Conference PaperPDF Available



Abstract and Figures

In this study, chemistry professors (experts) & undergrads (novices) view & categorize MERs. Using eye-tracking, we capture fine-grained data about participants' gaze patterns while they view given MERs, which we then correlate with the quality of categories they generate as well as justifications they provide for those categories. The professors tend to form chemically meaningful relationships between MERs than do undergrads. Eye-tracking data reveal differences between the two groups, in navigating chemical equations.
Content may be subject to copyright.
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
Prajakt Pande1, Prateek Shah2, Sanjay Chandrasekharan1
1HBCSE, TIFR, Mumbai, 2IIM, Ahmedabad, India.,,
In this study, chemistry professors (experts) & undergrads (novices) view & categorize MERs.
Using eye-tracking, we capture fine-grained data about participants' gaze patterns while they
view given MERs, which we then correlate with the quality of categories they generate as well as
justifications they provide for those categories. The professors tend to form chemically
meaningful relationships between MERs than do undergrads. Eye-tracking data reveal
differences between the two groups, in navigating chemical equations.
Chemistry deals with complex systems, entities & phenomena that often cannot be directly
perceived (e.g. atoms, chemical reactions, etc.) These imperceptible systems are understood at
multiple levels of detail (electronic configuration, stereo-chemistry, stoichiometric ratios etc.),
using multiple external representations (MERs), such as reaction mechanisms, molecular
diagrams, graphs & equations, at each level. The ability to generate & use these MERs in an
integrated fashion (for conceptualization, discovery & communication) is indicative of expertise
in chemistry. This skill-set is collectively known as representational competence (abbreviated as
RC, Kozma & Russell, 1997). Developing RC (expertise over MERs) is an important goal of
chemistry education. Problems & difficulties in teaching/learning chemistry are attributed to
difficulties in understanding the MERs in chemistry (Johnstone, 1991, 1993; Kozma & Russell,
1997; Gilbert & Treagust, 2009).
A significant strand of research in chemistry education reports descriptions of students’ use of
multiple representations, transformations of these representations, and the difficulties students
face while doing both of the above. Studies show that students fail to associate the symbols and
numbers with substances and phenomena (in other words relate MERs and the information they
convey; Yarroch, 1985; Herron and Greenbowe, 1986; Nurrenbern & Pickering, 1987; Hinton &
Nakhleh, 1999; Sanger & Phelps, 2007), primarily due to a lack of clarity on basic concepts such
as oxidation numbers, ionic charge, atoms and atomic structure, formal rules for writing
molecular formulae, as well as meaning of subscript numbers and brackets and coefficients
(Garforth, Johnstone & Lazonby, 1976; Savoy, 1988). Ben-Zvi, Eylon, & Silberstein, (1988)
propose that students' thinking about phenomena relies primarily on perceptual/sensory
information but since current pedagogical practices hardly provide perceptual/sensory assistance,
students do not understand chemical symbols in terms of their macro and micro-level
instantiations. Johnstone's model of three thinking levels (Johnstone, 1982) and versions thereof,
describe three different levels of chemistry MERs: (a) macro level, where one sees and handles
materials, observes and describes phenomena and their properties, such as color, flammability,
solubility, (b) symbolic level, where one represents chemical substances and phenomena using
symbols, formulas, equations and conventions, and (c) submicro level, at which one explains the
nature of chemical substances, mechanisms of reactions, and the underlying molecular/atomic
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
interactions. Johnstone (1991 & 2000) attributes students' difficulties in learning chemistry to the
difficulty in simultaneously handling MERs distributed across these three levels as a result of the
limited capacity of the human working memory (Ben-Zvi, Eylon & Silberstein, 1988; Justi &
Gilbert, 2002; Kozma & Russell, 1997; Mayer, 2002; Treagust, Chittleborough & Mamiala,
2003; Sirhan, 2007).
