Leonie
Lieber
and
Nicole
Graulich
*
Justus-Liebig University Giessen, Institute of Chemistry Education, Heinrich-Buff-Ring 17, 35392 Giessen, Germany. E-mail: nicole.graulich@dc.jlug.de
First published on 23rd August 2021
Building scientific arguments is a central ability for all scientists regardless of their specific domain. In organic chemistry, building arguments is a necessary skill to estimate reaction processes in consideration of the reactivities of reaction centres or the chemical and physical properties. Moreover, building arguments for multiple reaction pathways might help students overcome the tendency toward one-reason decision-making and offer them an authentic perspective on organic processes. Reasoning about multiple alternative organic reaction pathways requires students to build arguments and then judge and weigh the plausibility of these pathways. However, students often struggle to build strong arguments and use scientific principles appropriately to justify their claims. In the present study, the argumentation patterns of 29 chemistry majors students were analysed using a simplified version of Toulmin's argumentation model (claim–evidence–reasoning). The students solved various tasks related to alternative reaction pathways of a substitution reaction. They supported their claims with evidence and justified the evidence through reasoning. We investigated (a) the extent to which the students use evidence and reasoning in their argumentation (referred to as their argumentation approach), (b) how students with different argumentation approaches rationalised changes in their initial claims, and (c) how students used reasoning to justify their arguments. The results indicate that students need further support to appropriately use evidence and reasoning and to apply conceptual knowledge to build well-grounded arguments.
Building arguments or explanations is a core ability that must be learned by science students. This process includes evaluating claims and weighing evidence (Driver et al., 2000). Erduran (2019) purposefully linked these two aspects (claims and evidence), which were previously considered separate, by stating that claims need to be justified with evidence. Justifying and refuting claims using evidence is part of a scientific mode of thinking.
In organic chemistry, a student's success is often linked to the ability to build arguments (Bodé et al., 2019; Cruz-Ramirez de Arellano and Towns, 2014; Deng and Flynn, 2021) and reason about reaction mechanisms (Grove et al., 2012; Becker et al., 2016; Caspari et al., 2018; Watts et al., 2021). More explicitly, this means that students who solve tasks related to reaction mechanisms need to be able to argue and reason about the structural changes that occur and the causes of these changes. To explain these causes, students need to justify their evidence with reasoning (Cooper et al., 2016; Watts et al., 2020). However, students show various problems when building arguments and using reasoning. Students often struggle to consider alternative pathways since they tend to focus on one reaction product instead of the underlying processes (Popova and Bretz, 2018a). Thus, students are often triggered by single clues such as the presence of a leaving group, which often leads them to neglect other reactivities or multiple reaction pathways (Kraft et al., 2010). With regard to argumentation structures, forming a claim is often the easiest component for students (McNeill and Krajcik, 2012). Problems occur when students try to justify claims as they experience uncertainties regarding what counts as evidence (Sadler, 2004). Thereby, students explicitly struggle with the appropriate use of scientific principles while explaining why their given evidence supports a claim (McNeill et al., 2006; McNeill and Krajcik, 2012; Walker et al., 2019). Problems in students’ understanding of chemical concepts such as nucleophilicity and electrophilicity or acidity and basicity are well documented (Cartrette and Mayo, 2011; Anzovino and Bretz, 2015; DeFever et al., 2015; Akkuzu and Uyulgan, 2016). To avoid these problems in chemical concepts, students rely on personal views when judging the plausibility of a given context (Hogan and Maglienti, 2001). Even when students justify their claims, they often rely on one-reason decision-making based on single pieces of evidence when multiple supporting pieces are needed (McNeill et al., 2006; Talanquer, 2006; Kraft et al., 2010). This is primarily observed when students must argue about data that contradict their own ideas (Chinn and Brewer, 2001).
Many chemistry education researchers have specifically used Toulmin's argumentation model to analyse or support argumentation. For instance, Becker et al. (2013) analysed classroom discourse in which students made claims and particulate-level justifications about thermodynamic topics such as enthalpy and heat capacity. The authors suggested that while a conceptual understanding of the particular nature of matter is important, students must to be able to apply this knowledge about chemical and physical properties to build arguments (Becker et al., 2013). Lazarou and Erduran (2021) analysed science teachers’ instructional adaptions of Toulmin's argumentation pattern in the classroom and the quality of students’ written arguments. They revealed that some elements of Toulmin's argumentation pattern cause ambiguity and interpretative difficulties for students and science teachers (Lazarou and Erduran, 2021).
