Learning by comparing cases can be rationalized from a cognitive psychology perspective because it taps into several important cognitive processes, essential for learning and problem-solving. When comparing cases, a learner is engaged in a process called analogical reasoning, which involves finding similarities and differences between cases and using those similarities and differences to make inferences and draw conclusions. This analogical reasoning is a fundamental cognitive process that allows transfer knowledge and skills from one domain to another, or from one context to another (Gick and Holyoak, 1983). The structure mapping theory by Gentner (1989) and Gentner and Markman (1997) explains how this analogical reasoning works. When we compare two situations, objects, or reactions, we look for shared relationships. These relationships could either be similarities in surface features or relational features, such as causal or functional ones. Surface features are always visible features and details of a situation or object and, thus, are easy to discern. While relational structures refer to the abstract relationships between features and implicit information conveyed, they can, but do not necessarily share surface similarities. Comparing a set of correspondences between the surface or relational features of two cases leads to a structural alignment, i.e., discerning the information that two cases share. According to the structure mapping theory, the more shared relational features there are between two situations, the stronger the analogy, the easier to transfer our knowledge about one situation to reason about the other. For example, knowing that an electronegativity difference is needed to make a carbon-heteroatom bond polar, we can use that knowledge to infer that other carbon-heteroatom bonds might be polar as well, when there is a difference in electronegativity, even if the functional group looks different. However, attending to the relational similarity between cases is modulated by expertise. With increasing expertise, we can make use of abstract schemas and use them to categorize tasks based on implicit, conceptual aspects, whereas novice chemistry learners tend to focus on more explicit concrete features (Graulich et al., 2019; Lapierre and Flynn, 2020).

The Cognitive Load Theory (CLT) (Sweller and Chandler, 1994; Kalyuga et al., 1998) offers substantial insights into the use of case comparisons in learning chemistry, emphasizing how instructional design can manage cognitive resources to enhance learning (Paas et al., 2003). The CLT acknowledges the structure or extraneous load of a task (extraneous cognitive load), as well as the cognitive affordances that come with the content (intrinsic cognitive load) and the cognitive effort that a learner needs to activate for learning (germane cognitive load). When we compare cases, we activate our working memory system. However, the use of working memory and the associated capacity is limited, which is why sufficient available capacity must be accessible for effective learning or application of knowledge (Baddeley, 2010). CLT describes that learning is associated with cognitive load and that learning can be simplified or be more challenging depending on the circumstances. Intrinsic cognitive load is related to the difficulty or complexity of the learning material. Sweller (2003) focuses here on element interactivity. In concrete terms, this means that different elements must be processed simultaneously in the working memory during learning. This can happen sequentially, which causes a lower intrinsic cognitive load, or simultaneously, which results in an increased intrinsic cognitive load. If the elements are processed one after the other, e.g., in learning with single cases, this usually leads to memorization; if they are processed simultaneously, e.g., by comparing cases, links are created, which generates understanding but is also more demanding for the working memory (Sweller, 2010). The more prior knowledge learners have, the more links already exist and the lower the intrinsic cognitive load, even when processing elements simultaneously (Paas and Sweller, 2014). Two assumptions support the use of case comparison in light of the intrinsic cognitive load. On the one hand, as our working memory is limited in capacity, comparing cases instead of single cases helps us to be able to attend easily to differences and similarities and neglect other possibly irrelevant features of a situation or object (Schwartz and Bransford, 1998). Simultaneous processing of multiple and maybe irrelevant aspects can be challenging for learners; thus, the extraneous and intrinsic load can be reduced if cases help learners to focus on a reduced number of relevant aspects, as the one variable that needs to be compared can be focused on. This allows us to save capacity in our working memory. Furthermore, studying multiple cases allows learners to see how the same underlying principles apply to different contexts. This can help learners develop a deeper understanding of those principles and how they relate, which makes it easier to build conceptual chunks instead of memorizing single features (Schwartz and Bransford, 1998; Alfieri et al., 2013; Roelle and Berthold, 2015). Studying a single case in isolation may not give learners enough context or variation to understand the underlying principles involved fully (Alfieri et al., 2013). However, using case comparisons does not, per se, remediate mediocre ways of teaching. If the cases are not fully understood and the learner struggles to determine the relevant aspects, comparing cases might increase the intrinsic cognitive load compared to a single case, especially when multiple variables are involved (Schwartz and Bransford, 1998).