Another strand of research attempts to characterize and examine RC, and describes expert-novice
differences in terms of use of MERs. For instance, researchers demonstrate using eye-tracking,
that students mainly concentrate on graphical and model representations in animations and often
ignore equations, when interacting with a multi-representational molecular mechanics animation
(Stieff, Hegarty & Deslongchamps, 2011). While students face difficulties in producing static
representations (e.g. sketches; Madden, Jones & Rahm, 2011) of the (imagined) dynamic
particulate interactions, experts, on the other hand, seem to better transform between static (such
as equation & graphs) and dynamic representations (such as reaction mechanisms; Wu & Shah,
2004; Kelly & Jones, 2008; Nakhleh & Postek, 2008).
Kozma and Russell (2005), identify specific skills among chemistry experts, viz., (a) using
representations to describe chemical phenomena, (b) generating and/or selecting appropriate
MERs according to specific needs, (c) identifying and analyzing different features of MERs, (d)
comparing and contrasting different M E R s , (e) making connections across different
representations, relating/mapping features between MERs, (f) understanding that the MERs
correspond to phenomena but are distinct from them, and (g) using MERs to support claims,
draw inferences, and make predictions. Levy and Wilensky (2009) suggest that understanding
chemical phenomena involves building of internal (mental) models that simulate the behaviors of
many individual molecules/atoms, their collective behaviors and properties, and effects of
various parameters on such behaviors.
Current characterizations of student difficulties and/or RC in chemistry can summarily be
categorized into – cognitive load based explanations (expert is better able to handle the cognitive
load by employing cognitive strategies such as information chunking, whereas novices lack such
skills, Cook, 2006; Johnstone, 1982), context & practice based accounts (students lack exposure
to these while experts have had ample exposure, Ben-Zvi, Eylon, & Silberstein, 1987 & 1988;
Nelson, 2002; Tsaparlis, 2009), and conceptual understanding/prior-knowledge based
explanations (which say that students have superficial understanding and low prior knowledge
making it difficult for them to understand MERs; Cook, 2006; Cook, Wiebe & Carter, 2007; Nitz
& Tippett, 2012). Ultimately, all these accounts boil down to the classical information processing
framework emphasizing cognitive load and strategies to lower/handle it. Such accounts do not
seek to provide a detailed understanding of the cognitive mechanisms underlying the processing
of MERs, and thus offer only a rather superficial account of MER integration.
Our research attempts to characterize RC by developing models of the cognitive mechanisms
underlying the processing of MERs, particularly integration of MERs (which is how we define
RC), and suggest design principles for interventions. In this study, chemistry professors (experts)
& undergrads (novices) view & categorize MERs. Using eye-tracking, we capture fine-grained
data about participants' gaze patterns while they view given MERs, which we then correlate with
the quality of categories they generate as well as justifications they provide for those categories.
The professors tend to form chemically meaningful relationships between MERs than do
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
undergrads. Eye-tracking data reveal differences between the two groups, in navigating chemical
We used Tobii X2-60 static eye-tracker to capture fine-grained data on student eye-movement
and gaze patterns across MERs presented to (and handled by) them. Our preliminary analysis
confirms earlier reports on novices' surface-feature-based exploration of MERs, but adds details
of eye-gaze patterns.
An MER categorization task (from Kozma & Russell, 1997) was conducted with six chemistry
undergrad students (3 girls). We describe below the two phases of the study.
Preparing task material
Materials for the categorization experiment included different representations for five pre-
determined general chemical reactions. There were four representations corresponding to each
reaction – a chemical equation, a graph (except for the precipitation reaction), a video of
laboratory personnel performing the reaction in a laboratory, and a bare 3D molecular animation
(that depicted only the reaction mechanism at molecular level).
We developed bare 3D molecular animations for the five chemical reactions. Each animation
depicts only the molecular dynamics of that reaction, and does not have any other embedded
representations, such as text, narrative, graphs or equations; thus, only one kind of
representation. We downloaded free and open videos of the five chemical reactions (being
performed in laboratories) from on-line sources. Chemical equations and approximate graphs for
each reaction (except for the precipitation reaction that had no graph) were generated using an
image editing program. This resulted in 19 representations corresponding to five different
chemical phenomena. To make these representations more convenient for physical handling, the
image of each representation (for animation and video, snapshot of an important moment as an
image) was color printed and pasted on a 3x4 inch cardboard, generating 19 cards. Figure 1
depicts preparation and execution of the experiment in detail.
