Epistemic operations performed by high school students in an argumentation and decision-making context: Setrocia’s alimentary emergency

ABSTRACT

This study is framed in the social perspective of Epistemology of Science, and it aims to examine the epistemic operations performed by high school students while engaged in a role-play about food safety that requires them to engage in both argumentation and decision-making practices. The epistemic operations are examined on two different levels: 1) operations performed by each small group of students during their discussion as to how they intend to proceed in the general debate; and 2) operations performed in the general debate that prevail in students’ decisions for solving the alimentary emergency. The participants are 11^th grade high school students (16–17 years old). The data is examined through discourse analysis. The main findings suggest that the epistemic operations related to proposing knowledge are performed to a greater extent during the small group discussions, whereas those related to evaluating knowledge are more prevalent during the general debate. The main contribution of this work is the connections established between different areas of research, most notably the need to consider epistemic knowledge for the adequate engagement of students in the practice of argumentation as well as the use of role-playing in a socio-scientific setting to promote students’ argumentative discourse.

KEYWORDS:

• Epistemic practices
• argumentation
• high school
• discourse

Introduction

The purpose of this study is to examine the epistemic knowledge performed by high school students while undertaking a role-play activity about food safety.

Food safety is a global concern for consumers, as it affects both health and the economy. Moreover, issues related to food hygiene, such as the ways in which microbiological contamination can be controlled and the risk of contracting a foodborne illness can be reduced, are a concern for individuals.

The European Commission promotes several actions in order to assure food safety, and one of its main goals is to ensure a high level of protection of human health regarding the food industry (European Commission [EC], 2002). In addition, experts in this field consider food safety as everyone’s responsibility (Norton & Braden, 2007) and, as such, they emphasise the importande of consumer education. We translate this discourse to educational settings, considering food safety as a socio-scientific issue to be promoted in science lessons.

Socio-scientific issues involve a variety of social dilemmas with conceptual, procedural, or technological associations with science, which entail scientific content knowledge and that create social debate or controversy (Sadler & Zeidler, 2005). By addressing these issues in science lessons, it is possible to promote scientific literacy. According to Roberts and Bybee (2014), a scientifically literate person should be able to understand the impact of science and technology on everyday life, and as such, they should be able to make informed decisions about scientific-related issues, such as health.

This study adresses the concept of foodborne diseases as a socio-scientific issue, given that these diseases are still considered as a relevant public health problem. This is particularly significant in the Spanish context given the recent outbreak of listeriosis in the south of Spain, which was caused by manufactured meat, and which affected more than 200 people, resulting in three deaths.

The focus of our analysis is on the epistemic knowledge that is involved in argumentation in the form of epistemic practices, understood as the socially organised and interactionally accomplished ways that members of a group propose, communicate, evaluate, and legitimise knowledge claims (Kelly, 2008).

The practice of argumentation promotes students’ engagement in knowledge construction and evaluation practices, which require, for example, the use of criteria for the selection and evaluation of evidence, the provision of justifications and the construction of counter-arguments (Christodoulou & Osborne, 2014), the use of criteria for distinguishing between good and bad arguments (Zohar & Nemet, 2002) or the evaluation of the arguments constructed by others and their quality (Ryu & Sandoval, 2012).

According to Chinn et al. (2014), there is a need for investigating situated facets of epistemic cognition in authentic settings. In this study, the setting is a role-play about an alimentary emergency, which is occurring in a fictional country. By examining the epistemic practices performed in a socio-scientific argumentative environment, such the one used in this study, it is possible to obtain relevant information regarding the manner in which students engage in the characteristic practices of the scientific enterprise when dealing with a socio-scientific issue, as well as the challenges they face during the engagement. Although the results are not generalisable to other contexts, nonetheless they provide an insight into the practices and operations that entail more difficulties for students. It is also worth mentioning that this study, in which students are required to make informed decisions at different levels in order to solve this emergency, is in line with one of the new competencies proposed in the next PISA survey that will focus on science, called using scientific knowledge for decision-making and action (OECD, 2020).

The research questions that guide the investigation are:

1. Which epistemic operations are performed by the small groups of students during their conversations on how to proceed in order to solve the alimentary emergency?

2. Which epistemic operations prevail in the students’ decisions for solving the alimentary emergency?

Theoretical framework

The relevance of epistemic practices for a meaningful engagement in argumentation

Some recommendations from both science education research (Christodoulou & Osborne, 2014; Duschl & Grandy, 2013; Kelly, 2008) and recent policy documents (National Research Council [NRC], 2012) propose that science should be taught through students’ engagement in both practices and knowledge. The notion of practices moves away fron the idea of viewing science as a set of processes, emphasising instead the social interaction and discourse that accompany the building of scientific knowledge in classrooms (Reiser et al., 2012). However, there is evidence to suggest that successful learning using this approach does not depend solely on the practices, but also on the epistemic knowledge involved in them (Berland et al., 2016). In the context of argumentation, and in order to develop adequate practices, students must not only argue scientifically but also understand the epistemic demands of doing so. This means, for instance, to make a scientific claim and provide an argument for it and also knowing how to satisfy these demands through practices that meet specific epistemic criteria (Ryu & Sandoval, 2012).

