1 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... DOI: 10.17344/acsi.2024.9049 Review Using Eye-Movements to Explain Processing Triple Level of Chemical Miha Slapničar 1,2,* , Luka Ribič 1 and Iztok Devetak 1 1 University of Ljubljana, Faculty of Education, Kardeljeva ploščad 16, 1000 Ljubljana 2 BEC, Cesta v Mestni log 47, 1000 Ljubljana * Corresponding author: E-mail: miha.slapnicar@pef.uni-lj.si Received: 11-08-2024 Abstract The focus of this systematic literature review was to examine eye-tracking studies in the chemistry classroom and to eval- uate the usefulness of eye-tracking in studying students’ learning of complex chemical concepts. The reviewed studies primarily used small samples of students, with undergraduate students being the most studied population, and measured variables such as eye movements, fixation time, pupil dilation, and others. The results show that higher performing stu- dents have shorter fixation time and learn better with models, with prior knowledge being an important factor. However, both students rely mainly on the symbolic level of presentations with better performing students being more successful at giving explanations. Problems with the complexity of data analysis were reported in the studies. Keywords: Chemical education, eye-tracking, cognitive processes, visualisation, triple nature of chemical concepts 1. Introduction The education sector is one of the priorities of most member states of the European Union. 1 In Slovenia, too, the central national interest is to acquire quality knowl- edge, in which a quality educational process plays a key role. 2 Learning is a lifelong active social process. 3 Compe- tences also include the understanding of scientific (includ- ing chemical) concepts that enable individuals in today's society to develop effective learning strategies and conse- quently to successfully solve authentic scientific (includ- ing chemical) problems. 2 However, content of chemistry education is often abstract for learners in all educational settings and therefore difficult to understand. 4–6 One of the challenges of modern chemistry education is to focus on processes involved in understanding scientific (including chemical) phenomena and their application in solving au- thentic problems. These processes can be studied through reporting of problem solving or by observing their behav- iour during problem solving. Physiological indicators of the cognitive process of individual problem-solving strat- egies include skin conductance, heart rate, blood pressure, brain waves, blood flow to the brain (monitored by fMRA) and others, as well as eye movements, such as saccades, fix - ations and pupil size. These eye movements can be meas- ured with an eye-tracking device. These movements play a key role in the study of visual attention and the processing of information, and thus the learning process. 7 Research results show that there are important links between the cognitive of learning (including chemical learning) and eye movements. 8,9 1. 1. Why is Chemistry Difficult to Learn? The complexity of teaching and learning chemical concepts can be traced back to their triple nature. Chemi- cal concepts can be taught through experimental work or observation of phenomena – i.e. at the macroscopic level. This level represents the actual state of a chemical process that can be observed and perceived with the senses. At the second level, observations can be explained with theories based on atomic, molecular or ionic – i.e. submicroscopic particle levels. Both levels are represented by the symbolic level, for which it is essential to transform the first two lev- els into appropriate symbols with a specific meaning. 5,10–12 All three levels of the chemical concept (macroscopic, sub- microscopic and symbolic) were thus linked in a so-called triangle of the triple nature of a chemical concept. It is well known that students have difficulty distin- guishing between the description and interpretation of macroscopic phenomena at the submicroscopic level. The 2 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... submicroscopic level is the basis for understanding chem- ical concepts that students need to acquire before they can illustrate the concepts symbolically. 5 For students who have not yet developed sufficient abstract thinking skills, such content is difficult to understand and therefore unin- teresting and unpopular. 13 According to constructivist theory, students should initially only be introduced to concepts at the macroscopic level (the concrete level). Later, when they are capable of more abstract thinking, they are expected to move on to submicroscopic interpretations and symbolic representa- tions (abstract level). The model in Figure 1 shows the as- pects of students’ developing mental model as an expand- ing triangle (students’ depth of knowledge at each corner of the triangle increases), and at the same time the iceberg model serves as an analogy for students moving to higher levels of understanding as more of the symbolic and sub- microscopic level can be introduced, the horizontal line moves towards the submicroscopic and symbolic level. 4 From the model of teaching and learning chemistry (Figure 1), it can also be deduced that for a quality un- derstanding of chemical concepts, it is important that the teacher integrates the threefold nature of chemical con- cepts into the lesson by using different teaching materials and learning approaches. 14 As already mentioned, inade- quate prior knowledge on the basis of which the students cannot build increasingly sophisticated and abstract con- cepts plays an important role in the lack of understanding of chemical concepts and the creation of incorrect men- tal models of these concepts. The construction of mental models, including the assimilation of new and challenging concepts, also causes problems at all levels of education. 15 It is argued that in chemistry education in order to prop- erly build mental models of chemical concepts, the triple nature of chemical concepts must be integrated in the classroom using different teaching materials and learning approaches. 5 The gradual transition between the macro- scopic and submicroscopic levels of a given phenomenon could facilitate students understanding of the connections between the macroscopic and the submicroscopic levels, which is then added to the symbolic level. 16 It was shown that transition between the macroscopic and submicro- Figure 1. Adapted teaching and learning chemistry model. 4,14 3 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... scopic levels by presenting chemical concepts at the meso- scopic level helps to reduce the occurrence of misconcep- tions related to the shaping of the mental model of that chemical concept. 17 At the mesoscopic level, different lev- els of magnification are used until the level of molecules, atoms and ions (submicroscopic level) is reached. However, research show that the complexity of the triple nature of chemical concepts for students is not so demanding that it could not be understood. 18–22 1. 2. How Chemistry is Learned – from Processing Chemical Information to Forming Adequate Mental Models More than 80% of information is received and pro- cessed through a visual channel, so the use of different visualization tools is important for the quality of chemical knowledge and the development of problem-solving strat- egies. Abstract illustrations, especially of chemical con- cepts at the submicroscopic level, thus support the overall understanding of the concepts presented to students. 