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
Figure 1: Material development and experimental design details
Running the experiment
Six chemistry undergrads (3 females) as novices and seven chemistry faculty (4 females) as
experts from different university colleges in the city of Mumbai participated in the categorization
experiment. Each participant performed the experiment individually. The experiment had two
(a) On-screen phase
Participant was given each of the 19 cards (one after the other, in a pre-determined random order
maintained for all participants), and was shown the corresponding image/video on a laptop
screen. The participant could observe the images as long, and videos/animation as many times as
he/she wanted. Going back to a previously shown representation was not allowed.
(b) Off-screen phase
Once the participant viewed all the 19 representations and had all the cards, he/she was asked to
group the cards into meaningful categories. There was no time limit to this phase. They were also
asked to explain the different categories made and the basis of categorization (relationship
between the cards/representations). The researcher then asked the participant to perform another
round of categorization using a different grouping scheme, and explain the grouping criteria.
Data Collection
We used eye-tracking (Tobii X2-60, a static eye-tracker) during the on-screen phase of the task,
to obtain fine-grained data about participants’ eye-movements and gazes when they viewed the
representations (See Pande & Chandrasekharan, 2014b for details eye-tracker setups).
Sources of data collection: (a) for on-screen phase dynamic eye-movement and fixation data
superimposed on the screen-capture video, (b) for off-screen phase – categories made by the
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
participants, their verbal justifications, and side-view video recording of the categorization and
justification sessions. The entire session ranged from 40-60 minutes for each participant.
(1) Do experts make more chemically meaningful associations between MERs than novices?
(2) Are there any gaze-pattern differences between experts & novices over static
representations? If yes, what differences? How are they related to categorization?
(i) In graphs, the total fixation duration for experts would be higher for curves than the
axes, as the shape of the curves conveys dynamic information about the phenomena
(e.g. sigmoid behavior with time). Instead, novices are likely to spend more time on
the axes than curves, as they might find numerical information more relevant to
adhere to.
(ii) On equations, experts' total fixation duration would be more distributed across
reactants, arrow, and products, as they would systematically look through each part of
the equation and transit between the sub-scripts, super-scripts and coefficients.
Novices would either move randomly or tend to focus either on reactants or products
(iii) Experts would make more long-distance transitions over different parts of the
equations, than novices, who would tend to move between closely located elements.
We report preliminary results on (a) the nature of categories experts and novices make in the first
trial of categorization, and the justifications they provide for those categories, (b) statistical
analysis of the fixation data on static representations (graphs and equations), and (c) fine-grained
process data on how the two groups differ in the way they navigate chemical equations during
the viewing phase.
Nature of categories
We coded the categories of representations participants generated, based on the chemical
meaningfulness of relations/connections participants established between different
representations, into following five types. (1) Conceptual categories: Chemically meaningful
combinations of cards supplemented with correct conceptual description of grouping criteria
(e.g. associations of cards depicting equilibrium phenomena, precipitation reaction). (2) Mixed
categories: Categories with correct/plausible combinations of cards, with some associations
and/or representations explained using chemical concepts while others explained using visual
features (e.g. a category made with, say 4 cards depicting equilibrium reaction, of which two
cards are explained using the concept of equilibrium while the other two explained based on
similarity in features such as heating, or temperature-concentration axes of a graph). (3)
Categories based on similarity in visual-features between the representations: Associations of
cards explained purely on the basis of visual features of the representations grouped together
(e.g. animation showing settling of molecules and a laboratory demonstration exhibiting
precipitation; association explained in words such as, ‘both settling down’.) (4) Media-based
categories: Complete me d i a - b a s e d c o m b i n a t i o n s o f cards (e.g. all molecular
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
animations/simulations as a category, all graphs as another, etc.), and (5) Non-sense categories:
Incorrect or meaningless combinations of cards not employing falling under any of the above
category types (e.g. an association between a precipitation reaction equation with a video
showing effect of temperature on a chemical equilibrium)
Experts tend to form more number of mixed
as w e ll a s c on c e p tu a l (c h e m ic a ll y
meaningful) categories than do novices, who
tend to associate MERs more often based on
their visual features and their medium of
representation. This confirms the results
from a previous study by Kozma and
Russell (1997). The two groups do not seem
to differ from each other in terms of the
number of non-sense and media-based
categories they made. Figure 2 depicts the
mean percentage for each type of category
generated by experts and novices, during the
first round of categorization. A similar trend
is ob s er v ed o v er s e co n d r ou n d of
Figure 2: Distribution of participants’
categories across different types
Fixation/visit duration analysis
Fixation duration is a useful statistic to understand the total time spent by a participant viewing a
given area of interest (AOI) or part of the representation while viewing it. We found no expected
differences between experts and novices. They seem to spend their time viewing the different
AOIs roughly similarly, thus rejecting hypotheses (i) and (ii). Both the groups seem to fixate
slightly longer on the axes in the graphs, and reactants in the equations.