Epistemic knowledge can be defined as a set of knowledge, practices and motivations related to what is considered as knowledge and to how knowledge claims are justified (Chinn et al., 2014). This type of knowledge is really relevant in science learning, since, according to Lin and Chang (2018), a limited epistemic understanding of science, causes students to have limited motivation to learn it, and, likewise, they experience difficulties in producing and improving evidence-based ideas, or in constructing a new idea based on existing ones.

Different aspects of epistemic knowledge such as beliefs (Sandoval, 2005; Yang et al., 2018), criteria (Sandoval & Çam, 2011) or practices (Jiménez-Aleixandre et al., 2008; Kelly & Takao, 2002) have been examined in science education literature.

Our study focuses on epistemic practices, understood as the socially organised and interactionally accomplished ways in which members of a group propose, communicate, evaluate, and legitimise knowledge claims (Kelly, 2008).

Epistemic practices draw from social epistemology, a discipline that examines how knowledge claims are negotiated among members of a group within particular contexts (Kelly et al., 2012). This perspective is framed in sociocultural theory (Vygotsky, 1979), which considers that learning takes place through social interaction and that knowledge is conceived as an interactive process between an individual and their physical, social and cultural environment. This conception was translated to science education by Kelly et al. (1993), who introduced a new research approach framed in sociocultural perspectives, in which classroom discourse is analysed using linguistics-based tools. This idea emerged from the study of knowledge constructing communities in which knowledge claims are developed and evaluated (Cunningham & Kelly, 2017). This perspective focuses on the ways in which social groups learn together, therefore engaging in epistemic practices involves making meaning among people through discourse (Kelly & Licona, 2018). In this frame, scientific knowledge is considered to be construed, validated and communicated through culture and scientific institutions (Driver et al., 1994). Kelly (2008) points out that the students’ access to science is accomplished through their engagement in the social and symbolic worlds comprising knowledge and practices of specialised communities. Kelly’s assertion is consistent with the main goal of learning through scientific practices, which intends to develop students’ learning in line with the way in which science is produced, negotiated, accepted and communicated (NRC, 2012). This practice-based approach requires students to adopt a new role as learners, by being epistemic agents (Stroupe, 2014), which means, for instance, that by learning science through engaging in planning and carrying out scientific investigations, students are required, not only to learn to do science, but also to understand and consider the specific criteria that must be met in scientific research in order for it to be considered reliable. Under this sociocultural perspective, this knowledge is constructed through action and discussion.

Kelly and Licona (2018) propose four defining features for epistemic practices: 1) interactional, constructed among people through concerted activity; 2) contextual, situated in social practices and cultural norms; 3) intertextual, communicated through a history of coherent discourses, signs and symbols; and 4) consequential, legitimised knowledge that instantiates power and culture. These authors suggest that there is not a limited set of epistemic practices, however they have categorised them into four general dimensions associated with cognitive processes: ways of proposing knowledge, communicating knowledge, evaluating knowledge, and legitimising knowledge.

This approach involves a new way of exploring students’ knowledge and beliefs by studying their engagement in epistemic practices related to the scientific activity. Examining learning in science through the lens of social epistemology provides opportunities for analysing how scientific knowledge is developed through action (practice) and how it changes through experiences (Leach & Scott, 2003). Because the social notion of epistemic practices can be translated to educational settings in different ways, we follow Cunningham and Kelly’s (2017) criterion, which consider that actions are epistemic if they are patterned and concern adjudication of knowledge claims.

Applying this criterion to argumentation, epistemic practices are related to actions that aim to produce, evaluate and communicate the criteria used by students when choosing one explanation over another or the role of evidence in these explanations (Christodoulou & Osborne, 2014). More specifically, the epistemic operations involved in socio-scientific argumentation are related to the aspects of understanding multiple perspectives for developing coherent reasoning and supporting different positions (Kelly & Licona, 2018).

In science education literature, several studies addressed the epistemic practices that are involved in argumentation. Regarding the use of evidence, Kelly and Takao (2002) examined the epistemic status of knowledge claims and developed a framework of epistemic levels for understanding students’ use of scientific theory and data. The findings supported the potential of this framework for examining participants’ ability to develop evidence-supported arguments concerning the theory of plate tectonics with real earth data. Sandoval and Millwood (2005) explored the quality of written arguments in the explanations of high school students about natural selection, analysing the students’ epistemic understanding of argumentation practices. Their findings highlighted the students’ limitations to cite sufficient evidence for claims and to articulate how specific data related to particular claims. Sandoval and Çam (2011), examined the epistemic criteria used by children when evaluating competing justifications for causal claims, identifying a loose ordering of justification preferences from evidence to plausible mechanisms to authority.

