Developing deep learning in science classrooms: Tactics to manage epistemic uncertainty during whole-class discussion
Abstract
Science teachers usually view students' uncertainty as a barrier to overcome, a negative experience to be avoided, a deficiency in need of remedy. Building on the theory of deep learning in science as a generative and sensemaking process, the purpose of this design-based study is to identify tactics for teachers to manage their students' epistemic uncertainty as a pedagogical resource to develop student conceptual understanding during whole-class discussion. Classroom observations of whole-class discussion were collected from six teachers' classes ranging from third to eighth grade. A total of 18 whole-class discussions were collected, transcribed, and analyzed. A storyline talk to manage uncertainty during whole-class discussion was developed and consisted of three stages: (1) Raise epistemic uncertainty through creating ambiguous conditions; (2) Maintain epistemic uncertainty through preventing immature disclosure and discussing alternative explanations or conflicting ideas; and (3) Reduce epistemic uncertainty through making coherent connections among current uncertainty, prior knowledge, and familiar phenomena. Seven nuanced tactics used by teachers to achieve each stage of uncertainty management were identified. The results suggest that managing uncertainty goes beyond asking questions and problematizing phenomena. When engaging students in storyline-based whole-class discussion, teachers should focus on one specific uncertainty and establish a coherent, consistent storyline that raises, maintains, and reduces student uncertainty to horizontally and vertically construct a collective knowledge among students. The horizontal nature occurs within a stage of management, and the vertical nature of a storyline talk is related to moving along from stage to stage. Through the storyline talk focusing on students' epistemic uncertainty, students can truly become agents in the learning process when the lesson is centered on and driven by students' uncertainty.
1 INTRODUCTION
Scientific literacy is the desirable outcome of K-12 science education (Next Generation Science Standards [NGSS], 2013; Organisation for Economic Co-operation and Development [OECD], 2019). Pearson et al. (2010) recommended that one productive way of implementing literacy practices and advancing students' scientific literacy in K-12 classrooms is to create an environment for students to learn “how to think and practice like a scientist” (p. 459). The environment should be scientifically authentic enough for students to collect, analyze, interpret, and debate data as claim and evidence that lead to knowledge development.
Uncertainty is inherent in science and it helps science advance (Kampourakis & McCain, 2019; Oreskes, 2015; Star, 1985). Firestein (2015) stated, “For scientists, uncertainty is a norm. Experiments begin with uncertainty (why else do them?), and even when they are ‘successful’, the results contain only a range of certainty and a range of confidence about that certainty” (p. 638). He contended, “the very notions of incompleteness or uncertainty should be taken as the herald of science” (Firestein, 2012, p. 44). By reviewing the role of uncertainty in different fields of science (e.g., climatology, immunology, genetics), Kampourakis and McCain (2019) concluded that “it is uncertainty that drives this continual search for evidence” (p. 215). That is, uncertainty “leads to more research, which in turn serves the ultimate goal of science—understanding” (Kampourakis & McCain, 2019, p. x). Therefore, uncertainty is ubiquitous and fundamental to scientific practice, from framing a question to deriving a method of exploration, collecting and analyzing data, and presenting one's findings to the professional community and the public.
Incorporating the practice of managing uncertainty becomes a critical goal for preparing students to be scientifically literate (Britt et al., 2014; Kienhues et al., 2020). When students are provided opportunities to identify and resolve uncertainty in their own investigations, they learn how scientists achieve this practice, and they know how scientific knowledge evolves and develops. Incorporating the intentional practice of managing uncertainty in science classrooms is, however, challenging for both teachers and students because they are usually not familiar with the intentions and activities of scientists. Teachers and students frequently position scientific inquiry as facts and laws (Lemke, 1990). Rarely do teachers view uncertainty as a pedagogical resource for science teaching and scientific literacy.
Teachers often become anxious about the uncertainty of open-ended problems, causing them to purposefully avoid uncertainty by shutting down such problems immediately (Donnelly, 1999; Doyle, 1988; Kampourakis, 2018). Kampourakis (2018) contended, “Both teachers and students love the science of certainty that provides definitive answers and dislike the science of uncertainty where things can go wrong” (p. 830). Chen et al. (2019) found that when science teachers managed uncertainty in an argumentative environment, they tended to reduce or avoid student uncertainty during discussion and treated it as a threat to learning science. Lee et al. (2020) explored how South Korean ninth graders discussed a socio-scientific issue about genetically modified organisms. They found that both teachers and students exhibited a lack of tolerance of uncertainty caused by multiple perspectives and different interpretations. They found that students “often expressed uncomfortable feelings and resistance,” and “hastily attempted to reduce uncertainty” (p. 682).
Even if teachers attempt to integrate student uncertainty in their lessons and activities, they usually use it only as an “opening” to raise student interest in the topic. From there, teachers move quickly to provide solutions and answers, prematurely removing the uncertainty from the lesson (Reeve, 2009). Thus, teachers and students struggle to maintain, reduce, and resolve uncertainty in ways that are engaging and that contribute to conceptual development (Chen et al., 2019). As such, students have limited opportunities to engage in deep learning in which they can raise, maintain, and reduce uncertainty to generate a coherent knowledge system by integrating new information with existing knowledge. Beghetto (2013) argued that within this shallow and structured learning environment, student uncertainties are “killed softly,” not productively used.
One productive avenue to manage student uncertainty is to leverage it as a pedagogical resource to scaffold learning in the science classroom (Beghetto, 2017). As a pedagogical resource, teachers can use uncertainty to raise students' curiosity about a problematized phenomenon, and lead students to recognize what they know, what they do not know, and what they want to explore. Teachers also can maintain and further increase student uncertainty by fostering engagement in what Michaels and O'Connor (2015) consider “productive talk,” discursive opportunities to “dig deeper into their own reasoning with evidence and models, and to build on and critique the reasoning of others” (p. 334). As Kapur (2008) suggested, maintaining student uncertainty may situate students in the experience of “failure.” However, “to the extent that students are able to use their prior knowledge to generate suboptimal or even incorrect solutions to the problem, the process can be productive in preparing them to learn better from the subsequent instruction that follows” (Kapur, 2016, pp. 289–290). Eventually, teachers help students find solutions and answers to reduce their uncertainty, and that leads to establishing robust conceptual understanding. Student uncertainty is productively reduced not through giving them answers, but through scaffolding them to actively seek solutions by integrating “new information with existing relevant concepts and propositions in their cognitive structure” (Novak, 2002, pp. 557–558). The new knowledge is meaningful to students, rather than fragmentary and separated from their existing knowledge systems. In sum, the raising, maintaining, and reducing of uncertainty can engage the student in deep learning of science.
