Development of Representational Competence via Cognitively Activating Tasks for
Physics Experiments
Jochen Scheid1,1Andreas Müller2, Wolfgang Schnotz3, Rosa Hettmannsperger1,
Jochen Kuhn4, Sibel Telli1 & Patrik Vogt1,4
University of Koblenz-Landau, Campus Landau, Germany:
1
DFG-Graduate School “Teaching and Learning Processes”
3
Department of General and Educational Psychology
4
Institute for Science Education/Physics Education unit
University of Geneva, Switzerland:
2
Institute of Teacher Education and Faculty of Sciences / Physics Department
Abstract
Science education literature clearly shows that various mental representations and their
interplay are vitally important for a proper understanding of scientific experiments,
phenomena and concepts (Gilbert & Treagust, 2009). Additionally, research results indicate
that students remember and understand little by doing own experiments (Harlen, 1999) as
well as watching the lecturer doing experiments (Crouch, Fagen, Callan, & Mazur, 2004).
There is an increasing body of evidence, however, which shows that learning about and from
experiments can be enhanced by proper cognitive activation such as the predict-observeexplain sequence (Kearney Treagust, Yeo, & Zadnik, 2001; Crouch et al., 2004).
In general, external representations are used as problem solving tools in physics. In
particular, secondary level I students have problems to work correctly with these external
representations, e.g. to translate one type of representation in another type. Therefore, the
purpose of the present study was to develop and investigate cognitively activating tasks by
focussing on the essential role of representations in understanding of experiments.
A study (secondary level I, two classes, n = 59, domain: ray optics, duration 135 min.)
with quasi-experimental, one-factorial design (with vs. without explicit representational
exercises: completing, correcting, adapting & mapping of representations) was carried out.
Dependent variables were conceptual understanding and representational competence. The
treatment with representational analysis tasks (RATs, such as completing, correcting,
adapting & mapping of representations, and typically based on two or more forms of
representations) is described, as well as the operationalization of the target variables.
Results of an ANCOVA show no treatment effects on conceptual understanding, but a
considerable effect on representational competence (p < .001; ω² = .22, large effect size 1),
which is discussed in view of existing theory background. In sum, a relatively short
intervention can lead to a significant and practically important improvement of
representational competence. Limitations and perspectives of the present study are discussed,
including an outlook on future research.
1. Introduction
For a proper understanding of scientific experiments, phenomena and concepts, various
mental representations and their interplay are vitally important (Gilbert & Treagust, 2009).
However, it is not enough to know only one type of these representations at a time such as to
1
Corresponding author: [email protected]
1
According to the conventional categorization, cf. Cohen (1988)
handle an equation or draw and interpret a graph. It is a salient feature of scientific
understanding that it can only occur when students have the ability to build meaningful
connections between different representations (Mayer, 2005).
In particular, the understanding of a physics experiment and its interpretation is dependent
on a meaningful understanding of several relevant representations and their connections with
each other. However, research shows that students often remember and understand little by
doing their own experiments (Harlen, 1999) or watching the lecturer’s experiments (Crouch et
al., 2004). Nevertheless, there is an increasing body of evidence which demonstrates that
learning about and from experiments can be enhanced by proper cognitive activation as the
predict-observe-explain sequence (Kearney et al., 2001; Crouch et al., 2004).
Extending this line of research, the present study aims to develop and investigate
cognitively activating tasks by focussing on the essential role of representations in
understanding of experiments.
2. Research background
Representational competence can be defined as the ability to generate and use different
specific depictional representations or descriptions of a subject or a problem in a sophisticated
way (Dolin, (2007). Various functions and benefits of representations in science (and
mathematics) education are well known, in close parallel to their role in the disciplines
themselves.
Devetak, Urbančič, Grm, Krnel, & Glažar, (2004) pointed out that science (i.e. chemistry)
knowledge of students is fragmented, unless the pertinent representations are connected with
each other. Therefore, students’ drawings and annotated diagrams representing phenomena at
the sub-micro level can provide insight into students’ understanding at the macro level
(Davidowitz & Chittleborough, 2009). More specifically, if students are able to build
meaningful connections between representations, they can acquire a deeper understanding
than it is possible by learning by words or pictures alone (Mayer, 2005). On the other hand,
the necessity of the use of multiple representations imposes a high cognitive load experienced
by students (Sweller, 2005). Cognitive schemata are very important building blocks for the
development of expertise, because they may lower the cognitive load and set free mental
resources (attention, short term memory). Thus, it is very plausible (but not well investigated)
that cognitive schemata caused by cognitively activating tasks related to representations could
help the student’s understanding of science experiments with their underlying interplay of
multiple representations and the particularly high cognitive load resulting from it.