Focusing on the use of criteria, Ryu and Sandoval (2012) examined the potential of a instruction for improving children’s understanding and application of epistemic criteria for scientific arguments. Their findings showed that students are able to apply evidentiary criteria both when constructing and evaluating arguments.

Other research addressed how the practices of knowledge production, evaluation and communication are enacted in science lessons. Jiménez-Aleixandre et al. (2008) identified that the main practices in students’ discourse were related to articulating knowledge, identifying patterns in data, interpreting and constructing representations, developing reports and other texts, persuading peers, coordinating theories and evidence and contrasting claims with evidence. Christodoulou and Osborne (2014) examined the epistemic operations of secondary science teachers finding that the evaluative practices were not as present as the others in their discourse.

Our study is in line with this last perspective of examining the enacted epistemic practices. In particular, we analyse the epistemic operations associated with the practices of knowledge production, evaluation, communication and legitimation involved in a role-play game framed in socio-scientific decision-making.

The use of role-playing as a resource for promoting students’ engagement in epistemic practices

There has been considerable discussion regarding the most effective contexts for learning about argumentation and decision-making processes in a social context. Sadler (2004) suggests that these must require students to take a firm stand with respect to specific societal controversy, besides being relevant, authentic and openly debatable in the classroom. There is evidence to suggest that students show good argumentation skills when they are supported by a context that facilitates their argumentative discourse (Berland & Reiser, 2009; Cavagnetto, 2010), hence, role-playing can be a useful resource, given that it is described as a powerful tool for triggering the imagination of students, allowing them to participate in unfamiliar situations (Eilks et al., 2013). Recreations that take place during a role-play when portraying a specific event, time-period or personality are considered as an effective way of making students share different perspectives (Howes & Cruz, 2009). During role-playing practice, participants play a character in an improvised rather than scripted performance and under the direction of a guide. It involves a cooperative dialogue in which participants’ interventions mark the development and outcome of the game (Solís, 2012).

In science education, several studies use this resource in different contexts. Simmoneaux (2001) focused on the comparison of role-playing and classic classroom debate about animal transgenesis, describing a greater change in students’ opinions after role-playing than after a regular debate. Belova et al.’s (2013) evaluated students’ performances using different scenarios while teaching about climate change, highlighting the great capacity of role-playing to mimic societal debating and to help students make informed decisions. Additionally, studies that used role-play while teaching science and debating socioscientific issues reported improved argumentation skills (Agell et al., 2014; Squire & Jan, 2007).

In our study, we design and use a role-play for examining the epistemic practices that are performed by students in the context of an alimentary emergency when engaging in the practice of argumentation.

Methodology

The study is framed in qualitative research, which focuses on processes and meanings that are not experimentally examined or measured in terms of quantity, intensity or frequency (Denzin & Lincoln, 2000).

The methods draw from discourse analysis, an approach that considers language to represent the construction of social reality, especially within the social context of what is said, rather than assuming language only to represent what a person is thinking (Yin, 2011).

Study context

The examined task formed part of a larger study about argumentation and epistemic knowledge in the context of food safety that was conducted in a public high school located in the northwest of Spain with students from different grades.

In this paper, the participants are 14 11^th grade Biology and Geology students (16–17 years old), who worked in three small groups for two 50-minute sessions.

Description of the intervention

The task consisted of a role-play based on an alimentary emergency that arises in a fictional country. The students had to find out the cause of the emergency and stop it. This country is divided into different districts (1, 2 and 3), with fifteen resources that could have caused the emergency, such as the pharmaceutical industry, the water treatment plant and agriculture. Students had to decide wheter each of these should be activated or deactivated in order to stop the emergency. Activating a resource involves carrying out a certain action with it, while deactivation implies stopping all activity related to that resource.

The role-play rules that were explained to students were as follows:

• - The role-play is organised into several rounds. Each round involves a preliminary discussion in each small group about how to proceed, analysing data and developing the arguments that support their decision. After that, the small groups meet together to discuss their proposals in order to make a consensus decision about what to do with the resources in order to stop the emergency.

• - The activation and deactivation of resources has conditions: a maximum of two resources can be activated and one resource can be deactivated in each round.

• - In the case of deactivations, the final decision is to be made by the district that the specific resource belongs to.

• - The number of rounds is not defined, as this will depend on how long it takes for the students to solve the emergency problem.

The course of the role-play is summarised in Figure 1.

Figure 1. Different stages of the role-play.