To date, how uncertainty is approached and managed in learning environments has not been adequately researched (Chen, 2020; DeBoer, 2000; Hall, 1999; Tiberghien et al., 2014). The problem is the lack of understanding of the uncertainty inherent in science and of tactics1 to effectively manage uncertainty in the classroom. This study explores productive tactics that can be used to manage students' epistemic uncertainty as a pedagogical resource to develop conceptual understanding. Epistemic uncertainty refers to students' subjective awareness of their incomplete abilities to explain a phenomenon, interpret and present raw data as evidence, and understand what counts as scientific argument (Chen & Qiao, 2020; Hartner-Tiefenthaler et al., 2018; Tiberghien et al., 2014).
- In what ways do science teachers manage student epistemic uncertainty as a pedagogical resource to develop student conceptual understanding during whole-class discussion?
2 THEORETICAL BACKGROUND
This study is grounded in the notion that the use of students' epistemic uncertainty promotes deep learning in the science classroom. Therefore, we discuss the definition of deep learning, the role of uncertainty in deep learning, and its application in whole-class discussion. Next, we discuss why we focus on epistemic uncertainty for deep learning and source of epistemic uncertainty identified in the literature. Finally, a storyline talk approach is proposed to accomplish the productive use of epistemic uncertainty to horizontally and vertically contribute to students' coherent knowledge acquisition and processing.
2.1 Deep learning in the science classroom
Deep learning is a generative process that involves students actively making sense of new information by (1) activating existing knowledge that is related to the new information, (2) identifying the gap between existing knowledge and new information, and (3) integrating new information with existing knowledge into a coherent system (Fiorella et al., 2020; Wittrock, 2010). Deep learning is generative because students actively reshape their own knowledge by linking new information to existing schema (Fiorella & Mayer, 2016). Deep learning is a sensemaking process because students meaningfully construct explanations to resolve an identified gap or conflict in their existing knowledge and experience (Odden & Russ, 2019). The productivity of deep learning depends on the ways and processes that students synthesize new information and existing knowledge (Novak, 2002). If students do not activate relevant existing knowledge, identify the gap, and reshape a coherent knowledge system, learning may occur at a surface level, such as rote learning or verbatim memorization (Chi et al., 2018). They may keep new information isolated from existing knowledge (Vosniadou, 1994). As a result, new information does not make sense to them.
Deep learning in the science classroom means that students can identify a problem related to an encountered phenomenon, recognize the gap between what they know and what they do not know, and seek solutions to solve the problem in order to understand the phenomena, restructuring their existing knowledge (Chin & Brown, 2000). For this to happen, teachers must create an environment in which students can experience this process (Greeno & Gresalfi, 2008; Hand et al., 2020). For example, when students learn Newton's first law of motion (inertia) in a physics class, the teacher can ask students to talk about what happens when they are standing in a bus and it starts to move. Students may respond that their body leans backward. The teacher can ask why this happens. Such a problematized phenomenon activates students' existing knowledge related to the concept, as well as identifies the gap between what students know and do not know (Suárez, 2020). Loibl and Rummel (2014) suggested this leads to deep learning in the science classroom. Teachers can use this gap to motivate students to seek solutions and design experiments for solving the identified problem (Wickman & Östman, 2002) and restructuring of the existing knowledge system (Sharot & Sunstein, 2020).
2.2 The role of uncertainty in deep learning
Lamnina and Chase (2019) argued that this gap, identified, noticed, or highlighted by teachers or students, causes students' uncertainty, and that this motivates them to “seek information to fill knowledge gaps,” and “reduce uncertainty and regain coherence in their thoughts and understanding” (p. 2). At an individual cognitive level, uncertainty is the subjective awareness of wondering about or the lack of knowledge to make sense of a phenomenon (Jordan & McDaniel, 2014). Several education scholars suggest that uncertainty is a necessary experience to productively develop knowledge in academic settings, where students experience a problem (Engle & Conant, 2002), failure (Kapur, 2008, 2016), confusion (D'Mello et al., 2014), curiosity (Murayama et al., 2019), impasse (Munzar et al., 2021), or struggle (Fries et al., 2020). Uncertainty thus situates students in a cognitive state of what Piaget (1972) called “disequilibrium,” in which an individual recognizes the gap or conflict between what they already know and what they encounter. This experience of uncertainty positions students to learn new skills, knowledge, or solutions to resolve uncertainty. After this, students develop more sophisticated knowledge and skills that equip them to cope with more advanced and complex levels of uncertainty. Therefore, uncertainty is endemic to deep learning (Bradac, 2001).
The relationship between uncertainty and deep learning can be traced to John Dewey's (1933) notion of reflective thinking. Dewey suggested that reflection is a transforming process, starting with discomfort and uncertainty, then leading to a balanced and stable state. “The function of reflective thought is, therefore, to transform a situation in which there is experienced obscurity, doubt, conflict, disturbance of some sort into a situation that is clear, coherent, settled, harmonious” (p. 100). That is, uncertainty is not only central to deep learning, but also is an essential element driving the process of reflective thinking of learners (Jordan et al., 2012).
Recognizing the important role of uncertainty in deep learning, several science educators have designed and used tactics incorporating cognitive conflicts or discrepant events to elicit student awareness of the gap between their existing knowledge and the encountered situations they must explain (e.g., Chiu et al., 2002; Harrison & Treagust, 2001; Lombardi et al., 2016; Vosniadou, 1994). Approaches proposed by this line of research, aligning with the theory of conceptual change (Posner et al., 1982), emphasized the use of the cognitive approach to raise students' uncertainty and lead to Dewey's reflective thinking. Although incorporating cognitive conflict in a lesson generates opportunities for conceptual development and growth, productive social environments are needed to maximize student learning outcomes (e.g., Alexander, 2007; Chi et al., 2018; Roschelle, 1992).
2.3 Merging uncertainty of individual cognitive and sociocultural levels: Uncertainty during whole-class discussion
Consistent with the theory of social constructivism, in which knowledge is co-constructed with others and developed through social negotiation (e.g., Baker, 2009), students do collaborate with each other. Through comprehending, constructing, and critiquing each other's ideas they establish negotiated meanings (Wickman & Östman, 2002) and a mutual agreement and collective understanding (Chen et al., 2016; Forman et al., 2017). In such a learning environment, uncertainty reaches a sociocultural level, encompassing controversy, conflict, or difference in interpretation of a phenomenon, raw data, or target issue (Harker, 2015; Leitão, 2000). Kirch and Siry (2012) argued that uncertainty, at a sociocultural level, “originates and exists in dialog and is a product of interaction with others and the world” (p. 263). Radinsky (2008) suggested that “managing this uncertainty is accomplished through discourse in which meanings and goals are co-constructed, contested, and negotiated, as a range of communicative resources are brought to bear in assembling a representational state for reasoning about the world” (p. 147). That is, teachers should create a learning environment in which students have opportunities to identify or be attentive to their uncertainty at the individual cognitive level, and then express and negotiate, through social dialogue, to generate solutions to their uncertainty.