Figure 3: (a) Percent fixation duration on different parts (areas of interest - AOIs) across all four
graphs presented, (b) Percent fixation duration on different AOIs across all the five equations.
Since nothing conclusive can be said through the fixation duration statistics, we decided to delve
further into the viewing/thought process data. Below we report one aspect of such qualitative
data – nature of fixation transitions (jumps).
Nature of gaze transitions
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
Here we report transition data only for equations. We characterized two kinds of transitions viz.
long jumps (gaze transitions occurring within two distantly situated AOIs in the space) and short
jumps (gaze transitions happening over two closely situated AOIs). For instance, in figure 4, any
direct transition between the two reactants (R1 and R2) or between the two products (P1 and P2)
would be counted as short jumps, whereas, transitions between the reactant and the product side
would be long jumps.
Figure 4: long and short jumps
Experts performed more number of long jumps than novices on an average, while novices tended
to perform more number of shorter jumps than longer jumps in comparison to the experts (results
can be considered as partially significant at p = 0.05, as the extreme deviations from both groups
overlap slightly, apparent in the box plots in figure 5).
Figure 6 depicts a normalized distribution of long jumps performed by experts and novices
across all the equations. Experts make significantly higher number of longer jumps than novices.
Conversely, they make significantly less number of short jumps than the novices.
Figure 5: Box plots capturing (a) mean number of long jumps across all equations, (b) mean
number of short jumps across all equations.
Figure 6: Percentage long
jumps performed by experts
and novices across all the
equations, an inverse would be
percent short jumps performed.
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
Our findings confirmed some results from previous literature, and added further details about
how experts and novices move their eyes as they navigate (through) the MERs. Experts tend to
make chemically meaningful as well as mixed groups of MERs in the categorization task more
often than do novices, who tend to relate MERs based on their surface features. The eye tracking
data suggests that RC and expertise can be characterized in terms of eye movements and gaze
patterns across MERs. Further analysis would help isolate eyemovement and navigation patterns
related to RC.
Ben-Zvi, R., Eylon, B., & Silberstein, J. (1988). Theories, principles and laws. Education in chemistry, 25(3), 89-92.
Cook, M. P. (2006). Visual representations in science education: The influence of prior knowledge and cognitive
load theory on instructional design principles. Science education, 90(6), 1073-1091.
Cook, M., Wiebe, E. N., & Carter, G. (2008). The influence of prior knowledge on viewing and interpreting graphics
with macroscopic and molecular representations. Science Education, 92(5), 848-867.
Garforth, F. M., Johnstone, A. H. & Lazonby, J. N. (1976). Ionic Equations: Difficulties in Understanding and Use.
Education in Chemistry., 13, 72–75.
Gilbert, J. K., & Treagust, D. (2009). Towards a coherent model for macro, submicro and symbolic representations
in chemical education. In J. K. Gilbert & D. Treagust (Eds.), Multiple representations in chemical education,
333–350. The Netherlands: Springer.
Herron, J. D. & Greenbowe, T. J. (1986). What Can We Do About Sue: A Case Study of Competence. Journal of
Chemical Education, 63(6), 528.
Hinton, M. E., & Nakhleh, M. B. (1999). Students’ microscopic, macroscopic, and symbolic representations of
chemical reactions. The Chemical Educator, 4(4), 158–167.