In contrast to the intrinsic cognitive load, the extraneous cognitive load is about how learning materials are designed (Sweller, 2010). The more superfluous or irrelevant information learners are presented with, the greater the possibility that they will not be able to distinguish between relevant and irrelevant information and will be distracted, which increases extraneous cognitive load. To minimize extraneous cognitive load for learners, it is therefore advisable to use design principles such as Mayer’s, which are evidence-based and conducive to learning (Mayer, 2021). In relation to case comparisons, this means, for example, that in addition to reducing irrelevant information, the relevant information can be emphasized, e.g., by highlighting techniques (Rodemer et al., 2022).

The germane cognitive load describes the load that relates directly to learning as an activity and is considered productive (Paas and Sweller, 2014). The more a learner can focus on the learning itself, the more effectively links can be created. The germane cognitive load thus relates to the intrinsic cognitive load. Currently, there is an assumption “that germane cognitive load has a redistributive function from extraneous to intrinsic aspects of the task rather than imposing a load in its own right” (Sweller et al., 2019, p. 264). The lower the extraneous cognitive load is kept, the more space is given to the intrinsic cognitive load, which in turn results in an increased germane cognitive load (which is positive). However, this only becomes important with complex learning material, as the intrinsic cognitive load only becomes noticeable here. The simpler a task is, the lower the intrinsic cognitive load and the lower the germane cognitive load (Paas and Sweller, 2014). In relation to case comparisons, this means that the way in which the learning material is designed should be well considered so that there is more space for the germane cognitive load. Complex tasks can be chosen, whereby the complexity must match the prior knowledge and the capacity of the working memory to be able to generate effective learning and links (Sweller, 1994).

Overall, comparing cases as a task design can offload the working memory and engage multiple cognitive processes that are essential for learning and problem-solving when they match the capability of the learners (Roelle and Berthold, 2015).

While Cognitive Load Theory (CLT) focuses on the capacity of working memory and how instructional design can be optimized to avoid cognitive overload, Variation theory is a learning theory that emphasizes the importance of variation in the design of instructional materials and activities and places emphasis on the importance of experiencing variations in the learning material to understand and discern the critical aspects of the content. While CLT is more about managing the quantity and complexity of information, Variation Theory is about the quality and structure of learning experiences. According to this theory, learners need to experience variations in the material they are studying in order to fully understand the underlying concepts, i.e., to abstract the relational connections beside surface similarities. Variation theory is based on the work of Swedish researcher Ference Marton and his colleagues, who developed the theory in the 1970s and 1980s (Marton, 1981). Marton (1981) was interested in understanding how students develop their understanding of complex concepts, and he observed that learners often struggle to transfer knowledge from one context to another.

Lo and Marton (2011) proposed that the key to understanding complex concepts is to focus on the variations in the material. They argued that learners need to experience different examples of a concept in order to fully understand it and develop a flexible understanding that can be applied to new contexts, advocating for a deep understanding of the subject matter instead of surface-level memorization.

Variation Theory of Learning helps further to support the use of case comparisons in chemistry education, as it emphasizes the importance of discerning critical features of a concept being taught. Using case comparisons (like different chemical reactions) helps students notice and understand the essential characteristics of each case; for example, contrasting an acid–base reaction with a redox reaction can help students understand the unique features of each type of reaction. Second, Variation Theory suggests that exposure to a range of examples, prototypical and non-prototypical examples, can help students see beyond single examples and support the ability to discriminate between different entities and recognize the significance of these differences. Certain elements become more salient to the viewer through variation, while other elements are kept invariant (Lo and Marton, 2011; Bussey et al., 2013), which allows learners to notice critical features more quickly (Bussey et al., 2013). Using case comparisons helps in achieving this by requiring students to apply principles to different scenarios, thereby promoting a deeper understanding of the underlying concepts (Roelle and Berthold, 2015; Bego et al., 2023). By focusing on these variations, variation theory aims to help learners develop a more nuanced and flexible understanding of the concept they are studying, which can be applied to new situations and contexts. The theory highlights the importance of experiencing variations in the material being studied in order to develop a flexible understanding that can be applied to new situations.