The decisions made by students in each round partially determine the information they receive in order to continue with the task after reaching the final agreement. Some data, such as a chart specifying the food consumed in the days prior to the outbreak by those affected, is provided to students, regardless of their decisions. Other data is associated with the resources that they decided to activate and deactivate. Figure 2 represents an example of the information provided after activating the pharmaceutical industry and deactivating fishing.

Figure 2. Information generated by the students’ decision to activate the pharmaceutical industry and deactivate fishing.

Tools for data collection and analysis

Data collection includes audio and video taping of the lessons, as well as the researcher field notes. The first author adopted the role of the role-play conductor, because this task required familiarity with the game, and also conducted the observation of both sessions. The two sessions of the task were transcribed, coded and examined through discourse analysis. The teacher’s and students’ utterances were grouped by turns of speech with one number assigned to each intervention. Following this, the students’ epistemic operations were identified in the conversations. This analysis was carried out in interaction between literature and data. For the coding we drew from Kelly and Licona’s (2018) categorisation of epistemic practices, a set of four different dimensions associated with the cognitive processes that they entail: ways of proposing knowledge, communicating knowledge, evaluating knowledge and legitimising knowledge. Moreover, we used Christodoulou and Osborne’s (2014) distinction between epistemic practices and epistemic operations, in which epistemic operations are considered as discursive actions or talk moves whose function is to promote the creation and development of knowledge and understanding. Individuals that perform these operations engage in epistemic practices, considered as the sum of actions that contribute towards the same objective, for example, evaluating knowledge. In order to identify the epistemic operations performed by students in the context of argumentation we followed Jiménez-Aleixandre et al.’s (2008) categorisation, alongside other proposals such as Kelly and Licona’s (2018) and Crujeiras-Pérez’s (2014).

The final product consisted of a four-dimensional rubric representing the epistemic practices with several categories for each practice, representing the epistemic operations performed during the role-play. This rubric is reproduced in Table 1. Each epistemic operation is defined and described with examples.

Table 1. Definition and examples of the different epistemic operations identified in students’ talk during the task.

	EPISTEMIC OPERATION	DESCRIPTION	EXAMPLE (Students perform this operation when …)
PK	Inferring the scope of a decision	Students imagine a scenario that might occur once a decision has been made	… stating that the resource ‘pharmaceutical industry’ has the necessary equipment to analyse samples from those affected by the disease, which would result in more information
	Inferring a plausible cause	Students propose a certain cause for the outbreak or determine that it is impossible for a certain resource to be the cause	… proposing that the disease is generated by the incorrect operation of the resource ‘water treatment plant’
	Proposing explanation	Students provide an account of the implications that a certain circumstance may entail	… guessing that the progressive spread of the disease may be related to the fact that the districts share resources with each other
	Making sense of data	Students interpret or try to interpret the data that they possess	… examining the different categories in the chart showing the food consumed by those affected by the disease in the days prior to the outbreak
EK	Appealing to consistency with previous knowledge	Students use past experiences in their interventions or demand arguments that are consistent with the proven data	… stating that the following proposals have to be adapted to the fact that the disease is caused by food, therefore meaning that airborne transmission can be ignored
	Contrasting claims with available evidence	Students use the data provided to justify their claims	… saying that the fact that the disease originates in district 2 means that it is unlikely that the resource ‘agriculture’ was the cause, given that district 2 is the only district in which no vegetables are grown, therefore meaning that the disease would have appeared earlier in another district
	Acknowledging the absence of data	Students report a lack of sufficient data to be able to make a decision	… giving priority to obtaining the necessary information before making a precipitous decision
CK	Persuading other members	Students try to engage other members in their decision	… exerting pressure on the rest of the students to take action by emphasising the disease's high mortality rate
LK	Building consensus	Students try to reach consensus ir order to be able to make a decision	… trying to reach a common point when there is a disagreement between deactivating the resources ‘fishing’ and ‘water treatment plant’
	Recognising thevalue of other positions	Students recognise the usefulness of other interventions in arriving at a decision	… the district 2 students state that district 3's intervention is the most appropriate decision to resolve the emergency

Legend: PK: Proposing knowledge, EK: evaluating knowledge; CK: communicating knowledge; LK: legitimising knowledge.

Once developed, the rubric was applied to the data in order to obtain the results. This process was carried out in several cycles of analysis by the two authors until a 100% coincidence was obtained in the coding.

Ethical considerations

Both the students and their parents were previously informed of the goals of the study and how we intended to use the collected data. All of the students agreed to participate, and their parents signed an informed consent form. In order to protect their anonymity, all of the students were identified by pseudonyms.

Results

Analysis of the epistemic operations performed by students in the role-play in their conversations on how to proceed in order to solve the alimentary emergency

The general results in terms of the number of operations are summarised in Table 2.

Table 2. Epistemic operations performed by 11th grade students during discussions in small groups.