The claim–evidence–reasoning (CER) model, a simplified version of Toulmin's argumentation model, is currently the prevalent argumentation model used in chemistry education research. Walker et al. (2019) conducted a study in a general chemistry laboratory course based on the argument-driven inquiry instructional model. The participants worked in groups and collected data in the laboratory to build arguments that included claims, evidence, and reasoning as justifications. The authors found that students had problems changing their claims and considering alternatives. Instead, they stuck with their initial claims despite critique. Moreover, students struggle to justify their claims with scientific principles. Walker et al. (2019) reported that students may need additional guidance to revise or change their claims and to use scientific principles in their justification. Luo et al. (2020) investigated the effect of inquiry on students’ argumentation skills. They examined the influence of a reasoning flow scaffold in comparison to a control group that received conventional argumentation training. The students in the intervention group showed significant improvement in certain argumentation elements compared to the control group. Luo et al. (2020) concluded that dividing a claim into two parts (one at the beginning and one at the end as an opportunity for falsification) may assist students in building arguments by causing them to reflect on their decision more carefully. In organic chemistry, Bodé et al. (2019) investigated how students build scientific arguments when comparing two reaction mechanisms for nucleophilic substitution. They concluded that practicing argumentation skills are essential to make expectations clear for the students. These practices may include debating about justifications or comparing with prior knowledge.
Overall, the above studies indicate that students (1) struggle to revise their initial, possibly erroneous claims, (2) rarely consider alternatives in their argumentation, and (3) struggle to build well-grounded arguments.
Clarifying how students use evidence and reasoning in their arguments and rationalise their claims when assessing the plausibility of multiple alternative reaction pathways may provide further insights into their argumentation behaviours.
Toulmin determined that an argument consists of six elements, as represented in Fig. 1, which also illustrates the commonalities between Toulmin's argumentation pattern and the simplified CER model (Toulmin, 2003). More generally, Kuhn (1991) claimed that an argument is an assertion with justification, while Osborne and Patterson (2011) stated that an argument is a justified claim.
Fig. 1 Coherence of Toulmin's argumentation pattern (2003) and the CER model often used in chemistry education research. |
The following descriptions of the elements of an argument are taken from Toulmin's book The Uses of Argument (Toulmin, 2003).
The first step in building an argument is the expression of a claim. Asserting something is the key component of making a claim (Van Eemeren et al., 2014). The foundation of the claim is called data. Data are initiated by the question “What have you got to go on?”. Therefore, the claim based on the given data is the first step to justify the claim (and the second step of an argument). The given data consist of rules or principles and show that the claim is appropriate. The third step of an argument is called the warrant, which is the justification of the given data. The warrant is a general statement based on rules or principles. Since a warrant is a bridge between the data and the claim, the initiating question is “How do you get from your data to your claim?”. On the basis of data, the warrant gives permission to make a claim. The fourth element of an argument is the backing, which is based on the question “What entitles you to conclude from… to…?” and refers to the warrant. A backing is required when the warrant is not immediately accepted. Thus, the backing relies on the principles of ethics, values, or general norms. The fifth element of an argument is called the qualifier and is initiated by the question “Is that necessarily so?”. The qualifier provides information about the obligation of the claim and thus possibly weakens the claim. Words such as likely, probably, or necessarily can be applied to introduce a qualifier. The sixth and last element of an argument is the rebuttal. The rebuttal describes the circumstances under which the claim is not valid or only valid to a certain extent. A rebuttal is an exception of rules and is initiated with the question “When does the rule not apply?”. As soon as a rebuttal appears in an argument, the claim must be weakened with a qualifier. Conversely, just because there is a qualifier, there is not necessarily the need for a rebuttal (Toulmin, 2003).
Accordingly, CER is frequently used as a simplified version of Toulmin's argumentation pattern based on the terms “claim,” “evidence” (data), and “reasoning” (warrant).
In both the original and CER models, the main goal of an argument is the justification of a claim with the use of evidence and reasoning (Toulmin, 2003; Osborne and Patterson, 2011). The CER model consists of three parts (McNeill and Krajcik, 2012).The first part in both the CER and original models is the claim, which acts as a statement or a conclusion to a problem or question (McNeill et al., 2006; McNeill and Krajcik, 2012). Similar to Toulmin's argumentation pattern, the claim in the CER model is always in doubt and requires a justification based on known data and an explanation (Osborne and Patterson, 2011). The claim is supported with evidence based on scientific data (McNeill et al., 2006; McNeill and Krajcik, 2012). Thus, evidence in den CER model is similar to Toulmin's data (Van Eemeren et al., 2014). As scientific data contain information from various resources, multiple pieces of evidence can support a single claim (McNeill and Krajcik, 2012). Equally important is the reasoning (in the CER model) as it provides the bridge between the claim and the evidence; reason in the CER model combines Toulmin's warrant and backing (McNeill et al., 2006; McNeill and Krajcik, 2012). Reasoning indicates how the link between a claim and the given data can be justified and why they are justified by providing a logical connection (warrant) (McNeill and Krajcik, 2012; Toulmin, 2003; Van Eemeren et al., 2014). As this connection may need backing up, scientific principles should be included (backing) (McNeill and Krajcik, 2012). Reasoning is considered to be the most difficult step of an argument because a justification requires the bridging of the claim and evidence (McNeill and Krajcik, 2012).