One way to combine the use of case comparisons with lab work is to embed these case comparisons in the CPOE cycle (Graulich and Schween, 2018), an adapted form of the Predict-Observe-Explain cycle (White and Gunstone, 2014) with an added “Compare” step. The cycle is based on learners first receiving a case comparison where they need to compare two given reactions (C), to predict (P) by generating a hypothesis which of the two reactions, for example, is faster than the other. This hypothesis can then be tested experimentally. By experimentally testing the hypotheses that have arisen from the case comparison, the outcome of the reactions is observed (O). Once the data has been analyzed, the final step takes place, in which conclusions are drawn about the previously formulated hypothesis based on the experimental results (E). Figure 3 illustrates the theoretical CPOE cycle by giving concrete examples how each step can look like, which is described in more detail in the following section.

As the name suggests, however, this may not be a linear process with a defined end but a cycle that can be repeated based on new case comparisons. In this way, learners not only become familiar with scientific principles through independent experience, but the targeted choice of contrasting cases and experiments also enables a specific promotion of chemical concepts.

Schween’s group has developed numerous experiments that make intermediate stages “visible,” for example, based on conductivity measurements (cf. Trabert et al., 2023 for an overview). In each case, two or more reactions are compared with each other and learners are prompted to estimate the reaction with the higher reaction rate. Their work resulted in experimental case comparisons on electrophilic substitution on aromatic compounds, in which the sigma complexes were determined by conductivity measurements (Vorwerk et al., 2015), on the stability of carbenium ions, which makes intermediates directly and indirectly visible through color gradients as well as conductivity measurements (Schmitt et al., 2013), on the competition of primary and secondary haloalkanes in S_N2 reactions (Schmitt et al., 2018), as well as on electronic substituent effects in alkaline ester hydrolysis (Trabert and Schween, 2018). All these experiments can be used in a CPOE cycle. Figure 3 illustrates the linkage of Trabert and Schween’s (2020) case comparisons of an alkaline ester hydrolysis, which is focused on inductive effects and their experimental design to the CPOE cycle. Thereby, students first receive contrasting cases of ester hydrolysis, which differ in their substituents on the phenyl group (Figure 3, compare) Based on these two reactions, students have to predict which of the reactions proceed faster including a justification (Figure 3, predict). Students test their hypothesis afterwards in the laboratory with conductivity measurements (Figure 3, observe). Based on their observations, students are encouraged to explain the phenomenon and refer to their hypothesis (Figure 3, explain). When the shown cycle is used in teaching and learning, learners can transfer their knowledge of inductive effects into a second cycle. Therefore, learners can apply their knowledge of inductive effect on new reactions, which focus on the position of substituents. Thereby, learners complete the CPOE cycle a second time by comparing the position of substituents on aromatic compounds, predicting the reaction rate, observing the hypothesis by conducting experiments, and explaining the position dependency of inductive effects. The key aim of these experimental case comparisons is to engage learners in reflection about reaction rate, slowly increasing the sophistication of chemical concepts such as electronic effects that is not only supported by the experimental investigations but can also be advanced to other reactions and contexts. Those cases used in the lab and discussed in lecture might serve as a bridge between these two traditional course formats in organic chemistry.

Guiding learners to attend to the relevant features, i.e., important functional groups involved in a reaction, can be achieved by multiple means, such as simply signaling or highlighting the relevant areas of the representation [i.e., signaling principle as described by Mayer (2021)], e.g., by zooming in or out, spotlights, coloring, added on-screen text or symbols. Others used experts’ eye gaze as a model for the learner, as used in the context of medicine (Jarodzka et al., 2012; Gegenfurtner et al., 2017), whereas transferring this idea to learning organic reaction mechanisms has not yet been convincing (Graulich et al., 2022). By “signaling” (highlighting key structural features in a static or dynamic fashion) students can focus on these key features of the representation and reduce their attentional focus to the rest of the structure, thus, reducing their extraneous cognitive load, if they are not attending to everything all at once (Richter et al., 2016; Schneider et al., 2018). It can also allow us to model a certain sequence of comparing by highlighting, for example, a starting point of comparison and then the sequential decoding process. Although attending to the relevant features is a key step. Implicit chemical properties cannot be read out of the functional group but need to be linked to it. When the attention of the learner is on the relevant features of a representation, the respective implicit information needs to be added, either in terms of verbal or written information. This is in line with the dual channel assumption of the CTML, providing highlighting for the visual features and chemical information for the verbal channel, as well as presenting it at the same time, i.e., the contiguity principle (Mayer and Fiorella, 2014). Some instructors might intuitively use highlighting techniques by pointing toward the representational features on the blackboard and explaining simultaneously or by adding conceptual information, such as pK_a or partial charges on the board. Redirecting a learner’s attention to the relevant aspects, thus, can be complex, as decisions have to be made that cannot just be guided by the salience of a functional group, and conceptual information needs to be linked to make a purposeful selection.