	Epistemic Operation	D1	D2	D3	Total
PK	Inferring the scope of a decision	13	6	9	28
	Inferring a plausible cause	13	20	20	53
	Proposing explanation	10	14	24	48
	Making sense of data	10	8	6	24
	N	46	48	59	153
EK	Appealing to consistency with previous knowledge	4	2	6	12
	Contrasting claims with available evidence	4	2	5	11
	Acknowledging the absence of data	1	-	1	2
	N	9	4	12	25
CK	Persuading other members	1	-	1	2
	N	1	0	1	2
LK	Building consensus	1	1	-	2
	N	1	1	0	2
NT		57	53	72	182

Legend: PK: Proposing knowledge, EK: evaluating knowledge; CK: communicating knowledge; LK: legitimising knowledge; D1, D2, and D3: District 1, 2 and 3 respectively.

In general, students performed more diverse operations associated with proposing and evaluating knowledge than communicating or legitimising it, accounting for four and three operations versus one, respectively. Considering the decision-making context and the fact that that the task was designed to demand the communication and legitimation of knowledge, it was surprising that a greater variety of epistemic operations belonging to these two practices were nor found. We attribute this to the students’ lack of familiarity with argumentative tasks, as well as to other factors that have already been reported in the literature, such as the difficulties of changing established discourse practices in the classroom setting.

The majority of the epistemic operations performed by all of the groups corresponded to the practice of proposing knowledge, and this accounted for 153 of the 182 total operations, therefore representing 84% of the total. The other practices were far less frequent, such as evaluating knowledge, which represented 13.8% or even anecdotal, such as communicating and legitimising knowledge, which represented 1.1% of the total each. In small group discussions, the fact that the students focused more on proposing knowledge was not entirely unexpected, given that in this part of the task they had to interpret the information provided to them and also present ideas before coming up with their proposal for the general debate.

In terms of the single operations, the most frequent ones were inferring a plausible cause and proposing explanation, which accounted for 53 and 48 instances respectively during the students’ discussions, and both of these operations correspond to the practice of proposing knowledge. The goal of the role-play game is to find the resource that is causing the emergency, therefore, to a certain extent it was expected that the most common epistemic operation would be inferring a plausible cause. Regarding evaluating knowledge, the most frequent operations were appealing to consistency with previous knowledge and contrasting claims with available evidence, which appeared in 12 and 11 instances, respectively. The fact that students focused more on these two epistemic operations indicates their engagement in the scientific practice of argumentation, which requires the connection of evidence with statements in order to develop justifications (Jiménez-Aleixandre & Erduran, 2007). It is also a good indicator that students demanded justifications that were consistent with the previously provided data. In a socio-scientific context, such as the one presented in this task, it is important that students are able to perceive the difference between good and bad arguments, given that the consequences of incorporating these arguments into decision-making processes can make an impact in the real world.

In terms of the small groups (districts) and regarding the operations associated with proposing knowledge, district 1 had a similar performance among these operations, whereas districts 2 and 3 focused more on inferring a plausible cause and proposing explanations.

An example of how participants used the operations associated with proposing knowledge can be appreciated in the following excerpt.

Table

Round 2, District 1. Extract 1: epistemic operations associated with the practice of proposing knowledge
Turn	Transcript	Epistemic operation
74	Rafael: No, we promote this [railway transport] because we have to carry medicines.	Inferring the scope of a decision
75	Ismael: But …
76	Carla: But then we would also have to remove this [vegetables], because in this [railway transport] you are transporting them as well.	Inferring a plausible cause Proposing explanation
77	Ismael: No, that’s not like that …
78	Carla: Yes, but …
79	Ismael: The thing that most coincides in all of them [people affected by the outbreak] is vegetables, so …	Making sense of data
80	Carla: Yes, we should remove vegetables and …
81	Ismael: Promote railway transport.

In extract 1, Rafael inferred the scope of a decision by communicating to the group the need for railway transport to be activated to transport medicines. However, the group rapidly refocused their attentions on finding the cause of the disease thanks to Carla’s intervention in turn 76 and Ismael’s interpretation of the data in turn 79. This contrast between finding solutions for the current situation in Setrocia and focusing on the cause was a regular occurrence in the students’ discussions and, as previously mentioned, the students focused more often on finding the cause.

With regards to the practice of evaluating knowledge, the operation acknowledging the absence of data was performed on a much less frequent basis than the other two operations in this category in all three districts. Recognising the absence of information is relevant to socio-scientific issues, because if the students do not acknowledge this issue while elaborating proposals to tackle a problem, it is difficult for them to focus their efforts on carrying out actions to obtain data and therefore their conclusions will remain on a more propositional level. An example of the use of these operations is summarised in the following excerpt.