(1) To what extent do students use evidence and reasoning in their argumentation (which we refer to as the argumentation approach)?
(2) How do students with different argumentation approaches rationalise changes in their initial claims?
(3) How do students use reasoning to justify their arguments while assessing the plausibility of alternative reaction products?
To answer these three research questions, we conducted qualitative interviews with students majoring in chemistry prompting them to elaborate on serial tasks related to the plausibility of reaction products.
Eleven female and 18 male chemistry major students participated on a voluntary basis. Students were recruited at the beginning of the OC3 course via an announcement in the lecture. The students were between 20 and 27 years old (average 22 years). While Institutional Review Board is not required at German universities, interviews followed ethical guidelines, and the students had the opportunity to terminate their interview at any point. All recruited students were provided with information about their rights and the manner of data collection and gave their written permission for the following: (1) the questionnaire can be evaluated by the research team, (2) the transcripts, drawings, video data, and audio data can be used by the research team; and (3) the collected data can be analysed and published by the research team. All students created a pseudonym, and the video data only included voice and drawings. As all students were native German speakers, the interviews were conducted in German, and direct quotes were translated into English for publication.
The students completed a demographic questionnaire at the beginning of the interview. The data also included pseudonymised scans of the students’ work sheets as well as audio and video recordings. Verbal utterances were transcribed verbatim. Individual think-aloud interviews were led by the first author using a semi-structured protocol (Bernard and Bernard, 2013). After their interviews, all students were asked not to talk about the tasks with their colleagues to minimise the chances of other students preparing beforehand. After the interviews, the first author provided sample solutions of the task during the lecture.
The quotes used in this publication were checked multiple times by the research group and a German English student teacher to guarantee the validity of the translations.
The first subtask contained a typical reaction that the students had experienced several times throughout the organic chemistry courses and laboratories (see Fig. 2, task 1). This reaction was chosen to provoke the use of intuitive heuristics since nucleophilic substitution and elimination were familiar to the students. The second subtask (Fig. 2, task 2) contained a reaction that – at first sight – seems similar to the reaction from subtask 1 as it only differs in one surface feature. The aim was for the students to use the same problem-solving approach as in subtask 1; however, in this case, the same problem-solving approach would not be productive. In subtask 3 (Fig. 2, task 3), students were provided with alternative products for the reaction given in subtask 2. This step was intended to challenge students’ problem-solving approaches as the confrontation with alternative products might initiate a more analytical thinking process. This stepwise process of first generating possibly erroneous solutions by themselves and then comparing their solutions with comparable or different solutions (Loibl and Leuders, 2019) has been shown to have a beneficial effect on students’ reasoning. In the last subtask (Fig. 2, task 4), students had the opportunity to defend or revise their generated product from subtask 2 (i.e., revise their initial claim).
Category | Description of the code | Student example |
---|---|---|
Electronics | Students describe electronic aspects like polarisation, electronegativity, or charge | “This bond is more polarised.” |
Energetics | Students describe energetic aspects like bonds, energies, or thermodynamics | “The C–H bond is stronger. You have to spend energy to cleave the bond.” |
Kinetics | Students describe kinetic aspects like rates or statistics | “… that acid–base reactions react very fast.” |
Spatial arrangement | Students describe spatial aspects like steric effects, sizes, distances, or angles | “… because the groups are close to each other.” |
Analogies | Students describe examples that are not part of the task but compare or demonstrate aspects | “You can deprotonate phenols with hydroxide.” |
Strength | Students describe, for instance, the quality or strength of acids/bases, leaving groups, or nucleophiles | “… that this part is more acidic.” |
Conditions | Students describe aspects like concentrations, temperatures, or pressures | “It depends on how much base I add.” |
Stability | Students describe aspects of stability without an explicit explanation of the term (e.g., in terms of energetics) even after being prompted to do so | “… this product would be definitely not stable.” |
A special note is required for the coding of the Stability category. This code was only assigned if the students could not explain the term ‘stability’ after being prompted and relied on judgments such as “the reaction product is stable”.
Students were placed into group 2 when the ratio of reasoning to evidence statements was close to 1, and each of the frequencies of evidence and reasoning was between 10 and 20.
Lastly, group 3 contained students for which the ratio of reasoning to evidence was less than or equal to 1 (i.e., the students mostly used evidence than reasoning). For these students, the frequency of evidence statements was nearly 10, while that of reasoning statements was less than 10.
Argumentation approach 1 (reasoning-based approach): students’ argumentation patterns were classified as argumentation approach 1 when the ratio of reasoning to evidence statements was above 1. Additionally, the frequency of evidence statements had to be more than 10, and the frequency of students’ reasoning had to be equal to or greater than 20. Both the frequencies of students’ evidence and reasoning statements and the ratio has to be accomplished to be assigned to the approach. Ten out of the 29 students were assigned to the reasoning-based argumentation approach.