8,10 The models used as explanatory and learning tools to rep- resent an actual object, system, phenomenon or process as realistically as possible must be simple, logical and useful. In addition to the use of pre-prepared models, modelling plays an important role, in the classroom, with students creating their own models based on some known laws. 23 This improves students critical thinking and ability to solve chemical problems, which leads to an increase in learning success and situational interest in chemistry les- sons. 24 However, in linking the symbolic and submicro- scopic level to the macroscopic level of phenomena, the use of visualisation tools can also pose certain dangers and shortcomings due to the way they represent abstract con- cepts. To avoid further misconceptions of chemical con- cepts, the teacher should inform students of these dangers when using such tool. 25 The creation of an appropriate mental model de- pends to a large extent on interest in the learning con- tent. 26,27 The learning content should be designed to be contemporary and relevant to learning, to increase interest in chemistry and to promote higher order cognitive skills and research skills (e.g. asking research questions, plan- ning research design, selecting appropriate instruments to measure phenomena, knowing how to collect data, under- standing the interpretation of results, drawing conclusions and answering research questions). 28 Cognitive processes cannot be observed directly during learning, but only indirectly, through students’ answers (in achievement tests), observation of behaviour during learning (e.g. how quickly questions are answered, in what order tasks are solved) and other methods of self-reporting about the learning process. Due to the indi- rect measurement of cognitive processes and to methodo- logical limitations of the procedures for collecting data on cognitive processes during learning, questions often arise about the validity of the data obtained. Therefore, differ- ent research methods are needed to represent learning and problem-solving processes from different perspectives. 29 One of the techniques that has recently been used in the chemistry didactics research is the use of eye-tracking. There are important links between the cognitive process and eye movements. 7,30–32 The use of eye-tracking allows us to indirectly monitor the basic mechanisms of informa- tion encoding and integration. 7 Eye-tracking technology has been used in chemistry education research to investi- gate students’ understanding of the phenomena while they are working on a task as it provides information of stu- dents’ cognitive processes while working on activity. 33,34 The main question was how eye movements can be related to students’ understanding of chemistry. The eye tracker measures the number of fixations in the region of inter- est (i.e. periods of relatively stable gaze on the observed object during which visual information is processed), the total duration of fixations in the region of interest, number of visits on a particular source, which is useful in prob- lem-solving, and also some other measures of eye move- ments (such as pupil size). A typical fixation lasts 200–300 milliseconds. Successive fixations are separated by sacca- des (i.e., the rapid movement of the eyes to the centre of interest that provides new visual information). 35 A higher number of fixations and their longer duration on a particu- lar area of interest indicate which area in the visual field was of most interest to the person and where attention was directed. Eye-tracking system also allows measurements of pupil dilation which is related to person’s cognitive load. Eye-tracker measurements provide objective data about a person’s covert attention and offer additional support in the measuring the person’s cognitive activity and in stud- ying the process of problem solving with visual stimuli. 8,9 Previous studies on eye-tracking have shown that unsuccessful problem solvers have problem distinguishing between relevant and irrelevant factors also, they have a problem focusing on relevant factors. 36 2. The Purpose of this Research The aim of this study was to examine the findings from eye-tracking studies on cognitive processes during chemistry learning and solving chemistry problems. A review paper that examined papers which were published by recognized scientific journals from 2000 to 2012 and which employed eye-tracking method in studying learning in a general context were already published. Seven areas were examined in this study: patterns of information pro- cessing, effects of instructional design, review of existing theories, individual differences, effects on learning strate- gies, patterns of decision making, and conceptual develop- ment. 37 Another review focusing on the use of eye-track- ing method in chemistry education research analysed six studies (two of which were unpublished dissertations by 4 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... the authors). Authors examined various aspects of this method in chemistry education research, such as research questions, study type, sample size, protocol, and stimuli. 32 A more detailed analysis of published research in the field of problem solving in chemistry is needed, covering the complexity of chemical concepts, their influence on the processing of chemical information and learning chemis- try, and different methods for investigating cognitive pro- cesses in the performance of tasks in chemistry. Therefore, this review examines four themes in the analysed papers: (1) participants, (2) supported methods, (3) chemical con- tent, and (4) basic outcomes applicable in a school context. In addition, the reference lists of analysed papers were ex- amined to identify further studies that could be consid- ered for analysis. The results could help determine what aspects teachers should consider when teaching more ab- stract concepts at all levels of chemistry education. 3. Method The focus of this paper was to review eye-tracking studies in the chemistry education and evaluate the utility of eye-tracking in examining how complex chemical con- cepts (e.g. triple nature of chemical concepts) are learned by students. In conducting the systematic review of the published literature, the Preferred Reporting Items for Systematic Reviews (PRISMA) approach was followed in accordance with the definition of a systematic review. 38,39 Evidence from multiple studies were combined by identifying rele- vant research and assessing the quality of the studies. 3. 1. Search Strategy Papers from two databases Web of Science (WoS) and Scopus were identified using Boolean operators (Table 1) created for each database to identify published papers. To ensure that the references are current, and the results of this study are up to date, papers from 2010 to 2024 were included in this systematic review. 3. 2. Study Selection The inclusion criteria (IC) were established for the inclusion of a particular paper identified in both da- tabases in this review. These criteria are: (1) the paper was published between the year 2010 and 2024; (2) the paper is written in English; (3) it is not review study or meta-analysis; (4) the paper has been published in peer reviewed journal; (5) the study is not listed in another da- tabase; (6) the full text paper is available; (7) the study used eye-tracking technology for data collection; (8) the study was conducted with primary, secondary or univer- sity students and (9) the study was conducted in chemis- try education. Papers were selected in four phases. In each phase papers were excluded from the review based on the exclu- sion criteria. Forty-one papers were identified that explored con- cepts for science education using eye-tracking technol- ogy. The main reason for excluding the papers from this systematic review was that they were duplicates and did not meet the inclusion criteria IC7, IC8 and IC9. A total of twenty-two papers were included in this systematic re- view. In the analysed period, most articles (3) in the field of chemical education were published in 2015, 2019, 2020 and 2024. Ten papers were published in the Journal of Chemical Education, followed by the journal Computers in Human Behaviour with two papers. On average, the ana- lysed papers referred to about 60 units of literature. Papers included in this review were cited on average 25 times in WoS and 18 times in Scopus by the end of 2024, this shows that that using eye tracking technology in chemical educa- tion is trending internationally. 