Whole-class discussion provides a discursive environment for students to socially express, articulate, and negotiate their individual uncertainty. For example, Kirch and Siry (2012) explored how second-graders negotiated, during whole-class discussion, their uncertainty about the life of mealworms. They found that the students came to clearly know what they were uncertain about, to reflect on the source of the uncertainty, and to know how they could resolve and reduce it. In sum, whole-class discussion provides an opportunity for students to externalize their more or less vague individual uncertainty (Chen, 2020). When their peers require clarification, students have the opportunity to articulate any uncertainty they might have, reflect on what causes them to be uncertain, and find solutions to solve the uncertainty.
2.4 The sources of uncertainty when developing knowledge: Epistemic uncertainty
The first step to productively manage uncertainty is to identify its potential sources (Ford & Forman, 2015; Kirch, 2012). Sources may differ with type of uncertainty, and several types have been identified in learning science, such as relational (i.e., interactional challenge or lack of confidence with interpersonal relationships), ontological (i.e., doubt or lack of confidence with regard to disciplinary knowledge toward discussed topics), and epistemic (i.e., doubt or lack of confidence or knowledge about how to explain a phenomenon). Because this study focuses on knowledge generation rather than knowledge replication, we focus on and define epistemic uncertainty as: student subjective awareness of their incomplete skills or abilities for how to explain a phenomenon, to sort through data to produce trends, to interpret and represent raw data as evidence, and to generate potential solutions to solve problems (Chen & Qiao, 2020; Hartner-Tiefenthaler et al., 2018).
Several researchers in science education have identified significant sources of epistemic uncertainty. By studying second-, fourth-, and fifth-grade students' utterances of uncertainty in scientific inquiry, Metz (2004) identified five major sources of epistemic uncertainty: when attempting to produce the desired conclusion, when considering the quality of data, when identifying trends in the data, when generalizing an explanation from the data, and when attempting to align explanation to theory. Similarly, Lee and colleagues (Buck et al., 2014; Lee et al., 2014; Lee et al., 2019) conducted four survey-based studies to investigate high school student uncertainty toward argumentative science tasks. They showed that epistemic uncertainty resulted from challenges in coordinating theory and evidence, as well as using evidence to support claims. They also found that students that had a higher degree of uncertainty and were able to articulate it tended to engage more deeply in science tasks.
2.5 Storyline talk to horizontally and vertically construct a coherent knowledge system
Many scholars have advocated constructing a storyline-based curriculum, lesson, activity, and discussion to engage students in deep learning of science concepts and practices (e.g., Hanuscin et al., 2016; Haverly et al., 2020; McDonald & Kelly, 2007; Roth et al., 2011). A storyline can be defined as a conceptual sequence of learning activities in a coherent manner, in which each step is driven by students' ideas and questions that arise from their interaction with teachers, peers, and materials (Hanuscin et al., 2016). Roth et al. (2011) advocated that teachers should weave students' ideas to construct a coherent “storyline” rooted on the sequence of discussed concepts. They analyzed TIMSS (Trends in International Mathematics and Science Study) results from five countries and found that teachers in high-achieving countries tended to build a coherent storyline “with clear and explicit connections made between the opening focus question, the science ideas, the activities, the follow-up discussions of activities, and the lesson ending” (p. 120). Teachers in low-achieving countries tended to poorly organize students' ideas, science content, discussion, and activities to create a storyline.
Building on the literature, we applied the concept of storyline to the whole class discussion, called storyline talk, as the conceptual and methodological framework to guide data analysis and interpretation. Storyline talk, in this study, is defined as a coherently sequential discussion in which each stage is driven by student uncertainty arising from an encounter, phenomenon, or issue, and ending when this uncertainty is reduced. At each step, students progressively develop their understanding of uncertainty as they interpret raw data to shape evidence, propose hypothetical claims supported by evidence, and negotiate ideas with peers and teacher. Students may intensify or increase their uncertainty at each step that leads them to deep reasoning. Together, storyline talk helps students deal with their uncertainty through sequential pathways to build a conceptually coherent understanding of a problematized phenomenon.
Storyline talk goes beyond informal assessment (Furtak et al., 2016), responsive teaching (Robertson et al., 2016), and notice teaching (Chan et al., 2021; Jacobs et al., 2010; Luna, 2018) that often underpin teachers' in-the-moment decisions and actions related to elicitation, interpretation, and response to student ideas during discussion. Implementing a productive storyline talk requires teachers to adapt the skills mentioned above (e.g., informal assessment) and to manage students' uncertainty, horizontally and vertically. The horizontal nature of a storyline talk refers to systematically connecting, comparing, and synthesizing the differences and similarities of student uncertainties to shape a consensus or a conclusion for the next stages. The vertical nature of a storyline talk refers to a coherent sequence of content from raising student uncertainty of a phenomenon, maintaining the uncertainty to develop deep reasoning, and then to reducing the uncertainty in order to develop a coherent conceptual knowledge system. In short, the horizontal nature occurs within a stage of management, and the vertical nature of a storyline talk is related to moving along from stage to stage. Therefore, storyline talk is more than highlighting students' ideas or encouraging students to interact with each other at particular moments. A productive storyline can potentially scaffold students to develop deep learning in science through horizontally considering different ideas to understand strengths and weaknesses of peers' and their own ideas, as well as vertically developing knowledge from fuzzy ideas to robust concepts.
In order to productively manage student uncertainty in storyline talk, student epistemic uncertainty should be used not only in the beginning to raise student interest, but also continuously throughout the discussion to scaffold dialogic moves that help students solve their uncertainty and establish a consistent and meaningful knowledge system. Therefore, effective tactics to manage uncertainty will weave uncertainties as a pedagogical resource into a coherent storyline that is meaningfully understood from students' perspectives.
3 THE PERSPECTIVE OF THIS STUDY
This study is rooted in Vygotsky's sociocultural theory (1978), which considers knowledge to be generated and mediated through social interactions. So, although we consider that students' epistemic uncertainty is initially generated at the individual cognitive level and then negotiated at the sociocultural level, our data and analysis focused on whole-class discussion, which is at the sociocultural level. Applying sociocultural theory to the context of this study, when students have opportunities to express their individual uncertainty during whole-class discussion, their uncertainty can be discussed, debated, and resolved through peer interaction and teacher support. Thus, this study identifies tactics used by teachers during whole-class discussion that can productively support and manage students' epistemic uncertainty to develop conceptual knowledge.