The results reported so far, mainly related to chemistry education and educational
psychology, seem transferable and adaptable to physics education. For example, Lee (2009)
analyzed the usage of representations in 26 lessons in three physics classes on optics with
secondary level I students. Almost 150 stances of representation usage were found, most often
implicit and lasting on average less than three minutes. The most frequent source of
representations was the teacher, followed by the textbook or curriculum materials. Less than
ten percent of the representations were generated by students. This short, receptive and
therefore implicit work with representations leads to obvious restrictions and difficulties for
the learning process. For this reason, students lack cognitive activation and therefore they do
not have the possibility to develop relevant cognitive schemata. Moreover, even in the physics
courses at university level, students’ representational competence is low (Saniter, 2003).
Scheid & Müller (2010) investigated in the domain of electrodynamics how preservice
teacher students deal with representations in experiment related tasks. Qualitative analyses
revealed that students were more skilled in solving tasks by descriptive representations even
on a formal level (e.g. mathematical equations) than in solving tasks by depictive
2
representations (e.g. graphs). Moreover, the following difficulties were found even for
descriptive representations generated by students: on a basic level, some students, for
example, confused symbols for physical quantities with symbols for units (same symbol, but
different meaning), showing a lack of conceptual understanding. The students tried also to
establish connections between surface features (characters) without being aware what the
representation (symbol) symbolizes. On a higher level, most students had difficulties fitting
their knowledge into an integrated representation (i.e. a graph of entire time course of voltage
vs. time in a self induction experiment), even though most of them showed a proper
understanding of the most important elements of the phenomenon “induction” (such as
dI
and correct an application on an induction spark of a switch). Generally, even
U ind = L
dt
physics preservice teachers’ degree of representational competence in general and coherence
in particular was low and had to be improved.
3. Rationale, Research Questions and Hypotheses of the Study
Concluding from the preceding section, there is an urgent need for further research and
research based development of tasks on representations in science education. In particular, we
better need to know, why students’ representational competence is so unexpectedly low and
how it may impede learning from experiments. Accordingly, the rationale of the present study
is to improve this situation by explicitly integrating representations in the teaching-learningprocess. Specifically, it aims at designing and investigating cognitively activating tasks
fostering students’ representational competence (e.g. the ability to translate different types of
representations in each other and correct incoherent pairs of representations) and improving in
this way the understanding of physics experiments. For the purpose of our investigation, we
choose ray optics as representationally rich topic. On this basis, the study had the following
research aims and questions:
1. Is it possible to test for representational competence with items which are curricularly
valid for the chosen topic (ray optics)? To which degree are they reliable?
Hypothesis 1: The working hypothesis of the study is that a valid and sufficiently reliable
test of representational competence is possible.
2. What is the impact of the newly developed cognitively activating representational tasks on
students’ conceptual understanding and representational competence in ray optics?
Hypothesis 2: Based on the existing literature, we expect that an explicit focus on
representation related, cognitively activating tasks lead to improved learning.
4. Materials and Methods
The topic was ray optics (image formation by a convex lens, a standard topic in German
secondary level I). The sample comprised 59 students in two classes taught by the same
teacher (23 boys, 36 girls, 8th grade of German gymnasium (similar to UK grammar school),
age group 13 – 14 years; 4 students did not participate in one of two tests and were excluded
from analysis). The total duration of the intervention was 3 lesson hours (in sum 135 min).
Additionally, the pre- and post test demanded one class hour in each case.
The study had a quasi-experimental, one-factorial design. The instruction in the treatment
group (TG) was based on tasks explicitly asking for analysis of various representational
features (representational analysis tasks, RATs) , such as completing, correcting, adapting &
mapping of representations, and typically based on two or more forms of representations (see
e.g. Figure 1); the instruction in the control group was based on traditional tasks, e.g. dealing
3
Object G
Convex lens
Image B
Figure 1: A realistic picture and a schematic drawing of a cognitive activating task for formation of image by a
convex lens. Task type: mapping and correcting the schematic drawing; hint given: size of lenses and
perspective of drawing do not matter.
with construction of ray diagrams only, and typically dealing with only one type of
representation at a time.