4. Results and Discussion 4. 1. Characteristics of the Analysed Studies The research problem, the number of participants, the eye-tracker model and the data collection technique were analysed in the articles examined (Table 2). It can be seen from Table 2 that all the papers an- alysed investigated similar problems, related to different types of visualisation materials that students can use when learning certain chemical concepts or solving problems. Most of the papers focused on students’ fixation time, gaze and eye movements when viewing static/dynamic 2D/3D representations or simulations. The attention was also measured when using stick and ball models and when making connections between different representations. The different fixation times, gaze and eye movements of the students on the different models were also compared Table 1. Search strings. Database Search string Web of Science (TI=(eye-tracking) OR AB=(eye-tracking) OR AK=(eye-tracking) OR TI=(eye AND movements) OR AB=(eye AND movements) OR AK=(eye AND movements)) AND (TI=(chemical AND education) OR AB=(chemical AND education) OR AK=(chemical AND education)) Scopus (TITLE(eye-tracking) OR ABS(eye-tracking) OR KEY(eye-tracking) OR TITLE(eye AND movements) OR ABS(eye AND movements) OR KEY(eye AND movements)) AND (TITLE(chemical AND education) OR ABS(chemical AND education) OR KEY(chemical AND education)) 5 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... Authors (year) Research problem Participants Eye-tracker Data collection technique; instruments Stieff et al. (2011) 40 What are the students, abilities to coordinate representations using multi-representational displays (i.e. a molecular model, a general equation, a numerical equation and a graph). What are the complementary contributions of eye fixation data. 10 undergradu- ate chemistry students SMI EyeLink Eye-tracking system; interview, videotape-audio- tape Tang and Pienta (2012) 41 How students use online tools for prob- lem-solving. Effect of problem difficulty and cognitive processes on solving gas law word problems (i.e. the differences in problem-solving time and number of fixations between two groups of students). 12 high-school students Tobii T120 Eye-tracking system; questionnaire Williamson et al. (2013) 42 How students use ball-and-stick images versus electrostatic potential maps when asked questions about electron density, positive charge, proton attack, and hydroxide attack with six different organic molecules. 9 undergraduate chemistry students SMI EyeLink Eye-tracking system; interview Hinze et al. (2013) 43 How students use ball-and-stick images versus electron potential maps to solve specific questions about electronegativity, molecule polarity, and identification of oxidations states. 30 undergradu- ate chemistry students Tobii T60 Eye-tracking system; tests (pre-test basic chemical knowledge test; cognitive ability test; five spatial ability tests), interview, audiotape Tang et al. (2014) 44 What are students' chemistry problem-solving abilities. Which cognitive activities students use when solving gas law stoichiometry word problems. High-school students; number not reported. Tobii T120 Eye-tracking system; interview; O’Keefe et al. (2014) 45 How learners construct knowledge when learning environments that include multiple representations of ideal gas laws, as well as transition between them, are applied using a computer-based model. 26 high-school students SMI RED250 Eye-tracking system; tests, interview Chen et al. (2015) 46 What is the effectiveness of mental (spatial) abilities to construct mental models of atomic orbitals after learning with 3D static or dynamic representations. Can eye movements predict students' achievements on constructing the mental models of these concepts. 60 high-school students EyeLink 1000 Eye-tracking system; pre-post multiple-choice test, mental model construction test; spatial ability test Cullipher and Sevian (2015) 47 How students’ eye gaze patterns correlate with their implicit assumptions about structure property relationships using IR spectroscopy. 25 undergradu- ate and graduate chemistry students Tobii X2-60 Eye-tracking system; interview, audiotape Rau et al. (2015) 48 How students’ visual attention behaviours relate to their sense-making abilities and their reasoning about chemistry? Furthermore, identify specific concepts that Chem Tutor should target because helping students make connections between representations with respect to these concepts might lead to significant learning gains. 26 undergradu- ate and graduate chemistry students SMI RED250 Eye-tracking system; online test Topczewski et al. (2016) 49 How do students in organic courses use data to solve NMR spectral problems and how do these students develop effective strategies. 27 undergradu- ate and graduate chemistry students Tobii T120 Eye-tracking system; test Herrington et al. (2017) 50 How students understand solution process for ionic and covalent compounds? How and where do students allocate attention while interacting with a simulation, as compared to a screencast, when coupled with a guided assignment? 12 undergradu- ate students Tobii T60 Eye-tracking system; pre- and post-knowledge tests Table 2. Research problems and methods used in the analysed papers. 6 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... Authors (year) Research problem Participants Eye-tracker Data collection technique; instruments Karch et al. (2019) 51 Can gaze data and pupil measures be used to identify students’ solving strategies? How different tasks influence students’ level of cognitive load? 22 students before the beginning of first semester general chemistry Tobii X2-60 Eye-tracking system; chemical concepts invento- ry, retrospective interviews Connor et al. (2019) 52 To identify invalid assumptions of students when interpreting the results of IR and 1 H NMR synthesis. What invalid assumptions severely constrain organic chemistry students’ reasoning? 18 undergradu- ate organic chemistry students SMI RED Eye-tracking system; semi-structured retrospec- tive think-aloud interviews Baluyut and Holme (2019) 53 How students obtained and used information from particulate nature of matter diagrams as well as balanced chemical equations when asked questions about stoichiometry concepts such as limiting and excess reagents, and yield. 29 chemistry students of a two-semester general chemis- try sequence SMI RED Eye-tracking system; online test Nehring and Busch (2020) 54 To find out, whether students tend to observe demonstrations from left-to-right manner and to test if there is any difference in setting up the demonstration from left to right or not. 146 students, on lower and upper secondary school Eye gaze Edge Eye-Follower 2.0 with NYAN Eye-tracking system; tests Rodemer et al. (2020) 55 To find out to what extent do beginner and advanced students differ in in their visual processing behaviour in case-comparison tasks and to what extent does task characteristics influence beginner and advanced students’ visual processing behaviour. 37 undergradu- ate chemistry students Tobii Pro X3-120 Eye tracking system; tests Slapničar et al. (2020) 56 To determine if the justifications of successful and unsuccessful students differ according to their level of education and their age. Also, to check if successful students fix their gaze on correct 3D presentation for longer time. 30 seventh grade primary school students, 29 first year upper secondary school students and 20 students of the double-majors study pro- gramme of chemistry and biology/physics Eye- Link1000 Eye-tracking system; tests, think-aloud interviews VandenPlas et al. (2021) 57 How learning with PhET simulation effect students’ conceptual understanding, can screencast be most effective method for students’ learning outside of the class, and where do students allocate attention when using screencast or simulation. 