4 METHODS
This design-based study (Cobb et al., 2003) was conducted with six science teachers from third to eighth grade of public schools in the United States. A design-based study is “conducted to develop theories, not merely to empirically tune ‘what works’” (Cobb et al., 2003, p. 9). “Participants are not subjects assigned to treatments but instead are treated as co-participants in both the design and even the analysis” (Barab & Squire, 2004, p. 3). Therefore, we collaborated with the six teachers to tailor their individual instructional tactics, curricula, and classroom activities to our proposed teaching orientation.
This study was part of a preliminary, small scale, 1-year project funded by the public university of the authors. The goal of the project was to identify effective tactics that can be used to manage students' epistemic uncertainty to develop scientific knowledge across various contexts, content, and grade-levels. A diverse cases approach (Seawright & Gerring, 2008) was taken, aiming to explore and find common patterns among the various cases. The six teachers were recommended by coordinators, principles, and liaisons from different school districts. After the first author visited them individually, they volunteered to participant in this project. The diversity of the six teachers matched the goal of this project, as they possessed knowledge of effective tactics in various contexts and subjects (e.g., chemistry, earth science, life science, and physics).
4.1 Instructional context
- Design Principle 1: Design and embed epistemic uncertainty in inquiry.
- Design Principle 2: Engage students in using their epistemic uncertainty as a pedagogical resource to further knowledge development.
- Design Principle 3: Socially and coherently engage students to negotiate their claims to reach consensus and reduce uncertainty.
Communication is a key resource for managing uncertainty (Chen & Qiao, 2020; Jordan & McDaniel, 2014). Learning to negotiate with peers and garner peer support is critical for productive engagement in scientific inquiry. Negotiating with peers elicits consideration of the interrelationships between disparate or seemingly disparate ideas, and it establishes a consensus among individuals that experience uncertainty or hold contradictory ideas (Berland & Lee, 2012). Teachers must create space for students to discuss, debate, defend, support, and evaluate each other's ideas (Chen et al., 2017; Tang, 2021). Students need to work cooperatively as a community to find solutions that reduce their uncertainty and reach mutually acceptable consensus.
To help teachers adopt the three design principles, we conducted a one-day workshop that provided an overall view of the project. The content emphasized (1) the role of uncertainty in learning science, (2) instructional tactics to help students recognize (e.g., questioning, problematizing phenomena) and manage their uncertainties (e.g., design experiments, interpret data as evidence), and (3) the importance of social negotiation (e.g., whole-class discussion) in scientific processes. To implement this content, we used the unit of density (details see Chen et al., 2014) as an example to demonstrate how to integrate the three design principles into the lessons. The research team discussed with the teachers their concerns, questions, and future planning. Some exemplar video clips collected from the first author's previous project were provided to facilitate teachers' understanding of how to conduct productive conversation and negotiation with students during whole-class discussion.
In addition to the workshop, before lessons were implemented, the research team met with each teacher individually 3–4 times to focus on (1) identifying core concepts to be taught in the lessons, (2) identifying potential students' epistemic uncertainty, and (3) designing activities to use student uncertainty as a pedagogical resource. The meetings aimed to help each teacher modify their original lesson to align with the three design principles. While working with the six teachers, we did not identify the three stages of managing uncertainty—Raise, Maintain, and Reduce—and more nuanced tactics (see Section 4: Findings). Therefore, we did not use these concepts or encourage the teachers to maintain student uncertainty and avoid premature closure of discussion. We emphasized the use of student uncertainty to drive student learning and the importance of social negotiation (e.g., small group and whole-class discussions) by using claim and evidence to frame the discussion of ideas.
During the implementation of lessons, we assigned at least one member of the research team to visit and observe the teachers' implementation of their lesson plans at least twice per week. Because teachers usually only have 5–10 min between classes, we had brief conversations with them that focused on what student uncertainties they noticed and how they coped with them. Our research team scheduled with individual teachers a more thoughtful weekly meeting for after school or during lunch. These meetings focused on the fine tuning of lesson plans, activities, and worksheets. The research team also provided feedback for teachers on questioning skills based on our observations. The meetings lasted for 30–60 min.
4.2 Data collection
When the six teachers taught the lessons that were co-designed by teachers and the research team, we videotaped every lesson from the corner of the classroom. Due to the scope of the study, only whole-class discussion videos were selected for analysis. Next, the research team reviewed all videos and selected those for further analysis based on three criteria to prescreen the videos focusing on deep learning: (1) the conversation had to involve student epistemic uncertainty; (2) the conversation went beyond identifying the expectation of learning tasks, discussing experimental procedures, announcing daily logistics; and (3) the dialogic move went beyond the Initiate, Response, Evaluate (IRE) pattern (Mehan, 1978) that focuses on recall and checking of student answers. As a result, 18 videotaped classroom observations, ranging from 35 to 50 min, were selected for data analysis (Table 1).
| Unit | Grade | Teacher | # of transcripts | # of events |
|---|---|---|---|---|
|
3rd | Mrs. Weber | 2 | 6 |
|
5th | Mr. Meyer | 2 | 7 |
|
5th | Mr. Fischer | 3 | 8 |
|
6th | Mr. Huber | 3 | 5 |
|
7th | Mr. Smith | 5 | 6 |
|
8th | Ms. Chandler | 1 | 2 |
| 18 | 38 |
4.3 Data analysis
- Step 1. Read transcriptions and conduct narratives.
- Step 2. Identify events from each transcript.
We identified an uncertainty event as a main unit of analysis (Derry et al., 2010). From 18 transcripts, we identified 38 events. The length of each event ranges from about 10–25 min. Because the purpose of the study is to understand how epistemic uncertainty is used to develop knowledge, we defined the boundaries of an event as from the start of discussion of a specific core concept to the point when the core concept stops being discussed as consensus of understanding was reached. Therefore, each event focused on only one specific concept. For example, one event identified from Mr. Huber's heat transferring lesson started when he raised students' uncertainty about their prediction, feeling, and experience about touching woody and silver-plate trays. This focused the discussion on the concept of different materials having different thermal conductivities, and that this influences the efficiency of thermal conduction. In his class, all students said the finger felt colder when touching silver-plate trays. When Mr. Huber asked which tray had a lower temperature, students responded that it was the silver-plate tray. However, after measuring, Mr. Huber showed the two types of trays were at the same temperature. This created uncertainty and engaged the students in considering, reflecting, and discussing their experience, interpretations, and arguments. This event ended by students understanding that the silver-plated tray has a higher thermal transfer conductivity than the wood-plat tray, and that causes faster heat transfer from a warm hand to the silver tray than from the hand to the woody tray. Table 1 notes the distribution of 38 events identified from the 18 videotaped classroom observations.
- Step 3. Analyze each event using the constant comparative method.