Table

Round 2, District 3. Extract 2: epistemic operations associated with the practice of evaluating knowledge (other epistemic operations are omitted)
Turn	Transcript	Epistemic operation
110	Sara: I mean, before you paralyse something, you must analyse it.	Acknowledging the absence of data
111	Abel: We need to analyse it.
117	[…] Sara: Well, I think it’s the water, but …
118	Aníbal: It is.
119	Sara: It doesn’t say anything about the water [in the chart].
121	[…] Nerea: I don’t think it’s the water though, because if … I mean, if it is supposed to be infected, everyone was drinking water from the same … there, that is, from the same area.
122	Aníbal: But they [district 2] own the water treatment plant.
123	Nerea: But they [district 2] don’t have agriculture. By the same token they were using the other day, that they were infected by fish because they were the ones … that had fishing, then if it is by agriculture, districts one and three would have to be infected, and not district two.	Appealing to consistency with previous knowledge

Extract 2 reproduces one of the very few examples of the operation acknowledging the absence of data. This can be appreciated in Sara’s intervention in turn 110, in which she expressed that before deactivating any resource, it would be necessary to carry out an analysis. On the other hand, Nerea performed the operation appealing to consistency with previous knowledge in her reference to the previous session, in which students agreed that it would be more likely for the resource causing the illness to affect the district where it is located first.

Regarding the practice of communicating knowledge, only districts 1 and 3 performed the operation associated with this practice. It was not expected that the operation persuading other members would be present in the groups in which the members agreed on how to proceed, however this was expected in the groups in which consensus was not reached. The task promotes this by setting a limit on the resources that can be activated and deactivated. This also draws a parallel with socio-scientific issues, where the approach to address them is often limited by economic or ethical factors, for instance. The following excerpt summarises how students used this operation.

Table

Round 2, District 3. Extract 3: epistemic operations associated with the practice of communicating knowledge (other epistemic operations are omitted)
Turn	Transcript	Epistemic operation
196	Aníbal: You can, you can. And paralyse, what do we paralyse?
197	Sara: We don’t paralyse anything.
198	Aníbal: What do you mean, no? Esther, something has to be done.
199	Esther: People are dying …	Persuading other members
200	Aníbal: We suggest for the second time …
201	Abel: That people don’t eat.
202	Aníbal: … to paralyse the drinking water plant

In extract 3, Esther performed the epistemic operation persuading other members after Sara stated in turn 197 that they were not going to deactivate any resources, something that other members of the group did not agree with and which had a significant impact on the decision-making process.

Something similar occurred with knowledge legitimation, appearing only in two districts (1 and 2). The operation of building consensus appeared in district 1's discussion, when one of its members checked to see whether the rest of the group agreed with the idea of proposing the deactivation of the water treatment plant in the general debate. The same person consulted the justifications for that proposal with the rest of the group.

The following section analyses the epistemic operations performed during the general debate in order to make decisions that enable students to solve the emergency.

Analysis of the epistemic operations performed by students in the role-play during the general debate while making a final decision

In this section we focused on the general debate, in which the districts share their proposals in order to reach a joint decision, presented as the activation and deactivation of resources. Each group needed to present their proposals in order to make the most appropriate decision.

For this research question, the analysis focused on examining how the epistemic operations that were performed during the general debate contributed to the final decision in each round. These operations were classified according to their contribution to the deactivation or activation of a certain resource. It must be noted that not all epistemic operations considered in the framework appeared for every decision. In fact, the operation of persuading other members (communicating knowledge) did not appear in the general debate, which was rather unexpected. However, its absence could be due to the nature of the students’ performances, as they may have persuaded their colleagues in an implicit way, for example, by presenting an argument to challenge another group. Another reason could be related to the fact that in several cases the students' proposals for the activation and deactivation of resources were similar.

In rounds 1 and 2 the analysis was more complex due to the fact that a more diverse range of resources were considered and given the high number of epistemic operations performed. For this reason, epistemic operations that contributed to each decision of activating and deactivating resources are presented separately. The analysis for each round is presented below.

In round 1, four possible decisions were considered during the debate. Regarding the activation of the pharmaceutical industry, three epistemic operations of inferring the scope of a decision (proposing knowledge) were performed between districts 1 and 3, who were interested in activating this resource because of its potential to obtain more information, perform tests and manufacture medicines. This decision illustrates an example in which there was no disagreement between the groups and in which the scenario described through the operation inferring the scope of a decision was sufficient in order for the three groups to agree on activating this resource.

Regarding telecommunications, two epistemic operations of inferring the scope of a decision were performed by district 3, which were aimed at activating it, mentioning its potential to make people aware of the outbreak. District 2 proposed an explanation as to how that information can cause undesired panic in the population, thus making a rebuttal. District 3 provided another explanation stating that it would be worse for the population to receive unreliable information, contributing again to the activation of telecommunications. It is worth mentioning that students did not receive any data that would allow them to clearly decide whether to activate or deactivate telecommunications, so it was expected that epistemic operations that were identified in students’ discourse and that were associated with this resource corresponded to proposing knowledge.