302 undergradu- ate students overall, 16 students using eye-tracker Tobii T60 Eye-tracking system; pre-test and follow-up test Tóthová et al. (2021) 58 To examine students’ problem-solving skills by examining their visual attention (fixation duration time). Are students capable of using periodic table of elements (do they use it as an inductive tool, are they able to orient in the periodic table of the elements, what factors affect their use of the table). What strategies do students use when solving the tasks and what problems they face during their activity? 139 non-chemis- try vocational school students overall, 8 of them were selected to participate in eye-tracker study Gaze Point eye camera GP3 Eye-tracking system; think aloud interviews Sun et al. (2024) 59 The research problem is related to the questions of whether it can a microcomputer-based laboratory effectively promote the chemical triple representation of a preservice chemistry teacher and what cognitive mecha- nisms are demonstrated by students at different academic achievements. 35 second-year preservice chemistry teachers Eye-Link 1000 Plus Eye-tracking system; video experiment, knowledge test, semi structured interview 7 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... with the different learning outcomes, e. g. between low and high achieving students. In addition, the students’ learning strategies were determined based on their eye movements. In most (18) of the examined studies, the sample was relatively small, with the exception of four studies in which the sample comprised more than 100 participants. On av- erage, 43 participants were exposed to the eye-tracking procedure. In one study, there were only 9 students 42 and the highest number of participants using an eye-tracker device was included in the study by Nehring and Busch. 54 In one of the analysed papers no information was provided on the number of participants. It can therefore be conclud- ed that the number of participants in eye-tracking studies in chemistry education is generally low, and the results of these studies should therefore be interpreted with caution. As far as the educational level of the participants in the studies examined is concerned, university students were represented in most of the studies. Only five papers includ- ed high school students, and three papers included grad- uate students (Masters and PhD students) in the sample. Primary and secondary school students were least likely to be included in the research sample in the papers examined, with two studies including secondary school students and one study including primary school students in their re- search sample. Two studies were conducted with pre-ser- vice chemistry teachers. It can be therefore concluded that further eye-tracking studies should be conducted with sec- ondary school students and pre-service chemistry teachers to gain a more comprehensive understanding of the effects of using visual elements in learning chemical concepts at all educational levels. It is also important to obtain infor- mation about how future chemistry teachers understand chemical concepts and how this understanding develops as teachers’ knowledge influences their future teaching. Regarding to the instruments used in addition to eye-tracking technology, the results of the analysis showed that ten studies used interviews with the participants. Twelve of the studies analysed also used achievement tests to determine the participants’ level of knowledge of the specific chemical concepts used in the study. In addition, the majority of the papers analysed used different tests in a pre- and post-test research design to differentiate between pre- and post-activity knowledge levels. In three analysed papers, audio or video recordings of the participants solv- ing tasks the with the eye-tracking device were also used. In addition, a specific mental ability test, a spatial abili- ty test and a cognitive load questionnaire were used to measure some independent variables that were part of the research problem. The data suggest that researching an in- dividual’s cognitive activities requires the inclusion of ad- ditional instruments, such as interviews, questionnaires, mental ability tests, spatial ability tests, knowledge tests, in addition to the use of eye-tracking technology, to obtain a more comprehensive picture of the processes occurring during the chemical activities under investigation. 4. 2. Major Conclusions from Studies Using Eye Tracking in the Field of Chemistry Education Table 3 summarises the most important results of the studies examined. The most important results are printed in bold. As shown in Table 3, eye-tracking was used in most studies to provide visual feedback to students when they used models or different levels of representation when learning certain chemical content. Results show that stu- dents were better able to answer the questions with mod- els than with diagrams. However, the mathematical equa- tions posed the greatest problems for the students when they tried to solve the problem. 40 These results are con- sistent with the findings that students do best with models when solving chemistry problems. 41 In addition, students spend less time solving the problem when using models. Therefore, teachers should use models in classroom learn- ing as they allow students to form appropriate mental models. 24 Students’ prior knowledge and educational level are identified as important factors influencing the effec- tiveness of teaching with models and presentations, which indicates that the use of models in the classroom learning needs consideration and needs to be appropriate to the prior knowledge of the students in the group. Students with a higher level of knowledge were more successful in using representations for learning and had a longer fixa- Authors (year) Research problem Participants Eye-tracker Data collection technique; instruments Hamerská et al. (2024) 60 The pedagogical challenges in teaching chemical equations, focusing on the ability to balance chemical equations and understanding related basic chemical concepts among chemistry teacher students. 24 chemistry teacher students Tobii Fusion Pro Eye-tracking system; retrospective think-aloud interviews Huangfu et al. (2024) 61 How displayed teacher enthusiasm in video lectures affect students’ positive emotions, visual attention, cognitive load, and learning outcomes? 128 eighth-grade middle school students Eye Link 1000 Plus Eye-tracking system; experimental videos, self-report questionnaire, post-test; PAS question- naire; cognitive load questionnaire 8 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... Study The main conclusions Stieff et al. (2011) 40 – Students rely primarily on visual-spatial representations in the display and do not make use of accompany- ing mathematical representations. – Results support the complementary validity of verbal protocols and eye-tracking data; both data sets were highly correlated; both data sets provide insight into students’ problem-solving with multiple representa- tions, supporting the validity of both data sets for measuring students’ use of representations. – Students did not rely primarily on the visual representation of the molecules, which were rendered with bolder colours and included more striking animations of multiple atoms; they placed equal importance on the model and the diagram and switched frequently transitions between these two representations. – Participants performed well on questions that could be answered using the model and diagram and performed poorly on questions that required integration of information from the model, diagram, and mathematical equation. Tang and Pienta (2012) 41 – The unsuccessful students spent more time thinking about their solutions and focusing more on the questions than they did trying to solve the gas law problem. – The gas