Argumentation approach 2 (based-on-both approach): students were assigned to argumentation approach 2 when the ratio of reasoning to evidence statements was nearly 1. Moreover, each of the frequencies of evidence and reasoning statements had to be between 10 and 20. The frequencies of evidence and reasoning statements as well as the ratio has to be accomplished to be assigned to the second argumentation approach. Ten out of the 29 students were assigned to this group.
Argumentation approach 3 (evidence-based approach): students were assigned to approach 3 when the frequency of evidence in students’ argumentation was nearly 10 while the number of reasoning was lower. The ratio of reasoning to evidence was less than or equal to 1. In contrast to the based-on-both argumentation approach, the frequency of reasoning statements must be lower than 10 and the ratio of reasoning to evidence statements must not be greater than 1 to be assigned to the evidence-based argumentation approach. Thus, nine out of the 29 students were assigned to this approach.
Fig. 4 illustrates example argumentation patterns for the alternative reaction product card C (“alkene”) in subtask 3 (see Fig. 2). A complete argumentation pattern map for each argumentation approach is shown in Appendix 2. In each argumentation pattern, the claim (plausible or implausible) and, when present, the support by evidence and the justification by reasoning are shown moving from top to bottom.
All three students whose argumentation patterns are shown in Fig. 4 for product card C initially provided the same correct claim: the alkene is implausible as a reaction product. However, the students differed in the way they supported the claim with evidence or justified it with reasoning. For instance, Reuben, who was assigned to the reasoning-based argumentation approach, supported his claim with three pieces of evidence and justified each piece of evidence with at least one reasoning statement. The reasoning statements were classified into different reasoning categories including Electronics, Spatial Arrangement, and Strength.
Sydney, who was classified into the based-on-both argumentation approach, supported her claim with three pieces of evidence, just as Reuben did. However, unlike Reuben, she did not support each of her evidence statements with reasoning statements.
Frank, who was classified into the evidence-based argumentation approach, provided the same claim as Reuben and Sydney. However, although he supported his claim with a piece of evidence, he was unable to justify it with reasoning even after being asked several times to do so. These three argumentation approaches are illustrated in more detail in the following section with examples provided for each approach.
Fig. 5 illustrates the reasoning-based argumentation approach with a quote from Phil. In general, Phil supported his five claims for the five reaction product cards with a total of 14 pieces of evidence and 28 reasoning statements (see Appendix 1). In the shown interview excerpt, he argued for reaction product card D, the reaction of 4-chlorobutanol and hydroxide to alkoxide and water, claiming correctly that the reaction product is plausible. He then supported his claim with two pieces of evidence (“If only a very small amount of base is added” and “the acid–base reaction is much faster [compared to the nucleophilic substitution]”). These two statements were categorised as evidence since they refer directly to the claim. Furthermore, he used reasoning to describe statements based on chemical concepts such as electronic or energetic aspects that support the evidence (e.g., why an acid–base reaction is faster compared to a SN reaction).
Fig. 5 Phil's argumentation for the alkoxide (product card D) as an example for the reasoning-based argumentation approach. |
To further justify the correct evidence “the acid–base reaction is much faster”, Phil provided four reasoning statements. First, Phil used the HSAB principle by reasoning that the hydroxide ion is a hard base, while the proton of the hydroxyl group is a hard acid (Strength category). Second, he emphasised that the energy gain is best for the reaction between a hard base and a hard acid (Energetics category). Third, Phil referred to the charge density that can be shifted into an empty orbital in the acid–base reaction instead of an antibonding orbital in the nucleophilic substitution (Electronics category). Finally, he referenced steric effects related to the ability of the hydroxide ion to attack from any direction in the acid–base reaction (Spatial Arrangement category). Overall, Phil used many reasoning statements to justify his evidence, which in turn supported his claim regarding the reaction product card. Thus, Phil was assigned to the reasoning-based argumentation approach.
Dylan serves as an example of the based-on-both argumentation approach. He gave 15 pieces of evidence in total and 12 reasoning statements (see Appendix 1). An interview excerpt of his argument regarding reaction product card E, the reaction of 4-chlorobutanol and hydroxide to tetrahydrofuran (THF) and water, is provided in Fig. 6. Dylan was asked if the reaction product card was plausible or implausible. He mentioned evidence five times correctly (“here [α-carbon atom] is a partial positive charge”, “there [hydroxide] is a negative charge”, “just a little ring strain”, “five-membered ring is quite stable”, and “chloride is a good leaving group”). However, only one piece of evidence (“chloride is a good leaving group”) was justified with three reasoning statements.
Fig. 6 Dylan's argumentation for THF (product card E) as an example for the based-on-both argumentation approach. |
First, he used reasoning in the Electronics category when he said that chloride has an octet. The second reasoning statement is categorised in the Electronics category because he indicated that chlorine has a higher electronegativity than carbon. The final reasoning statement was related to the thermodynamic stability of chloride ion (Energetics category). Dylan was assigned to the based-on-both argumentation approach since he only supported one piece of evidence with reasoning.