Once events were identified, the constant comparative method (Strauss & Corbin, 1990) was utilized to identify and analyze (1) the sources of student epistemic uncertainty; and (2) the tactics used to manage uncertainty in the given events. To identify what uncertainties were involved in the discussion and how these uncertainties were used as a pedagogical resource to develop knowledge, we examined utterances situated in the context, and interpreted potential epistemic uncertainty within the sequence of interactive utterances. For the first event analyzed, we developed tentative codes. We applied these tentative codes to other analyzed events, and we refined, edited, and modified the codes as we went. The process to develop a code scheme was inductive and iterative until a comprehensive scheme emerged that covered all 38 events (Lincoln & Guba, 1985). Table 2 summarizes four sources of student epistemic uncertainty identified in the events. Building on the definition of epistemic uncertainty and the purpose of the study focusing on student development of scientific understanding, the four sources especially focus on the process of how students construct and represent scientific understanding from raw data, prior knowledge, and ideas to scientific argument (Aguirre-Mendez et al., 2020; Kelly & Takao, 2002; McNeill & Berland, 2017).
| Source | Interpretation of the data | Representation of the idea, concept, and argument | Components of the argument | Coherence of the argument |
|---|---|---|---|---|
| Example | Explanations of the meaning of raw data for evidence and claim | Drawing, table, figure, diagram, material use | Claim, data, evidence, explanation | The relationship between claim, evidence, and the question |
- Step 4. Generate themes.
| Tactics | Description | When used | # of event |
|---|---|---|---|
| Stage 1: Raise uncertainty | Elicit awareness of uncertainty toward the discussed topic/concept/issue/event | To begin a discussion | |
|
Pose that current knowledge is limited in explaining an encountered phenomenon, allowing new ideas, viewpoints, and action to emerge | To point out a problem and open a discussion space enabling students to discuss, negotiate, analyze, and explore | 20 |
|
Pose that trends or patterns in data are not consistent or expected from a hypothesis or theory | To encourage and foster students to interpret data and find patterns in them | 18 |
| Stage 2: Maintain uncertainty | Prolong discussion of uncertainty to engage students in deep reasoning | To ensure students continue the discussion meaningfully and critically | |
|
Negotiate conflicting ideas in order to weave those ideas together into a more coherent scientific concept | To help students comprehend another's ideas and elaborate on their own | 11 |
|
Pose different or opposite ideas and require articulation and defense | To encourage students to seek and identify the differences, similarities, or weaknesses of a peer's ideas | 6 |
|
Point out the blind spots or errors of an argument/model to stimulate alternative thinking | To foster students to think more deeply and from different perspectives about a discussed uncertainty | 5 |
| Stage 3: Reduce uncertainty | Resolve discussed uncertainty | To conclude and shape a shared and coherent understanding | |
|
Converge disparate ideas into one mutual agreement | To guide students to clarify ideas toward the construction of canonical scientific knowledge | 15 |
|
Use what is known (e.g., prior knowledge, everyday experience) to solve the uncertainty for building a coherent knowledge system | To help students integrate different concepts and sources into a conceptual framework of meaningfully interconnected concepts | 9 |
- Notes: The total number of events is 38. One event may include more than one tactic at each stage.
Once all events were analyzed, we began to seek the common themes that emerged across the 38 events. First, we generated multiple tentative themes per each event. Then, we took the themes generated from one event and applied them to all other events to see if they could be used to interpret those events. The tentative themes were recursively reviewed and checked until we felt we had “tuned the picture” (Do & Schallert, 2004) of effective tactics to manage student epistemic uncertainty to develop scientific knowledge.
To establish the credibility of data analysis and interpretation, the two authors independently analyzed the data following the four steps discussed above and had weekly meetings to discuss any incongruities until an agreement was established and the initial coding was refined accordingly.
5 FINDINGS
We identified and present in this section three consequential stages to complete a storyline talk that manages uncertainty: Raise, Maintain, and Reduce. We identified seven nuanced tactics used by the six teachers to achieve the goals of each stage. The tactics identified in this study emphasize ways to scaffold students to restructure their existing knowledge system.
The three stages have different functions that are integrally related to one another, requiring teachers to use tactics during the three stages to build a horizontal and vertical storyline to scaffold students to manage uncertainty for knowledge development. For each stage, we present examples from various lessons to demonstrate that tactics can be applied in a variety of contexts.
5.1 Stage 1: Raise student epistemic uncertainty by creating ambiguity
The first stage in managing uncertainty is to raise it, which can elicit student curiosity, motivation, and interest in discussing and learning about the topic. Rather than merely eliciting and probing students' prior knowledge, this stage aims to identify the gap between their current understanding and their targeted understanding of the concept, thereby eliciting awareness of what they do not know and what they want to know. This stage functions as an initial step to encourage students to express, explicate, and elaborate their individual uncertainty that may initially be fuzzy and unclear. Among the 38 events, we identified two tactics used by teachers to raise uncertainty: problematizing a phenomenon and highlighting inconsistent data patterns.
5.1.1 Problematizing a phenomenon
Okay, so here's the problem. You wanna bring a cold soda on a field trip and drink it for lunch, but you want it to still be cold when you drink it, and you don't know what the best thing would be to keep it cold. So, the question is, which material would work best for keeping the soda cold until lunch?
After problematizing the phenomenon, Ms. Chandler directed the students' attention to a table with seven cans set up earlier. Six cans were wrapped with different materials: plastic, paper, towel, aluminum foil, wool, or cotton, with an unwrapped can as the control. A thermometer was attached to each can to read its temperature. Instead of letting students brainstorm solutions to the problem, Ms. Chandler constrained the situation to six specific materials, and asked students to decide which materials are the best heat isolators and provide explanations for their claims. This problematized the phenomenon, thus triggering student uncertainty and a heated debate. In the beginning, students had their own explanation in their mind. However, when they had opportunities to express their ideas, the uncertainty became clear at a social level due to their different explanations and claims. While some students said “aluminum foil,” others said “wool sock because you could consider when you wear like a sweater—sweaters or socks or something, and if can be used to trap heat in, it could also be used to trap, uh, the cool in.” After ideas were shared, Ms. Chandler arranged students into groups to discuss which material was the best insulator in the experiment, then present their claims and evidence to the class.