law problem represented a greater cognitive load for the unsuccessful students. – Both successful and unsuccessful participants spent the same amount of time on the initial reading, but the unsuccessful students spent more time on the problem planning phase. – Students with varying degrees of success in solving the gas law problem showed different eye movement patterns that corresponded with their cognitive effort. Williamson et al. (2013) 42 – Students spent significantly more time with the ball-and-stick image than with the potential map when asked questions about proton attack or hydroxide attack, but same amount of time with both images when asked about electron density or positive charge. – Students who spent more time with the ball-and-stick image when asked about positive charge were less likely to be correct, while those who spent more time with the potential map were more likely to be correct. Hinze et al. (2013) 43 – Eye-tracking and explanation measurements showed that attention to and explicit use of EPMs (electron potential maps) was associated with higher accuracy. – When answering questions that required identifying the format and features of a representation, participants were more likely to adopt the novel visualizations than when answering questions that required drawing inferences. – Students’ adoption or avoidance of EPMs was caused, in part by the difficulty of understanding unfamiliar visualizations, although their designs was explicitly specifically intended to promote learning and was often praised by the designers. – Prior knowledge facilitated successful use of EPMs, as reflected in accuracy and eye-tracking data, but not in explanations. – Experts tended process external representations more efficiently than did novices. – Even when the features of a representation were clear, expertise played a necessary supporting role in understanding the relevance of these features to the domain of interest. – The development of representational competence during early learning and comprehension experiences depended not only on practice with visualizations, but also on task and learner characteristics. – Interestingly, participants with greater prior knowledge began to use EPMs effectively in inference problems with some practice; participants with little prior knowledge, on the other hand, stuck with their ball-and- stick representations and were more likely to answer questions incorrectly. Tang et al. (2014) 44 – Students were able to find and interpret relevant information needed to solve the tasks (the reading phase in the eye-tracking experiment), but their working memory load was increased by the subsequent procedures (post-reading phases). – Almost all students applied or attempted to apply dimensional analysis (step-by-step procedure) to solve the stoichiometry problem. O’Keefe et al. (2014) 45 – Students can integrate multiple representations through sequential fixations over related elements of a simulation, and transitions between different simulation elements are associated with different learning outcomes. Chen et al. (2015) 46 – Learning with dynamic 3D representations leads to a better understanding of atomic orbitals. – The construction of correct mental models is positively predicted by the number of saccades and rereading and negatively predicted by the number of fixations. – Students spent similar amounts of time on 3D representations. Table 3. The main conclusions of the included studies. 9 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... Study The main conclusions Cullipher and Sevian (2015) 47 – Analysing fixation sequences during eye-tracking can provide useful information about what students are thinking when they associate molecular structures with spectroscopic reactions, and that this information goes beyond what can be learned from thinking aloud. – Viewing patterns differed depending on students’ implicit assumptions about how spectroscopic reactions relate to molecular structures. – The initial results provide partial validation of a newly developed methodology for analysing eye tracking data to uncover reasoning patterns that appear to correspond to identifiable underlying assumptions. Rau et al. (2015) 48 – Sensory perception ability: Students have a higher sense-making ability than fluency. It can be assumed that the chemical instructions do not provide sufficient support for making the connection. – Sense-making ability: difference-connections were associated with significantly more inferences. However, there were no correlations between surface-connections and inferences, or between similarities-connections and misconceptions or between differences-connections and misconceptions. The Frequency of switching was significantly related to surface connections. However, no significant correlations were found between eye-tracking and similarity connections. There were also no correlations between eye-tracking and misconceptions. – Difficult concepts: Students with a university degree mentioned colour coding, bond angles, atomic radii, etc. more frequently than students without a university degree. Graduate students were more likely to use these concepts to describe the behaviour of electrons, atoms, bods, etc. Undergraduate students were more likely to describe the differences between the concepts as shown in the animations without going beyond what the presentation shows. Graduate students were more likely to present complementary features of the representation. Topczewski et al. (2016) 49 – Two groups: Statistically significant differences were found between experts and novices in the correctness of their responses to the NMR tasks. Experts had a much shorter fixation time than novices. – Three groups: Statistically significant differences were also found between advanced and early beginners and experts in answering the NMR tasks. In addition, statistically significant differences in fixation time were found between all three groups. Experts have a longer fixation time for the last three items. Herrington et al. (2017) 50 – Student responses: Most students were able to recognise the differences between sugar and NaCl dissolving in water. However, further analysis shows that students who used the screencast were better able to identify electrolytes and recognise how water interacts with covalent compounds. The pre- and post-test analysis revealed statistically significant differences between the groups of students. The students who used screen - cast were able to significantly improve their understanding. – Eye-tracking: Significant differences were found between the students in the screencast and simulation groups. The number and duration of fixations were significantly different in the two groups. In addition, the screencast group focused significantly more on the electronic resource. Karch et al. (2019) 51 – Pupil dilatation and gaze: In complex multiple-choice tasks, pupil dilation fluctuated while solving the task. In a simpler true/false task, there was a clear difference in pupil signal between participants who answered the question correctly and those who answered it incorrectly. Those who answered the question correctly had linearly increasing pupillary signals, while those who answered the question incorrectly had pupillary signals more like those observed in the multiple-choice tasks. Connor et al. (2019) 52 – Invalid assumptions: 20 invalid assumptions were identified as part of the research. In addition, eight heuristic reasoning strategies were identified (1) basic associative processes, (2) inductive judgments and (3) affective judgments. Baluyut and Holme (2019) 53 – Statistically significant differences were found between the visual behaviours of high-achieving and low-achieving students. – High-achieving students spent more time examining diagrams of the nature of matter and switched between parts of the diagrams and other areas of interest more frequently than low-achieving students. – The study is an example of how the underlying strategies students use when answering conceptual stoichi- ometry questions can be triangulated using quantitative and qualitative techniques related to visual eye-tracking behaviour. Nehring and Busch (2020) 54 – Student observations: Students who look at an object more often tend to look further at the object on the left side. The left-to-right setting reduces right-to-left eye movements, but the right-to-left setting does not reduce left-to-right eye movements. The effects found were small. Thus, the results indicate that setting the demonstration from left to right may reduce eye movements and support the reaction flow. 10 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... tion time on these models. 