Amber serves as an example of the evidence-based argumentation approach. She used a total of seven pieces of evidence and four reasoning statements to support her claims (see Appendix 1). In the interview, she argued the plausibility of product card A, the reaction of 4-chlorobutanol and hydroxide to 1,4-butanediol and chloride. Here as well, Amber was asked if she thought the reaction product to be plausible or implausible. She answered erroneously that the reaction product is plausible and supported her claim with the evidence that chloride is a good leaving group. Fig. 7 shows an excerpt of Amber's explanation for why she thinks chloride is a good leaving group. At first, she answered that she has no idea why chloride is a good leaving group and that she learned or accepted the fact without questioning it. After the interviewer prompted her to try to explain her evidence, she was quiet for a while before responding incorrectly that chloride could be smaller than hydroxide ion but that she did not know the correct answer. This reasoning statement was assigned to the Spatial Arrangement category. Amber was one out of nine students assigned to the evidence-based argumentation approach because she hardly supported her claims with any evidence and barely used reasoning to justify her evidence.
Fig. 7 Amber's argumentation for the diol (product card A) as an example for the evidence-based argumentation approach. |
The three argumentation approaches identified illustrate that the students differed in the amount of evidence and reasoning statements used in their arguments. The more evidence and reasoning a student uses, the higher the probability that a large amount of influential variables are included in the decision-making process. This means that the probability of considering different chemical concepts such as kinetics, energetics, and energetic processes also increases. Since the course of a chemical reaction does not only depend on one factor but rather on various influencing factors, the inclusion of a larger number of evidence statements results in a more comprehensive decision-making process by increasing the breadth of the argumentation process. The number of reasoning statements can then influence the quality or depth of the argument. The more reasoning statements a student uses, the better the evidence can be justified. Reasoning serves to strengthen the claim through the justification of evidence; thus, the students’ arguments gain depth and are not just “empty envelopes”.
Each student made a claim about each of the five reaction product cards he or she discussed during the interview. Fig. 8 summarises “most plausible” claims made by the students; claims are given as percentages and separated into the three argumentation approaches. In task 2, students formed products independently as they only received the reactants. In contrast, in task 3, five cards with alternative reaction products were provided. All products that were claimed to be “most plausible” by the students in task 2 and task 3 were aggregated so that the total represents 100%. Since the students were able to claim more than one product as plausible, the number of products differs for the tasks and argumentation approaches. For example, 10 students in the reasoning-based argumentation approach claimed the formation of 14 products in task 2 and 15 products in task 3, which corresponded to product cards A-E. The 10 students with the based-on-both argumentation approach built 15 products in task 2 and 21 products in task 3. The nine students with the evidence-based argumentation approach claimed the formation of 12 products in task 2 and nine products in task 3. Regardless of the argumentation approach, most claimed incorrect products such as the diol (product card A) and the alkene (product card C) in task 2. This outcome is not surprising since tasks 1 and 2 were purposely designed to provoke the use of the same problem-solving approach used by students in task 1, leading to an SN2 reaction (Lieber and Graulich, 2020). After discussing the reaction product cards in task 3, an increase in claims for the correct product THF (product card E) and its precursor (alkoxide, product card D) along with a decrease in claims of the incorrect product diol (product card A) were observed for all argumentation approaches. Nevertheless, there were noticeable differences in how students with different argumentation approaches rationalised the changes in claims.
Considering only the percentage of THF (product card E) chosen as the most plausible product, it seems that the students who used an evidence-based argumentation approach claimed more correct products (56% THF) compared to those who used a reasoning-based argumentation approach (40% THF) or a based-on-both argumentation approach (24% THF). However, it should be noted that the alkoxide (product card D) is the precursor of THF. Many students who applied reasoning-based and based-on-both argumentation approaches recognised this fact and chose both THF and the alkoxide as the most plausible reaction products. When looking at the total of THF (product card E) and alkoxide (product card D), it is apparent that the students who used an evidence-based argumentation approach did not consider alkoxide as a plausible product or precursor. For example, Charlie who applied a reasoning-based argumentation approach chose THF and the precursor alkoxide as the most plausible products in task 3 when asked if he thought THF is plausible or implausible.
Charlie: “I don’t think it's not plausible (he refers to THF). I wouldn’t have intuitively guessed that something like this (THF) would happen if I am honest. But I would have taken an easier way and said it is a normal nucleophilic substitution, at this point, if I look at it that way I would not necessarily say it (THF) could be a by-product of what is being created. But the main product, the more I think about it, the more fascinating I think it is.”
Interviewer: “Why?”
Charlie: “Because I really didn't think that this (formation of THF) was a possibility. Well, I didn't think about it at all, I even think that what was there before (nucleophilic substitution) is wrong, because I just saw that I have a nucleophile and a good leaving group and nucleophilic substitution, but I didn't see that you can build another nucleophile (alkoxide).”