Johnstone, A. H. (1982). Macro and microchemistry. School Science Review, 64(227), 377-379.
Johnstone, A. H. (1991). Why is science difficult to learn? Things are seldom what they seem. Journal of computer
assisted learning, 7(2), 75-83.
Johnstone, A. H. (1993). The development of chemistry teaching: A changing response to changing demand.
Journal of Chemical Education, 70(9), 701–705.
Johnstone, A. H. (2000). Teaching of chemistry: Logical or psychological? Chemistry Education: Research and
Practice in Europe, 1(1), 9–15.
Justi R. S. & Gilbert J. K., (2002). Models and modelling in chemical education. In J. Gilbert, O. De Jong, R. Justi,
D. F. Treagust & J. H. Van Driel (Eds.), Chemical education: towards research-based practice, 213-234,
Dordrecht: Kluwer Academic Publishers.
Kelly, R. M. & Jones, L. L. (2008). Investigating students' ability to transfer ideas learned from molecular
animations of the dissolution process. Journal of Chemical Education 85 (2), 303-309.
Kozma, R. B., & Russell, J. (1997). Multimedia and understanding: Expert and novice responses to different
Tsaparlis, G. (2009). Learning at the macro level: The role of practical work. In Multiple representations in
chemical education (pp. 109-136). Springer Netherlands.representations of chemical phenomena. Journal of
research in science teaching, 34(9), 949-968.
Kozma, R., & Russell, J. (2005). Students becoming chemists: Developing representationl competence. In J. K.
Gilbert (Ed.), Visualization in science education, 121-145. Springer: 2005.
Pande, P., Shah, P. & Chandrasekharan, S. (2015). How do experts and novices navigate chemistry representa"ons –
an eye tracking inves"ga"on, In S. Chandrasekharan, S. Murthy, G. Banarjee, & A. Muralidhar (Eds.), Proceedings of
EPISTEME-6 (pp. 102-109), HBCSE-TIFR, Mumbai, India.
Levy, S. T. & Wilensky, U. 2009. Crossing levels and representations: The connected chemistry (CC1) curriculum.
Journal of Science Education and Technology, 18(3): 224–242.
Madden, S. P., Jones, L. L., & Rahm, J. (2011). The role of multiple representations in the understanding of ideal gas
problems. Chemistry Education Research & Practice., 12(3), 283-293.
Mayer, R. E. (2002). Multimedia learning. Psychology of Learning and Motivation, 41, 85-139.
Nakhleh M. B. & Postek B. (2008). Learning Chemistry Using Multiple External Representations .In J.K. Gilbert et
al., (eds.), Visualization: Theory and Practice in Science Education, 209–231. Spinger: 2008.
Nelson, P. G. (2002). Teaching chemistry progressively: From substances, to atoms and molecules, to electrons and
nuclei. Chemsitry Education Research & Practice, 3(2), 215-228.
Nitz, S., & Tippett, C. D. (2012). Measuring Representational Competence in Science. In E. de Vries & K. Scheiter
(Eds.), Staging Knowledge and Experience: How to Take an Advantage of Representational Technologies in
Education and Training?, 163-165.
Nurrenbern, S. C. & Pickering, M. (1987) Concept Learning versus Problem Solving: Is There a Difference?
Journal of Chemical Education, 64(6), 508.
Pande, P., & Chandrasekharan, S. (2014) Eye-tracking in STEM education research: limitations, experiences and
possible extensions, In Kinshuk & Murthy, S. (Eds.), Proceedings of The 6th IEEE International Conference on
Technology for Education (pp. 116-119), Amrita University, Kerala: IEEE Conference Publications/ Conference
Publishing Services.
Sanger, M. J., & Phelps, A. J. (2007). What are students thinking when they pick their answer? A content analysis of
students’ explanations of gas properties. Journal of Chemical Education, 84(5), 870–874.
Savoy, L. G. (1988). Balancing Chemical Equations. School Science Review, 69(249), 713-720.