It is important to problematize a phenomenon, stimulate student uncertainty, and engage them to discuss different and contradictory ideas. The problematized phenomenon creates an argumentation space for students to discuss, express, and reexamine their common-sense understanding related to the problematized phenomenon (Wickman & Östman, 2002). As Suárez (2020) stated: “problematizing is more concerned with the complex questions and wonderments that arise when students choose to engage with disciplinary content and make progress towards figuring out what is going on” (p. 795). We consider that authenticity and meaningfulness should be emphasized when problematizing a phenomenon. Ms. Chandler used an authentic phenomenon to connect students' everyday life to the concept they were going to explore in class. The phenomenon happens often in student everyday life, but they perhaps did not notice the problem. In addition, the phenomenon was meaningful enough to help students make sense of how the target concept applies in their out-of-school life. Thus, students were able to recognize the value and need to learn about the concept and discuss uncertainty related to the concept. In this event, Ms. Chandler did not encourage students design an experiment. Instead, she purposefully limited the situation to six materials, and that focused the problem and discussion on two variables: materials and temperature. Her approach might have helped students tie the concept of the material-heat transfer relationship to their claims and explanations, rather than simply promote an empiricist orientation (McComas, 1996; Tsai, 1998) that emphasizes “truths” discovered through hands-on activities, experiments, and observation. That is, focusing on the uncertainty related to students' claims and explanation promotes students' competence in the interpretation of the phenomenon, and moves away from the orientation of scientific knowledge as unproblematic and data as representing a correct and unchanging answer.
5.1.2 Highlighting inconsistent data patterns
Sometimes, students ignore or deny data patterns that are anomalous or inconsistent with their hypotheses. They tend to “protect” their hypothesis and avoid dealing with the uncertainty caused by inconsistent or unexpected results (Chinn & Brewer, 2001; Mason, 2001). Teachers need to explicitly point out anomalies to raise student uncertainty. Therefore, this tactic is different from the previous one, that aims at eliciting the knowledge gap between what students know and do not know. This tactic focuses on the conflict between students' hypotheses/prior knowledge and the experimental results. As such, this tactic encourages students to confront their “misinterpretation” of a phenomenon or results. In a fifth-grade plants unit, students explored the question: what do seeds need to germinate? One group claimed that “sunlight or darkness” was needed, but it had inconclusive data to present to the whole class. Their teacher, Mr. Meyer, took advantage of the moment and asked students to think about how to explain the inconclusive results. He said, “Five seeds germinated in sunlight and three seeds germinated in darkness. So, seeds need sunlight or darkness to germinate. Agree? Disagree?” Bailey said, “I disagree with darkness because not every single plant needs darkness.” However, Salena immediately proposed an alternative idea and said, “But some plants do!” Another student, Kelly, rejected Bailey and Salena's claims and said, “Based on the data, I don't believe that they need sunlight or darkness.” Guided by these three different explanations of the data, students engaged in sharing and debating their ideas about how to explain the data, and how the data aligned with their explanations.
Consequently, the conflict between different ideas caused uncertainty among the students. For example, Olivera said, “I don't understand because Journey's saying it doesn't need darkness, and Noah's saying it doesn't need sun. We need something.” In response to Olivera, Kailee said, “We do not need darkness because darkness gives off coldness! So, we can't exactly ….” However, Emma disagreed and immediately said, “Darkness does not give off cold.” Clearly, the discussion of inconsistent data raised student uncertainty about how to interpret data that violated the hypothesis that seeds need sunlight to germinate.
In this event, Mr. Meyer used inconsistent data patterns to raise uncertainty and engage students in discussing, debating, and elaborating their interpretation of the data. In this tactic, the source of uncertainty lay in the students' interpretation of the data. Mr. Meyer did not directly tell them how to interpret the data correctly and then teach them the concept (i.e., seeds need appropriate temperature to germinate). Instead, Mr. Meyer used the moment to raise uncertainty, then utilized it as a pedagogical resource to “get the student to explicate his or her reasoning so the teacher gains a better sense of the student's understanding and all students can work with it” (Michaels & O'Connor, 2015, p. 334).
5.2 Stage 2: Maintaining epistemic uncertainty through preventing the premature closure of discussion
The second stage, Maintain, focuses on encouraging students to socially discuss, debate, debunk, defend, compare, critique, and comprehend their epistemic uncertainty with peers and the teacher. This stage intends to sustain or increase students' epistemic uncertainty to socially engage them in deep reasoning and prevent a premature closure of discussion. After raising uncertainty, teachers prolong discussion to engage students to explain their reasoning. In this way teachers and students gain a better sense of the classroom's understanding and interpretation of the discussed concepts. This stage functions to help students understand and explore their uncertainty deeply, as well as encourage them to think from alternative perspectives. We identified three tactics used by teachers to maintain uncertainty: strategically selecting and comparing conflicting ideas, inviting students to critique each other's arguments, and challenging students' ideas to clarify and stimulate thinking.
5.2.1 Strategically selecting and comparing conflicting ideas
To maintain the uncertainty raised in the beginning of a discussion, teachers strategically compared conflicting ideas to prompt students to think critically and comprehend alternative ideas. This tactic does not suggest teachers “randomly” compare student ideas, but strategically select different ideas that can lead to desired learning outcomes. For example, in Mr. Huber's sixth-grade heat transferring unit, students discussed the three ways that heat is transferred. Students had different ideas about the way sun energy transfers, and at first their discussion was chaotic and without agreement. Mr. Huber then compared two students' ideas that conflicted about how sun energy was transferred to Earth (Table 4).
| Turn | Person | Dialogue |
|---|---|---|
| 1 | Emily: | The sun's energy is absorbed into the ground, warms the atmosphere through conduction, as energy is radiated into the atmosphere. |
| 2 | Mr. Huber: | Matt, what are you thinking? You're shaking your head. |
| 3 | Matt: | Well, we had the sun's energy is absorbed into the ground, warms the atmosphere through radiation as energy is conducted in the atmosphere. |
| 4 | Mr. Huber: | Okay. So, we have a little impasse. Anyone else agree with Matt? … [Inaudible] So, Matt, defend yours. Look at what you have up there, defend that. That's your claim. Defend it. |
| 5 | Matt: | Well, because the sun's energy, we talked about yesterday how it's absorbed into the ground. And then, it forms, from that, it goes up into the atmosphere. And then, I was thinking through radiation – radiated, the ground is radiating the heat off of that going into the atmosphere. And then, that energy is being conducted into the atmosphere. |
| 6 | Mr. Huber: | Okay. Emily? |
| 7 | Emily: | The sun's energy is radiated, but conduction is the thing that's happening to get it there, basically. |
| 8 | Mr. Huber: | When we conduct something, what kind of contact is that? Can I conduct electricity from media right now? |
| 9 | Emily: | No! |
| 10 | Mr. Huber: | Why? |
Mr. Huber explicitly picked two students and encouraged them to compare their conflicting ideas about the concept. First, after Emily expressed her ideas about how heat is transferred from the Sun to the Earth, Mr. Huber noticed Matt shaking head, showing his disagreement. Mr. Huber pointed this out and asked him to express his ideas (Turn 2). After Matt expressed his ideas, Mr. Huber explicitly let the students know there were two interpretations and asked who supported Matt's ideas. Mr. Huber strategically selected two different ideas to maintain the uncertainty that may lead to more student conversation and reasoning. Second, Mr. Huber increased the uncertainty by asking Matt to “defend” his ideas (Turn 4). He did not directly lecture on the correct answer, but he encouraged them to elaborate their ideas (Turn 6). He asked both Emily and Matt to defend their claim and explanations. Mr. Huber even used the word “impasse” (Turn 4) to indicate that the two conflicting ideas caused uncertainty among the students. Through comparing two conflicting ideas, students have an opportunity to deeply examine their reasoning and understand each other's explanation.