Another decision was whether to deactivate fishing or the water treatment plant. An example of the operations performed is summarised in Figure 3. This figure brings a new dimension to the analysis in the sense that it allows us to monitor the appearance of epistemic operations in relation to the moment in which they occurred (through the visualisation of turns of speech) along with the content of the discourse (contribution to the decision). Therefore, it is possible to observe the evolution of epistemic practices in the course of the debate and also examine how the stance of a certain group regarding a decision is able to influence the sort of epistemic operations found.

Figure 3. Evolution of the epistemic operations involved in deciding which resource to deactivate in round 1. Legend:

: operations performed by district 1,

: operations performed by district 2,

: operations performed by district 3,

: proposing knowledge,

: evaluating knowledge.

Figure 3 describes a clear confrontation between district 2 (aiming at the deactivation of fishing) and district 3 (aiming at the deactivation of the water treatment plant). Both districts mainly performed evaluating knowledge operations, both for adding to their own proposal and for rebutting the other. Epistemic operations belonging to the practice of evaluating knowledge appeared from the first disagreement, in district 2's intervention in turn 258. District 1 focused more on the practice of proposing knowledge and did not express a clear preference for either one of the decisions.

In round 2, four new decisions were considered and the course of the debate for the two possible activations that were discussed during this round is described in Figure 4.

Figure 4. Evolution of the epistemic operations involved in deciding which resources to activate in round 2. Legend:

operations performed by district 1,

: operations performed by district 2,

: operations performed by district 3,

: proposing knowledge,

: evaluating knowledge, ..... : legitimising knowledge.

Regarding the activation of railway transport and telecommunications, no disagreement was observed between districts, and once again almost all of the operations corresponded to proposing knowledge, although district 2 did recognise the value of activating railway transport, whereas district 3 appealed to consistency with previous knowledge in relation to activating telecommunications.

Regarding the deactivation of the water treatment plant and agriculture, districts 1 and 2 began by performing three operations of proposing knowledge while they remain undecided about the cause. On the contrary, district 3 started the intervention by inferring that the water treatment plant was the only plausible resource that could be causing the outbreak. They then performed two operations of contrasting claim with available evidence (evaluating knowledge), that constituted a rebuttal for deactivating agriculture, using the data that the disease arises in district 2 (the only district without that resource) and appealing to consistency with previous knowledge, stating that even if resources are shared between districts it is more likely for a resource to be used first in the district in which it originates. District 1 performed two more operations of proposing knowledge and subsequently districts 1 and 2 performed two epistemic operations of recognising the value of other positions, in this case district 3’s. However, district 1 and 2 both expressed awareness of the bad consequences that could be provoked through the deactivation of the water treatment plant, thus performing the operation inferring the scope of a decision. It is worth mentioning that the two operations of recognising the value of other positions were directed to the only district that performed operations that belonged to the practice of evaluating knowledge. And also, that in interventions with no clear inclination towards a decision, the epistemic operations related to proposing knowledge predominated, which suggests that these operations are suitable for the study of decision-making processes, such as those that occur in socioscientific issues.

In round 3, the final round, students discussed the deactivation of the petrochemical industry versus the deactivation of the food industry. The round concluded with the deactivation of the food industry, which ended up solving the alimentary emergency. In this round, no epistemic operations related to communicating and legitimising knowledge were found, which may be due to the fact that the students had similar views on this decision. Likewise, only three distinct operations that are classified into the practices of proposing and evaluating knowledge appeared. Figure 5 summarises the course of the debate that led to the final decision.

Figure 5. Evolution of the epistemic operations involved in the final decision (round 3). Legend:

operations performed by district 1,

: operations performed by district 2,

: operations performed by district 3,

: proposing knowledge,

: evaluating knowledge.

The contributions from districts 1 and 3 were aligned with the final decision (deactivating the food industry) whereas district 2 explored a different idea (deactivating the petrochemical industry) although these resources are closely related. The epistemic operations that were performed during this round corresponded mostly to the epistemic practice of proposing knowledge except for the operation of appealing to consistency with previous knowledge that was performed by district 3.

Discussion and conclusions

This paper examines the epistemic operations that were involved in the socio-scientific argumentation carried out by 11^th grade students during a role-play about food safety.

With regards to the epistemic operations performed by small groups, those related to the practice of proposing knowledge clearly dominated the students’ discourse. This finding is in line with Christodoulou’s (2012) study. However, another study about epistemic practices in inquiry settings (Crujeiras-Pérez, 2014) identified evaluating knowledge as the most frequent practice. These similarities and differences can be associated with the different nature of the tasks, that require different epistemic operations, given that planning and carrying out an experiment is not the same as making a decision to solve a socio-scientific issue. Another factor that can influence the dominance of the epistemic practice of proposing knowledge is the students’ creativity. Role-playing is regarded as a tool for triggering imagination (Eilks et al., 2013) so it may be anticipated that students would imagine various scenarios in their heads and share those proposals with their peers.