Sirhan, G. (2007). Learning Difficulties in Chemistry: An Overview. Journal of Turkish Science Education, 4 (2), 2-
Stieff, M., Hegarty, M., & Deslongchamps, G. (2011). Identifying representational competence with multi-
representational displays. Cognition and Instruction, 29(1), 123-145.
Treagust D. F., Chittleborough G. D. & Mamiala T. L., (2003). The role of sub-microscopic and symbolic
representations in chemical explanations. International Journal of Science Education, 25, 1353-1369.
Wu H. K., Shah P. (2004). Exploring visuospatial thinking in chemistry learning. Science Education, 88(3):465–492.
Yarroch, W. L. (1985). Student understanding of chemical equation balancing. Journal of Research in Science
Teaching, 22(5), 449-459.
... Indicative evidence from our pilot studies in the past has shown that this expert-novice difference in perceptual navigation could be a function of growing expertise. Undergraduate students who performed categorisation in ways similar to experts (Pande et al., 2015), in one such study, tended to exhibit expert-like gaze patterns across equations and graphs, in contrast to relatively novice candidates who scanned equations linearly with long pauses in between, with nearly twice as many short transitions as long transitions ( fig. 7). ...
... Basu, Sengupta & Biswas, 2015;Kothiyal et al., 2014;Majumdar et al., 2014;Virk & Clark, 2019). These new evidences demonstrate how one's richness of understanding of concepts and representations in Fig. 7 An instance of gaze behaviour of an expert-like undergraduate student (Pande et al., 2015). Arrows indicate direction of transitions. ...
... Our results extend findings from previous research, indicating that distinct perceptual navigation (eye movement) marks expertise in chemistry. Furthermore, we show how this marker is triggered specifically while solving problems that demand representational competence, and how this could be related to the growth of expertise (Pande et al., 2015). The specific activation of experts' perceptualsensorimotor behaviour indicates that their perceptual system could be 'tuned' to support representational competence, allowing them to seamlessly integrate perception and imagination processes, as well as (epistemic) actions. ...
Full-text available
Representational competence in science is the ability to generate external representations (e.g. equations, graphs) of real-world phenomena, transform between these representations, and use them in an integrated fashion. Difficulties in achieving representational competence are often considered central to difficulties in learning science. Representational competence is indicative of domain expertise and is characterised by distinct problem-solving strategies. Eye-tracking studies have consistently demonstrated that experts employ unique perceptual attention (e.g. gaze fixation) patterns while solving problems that involve different external representations. Here, we present a different strand of evidence, indicating that perceptual navigation patterns (eye movements) mark representational competence in science, in more specific ways than attention. Gaze behaviours of chemistry professors (experts) and undergraduate students (novices) were tracked as they individually performed a multi representational-categorisation task and a chemical equation-balancing task. The following three-step analysis was performed on these data: (i) First, we independently calibrated the levels of representational competence of our participants through their performance in the categorisation task. (ii) Then, we compared these competence levels with the participants’ perceptual patterns (gaze behaviour) exhibited during the categorisation task. (iii) Finally, we analysed whether the identified perceptual patterns were specific to representational competence, or more general, through the results of the equation-balancing task. Our analysis of perceptual navigation (eye movements) provided further support to previous findings showing gaze-behaviour differences between experts and novices. Going further, our analysis indicated that experts deploy distinct eye movement patterns, but specifically during representational competence-related problems. This suggests that representational competence is an embodied skill that fundamentally changes the tuning of the perceptual system, as argued by recent ‘field’ theories of cognition.
... Stieff, Hegraty and Deslongchamps (2011) examined students' use of a multi-representational molecular mechanics animation using eye-tracking, and observed that students mainly used graphical and model representations in animations, and often ignored the equation. Based on an eye-tracking exploration of participants' chemistry MER viewing as well categorization processes, Pande and Chandrasekharan (2014), and Pande, Shah and Chandrasekharan (2015) concluded that the richness of transitions, as well as the nature of transitions, between different parts of a representation (and/or different representations) could be considered a good marker of MER integration. ...