Dependent variables were operationalized as follows: Conceptual understanding was
measured by a concept test, which was designed as multiple choice test and consisted of 30
items. All the items had known misconceptions in ray optics as distractors, based on existing
research (Goldberg, & McDermott, 1987; Wiesner, 1992; Reiner, Slotta, Chi, & Resnick,
2000). The reliability, as measured by αC, was .73.
Representational competence was tested with a 16 items test on completeness, consistency
and coherence of domain specific representations. Development and validation (by expert
rating) were carried out in cooperation between physics teachers and the local PER group.
Reliability (and other characteristics) of this test will be reported below; more details about
both tests will be reported elsewhere (Hettmannsperger, Scheid, Müller, Schnotz, Kuhn, Telli,
& Vogt, 2011a).
Cognitive activation by RATs occurs if students think more often or more deeply about
physics related representations, express them verbally and draw conclusions from them as
would be the case without adequate instructional means. Figure 1 is an illustrative example of
such task designed for the purposes of this study. This task is cognitively activating
(according to hypothesis 2) because of three reasons: first, students were asked to analyze
whether the realistic and the schematic picture belong to exactly the same experimental
situation (which is not the case). Furthermore, they should identify and name the differences
between the arrangements of optical items of both pictures. Second, the students had to
correct the schematic drawing (in order to be equivalent to the realistic one). Third, students
were asked to express verbally why and how they made the correction. In this way, students
worked with three different types of representations (realistic picture, schematic drawing and
words). In general, the “translation” between these three types of representations is a big
challenge, to be explicitly and actively practiced by students using such a kind tasks. In
particular, coherence between the three different forms of representations is established, if
they correspond in conjoint information (the pictures presented in Figure 1 are not coherent).
5. Results
Preliminary analyses showed that three items of the representational physics test had to be
changed before using them in further studies. Item difficulties of the residual test are .32 - .56
(M = .41, SD = .13). Corrected scale correlation is .42-.56 (M = .50; SD = .08) and αC is .75.
To answer the second research question, it is necessary to investigate main (and possible
interaction) effects of the treatment on conceptual understanding and on representational
competence. For this purpose an ANCOVA for each dependent variable was carried out.
For the concept post test score as dependent variable of the first ANCOVA, results
showed an effect of the pre test score (p = .006; ω² = .12, medium effect size), but no effect of
the RAT intervention compared to the control group.
4
For the representational competence post test score as dependent variable of the second
ANCOVA, again an influence of the pre test score was found (p = .001; ω² = .19, large effect
size). Moreover, in this case a considerable effect in favour of the treatment group could be
established (p < .001; ω² = .23, large effect size).
6. Discussion and Outlook
The results indicate that the test is acceptable for further use, supporting hypothesis 1 as
working assumption of the study.
One the one hand, the treatment group considerably (ω² = .23) outperformed the control
group in representational competence, which partially confirms Hypothesis 2. This increases
confidence that students’ representational competence, as an important part of physics
understanding in general, indeed can be supported by the kind of specific cognitive activation
adopted in this study.
On the other hand, the RAT treatment had no effect on conceptual understanding, so this
part of hypothesis 2 is not confirmed. A possible explanation is that the concept test assessed
students’ conceptual difficulties beyond the scope of the intervention, in a broader area of ray
optics and beyond representational difficulties. So the results leave us with the message, that
representational analysis tasks indeed can support representational competence, but that in
order to overcome misconceptions, they have to be combined with other instructional
measures (Hettmannsperger, Schnotz, Müller, Scheid, Kuhn, Telli & Vogt, 2011b).
In summary, a relatively short intervention in the area of ray optics (135 min. altogether)
on the basis of representational analysis tasks can lead to a significant and practically
important improvement of representational competence. Both the intervention and assessment
are compatible with curricular and classroom conditions (at least in the author’s country), in
order to ensure practical applicability of the approach.
Further research is under way in order to increase sample size, statistical power, and in
this way also the possibilities of a more detailed investigation of possible ATI effects, such as
students’ motivation and intelligence (verbal, numerical, reasoning) and academic level of
pertinent school subjects (German language, physics and mathematics).
References
Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Erlbaum.