The fact that the most commonly performed epistemic operation was ‘inferring a plausible cause’ suggests that more importance was placed on finding the cause of the alimentary emergency than on adressing its consequences. This is in line with Crujeiras-Pérez and Jiménez-Aleixandre (2015) study, in which participants dedicated more time to identify the cause of browning in apples than in proposing a way for preventing it.

A surprising finding was that very few operations related to communicating and legitimising knowledge were performed, and this was not what was expected from a task in which discussion between groups is promoted. We associate this finding with aspects related to the design of the task and the content knowledge involved. With regards to the design of the task, the necessity to reach consensus between the members of the small group before bringing their proposal to the general debate was mentioned in the role-play rules, thereferore meaning that although the students may not have expressed this verbally they worked towards this objective through their interventions during the small group discussions. With regards to the content knowledge, which was unfamiliar to students, perhaps it was difficult for them to recognise the ideas that would be more useful to bring to the general debate, and this could explain why the epistemic operation ‘recognising the value of other positions’ (legitimising knowledge) was not identified during the small group discussions. The limited presence of the operation ‘persuading other members’ may be the result of the students’ lack of experience within the area of food safety. If students are not sure of something, they are unlikely to want to pressurise their peers to go in a certain direction. Our study was set in a fictional context with which the students were not very familiar with, and therefore we anticipate that a greater number of epistemic operations will likely appear in more familiar contexts, such as a food emergency that arises in a school cafeteria.

With regards to the epistemic operations that were performed during the general debate for making the final decision, those related to the practice of evaluating knowledge were more frequent in comparison with the number of epistemic operations performed in small discussions. In the cases in which there were disagreements as to how to proceed (for example, during the discussion about the possible deactivations in round 1), ‘appealing to consistency of previous knowledge’ and ‘contrasting claim with available evidence’ were the most frequently performed operations. However, in the discussions in which the students were largerly in agreement, the predominance of proposing knowledge operations was still observed.

In terms of the evolution of the debate, the epistemic operations that were performed decreased as the game progressed. In Figures 3 and 5 (both representing the discussion about the deactivation of two resources in rounds 1 and 3 respectively) we can observe that the number of epistemic operations decreased from fourteen in round 1 to nine in round 3. The same pattern was observed for the total number of operations performed in each round. The variety of the epistemic operations also decreased, therefore meaning that the least diverse epistemic operations were performed in round 3. This tendency went against the desired outcomes, given that it was anticipated that once the students are familiar with the dynamics of the game and with the practice of argumentation, more instances of operations are expected to appear, especially those related to the practice of evaluating knowledge. We anticipate that this tendency was due to the students’ lack of experience with the scientific practice of argumentation as well as the fact that the time allocated for discussions decreased over the course of the game to mimic the growing sense of urgency for the resolution of the alimentary emergency, as the situation in the fictional country becomes more concerning.

Our findings suggest that by examining students’ epistemic operations performed in an argument-based role-play situation it is possible to get a real sense of how students manage with this scientific practice. In this case, the operations concerning proposing knowledge can be associated to those related to producing arguments, such as drawing a conclusion, using evidence and justifying claims. The same relationship can be established with the operations of evaluating knowledge.

Addressing a socio-scientific issue by using a role-play situation that involves a final decision-making process is a good resource for promoting collaborative argumentation. This is in line with Evagourou and Osborne (2013), who indicated the need of framing the tasks in order for students to be able to understand them as a whole.

Engaging students in learning environments that enable them to perform epistemic operations associated with argumentation has proved beneficial for the students’ learning. However, it is important that they are provided with more opportunities to engage in these environments and to learn how to participate in effective discussions that lead to well-informed decisions, which would mean to become skilled argumentatives (Walton, 1989).

To conclude, the main contribution of this work emerges from the many connections that have been taken into consideration following the review of the literature. Firstly, the need to consider the importance of epistemic knowledge for the students' adequate engagement in scientific practices, in this case argumentation, as well as the ways in which epistemic knowledge can be effectively examined in the form of epistemic practices. And secondly, the use of role-plays to support argumentative discourse, which is directly applicable to socio-scientific issues, such as food safety.

The detailed analysis that has been presented in this paper is useful for identifying the epistemic operations that are relevant for an adequate performance of argumentation practices in a role-play game. Our analysis has shown that it is important to promote these epistemic operations from the teaching practice. This can be achieved by designing activities to promote these operations or by formulating concrete questions that will prompt students to perform these operations.

Acknowledgements

The authors would like to thank the students and teachers who participated in the study.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This study was funded by the FEDER/Ministry of Science, Innovation and Universities-National Agency of Research/Project EDU-2017-82915R.