Charlie was one of the students who initially chose the diol (product card A) as the main product of the reaction of 4-chlorobutanol and hydroxide in task 2. When he received the alternative reaction product cards, he focused on THF as a reaction product. After he argued about both the alkoxide (product card D) and THF (product card E), he claimed both as the most plausible products of the reaction and explained that the alkoxide is the precursor of THF.
Among the students who applied evidence-based argumentation approaches, most did not provide a valid explanation for choosing THF as their most plausible product. Sonia, for instance, tried to build the mechanism for the formation of THF and could not find an acceptable explanation for her decision. As a result, the interviewer attempted to discuss the product THF in another way.
Interviewer: “We can do it differently. When you see the product, would you describe the product as plausible in principle?”
Sonia: “Yes, but I don't know why. My feeling tells me that again, I don't know. It looks so right somehow.”
Interviewer: “And are there any factors that help you determine why you think the product could be right?”
Sonia: “We once had a similar task in an exercise and I don't know, when I saw that, it kind of clicked in my head, somewhere in the back corner where I thought, I think that's it.”
Sonia later identified THF as the most plausible product of the reaction of 4-chlorobutanol and hydroxide. Nevertheless, she could not support her claim with any evidence or reasoning. She justified her claim based only on her feelings and on her memory of a similar task she completed sometime before. Like Sonia, no student with an evidence-based argumentation approach mentioned that the alkoxide (product card D) is the precursor for THF (product card E). Still, five out of nine students claimed THF as the most plausible product of the reaction with either little justification or without further justification. The following quote from Andy illustrates such a missing link from the precursor alkoxide to the reaction product THF. Andy built the diol (product card A) as a product in task 2 and chose the alkoxide as his first reaction product card to argue in task 3. He started to describe how the alkoxide was built.
Andy: “So here, the hydroxyl group was simply deprotonated.”
Interviewer: “How plausible do you think is this?”
Andy: “That's a good question. Actually, so it may happen. But a negative charge only on oxygen is very, I would say, unlikely.”
Interviewer: “Why?” […]
Andy: “So, that charge is on the oxygen and then is not somehow stabilised. That is a bit strange to me. Because, charge on an ester group would be kind of, I think, normal, because that's relatively stabilised by resonance. But only at the oxygen is a bit strange to me.”
Andy claimed incorrectly that the alkoxide is implausible as the reaction product. Since he could not explain why the charge on the oxygen atom seemed “strange” to him, he used an analogy with a molecule known to him, namely an ester, to support his claim. After Andy argued about the reaction product cards of the alkene (C) and the carbanion (B), he started to argue about THF (E) by drawing his idea of the reaction mechanism for the formation of THF (see Fig. 9).
Subsequently, the interviewer discussed the plausibility of the formation of THF (product card E) with Andy.
Interviewer: “How plausible do you think is it?”
Andy: “That might even be more plausible than this (diol).”
Interviewer: “Okay. Why?”
Andy: “The O has lone pairs, a five-membered ring is not as strained as a four-membered-ring, for example. I would say that could be possible. So, I would even consider it relatively plausible.”
Both in Andy's drawing of the reaction mechanism and his explanation of the plausibility, he did not consider the alkoxide (product card D) as a precursor of THF formation. Andy proposed falsely that the hydroxyl group of 4-chlorobutanol attacks intramolecularly to form a five-membered ring that is deprotonated by the hydroxide ion in the last step. Since Andy needed several attempts to complete the reaction mechanism to his satisfaction, the interviewer asked him how he proceeded in setting up the reaction mechanism.
Andy: “I noticed that the O has lone pairs. And then I thought, in order to reconstruct the five-membered-ring, it must attack here. That's where I first drew the Cl. But it would be a pentavalent carbon atom, which does not exist. So, the Cl has to get out somehow. And in the last step the H is taken by the OH. But only because the O is positive and it wants to become relatively neutral. And somehow furan (Andy confounded THF and furan) has to be the product.”
Interviewer: “That means, you basically made sure that you knew the product and how to put it together?”
Andy: “Yes, somehow. Exactly.”
Comparable statements to those made by Andy, are well known in the literature. Bhattacharyya and Bodner (2005) showed that students tend to postulate a mechanism with the goal of forming a given product. However, in doing so, students are likely to build intermediates that are implausible.
In summary, Andy was not the only student who did not associate the alkoxide (product card D) with THF (product card E). Nevertheless, while most students changed their initial claim, the rationale for the change varied greatly. In particular, students who used a reasoning-based argumentation approach identified the alkoxide as the precursor of THF and provided a rationale for it. Although five out of nine students who applied an evidence-based argumentation approach also chose THF as the most plausible product, they were not able to (sufficiently) justify their decision.