43,45,48,53 This was also the case in the study where graduate students paid much more at- tention to the models than undergraduate students. They are also better able to use these models when explaining certain chemical concepts than undergraduates and also understand the background of the presentation. 48 More Study The main conclusions Rodemer et al. (2020) 55 – Test results: No statistically significant differences were found between advanced students and beginners in their test scores at the baseline. This could be due to the small sample size. However, the effect size was .32, – Visual processing behaviour: Advanced students require less time and fewer fixations for each task than beginners. Both groups of students did not differ in their attention distribution. – Task characteristics: Task duration and number of fixations are significantly influenced by the task. No significant effect was found for advanced students. Advanced students also work more homogeneously through the task than beginners. Beginners appear to be more strongly influenced by the task characteristics. Overall, the tasks appear to have a significant influence on visual behaviour. It was also shown that viewing behaviour is much more influenced by the amount of information a presentation contains than by the type of information. Slapničar et al. (2020) 56 – Justification of correct presentation: Students’ successful justification of correct 3D presentation increased with educational level. Students have problems describing the submicroscopic level with the macroscopic level. Primary school students have problems with understanding 3D presentations and with the transition between them. The number of students who use macroscopic presentation to describe different states of water decreases with age. – Differences between successful and unsuccessful students in terms of total fixation duration (TFD) and area of interest (AOI): It can be concluded that successful and unsuccessful students of the same age group look at 3D presentations for a similar amount of time. VandenPlas et al. (2021) 57 – Learning gains: In all participant groups (simulation, screencast and extended screencast), statistically signif- icant differences were found between the results of the pre-test and post-test. Furthermore, no statistically significant differences were found between the students who used using simulation and screencast. it can therefore be assumed that the simulation did not significantly reduce the students’ cognitive load. The extended screencast proved to be the best presentation to improve students’ knowledge and understanding. – Eye-tracking: Students using the simulation spend more time on the task than students using the screencast. This suggests that student do not need as much time to decipher the meaning of the simulation. However, students using screencasts spend less time reading and responding to the tasks, proving that screencasts can reduce students' cognitive load. Tóthová et al. (2021) 58 – Use of periodic table: The students tried to use the periodic table of the elements to solve the task. The same applies to high-achieving and low-achieving students. There are statistically significant differences between the students’ fixation on the table and on the text. However, success in solving the task was not related to the use of the periodic table, but to several other factors. It was also found that students spent less time on the legend, which could indicate that students did not understand the reason for the table. – Factors influencing students’ engagement with the periodic table: Understanding the task proved to be the main reason why students could not successfully complete the task. Their motivation, unfamiliarity with the content and unfamiliarity with the basic concepts also proved to be important factors in solving the task. – Students’ strategies for solving the tasks: The main problem in unsuccessfully solving the task was not the strategies they used but the problems they faced. In addition, the students used extensive strategies when solving the tasks. Sun et al. (2024) 59 – The comparison of the results of the triple representations test and the eye-tracking indicators of the two groups shows that a microcomputer-assisted laboratory has a positive effect on triple representation. – Students in the experimental group showed significantly better performance in the recording of image information under the influence of the microcomputer-assisted laboratory. Hamerská et al. (2024) 60 – The students could be divided into four different groups based on their total score and solution time: efficient, persistent, impulsive, and inefficient solvers. – The results of an eye-tracking study show that students predominantly rely on symbolic representations, with additional sub-microscopic representations provided by digital applets not significantly improving equation solving skills. This reliance potentially hinders students’ ability to conceptualize reactions at the particle level, impacting their overall understanding of chemistry. – The study highlights the need for a pedagogical shift toward integrating multiple representations into chemistry education to bridge the gap between procedural skills and deep conceptual understanding. Huangfu et al (2024) 61 – The results show that an enthusiastic teacher positively influences students’ positive emotions, reduces students’ cognitive load and makes students focus more on the content learning. – More enthusiasm shown by the teacher improved the learning outcomes of the students. 11 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... successful students are also able to focus on different parts of the diagram when analysing them, than less success- ful students and are therefore able to complete the task successfully. 53 These results are also consistent with the findings that higher performing students were more like- ly to be able to choose the correct representation than lower performing students. 56 Low-performing students also spend more time on questions than on models. 41 There were no statistically significant differences between high-performing and low-performing students in terms of their fixation time on a particular model. 56 These re- sults contradict to the findings, that there were statisti- cally significant differences between high-performing and low-performing students in terms of fixation time, with low-performing students spending more fixation time on a particular model than high-performing students. 55,57,51 Successful students also spend less time fixating on the symbolic level of presentation. 58 However, successful and unsuccessful students rely on this level of representa- tion. 60 Results in the analysed papers support the idea, that all levels of representation need to be implemented in the classroom learning alongside with different learning approaches. 5 When using different models, students spend an equal amount of time focusing on all models and that us- ing these models improves their knowledge. 46 However, results also show that fixation duration does not predict the correctness of the mental models that students formed when using static and dynamic 3D presentations, 46 but us- ing of specific models can show how students think. 