In a quantitative study, we tested if a highlighting technique actually supports students to attend to relevant areas of organic chemistry case comparisons and solve them more successfully. Thus, we created tutorial videos with case comparisons and used a dynamic moving dot highlighting representational features, which was synchronized with the information given as a verbal explanation in parallel (Rodemer et al., 2020; Eckhard et al., 2022). The study could document that all students in the study were profiting from the given verbal explanation, but especially low performing students profited from the highlighting. Following students while watching the videos with highlighting with the help of eye-tracking could show that the attention to relevant areas is focused over the entire time of the video, and the perceived extraneous cognitive load is decreased (Rodemer et al., 2022). These overall results illustrated that beginners need more support in decoding the molecular structures that we use in organic chemistry, and guiding their attention is key for a decreased extraneous cognitive load. Besides using eye-tracking as an analytical lens to track students’ attention, using it in instruction might help students understand their own viewing behavior. In an eye-tracking study conducted by Hansen et al. (2019), they investigated how students view and critique different animations of redox reactions and precipitation reactions. After their reasoning process, students received visual feedback on their own viewing behavior. Hansen et al. (2019) revealed that viewing this feedback helped the students to be critical about their own viewing behavior and to deepen the critique regarding the animations shown.

Further breaking down the reasoning process with case comparisons into manageable parts can help students process the information more effectively (Belland, 2017). A simple nucleophilic substitution, taught in an introductory organic chemistry course, for instance, requires the consideration of three main influential factors, i.e., leaving group ability, nucleophilicity, substrate effects, and the cause-effect relationships that determine the reactivity in this type of mechanism. Thus, a lot needs to be considered by the learners. Using case comparison can have positive effects on students’ engagement with the conceptual knowledge, as it shifts the focus onto implicit and influential factors of the organic reaction mechanism (Watts et al., 2021). However, if we expect students to reason in a particular way, i.e., building cause-effect relationships, and connect different concepts and properties, we need to be explicit how students should integrate these multiple pieces of knowledge. Developing mastery requires explicit learning of how to create those mechanistic explanations (Cooper, 2015). Thus, supporting students in solving case comparisons should acknowledge the complexity and reasoning steps required and ideally make these steps transparent through a scaffold (Caspari et al., 2018; Kranz et al., 2023). Scaffolding is a known technique widely used as an instruction in science education (cf. Lin et al., 2012; Wilson and Devereux, 2014) and helps students to slow down the decision-making process and gives students the opportunity to activate necessary conceptual and procedural knowledge (Rittle-Johnson and Star, 2007; Rittle-Johnson and Star, 2009; Shemwell et al., 2015; Chin et al., 2016). A scaffold for the case comparisons illustrated therein thus can guide the learner through the different considerations necessary to make a claim about the outcome of a case: (1) describing the chemical changes in the given cases; (2) explicitly stating the overall goal of comparison (task prompt); (3) naming the similarities and differences; (4) stating the role of the influential factors (i.e., implicit properties); (5) explaining and contrasting the influences of the implicit properties; (6) stating how the transition state is affected to refer to the energetic account and (7) making a final claim about the reactivity of both reactions (Bernholt et al., 2023).

Various studies already documented the positive effect of using scaffolding with case comparisons on students’ reasoning. In prior studies, we used a scaffold grid, represented by a worksheet with empty boxes, which visually connects the structural differences, changes, and cause-effect relations (Caspari et al., 2018). By utilizing this grid, students can systematically relate each structural difference to each ongoing change, verbalizing the influence of the structural difference on the change. We compared how students are reasoning through contrasting cases with and without a scaffold and could observe that students’ reasoning is more guided and includes the consideration of more implicit properties and influential effects when solving a contrasting case with a scaffold (Caspari et al., 2018). This structured approach helps students avoid jumping to the final answer without considering the underlying reasons. A mixed-methods study could confirm that especially students with a low prior knowledge profited from working with a scaffold and had a higher learning gain, whereas it does also not harm those with higher prior knowledge (Kranz et al., 2023). Lieber et al. (2022) advanced a scaffold further by acknowledging students’ individual needs when arguing about alternative reaction pathways. Those adaptive scaffolds could show that more individualized instruction when using different cases in organic chemistry might be a new avenue to improve teaching.