Full-text available
Multiple external representations (MERs) are central to the practice and learning of science, mathematics and engineering, as the phenomena and entities investigated and controlled in these domains are often not available for perception and action. MERs therefore play a twofold constitutive role in reasoning in these domains. Firstly, MERs stand in for the phenomena and entities that are imagined, and thus make possible scientific investigations. Secondly, related to the above, sensorimotor and imagination-based interactions with the MERs make possible focused cognitive operations involving these phenomena and entities, such as mental rotation and analogical transformations. These two constitutive roles suggest that acquiring expertise in science, mathematics and engineering requires developing the ability to transform and integrate the MERs in that field, in tandem with running operations in imagination on the phenomena and entities the MERs stand for. This core ability to integrate external and internal representations and operations on them – termed representational competence (RC) – is therefore critical to learning in science, mathematics and engineering. However, no general account of this core process is currently available. We argue that, given the above two constitutive roles played by MERs, a theoretical account of representational competence requires an explicit model of how the cognitive system interacts with external representations, and how imagination abilities develop through this process. At the applied level, this account is required to develop design guidelines for new media interventions for learning science and mathematics, particularly emerging ones that are based on embodied interactions. As a first step to developing such a theoretical account, we review the literature on learning with MERs, as well as acquiring RC, in chemistry, biology, physics, mathematics and engineering, from two perspectives. First, we focus on the important theoretical accounts and related empirical studies, and examine what is common about them. Second, we summarise the major trends in each discipline, and then bring together these trends. The results show that most models and empirical studies of RC are framed within the classical information processing approach, and do not take a constitutive view of external representations. To develop an account compatible with the constitutive view of external representations, we outline an interaction-based theoretical account of RC, extending recent advances in distributed and embodied cognition.
Conference Paper
Full-text available
Eye-tracking technology is being increasingly used in science, technology, engineering and mathematics (STEM) education research. However, most available eye-tracking devices are oriented towards research problems focusing on attention, particularly in areas such as advertising, linguistics, human factors, human-computer interaction, training simulators, sports and virtual reality. Problems in these areas are fundamentally different from those in STEM education, where attention is only one of the many important variables in teaching-learning. Since learning is a process happening over time, STEM investigations focus on understanding the learning process, and this requires moving beyond attention information, doing sophisticated analysis of fixation data, and ways of collecting richer data that support such process analysis. In this paper, we present our difficulties and experiences in using standard eye-tracking systems to understand the learning process. We also describe some new methods developed to collect and analyze process data, and propose several possible extensions to eye tracking systems that would make eye-tracking more useful for STEM education research.
Full-text available
This study examined the representational competence of students as they solved problems dealing with the temperature-pressure relationship for ideal gases. Seven students enrolled in a first-semester general chemistry course and two advanced undergraduate science majors participated in the study. The written work and transcripts from videotaped think-aloud sessions were evaluated with a rubric designed to identify essential features of representational competence, as well as differences in student use of multiple representations. The data showed that both beginning and advanced chemistry students tend to prefer one type of representation. However, advanced students were more likely to use their preferred representations in a heuristic manner to establish meaning for other representations. Students were found to build conceptual understanding most easily when using familiar types of representations. Molecular-level sketches representing dynamic concepts not easily represented as static images, such as an increase in average molecular velocity, were the most difficult type of representation for students to interpret. These results suggest that students may benefit from instructional strategies that emphasize the heuristic use of multiple representations in chemistry problem solving.