47 In addition, the correctness of students’ mental model can be predicted by the number of cascades when working with physical models of chemical structures 46 and the use of a computer-based laboratory also improves students’ under- standing of the triple nature of chemical concepts. 59 In some papers, the authors examined students’ eye movements in relation to their level of cognitive load. In these studies, more eye movements and thus shorter fix- ation time were found in low-performing students, which is likely due to the higher cognitive load resulting from a more difficult task for the low-performing students. 40,44 Cognitive load can also be correlated with students’ pupil dilation during solving specific tasks and research showed that it varied in multiple-choice tasks. In tasks where stu- dents had to decide whether the statement was correct or incorrect pupil size changed linearly when students an- swered correctly. 51 The analysis of 22 papers published in respected peer review journals on the topic of using eye tracking technol- ogies in chemical education has shown that eye-tracking can provide important insights into students’ use of visual models in chemistry. Students’ fixation patterns vary with their performance level, cognitive load and task complex- ity. Eye-tracking metrics, such as fixation time and pupil dilation can provide clues to cognitive processes and un- derstanding in chemistry learning. 5. Conclusion The focus of this paper was to review eye-tracking studies in chemistry education and evaluate the usefulness of eye-tracking methodology in investigating how com- plex chemical concepts are learned by students at different educational levels. The results could help determine which aspects teachers should consider when teaching abstract chemistry triplet concepts at all levels of chemistry edu- cation. Among the 22 analysed studies on chemistry edu- cation published between 2010 and 2024 in which the eye-tracking method was used, most students were rep- resented as study participants. The least included student population in the analysed studies were lower and upper secondary school students. Future studies should there- fore include samples at the lower levels of chemistry edu- cation so that the cognitive processes involved in solving chemical problems with younger students can be investi- gated. Furthermore, only two studies were conducted with pre-service chemistry teachers, so it could be recommend- ed that future studies be conducted with pre-service chem- istry teachers, as it is important to know how these stu- dents learn chemical concepts. This is especially important because pre-service chemistry teachers’ educators should know how to design courses to develop appropriate men- tal models of chemistry concepts for the students who will be responsible for teaching these concepts in their future careers. This study examined papers that were mainly con- cerned with students’ eye movements when using various visualisation methods in chemistry lessons. The studies measured the fixation time, eye movements, pupil dilation, gaze, knowledge retention, knowledge gain, attention, etc. of the participants. In most cases, low-performing and high-performing students were compared with each other based on these measured values. An eye tracker was gen- erally used to collect data to obtain information about eye movements, gaze, fixation times, etc. Knowledge tests in the form of pre- and post-tests and follow-up tests were used to obtain information about the students’ learning gains or retention of knowledge. Other instruments such as cognitive load questionnaires, mental ability tests, spa- tial ability tests etc. were rarely used in the studies exam- ined. Small sample sizes are not uncommon in studies that provide detailed analyses of verbal explanations during problem solving or eye movements due to the complexity of data collection and analysis. The complexity of data col- lection is also related to the triangulation of instruments used to collect data on students’ learning processes and the large number of different types of data collected with the eye tracker. The complexity of data analysis is related to the difficulty of determining which feature of the re- corded signals to extract and analyse and how to correlate the eye movement data with other measured variables. It is important to be aware that the conclusions drawn from eye-tracking studies are usually subject to certain limita- 12 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... tions due to the small sample sizes and the complexity of the data analysis. As mentioned above, eye-tracking metrics such as fixation time and pupil dilation can provide clues to cog- nitive processes and comprehension during chemistry learning. The reviewed papers concluded that teaching with models can improve chemistry learning and lead to better learning outcomes. However, prior knowledge is an important factor in students’ ability to learn using this method. Higher performing students can use models with ease and give explanations based on these models. They can also focus on different parts of the diagram, spend less time fixated on certain parts of the models and can distin- guish between good and bad representations. In addition, low- and high-achieving students rely on the symbolic level of the presentations. However, fixation time cannot predict the correctness of students’ mental models, but it can provide information about how students think. Stu- dents’ eye movements can also provide information about the cognitive load they are exposed to. A higher number of eye movements can predict higher cognitive load. There- fore, more eye movements are detected in low-achieving students. The possibilities for studying chemistry learning using eye-tracking methods in the future are enormous. Eye-tracking data can be considered more objective and valid compared to other methods of studying chemistry learning, such as self-reporting. The use of eye-tracking data as additional information about the learning of chem- ical concepts could potentially lead to many new insights and the development of more effective educational strate- gies that could be applied in the school environment. Acknowledgements The authors acknowledge the project »Augmented reality to achieve better understanding of the triple nature of chemical concepts« (grant No. J5-50155) that was finan- cially supported by the Slovenian Research and Innovation Agency (ARIS). 6. References 1. F. Buchberger, B. P. Campos, D. Kallos, J. Stephenson, 2001, Zelena knjiga o izobraževanju učiteljev [The green book on teacher education]. Ljubljana: Ministrstvo za šolstvo, znanost in šport. 2. European Reference Framework, 2007 , Retrieved from http:// www.ec.europa.eu/dgs/education_culture/publ/pdf/lllearn- ing/keycomp_sl.pdf 3. B. Marentič-Požarnik (2000) Psihologija učenja in pouka [Psychology of learning and teaching]. Ljubljana: DZS 4. G. Chittleborough, in: I. Devetak, S. A. Glažar (Eds.), The development of theoretical frameworks for understanding the learning of chemistry: Learning with understanding in the Chemistry Classroom, Dordrech, Netherland: Springer, 2014, pp. 25−40. DOI:10.1007/978-94-007-4366-3_2 5. I. Devetak, 2012, Zagotavljanje kakovostnega znanja nara- voslovja s pomočjo submikroreprezentacij. Analiza ključnih dejavnikov zagotavljanja kakovosti znanja v vzgojno–izo- braževalnem sistemu [Providing quality science knowledge with the help of submicrorepresentations. Analysis of the key factors of ensuring the quality of knowledge in the educa- tional system], Ljubljana, University of Ljubljana, Faculty of Education. 6. H. K. Wu, J. S. Krajcik, E. Soloway, J. Res. Sci. Teach. 2001, 38, 821‒842. DOI:10.1002/tea.1033 7. Y . F. Y ang, C. Y . Chang, W . R. Chien, Y . T. Chien, Y . H. Tseng, Comp. Educ. 2013, 62, 208–220. DOI:10.1016/j.compedu.2012.10.009 8. H. H. Chuang, H. C. Liu, J. Sci. Educ. T echnol. 