •  educación química de aniversario 121

 Figure 2. Data from student work.
V PELTIER(v) Tª CAL - Tª FRIA

0,5   9,57

1 18,38

1,5 26,43

2 33,71

2,5 40,23

3 46

3,5 51

4 55,23

4,5 58,71

5 61,43

justification or warrant for this claim. When the student side’s temperature” as a warrant and “DT has a determinate
writes, “In the previous tables (tables I and II) we can observe value for each value of the voltage” as data.
how the temperature changes similarly between the cold and
the hot side of the cell. One can realize that when we in- Analysis of the inferences made between
crease the applied voltage, we increase the absolute value of facts and observed phenomena
both sides, this is, the hot side increases its temperature and Throughout the text, from the initial thesis to the final con-
the cold side decreases its temperature”, he is referring to the clusion there needs to be coherence in order to validate the
empirical data from his experimental work. In summary, in whole argument. Facts are established from experimental
this example, there is an argument with a main claim, refer- data having a meaning in the conclusions due to laws, theo-
ence to empirical data and a warrant to justify how the data ries, principles, models and so on. Hence concordance be-
relate to the main claim. tween experimental evidence and established conclusions
were analyzed in the students’ inferences. Conclusions were
Example 2: Algebraic equation as a claim made from three different points of view: (a) a theoretical
In this example, the experimental data are used to substanti- view inserted in a scientific context; (b) the experimental
ate a claim made through a common algebraic equation in facts themselves; (c) a descriptive account. Only 35% of ar-
the conceptual domain. gued texts present a concordance between facts and conclu-
sions (e.g. Figure 3). Most of the texts used evidence that did
The value of the lower temperature depends on the temperature of the not reach the conclusions using the evidence. For example, as
hot side. When we apply a voltage, a difference of temperature between illustrated in Figure 4 (from one report consisting of the main
both sides is created, not a fixed value of temperature. Therefore, DT claim, the experimental data used and the justification for the
has a determinate value for each value of the voltage. That is, if the hot conclusion) many students’ arguments did not connect the
side has a high temperature (as the DT is fixed), the cold side will be
experimental data and any theoretical warrants to justify the
a higher temperature than if the hot side has a lower temperature.
Then increasing the temperature of the hot side means that the conclusions reached. Instead, the main claims made were ir-
temperature of the cold side will be higher for a determinate voltage respective of the data collected. The warrants used were rep-
DT = Th – Tc = cte. etitions of formulae used in the content domain but not nec-
In summary, essarily leading to the conclusions reached in the particular
The higher T, the Tcold ; The lower Thot, the Tcold
example.

Design of rubrics to support writing

Here the main conclusion is that “increasing the temperature and evaluation of arguments
of the hot side the temperature of the cold side will be higher The results of the study suggest that the writing frame
for a determinate voltage DT = Th – Tc = cte”. The student is with the questions provided to the students (Table 1) to fa-
using “the value of the lower temperature depends on the hot cilitate their writing needs to be supplemented with further

122 de aniversario educación química  •  abril de 2009

Figure 3. Example of data and warrants being used to justify the conclusion reached.

Conclusion

The observed behaviour is logical, since when the voltage is higher in the Peltier cell, ∆Tª(Tc – Tf) is higher, and to increase Tc, as we
have seen in b section, higher should be the heat that must be dissipated at hot side, so a high intensity that flow

Ley de Kirchoff    sd   V = I · R / R = cte.    sd fV – fI

Therefore, it is consistent with the consumed power to be higher.

Experimental data Justification or warrant

Medición Vmód [V] Imód [A] Tc [ºC] Tf [ºC] At practical level, due to the loss by the heat transferred
between the cell and refrigeration fin, it is difficult to get the
1 1 2,9 22,09438   3,71094 thermal jump. Neither does it have a linear performance and
2 2 8 26,42923 – 7,28944  there are very heavy elements, so the working performance
obtained is very low.

3 3 8,7 32,64194 – 13,36374

4 4 11,6 40,73251 – 14,51196

5 5 14,5 50,70095 – 11,7341

f Tensión  sd  f Potencia

f Potencia   sd   fTc   sd  gTf

abril de 2009  •  educación química de aniversario 123

 Figure 4. Conclusion reached being disconnected from data and warrant.
Conclusion
The main use of Peltier cell are as refrigerant, therefore must be the cold side to the lowest temperature possible, we can get this dissipating
heat from hot side with the higher performance possibly in a fun manner! In this respect, the results are environmentally important since
at hot room temperature we will have a higher Theat and higher Tcold. That is what we don’t want that are that cold side make a refrigeration
higher possible.

Experimental data Justification or warrant

En cuanto a la cesión y absorción de calor en los bloques debido a la apli-
Los datos obtenidos por simulación son los siguientes: cación de una tensión en la célula cumplen la siguiente relación:

QTC = 2α TC I poder termoeléctrico calentando

QTF = 2α TF I poder termoeléctrico enfriando.

Donde:
TC: es la temperatura de la cara caliente.
TF: la temperatura de la cara fría
α: es el coeficiente Seebeck
I: la corriente que atraviesa al circuito.