Clear differences in the numbers of reasoning statements between argumentation approaches can be seen in Fig. 11. The reasoning-based argumentation approach (blue) was associated with the use of considerably more reasoning statements compared to the based-on-both argumentation approach (orange) and the evidence-based argumentation approach (purple), although a distinction is also visible between the latter two approaches. However, the percentages of reasoning statements in the different reasoning categories were similar across the three argumentation approaches (see Appendix 3). For example, the Electronics category accounted for 39% for the reasoning statements for the reasoning-based argumentation approach, 43% for the based-on-both argumentation approach, and 40% for the evidence-based argumentation approach. The percentages were also similar for the Energetics category (17% for the reasoning-based argumentation approach, 17% for the based-on-both argumentation approach, 16% for the evidence-based argumentation approach) and the Kinetics category (3% for the reasoning-based argumentation approach, 2% for the based-on-both argumentation approach, and 3% for the evidence-based argumentation approach). The students not only differed in their number of reasoning statements, but also in the way they justified their statements. For example, Cameron (reasoning-based argumentation approach) and Gloria (evidence-based argumentation approach) both mentioned the correct evidence that chloride is a good leaving group. However, these two students differed in how they justified this evidence.
Cameron: “Chloride is a good leaving group because the chlorocarbon bond is not very stable, unlike leaving groups such as OH - . Then it can be stabilised in an aqueous solution. […] There is also a high electronegativity. It's also very electron-withdrawing, chlorine as an atom.”
Interviewer: “Could you explain to me why chloride is a good leaving group and hydroxide is a weak leaving group?”
Cameron: “[…] So theoretically it should be that the electron that is added in chlorine is stored in an orbital, which is energetically higher than hydroxide. But I’m not really getting anywhere with it (laughs). So, I would say that this is not the cause because of the stability of the two ions, but because of the stability of the bond. That the splitting, here with chlorine and then there is another bond with carbon, that the splitting, if I were to occupy that as well, of the orbitals is lower with chlorine than with oxygen. And therefore, if this energy gain, through the bond or the energy loss through the cleavage of the bond, is then lower with chlorine than with OH.”
Cameron provided multiple reasoning statements for his evidence, including statements in different reasoning categories. In the Energetics category, for example, he stated, that the energy loss or energy gain in the case of the cleavage of the C–Cl bond is energetically more favourable than the cleavage of the C–O bond. Furthermore, Cameron stated that the orbital of the chlorine atom involved in the reaction is higher in energy than the hydroxide orbital. In general, few students included orbitals in their reasoning. Cameron also mentioned the stabilisation of the chloride ion in water (Stability) and the electronegativity of chlorine (Energetics) as factors that make chloride a better leaving group than hydroxide. Gloria also cited the fact that chloride is a good leaving group as evidence.
Gloria: “Halogens are generally a good leaving group. That's why I suspect a nucleophilic substitution.”
Interviewer: “Can you tell me why halogens are good leaving groups?”
Gloria: “Because Cl - in particular is relatively stable and obeys the octet rule.”
Gloria made two reasoning statements. On one hand, she described chloride as a relatively stable ion (Stability). On the other hand, she mentioned the obeyed octet rule (Electronics) as a factor to justify her evidence. In contrast to Cameron, Gloria did not provide any further information on the stabilisation of the chloride ion, even when asked. In addition, she did not draw a comparison to the hydroxide ion, which competes with the chloride ion as a possible leaving group. When prompted, she could not provide any reasoning for why hydroxide is a worse leaving group compared to the chloride. This is in line with the analysis of Popova and Bretz (2018b), who found that students often have problems when asked to justify why a certain leaving group is good or bad. Directly comparing the statements of Cameron and Gloria regarding the same evidence indicates that both the number of reasoning statements and the depth of the reasoning statements are important for substantiating an argument.
The analysis revealed that students can be categorised into three argumentation approaches (reasoning-based, based-on-both, and evidence-based) based on the frequencies of evidence and reasoning statements and their ratio. There are different ways to form arguments, and it should be emphasised that there is no ideal argumentation approach. Instead, the numbers of evidence and reasoning statements were used in this study as an indicator of the potential quality of argumentation. A large number of evidence statements suggests that the student has recognised and included many different aspects in the decision-making process and considered various chemical variables influencing the reaction process. A large number of evidence statements can thus increase the quality and breadth of an argument since the course of a chemical reaction is influenced by a variety of factors. Recognising different influencing factors can also enhance the quality of the weighing process if the evidence is supported by reasoning. The number of reasoning statements is an indicator of how precise or elaborated the justification of evidence is. The more precise the justification, the greater the depth of the built argument. Proper justifying statements increase the quality of an argument by creating a stronger link between the components of the argument and the required conceptual understanding, which is a crucial aspect of the quality of an argument (Sandoval and Millwood, 2005; Choi et al., 2013). A well-grounded foundation of evidence and reasoning gives weight to an argument and should be an integral part of the classroom culture (Walker et al., 2012; Towns et al., 2019).