Mr. Huber's tactic also enabled him to monitor student understanding of the discussed topic through engaging students in elaborating on their ideas and critiquing each other's. In this case, he found that students had an incomplete understanding of the concepts of conduction and radiation. Thus, he focused the discussion on defining the concept of conduction (Turns 8 and 10). This conversation was meant to maintain it to understand what “misconception” students had, and then to frame the discussion around the misconception.
With this tactic, student claims and evidence serve as a source of uncertainty to be used as a pedagogical resource. To maintain uncertainty, Mr. Huber explicitly and strategically elicited the conflict between different interpretations and arguments of a single concept. Students thus had an opportunity to think through how well their argument could explain a phenomenon.
5.2.2 Inviting students to critique each other's arguments
This tactic maintains uncertainty by eliciting different ideas from students and having them critique each other's ideas. This tactic goes beyond the previous tactic, elaborating conflicting ideas, as it focuses on seeking weaknesses in different ideas. The primary goal is to help students notice and comprehend the difference between their ideas and those of others, leading them to recognize the weaknesses and strengths of their argument. In Mr. Smith's seventh-grade ecosystem unit, he asked students to discuss the article about the ecosystem in Yellowstone National Park. He used this article to help students discuss and understand the interdependent relationships between biotic and abiotic factors of an ecosystem. He began by discussing how the absence of wolves impacted the Yellowstone National Park ecosystem. He asked, “When grey wolves all died, was that good for Yellowstone? Or was that bad for Yellowstone?” This question required students to select one side of the argument. Some students claimed the wolves' absence was good for the ecosystem, whereas other students had an opposite opinion. Mr. Smith used this opportunity to invite students on one side of the argument to critique the other side (Table 5).
| Turn | Person | Dialogue |
|---|---|---|
| 1 | Mr. Smith: | You think it was good? That is your claim. What is your evidence to support your claim? |
| 2 | Jack: | I said that maybe the more wolves there are, the more animals there are, and the least amount of wolves that are there will be the least amount of elk, deer and bison. |
| 3 | Mr. Smith: | Kate? |
| 4 | Kate: | I disagree with you, Jack. If there were more wolves, let's just say there is an equal amount of wolves and they eat all the elk, deer and bison … It would be less because wolves are predators and they will most likely eat and kill them, so there would be a decline in numbers. The less wolves there are, probably that number will go up because there will be nobody to stop them—should I say impede them—going up in numbers. |
| 5 | Mr. Smith: | One thing that I want to go ahead and recognize about what Kate said, that she did a really good job of explaining that she disagrees with what Jack is thinking, and she referred to the data that we collected. She has a specific piece of evidence to back up what it is that she is thinking …. Do we have a question for Jack? Jack and Kate are saying opposite things. |
| 6 | Jill: | Jake, why do you say there are more wolves than other animals? |
The conversation in Table 5 demonstrates that Mr. Smith purposefully brought attention to two opposing opinions regarding the discussed question. He called upon one student, Kate, to express her ideas about the topic (Turn 3). In addition, he praised Kate as an example to encourage students to use evidence to support their claims (Turn 5). These dialogic moves, bringing attention to conflicting arguments and encouraging use of evidence to support claims, thus increasing student debate about the discussed topic and uncertainty. In addition, Mr. Smith explicitly highlighted how to productively maintain uncertainty through the use of evidence to support a claim (Turns 1 and 5). As such, students did not just state that they disagreed with a different argument, but their dialogue indicated they understood that using evidence to support a claim is critical to critique opposing arguments. This leads them to recognize the difference between/among conflicting ideas. By requiring students to confront conflicting opinions and defend their claims, this tactic is effective in getting students to use evidence to effectively support their ideas.
5.2.3 Challenging student ideas to clarify and stimulate thinking
This tactic goes beyond the previous two tactics, which focus on comparing and critiquing ideas. This tactic aims to inspire students to generate new ideas and recognize the weaknesses of their interpretations so they can be improved. In Mr. Fischer's fifth-grade Day and Night Cycle unit, students discussed Kailee's group's model of how the dynamic relationship among the Sun, Earth, and Moon causes days and nights. After students discussed the model, the conversation became stagnant. Mr. Fischer seemed not to want the conversation to stop, so he jumped into the conversation and asked Kailee to demonstrate her group's model by using a “ball” (i.e., a ping pong ball) to represent the Moon and a globe to represent the Earth (Figure 2). Kailey said, “When it's night, this is the Sun, and this is the Moon … Then it's day over here. Then it just keeps going, and those are day periods.” Based on the demonstration, Mr. Fischer observed that there was a blind spot in their model regarding the relationship between the Moon and Earth. He challenged Kailee's group's model by asking several questions focusing on the role of Moon in creating day and night. Table 6 shows the conversation started by Mr. Fischer as he challenged Kailee's model, stimulating several new ideas among the students.
| Turn | Person | Dialogue |
|---|---|---|
| 1 | Mr. Fischer: | I just want to know what the moon is doing. |
| 2 | Kailey: | It's just staying here so when the earth … It's spinning, because like what we said, there are different stages of the moon, which means it has to spin because … |
| 3 | Mike: | No! It doesn't. |
| 4 | Kailee: | Yes! It does. It's the shadow of the earth. Then it goes around its orbit. |
| 5 | Lexi: | The earth is just spinning a little bit faster than the moon, but the moon is spinning. The earth is just spinning faster. |
| 6 | Mr. Fischer: | Show it like high-speed action. What's going on? (Kailey demonstrated their model by revolving the ping pong ball [Moon] around its own axis above the northern hemisphere, see Figure 2). Keep going, keep going, keep going. Stop. What are you seeing right now? |
| 7 | Class: | The moon and the sun. |
| 8 | Mr. Fischer: | But you said at night you see the moon, but not the sun. |
| Lexi: | It's called a solar eclipse. | |
| 9 | Kailee: | Sometimes you see the moon during the day. So then … Yeah, but sometimes you just see the moon during the day sometimes. |
| 10 | Kaleb: | Lexi said it's a solar eclipse. What's a solar eclipse? |
| 11 | Lexi: | A solar eclipse is when the moon is blocking the sun. It's when the moon is right in front of the sun, blocking it. It happens every like a thousand years. |
| 12 | Wendy: | What Kailee was showing, it was a solar eclipse all the time when you were spinning it, because … |
| 13 | Kailee: | Because the moon was up here. It wasn't here blocking. It was up here. |
| 14 | Bailee: | Why are there solar eclipses? |
| 15 | Noah: | They can't answer that because they don't know. |
| 16 | Andy: | You guys said the moon is up above and it goes around, but how does it go down to go and block the sun? |
| 17 | Lexi: | We don't quite know that. |
| 18 | Mr. Fischer: | The way you're showing it, wouldn't it only make sense if the northern hemisphere would see the moon and the southern hemisphere would never get to see the moon? |
During the conversation, Mr. Fischer asked Kailee's group to explain “what the moon is doing” (Turn 1), and used Kailee and her group members' responses (e.g., Lexi) to challenge them to think about the weaknesses of their model. Mr. Fischer's challenge not only “pushed” Kailee and her group to elaborate on their model, but also helped other students comprehend the model. Once they comprehended it, students were able to understand how the model failed to explain the phenomena (e.g., solar eclipse; Turns 12, 14, and 15) they understood from prior knowledge. Therefore, uncertainties were increased and discussed through the challenges from Mr. Fischer and student peers. For example, Kaleb challenged that Kailee's group was not able to demonstrate a solar eclipse (Turn 10). This challenge was supported by other students (Turns 15 and 16).