Y el flujo neto de calor intercambiado el siguiente, según las expresiones

anteriores

P = QC − QF = α (TC − TF ) I = α Δ T I

Por tanto para modificar la temperatura en el lado caliente TC depende

de la intensidad que atraviesa el circuito y del coeficiente de Seebeck.

scaffolds in order to improve the quality of students’ argu- result provide some indicators for guidelines as to how future
ments. Given that most issues related to the quality of the support structures could be designed to help students in their
formal validity of an argument and the inferences made be- collection, analysis, interpretation and presentation of experi-
tween the components of an argument, we consider that ex- mental evidence. For instance, we have illustrated how an
plicit support of these aspects would help to improve stu- evaluation rubric could supplement students’ writing. The
dents’ writing. Hence we have devised a rubric (Table 2) to study has implications for how students can be introduced to
supplement the report writing. The rubric again uses the the writing of arguments in science and engineering contexts
Toulmin model to target particular aspects of argument and earlier on in their education in order to minimise difficulties
are accompanied by a set of questions that the students can at a more professional level. As the task context illustrates,
reflect on with respect to their own data collection and inter- understanding the rationale for and the structure of argument
pretation. There is also a section on evaluation of the feature are prerequisite to tertiary students’ satisfactory performance
of argument suggesting the self- and peer- assessment of the in both basic and applied science. These skills are unlikely to
arguments produced in groups. be acquired effectively at university level without any earlier
background on scientific reasoning with argument.
Conclusions and educational implications
The study presented illustrates engineering students’ written References
arguments in the context of Peltier effect in thermodynamics Chamorro, C.R., Segovia, J.J., Villamañán, M.A., Martín, M.C.
of refrigeration. The particular task context provides exam- and Villamañán, R., Developing computer simulation as com-
ples for the nature of arguments in the applied field of engi- plementary tool in refrigeration labwork. International Con-
neering where appeals to scientific principles are made to ference of Interactive Computer Aided Learning. Villach,
justify the design of an industrial product. The results high- Austria, 2004.
light the difficulties that tertiary students face with the writ- Driver, R., Newton, P. and Osborne, J., Establishing the norms
ing of arguments. Considering that the student sample in the of scientific argumentation in classrooms, Science Educa-
study were third year engineering students, it is particularly tion, 84(3), 287-312, 2000.
surprising that the majority of the conclusions were not de- Erduran, S. and Jimenez-Aleixandre, M.P. (eds.), Argumenta-
rived from the experimental evidence despite the question- tion in Science Education: Perspectives from Classroom-Based
ing support provided for the writing of the final report. The Research. Dordrecht: Springer, 2008.

124 de aniversario educación química  •  abril de 2009

 Table 2. Rubric to support and evaluate written arguments.
Feature of arguments Analysis criteria Scores
(0 = minimum,
2 = maximum)
Formal validity Are the essential components of an argument (claim, data, warrant)
present and are they accurate?
Structure of reasoning Are there multiple lines of reasoning? What are the different lines of
reasoning?
Are the lines of reasoning plausible given the scope of the thesis?
Do the lines of the reasoning converge to a conclusion?
Observational data Are the data appropriate? Are they based on or derived from observation?
Are the data presented in the right scientific language?
Are the representations well expressed?
Are the data used relevant?
Are the data sufficient to reach the conclusion?
Warrants Do the warrants used relate to the data and the conclusions?
Are the warrants used to justify the use of data to reach the conclusion?
How are the warrants related to backings? Are there any backings to
support and justify the warrants?
Qualifiers Are qualifiers used?
Are qualifiers used of different kinds?
Conclusion Are the inferences valid made between the data and the conclusion valid?
Is the conclusion supported by the data, warrants and backings?

Erduran, S., Simon, S. and Osborne, J., TAPping into argu- 691–718, 2000.
mentation: Developments in the use of Toulmin’s Argu- Kelly, G. and Takao, A., Epistemic levels in argument: an anal-
ment Pattern in studying science discourse, Science Educa- ysis of university oceanography students’ use of evidence
tion, 88(6), 915-933, 2004. in writing, Science Education, 84(6), 757-792, 2002.
Erduran, S., Breaking the law: promoting domain-specificity Keys, C.W., Revitalizing instruction in scientific genres: con-
in science education in the context of arguing about the necting knowledge production with writing to learning in
Periodic Law in chemistry, Foundations of Chemistry, 9(3), science, Science Education, 83, 115-130, 1999.
247-263, 2007. Sandoval, W.A. and Reiser, B.J., Explanation-driven inquiry:
Hand, B., Prain, V., Lawrence, C. and Yore, L.D., A writing in integrating conceptual and epistemic scaffolds for scien-
science framework designed to enhance science literacy, tific inquiry, Science Education, 88, 345-372, 2004.
International Journal of Science Education, 21, 1021-1035, Suppe, F., The structure of a scientific paper, Philosophy of Sci-
1999. ence, 65, 381–405, 1998.
Jimenez-Aleixandre, M.P., Rodriguez, A.B. and Duschl, R.A., Toulmin, S., The uses of argument. Cambridge, UK: Cambridge
“Doing the lesson” or “doing science”: argument in high University Press. 1958.
school genetics, Science Education, 84(6), 757–792, 2000. Zohar, A. and Nemet, F., Fostering students’ knowledge and
Kelly, G.J., Chen, C. and Prothero, W., The epistemological argumentation skills through dilemmas in human genetics,
graming of a discipline: writing science in university ocean- Journal of Research in Science Teaching, 39(1), 35-62,
ography, Journal of Research in Science Teaching, 37(7), 2002.

abril de 2009  •  educación química de aniversario 125