In addition to examining students’ use of evidence and reasoning, we assessed whether students changed their initial claims and their ability to rationalise these changes. Solving tasks that require the student to build multiple arguments can lead to more thorough decision-making, possibly enabling the student to justifiably defend a claim while also providing the opportunity to revise the claim. Based on the analysis of students’ argumentation patterns, 21 out of the 29 students in this study questioned their decisions and revised their claims during the argumentation process. Still, it seems that some students reached the limits of their conceptual understanding when rationalising their decisions; these students were unable to elaborate on their decisions beyond a certain individual point, even when specifically prompted. This issue was particularly noticeable for students who used the evidence-based argumentation approach, as these students were often unable to justify their claims. Luo et al. (2020) and Walker et al. (2019) found that students struggled to change their initial claims and thus stuck to their often erroneous claims. In the present study, the students were likely to question their initial claims after being specifically prompted to defend or revise them. Luo et al. (2020) analysed written arguments, whereas our study was based on interviews, which may lead students to be more willing to change their claim after prompting.
The findings of this study have some implications for teaching. In accordance with former studies, when introducing the building of arguments, teachers should ensure that students do not stop at supporting the claim with evidence, but also justify the evidence with reasoning (Bodé et al., 2019; Crandell et al., 2020). This part of argumentation is where students are likely to have the most problems; thus, it is important that teachers value justification through reasoning, even if it may be erroneous, to motivate students to build complete CER argumentation patterns. To help students learn to support and justify claims, argumentation patterns similar to concept maps can be used in teaching. As shown in Fig. 4 and Appendix 2, these argumentation structures can be used to create “argumentation trees” in the form of CER trees. This might help learners and teachers visualise both the depth and breadth of arguments and highlight gaps.
Furthermore, building arguments for alternative reaction pathways can be helpful as it helps students recognise that chemical reactions in the field of organic chemistry can lead to by-products (Popova and Bretz, 2018a) or may not form the desired product due to the choice of reaction conditions and reagents. In this study, we extended a common task format (i.e., comparing two reaction paths or reaction products) to the consideration of five alternative reaction pathways.
Students in this study were divided into three argumentation approaches based on the number of reasoning and evidence statements and their ratio. Based on the individual arguments formed by the students, the students need further support to develop strong arguments. Compared to the other two argumentation approaches, students who applied a reasoning-based argumentation approach provided a larger number of evidence and reasoning statements. To encourage this level of argumentation in future tasks, students should be supported in justifying their claims (e.g., by being trained to use several pieces of evidence for each claim and to justify each piece of evidence with several reasoning statements). In addition to the breadth of an argument, the depth of the argument is also of interest. In this regard, students could be provided with conceptual support to improve the quality of their evidence and reasoning statements. Additional support could focus on the weighing process through which students consider their formed evidence and reasoning statements. Students using the based-on-both argumentation approach supported their claims with multiple pieces of evidence and reasoning. However, the ratio of evidence to reasoning statements was close to 1, meaning that each evidence statement was justified on average with one reasoning statement. Thus, students in this group could receive targeted reasoning training to improve the breadth of their arguments and increase the number of reasoning statements for each piece of evidence. Students who used the evidence-based argumentation approach hardly provided any evidence or reasoning statements on their own and often failed to justify them, even after targeted prompting. On the one hand, this may be due to the fact that the students were conceptually unable to support their statements. Therefore, these students would benefit from additional conceptual support (e.g., providing information that they should include in their arguments). On the other hand, this group of students might benefit from examples highlighting the difference between evidence and reasoning statements or additional training on structuring an argument.
Scaffolds such as those used by Luo et al. (2020) and Caspari and Graulich (2019) to slow down students’ decision-making process (Caspari and Graulich, 2019) have proven to be helpful and could be adapted to help students build arguments for multiple alternative pathways. Students could be engaged to collect evidence as a first step and then be prompted to purposefully provide reasoning linked to the evidence. However, when developing evidence and reasoning to assess the plausibility of alternative reaction pathways, the quality of the evidence can vary, leading students to struggle in justifying their claims. Some students may profit from conceptual support in addition to support in building of arguments. An ongoing study is investigating whether students’ argumentation patterns in organic chemistry can be supported with an adaptive scaffold that targets the student's level of conceptual knowledge and argumentation skills (i.e., the use of evidence and reasoning).
Fig. 12 Example argumentation pattern (CER structure) for all product cards of the reasoning-based argumentation approach from Reuben. |
Fig. 13 Example argumentation pattern (CER structure) for all product cards of the based-on-both argumentation approach from Sydney. |
Fig. 14 Example argumentation patterns (CER structure) for all product cards of the evidence-based argumentation approach from Frank. |
Electronics (%) | Energetics (%) | Kinetics (%) | Spatial arrangement (%) | Analogies (%) | Strength (%) | Conditions (%) | |
---|---|---|---|---|---|---|---|
Approach 1 | 39 | 17 | 3 | 12 | 3 | 14 | 2 |
Approach 2 | 43 | 17 | 2 | 10 | 4 | 10 | 1 |
Approach 3 | 40 | 16 | 3 | 22 | 3 | 5 | 0 |
This journal is © The Royal Society of Chemistry 2022 |