Full-text available
Chemistry is regarded as a difficult subject for students. The difficulties may lie in human learning as well as in the intrinsic nature of the subject. Concepts form from our senses by noticing common factors and regularities and by establishing examples and non-examples. This direct concept formation is possible in recognising, for instance, metals or flammable substances, but quite impossible for concepts like ‘element’ or ‘compound’, bonding types, internal crystal structures and family groupings such as alcohols, ketones or carbohydrates. The psychology for the formation of most of chemical concepts is quite different from that of the ‘normal’ world. We have the added complication of operating on and interrelating three levels of thought: the macro and tangible, the sub micro atomic and molecular, and the representational use of symbols and mathematics. It is psychological folly to introduce learners to ideas at all three levels simultaneously. Herein lies the origins of many misconceptions. The trained chemist can keep these three in balance, but not the learner. This paper explores the possibilities, for the curriculum, of a psychological approach in terms of curricular order, the gradual development of concepts, the function of laboratory work and the place of quantitative ideas. Chemical education research has advanced enough to offer pointers to the teacher, the administrator and the publisher of how our subject may be more effectively shared with our students. [Chem. Educ. Res. Pract. Eur.: 2000, 1, 9- 15]
Full-text available
Chemistry is commonly portrayed at three different levels of representation – macroscopic, submicroscopic and symbolic – that combine to enrich the explanations of chemical concepts. In this article, we examine the use of submicroscopic and symbolic representations in chemical explanations and ascertain how they provide meaning. Of specific interest is the development of students' levels of understanding, conceived as instrumental (knowing how) and relational (knowing why) understanding, as a result of regular Grade 11 chemistry lessons using analogical, anthropomorphic, relational, problem‐based, and model‐based explanations. Examples of both teachers' and students' dialogue are used to illustrate how submicroscopic and symbolic representations are manifested in their explanations of observed chemical phenomena. The data in this research indicated that effective learning at a relational level of understanding requires simultaneous use of submicroscopic and symbolic representations in chemical explanations. Representations are used to help the learner learn; however, the research findings showed that students do not always understand the role of the representation that is assumed by the teacher.
Full-text available
In two experiments, we examined how professional chemists (i.e., experts) and undergraduate chemistry students (i.e., novices) respond to a variety of chemistry representations (video segments, graphs, animations, and equations). In the first experiment, we provided subjects with a range of representations and asked them to group them together in any way that made sense to them. Both experts and novices created chemically meaningful groupings. Novices formed smaller groupings and more often used same-media representations. Experts used representations in multiple media to form larger groups. The reasons experts gave for their groupings were judged to be conceptual, while those of novices were judged to be based on surface features. In the second experiment, subjects were asked to transform a range of representations into specified alternative representations (e.g., given an equation and asked to draw a graph). Experts were better than novices in providing equivalent representations, particularly verbal descriptions for any given representation. We discuss the role that surface features of representations play in the understanding of chemistry, and we emphasize the importance of developing representational competence in chemistry students. We draw implications for the role that multiple representations—particularly linguistic ones—should play in chemistry curriculum, instruction, and assessment. © 1997 John Wiley & Sons, Inc. J Res Sci Teach 34: 949–968, 1997.
The hypermedia tools developed for case- and problem-centered learning help foster significant learning outcomes, however, research has focused on the extent, where students are able to transfer conceptual understanding of the particulate nature of matter gained from viewing animations of molecular processes to new situations. A study was conducted in order to investigate students' ability to transfer ideas learned from two video animations of molecular level sodium chloride dissolution process. The result of the study showed that students learn to incorporate some features seen in animations in to own explanations, though encountered difficulty in transferring its understanding to new situations.
In most chemistry courses today, students are introduced to atoms, molecules, ions, and electrons early in the course, and have to accept the teacher’s word that these exist. A better method is to teach chemistry progressively, starting with observations at a macroscopic level, interpreting these at an atomic and molecular level, and then at an electronic and nuclear level. A modern way of doing this is described. [Chem. Educ. Res. Pract. Eur.: 2002, 3, 215-228]
In this article, we examine the role of visuospatial cognition in chemistry learning. We review three related kinds of literature: correlational studies of spatial abilities and chemistry learning, students' conceptual errors and difficulties understanding visual representations, and visualization tools that have been designed to help overcome these limitations. On the basis of our review, we conclude that visuospatial abilities and more general reasoning skills are relevant to chemistry learning, some of students' conceptual errors in chemistry are due to difficulties in operating on the internal and external visuospatial representations, and some visualization tools have been effective in helping students overcome the kinds of conceptual errors that may arise through difficulties in using visuospatial representations. To help students understand chemistry concepts and develop representational skills through supporting their visuospatial thinking, we suggest five principles for designing chemistry visualization tools: (1) providing multiple representations and descriptions, (2) making linked referential connections visible, (3) presenting the dynamic and interactive nature of chemistry, (4) promoting the transformation between 2D and 3D, and (5) reducing cognitive load by making information explicit and integrating information for students.