Crouch, C. H., Fagen, A. P., Callan, J. P. & Mazur, E. (2004). Classroom demonstrations: Learning tools or
entertainment? American Journal of Physics, 72 (6), p. 835-838.
Davidowitz, B. & Chittleborough, G. (2009). Linking the Macroscopic and Submicroscopic Levels: Diagrams.
In John, K., Treagust, D. (Eds.). Multiple representations in chemical education. In models and modeling
in science education. Volume 4, New York: Springer, p. 169.
Devetak, I., Urbančič, M. W., Grm, K. S., Krnel, D. & Glažar, S. A. (2004). Submicroskopic representations as a
tool for evaluating students´ chemical conceptions. Acta Chimica Slovenica, 51, p. 799−814.
Dolin, J. (2007). Science education standards and science assessment in Denmark. In D. Waddington, P.
Nentwig & S. Schanze (Eds.). Making it comparable. Standards in science education Münster: Waxmann,
p. 71-82, p. 77.
Gilbert, D. & Treagust, J. K. (2009). Towards a Coherent Model for Macro, Submicro and Symbolic
Representations in Chemical Education. In Treagust, J. K., Gilbert, D. (Eds.). Multiple representations in
chemical education. In Models and modeling in science education, Volume 4, New York: Springer, p. 333.
Goldberg, F.M. & McDermott, L. C. (1987). An investigation of student understanding of the real image formed
by a converging lens or concave mirror. American Journal of Physics, 55 (2), p. 108-119.
Harlen, W. (1999). Effective teaching of science - A review of research. Edinburgh: Scottish Council for
Research in Education, p. 7.
Hettmannsperger, R., Scheid, J., Müller, A., Schnotz. W., Kuhn, J., Telli, S., Vogt, P. (2011a). A concept test
and representational competence test in the area of ray optics (manuscript in preparation).
5
Hettmannsperger, R., Schnotz, W., Müller, A., Scheid, J., Kuhn, J., Telli, S., Vogt, P. (2011b). Fostering
representational and experimental competence considering students’ prior knowledge in middle school
physics
classes.
Paper
presented
at
the
GIREP
Conference
in
Reims,
2010.
http://www.univ-reims.fr/site/evenement/girep-icpe-mptl-2010-reims-international-conference/list-ofsubmitted-full-papers-for-proceedings,13181,22950.html? (follow menu “posters”; access 11-09-29).
Hoffmann, L., Häußler, P. & Peters-Haft, S. (1997). An den Interessen von Mädchen und Jungen orientierter
Physikunterricht. Ergebnisse eines BLK-Modellversuches. Kiel: Leibniz-Institut für die Pädagogik der
Naturwissenschaften (IPN) an der Universität Kiel.
Kearney, M., Treagust, D. F., Yeo, S. & Zadnik, M. (2001). Student and Teacher Perceptions of the Use of
Multimedia Supported Predict-Observe-Explain Tasks to Probe Understanding. Research in Science
Education, 31, p. 589-615.
Lee, V. (2009). Examining patterns of visual representation use in middle school science classrooms.
Proceedings of the National Association of Research in Science Teaching (NARST) Annual Meeting
Compact Disc, Garden Grove, CA: Omnipress.
Mayer, R. E., ed. 2005. Cambridge Handbook of Multimedia Learning. Cambridge: Cambridge University Press.
Scheid, J, Müller, A. (2010). Preservice teacher student’s usage of representations in an Electrodynamics course
at university level. Unpublished document.
Reiner, R., Slotta, J.D., Chi, M.T. H. & Resnick, L.B. (2000). Naive Physics Reasoning: A Commitment to
Substance-Based Conceptions. Cognition and Instruction, 18 (1), p. 1-34.
Saniter, A. (2003). Spezifika der Verhaltensmuster fortgeschrittener Studierender der Physik. In Niedderer, H.,
Fischler H. [Hrsg.]. Studien zum Physiklernen Band 28, Berlin: Logos, p. 289.
Sweller, J. (2005). Implications of cognitive load theory for multimedia learning. In Mayer, R. E. (Ed. 2005), p.
19-30.
Wiesner, H. (1992). Verbesserungen des Lernerfolgs im Unterricht über Optik. Schülervorstellungen und
Lernschwierigkeiten. Physik in der Schule, 30 (9), p. 286-290.
6