2011, 30, 239–261. 9. T. Meng-Jung, H. Huei-Tse, L. Meng-Lung, L. Wan-Yi, Y. Fang-Ying, Comp. Educ. 2011, 58, 375−385. 10. A. H. Johnstone, Sch. Sci. Rev. 1982, 64, 377–379. 11. A. H. Johnstone, J. Comput. Assist. Learn. 1991, 7, 75–83. DOI:10.1111/j.1365-2729.1991.tb00230.x 12. S. K. Taber, Chem. Educ. Res. Pract. 2013, 14, 156−168. DOI:10.1039/C3RP00012E 13. R. B. Thiele, D. F . T reagust, J. Res. Sci. T each. 1994, 31, 227‒242. DOI:10.1002/tea.3660310304 14. I. Devetak, S. A. Glažar, in: I. Devetak, S. A. Glažar (Eds.), Ed - ucational models and differences between groups of 16-year- old students in gender, motivation, and achievements in chemistry. In Learning with understanding in the Chemistry Classroom, Dordrech, Netherland, 2014, pp. 103−126. DOI:10.1007/978-94-007-4366-3_6 15. D. F. Treagust, A. G. Harrison, G. J. Venville, J. Sci. Teach. Educ. 1998, 9, 85−101. DOI:10.1023/A:1009423030880 16. M. R. Meijer, A. M. W . Bulte, A. Pilot, in: G. Tsaparlis, H. Se - vian (Eds.), Macro–micro thinking with structure–property relations: Integrating 'meso-levels' in secondary education. In Concepts of Matter in Science Education, Dordrecht, Nether- land, 2013, pp. 419−436. DOI:10.1007/978-94-007-5914-5_20 17. M. R. Meijer, A. A. M. Bulte, A. Pilot, in: J. K. Gilbert, D. Treagust (Eds.), Structure–property relations between macro and sub-micro representations: Relevant meso-levels in au- thentic tasks. In: Multiple representations in chemical educa- tion, models and modeling in science education, Dordrecht, Netherland, 2009, pp. 195–213 Springer. DOI:10.1007/978-1-4020-8872-8_10 18. D. M. Bunce, D. Gabel, J. Res. Sci. Teach. 2002, 39, 911−972. DOI:10.1002/tea.10056 19. G. Chittleborough, D. F. Treagust, M. Mocerino, 2002, Con- straints to the development of first year university students’ mental models of chemical phenomena. Teaching and Learn- ing Forum, Focusing on the Student. Retrieved from: http:// www.ecu.edu.au/conferences/tlf/2002/pub/docs/Chittlebor- ough.pdf. 20. O. Eskilsson, G. Hellden, Chem. Educ. Res. Pract. Eur. 2003, 4, 291–304. DOI:10.1039/B2RP90046G 13 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... 21. A. Georgiadou, G. Tsaparlis, Chem. Educ. Res. Pract. Eur. 2000, 1, 217−226. DOI:10.1039/A9RP90023C 22. N. Valanides, Chem. Educ. Res. Pract. Eur. 2000, 1, 249–262. DOI:10.1039/A9RP90026H 23. J. H. Mathewson, Sci. Educ. 1999, 83, 33–54. DOI:10.1002/(SICI)1098-237X(199901)83:1<33::AID- SCE2>3.0.CO;2-Z 24. S. Turkoguz, Chem. Educ. Res. Pract. 2012, 13, 401–409. DOI:10.1039/C2RP20046E 25. S. K. Abell, M. Roth, Int. J. Sci. Educ. 1995, 17, 59−74. DOI:10.1080/0950069950170105 26. A. B. Morrison, J. M. Chein, Bull. Psychon. Soc. 2010, 18, 46– 60. DOI:10.3758/s13423-010-0034-0 27. A. Woolfolk, 2002, Pedagoška psihologija. [Pedagogical psy- chology]. Ljubljana: Educy. 28. R. Marks, S. Bertram, I. Eilks, Chem. Educ. Res. Pract. 2008, 9, 267–276. DOI:10.1039/B812416G 29. C. W. Anderson, in: S. K. Abell, N. G. Lederman (Ed.): Per- spectives on science learning. In Handbook of research on science education Lawrence Erlbaum Associates. Mahwah, New Jersey 2007, pp. 3−30. 30. K. Rayner, Bull. Psychon. Soc. 1998, 124, 372−422. DOI:10.1037//0033-2909.124.3.372 31. K. Rayner, Q. J. Exp. Psychol. 2009, 62, 1457−1506. DOI:10.1080/17470210902816461 32. K. L. Havanki, J. R. VandenPlas, in: D. M. Bunce, R. S. Cole (Eds.): Eye tracking methodology for chemistry education research. In Tools of Chemistry Education Research American Chemical Society, Washington: DC, 2014, pp. 191–218. DOI:10.1021/bk-2014-1166.ch011 33. K. Rayner, Psych. Bull. 1998, 124, 372–422. DOI:10.1037//0033-2909.124.3.372 34. B. B. Koning, H. K. Tabers, R. M. J. P. Rikers, F. Paas, Educ. Psych. Rev. 2009, 21, 113–140. DOI:10.1007/s10648-009-9098-7 35. A. T. Duchowski, Instrum. Comput. 2002, 34, 455–470. DOI:10.3758/BF03195475 36. R. B. Kozma, E. Chin, J. Russell, N. Marx, J. Learn. Sci. 2000, 9, 105–143. DOI:10.1207/s15327809jls0902_1 37. M.–L. Lai, M.–J. T sai, F .–Y . Y ang, Hsu, T .–C. Liu, S. W .–Y . Lee, Educ. Res. Rev. 2013, 10, 90–115. DOI:10.1016/j.edurev.2013.10.001 38. V . Smith, D. Devane, C. M. Begley, M. Clarke, BMC Med. Res. Methodol. 2011, 11, 15−16. DOI:10.1186/1471-2288-11-15 39. M. J. Page, J. E. McKenzie, P. M. Bossuyt, I. Boutron, T. C. Hoffmann, C. D. Mulrow, BMJ. 2011, 372, 1−9. 40. M. Stieff, M. Hegarty, G. Deslongchamps, Cogn. Instruct. 2011, 29, 123–145. DOI:10.1080/07370008.2010.507318 41. H. Tang, N. Pienta, J. Chem. Educ. 2012, 89, 988−994. DOI:10.1021/ed200644k 42. V . M. Williamson, M. Hegarty, G. Deslongchamps, K. C. Wil - liamson, M. J. Shultz, J. Chem. Educ. 2013, 90, 159−164. DOI:10.1021/ed200259j 43. S. R. Hinze, D. N. Rapp, V. M. Williamson, M. J. Shultz, G. Deslongchamps, K. C. Williamson, Learn. Instruct. 2013, 26, 12–21. DOI:10.1016/j.learninstruc.2012.12.002 44. H. T ang, J. Kirk, N. J. Pienta, J. Chem. Educ. 2014, 91, 969−975. DOI:10.1021/ed4004113 45. P . A. O’Keefe, S. M. Letourneau, B. D. Homer, R. N. Schwartz, J. L. Plass, Comp. Hum. Behav. 2014, 35, 234–242. DOI:10.1016/j.chb.2014.02.040 46. S. C. Chen, M. S. Hsiao, H. C. She, Comp. Hum. Behav. 2015, 53, 169–180. DOI:10.1016/j.chb.2015.07.003 47. S. Cullipher, H. Sevian, J. Chem. Educ. 2015, 92, 1996−2005. DOI:10.1021/acs.jchemed.5b00529 48. M. A. Rau, J. E. Michaelis, N. Fay, Comp. Educ. 2015, 82, 460−485. DOI:10.1016/j.compedu.2014.12.009 49. J. J. Topczewski, A. M. Topczewski, H. Tang, L. K. Kendham - mer, N. J. Pienta, J. Chem. Educ. 2016, 94, 29−37. DOI:10.1021/acs.jchemed.6b00528 50. D. G. Herrington, R. D. Sweeder, J. R. V andenPlas, J. Sci. Educ. Tech. 2017, 26, 359−371. DOI:10.1007/s10956-017-9684-2 51. J. M. Karch, J. C. Garcia Valles, H. Seviah, J. Chem. Educ. 2019, 96, 830–840. DOI:10.1021/acs.jchemed.9b00014 52. M. C. Connor, A. Solaire, Finkenstaedt-Quinn G. V . Schultz, Chem. Educ. Res. Pract. 2019, 20, 522–541. DOI:10.1039/C9RP00033J 53. J. Y . Baluyut, T. A. Holme, Chem. Teach. Int. 2019, 28, 1−13. DOI:10.1515/cti-2018-0003 54. A. Nehring, S. Busch, J. Chem. Educ. 2020, 97, 275–282. DOI:10.1021/acs.jchemed.9b00102 55. M. Rodemer, J. Eckhard, N. Graulich, S. Bernholt, J. Chem. Educ. 2020, 97, 3530–3539. DOI:10.1021/acs.jchemed.0c00418 56. M. Slapničar, V. Tompa, S. A. Glažar, I. Devetak, J. Pavlin, Acta Chim. Slov. 2020, 67, 904–915. DOI:10.17344/acsi.2020.5908 57. J. R. VandenPlas, D. G. Herrington, A. D. Shrode, R. D. Sweeder, J. Chem. Educ. 2021, 98, 730–744. DOI:10.1021/acs.jchemed.0c00470 58. M. Tóthová, M. Rusek, V. Chytry, J. Chem. Educ. 2021, 98, 1831–1840. DOI:10.1021/acs.jchemed.1c00167 59. X. Sun, G. Lin, M. Zhan, Y . Zheng, J. Y e, D. Chen, J. Chem. Educ. 2024, 101, 858−867. DOI:10.1021/acs.jchemed.3c00075 60. L. Hamerská, T. Matecha, M. Tóthová, M. Rusek, Educ. Sci. 2024, 14, 570–596. DOI:10.3390/educsci14060570 61. Q. Huangfu, H. Li, Y. Ban, J. He, J. Chem. Educ. 2024, 101, 259–269. DOI:10.1021/acs.jchemed.3c00498 14 Acta Chim. Slov. 2025, 72, 1–14 Slapničar et al.: Using Eye-Movements to Explain Processing Triple Level ... Except when otherwise noted, articles in this journal are published under the terms and conditions of the  Creative Commons Attribution 4.0 International License Povzetek Glavni poudarek sistematičnega pregleda literature je bil preučiti študije očesnih premikov pri pouku kemije in ocen- iti uporabnost očesnih premikov pri raziskovanju učenja zapletenih kemijskih pojmov pri učencih, dijakih in študen- tih. Pregledane študije so večinoma uporabljale majhne vzorce študentov, pri čemer so bili pogosto vključeni študenti dodiplomskega študija, ter so merile spremenljivke, kot so očesni premiki, čas fiksacije, premer zenic itd. Rezultati kažejo, da imajo uspešnejši študenti krajši čas fiksaciji in se bolje učijo s pomočjo modelov, pri čemer je predznanje pomemben dejavnik. Obe skupini študentov se večinoma opirata na simbolno raven predstavitev, pri čemer so uspešnejši študenti boljši pri podajanju razlag. Študije so poročale tudi o težavah, povezanih z zapletenostjo analize podatkov.