In a dialogic move, Mr. Fischer continuously asked challenging questions to help Kailee's group clarify their model and understand its weakness. These challenging questions maintain and further increase students' uncertainty by providing students an opportunity to reflect on their own models, stimulating peers to think alternatively and critique the model, and driving the whole class to discuss and refine the model. For example, Lexi acknowledged that their model could not demonstrate the phenomenon of a solar eclipse (Turn 17).
In this example, the students' explanation of the dynamic relationship among Sun, Earth, and Moon and their use of materials to represent their mental model, used as a pedagogical resource, are what caused the uncertainty. The pedagogical resource used here functions not only to maintain and increase uncertainty, but also to prolong discussion and help students comprehend a mental model, and begin to identify its weaknesses.
5.3 Stage 3: Reduce epistemic uncertainty through making coherent connections among current uncertainty, prior knowledge, and familiar phenomena
The third stage, Reduce, focuses on resolving epistemic uncertainty through synthesizing and integrating students' existing knowledge and new information to shape a coherent knowledge acquisition and processing system. In the last stage, students already have had sufficient opportunities to discuss their uncertainty and understand the weakness and strengths of their ideas. After creating a space to raise and maintain uncertainty, teachers need to reduce uncertainty and consolidate ideas to develop shared scientific understanding within the community of students. Therefore, to foster shared understanding, teachers can utilize students' previous knowledge or experience to help them process new information and adjust their knowledge base to accommodate or assimilate this new information. We identified two tactics teachers used to reduce uncertainty: leading students to collectively find a solution, and purposefully using familiar phenomena-based evidence to explain the target concept.
5.3.1 Leading students to collectively find a solution
This tactic aims to converge disparate ideas into an agreement. It is often used after students have had opportunities to express their uncertainty and critique each other's ideas. During Mrs. Weber's third-grade unit about the three states of water, students presented their group's claim and evidence to get feedback from peers. After Brock's group presentation, students discussed if the group's evidence appropriately and sufficiently supported the claim. Some students pointed out that their evidence did not show “how water evaporates into the air.” After a heated debate among students, Mrs. Weber intervened and attempted to arrive at a mutual agreement. Table 7 shows how Mrs. Weber guided students to reduce their uncertainty by establishing a mutual agreement between challengers and defenders. In the conversation, Brock's group presented its argument and the rest of class provided feedback to their presentation.
| Turn | Person | Dialogue |
|---|---|---|
| 1 | Jason: | My question is that it said that it dries up into thin air, but they never read a part about how it could dry up. |
| 2 | Mrs. Weber: | What do the rest of you think? You agree with Jason or you agree with Brock's group? What do you think? Grace, what do you think? |
| 3 | Grace: | I agree with Jason because … [Grace stops talking] |
| 4 | Mrs. Weber: | So, from what they said for their evidence—that doesn't prove to you to help you to think … agree with them? What kinds of things do you think they might want to put in their evidence? They said that the claim was—it didn't what? Who remembers what the claim was? Christian? |
| 5 | Christian: | Well, that is true because the claim, it says it evaporates but when it says the evidence, it doesn't give you the information that tells you that's what happened. |
| 6 | Mrs. Weber: | So, what do you think—guys up here—what do you think could tell us that it went into thin air? |
| 7 | Brock: | Mrs. Weber, I kind of agree with Jason that we don't have enough evidence. I think we need a little bit more. But I don't really know what to use for our evidence. |
| 8 | Mrs. Weber: | So, what do you think, Shelby? Do you think that your pictures that say it went from 60 to 60 and a half to 50 and to 40 is showing that it went somewhere? |
| 9 | Shelby: | Yeah! |
| 10 | Mrs. Weber: | Yes? And you're thinking, then, that it went where? |
| 11 | Shelby: | The air came down and it brought some water up. |
| 12 | Mrs. Weber: | So that the air had something to do with the water? Not being there anymore? So, do you think that the numbers that you had by your pictures might be your evidence? Is that what you're thinking, Brock? |
| 13 | Brock: | Yeah! |
| 14 | Mrs. Weber: | What do you think about that, Jason? |
| 15 | Jason: | I do agree with him on that. His pictures do show it on that, but still, on the evidence, you don't have anything to prove that it just evaporated into the air. I still disagree with you on that part, but on the rest of it, I think you did a really good job! |
During the conversation, Mrs. Weber asked whether the rest of the students agreed with Jason's critique of Brock group's argument. Her questions guided them to consider which parts of argument needed to be improved (Turns 2, 4, and 6). After hearing other students' explanations, Brock's group realized the weakness of their argument, and Brock expressed his uncertainty about how to use good evidence to support his claim (Turn 7). Instead of directly telling Brock how to generate good evidence based on their data, Mrs. Weber used a series of questions to scaffold students to collectively think about a solution (Turns 8–12). Mrs. Weber gave students hints that their pictures could be used as evidence to support their claim. In the end, students, including the defender Brock and challenger Jason, agreed with the solution.
The uncertainty used in this event is the students' disagreement about what evidence should be constructed and included in the argument. Mrs. Weber did not lecture them about correct answers, but instead used their uncertainty to drive them to think about what could be included as evidence and how to represent it. In the end, she made sure all students agreed with the solution.
5.3.2 Purposefully using familiar phenomena-based evidence to explain the target concept
This tactic to reduce student uncertainty focuses on using evidence that comes from student everyday life and that students are familiar with to understand the discussed concept. This tactic can be used when students do not understand the explanation of a concept that appears to be abstract and isolated from their experience. In Mr. Meyer's fifth-grade plant unit, some students had grasped the concept that seeds need warmth, not sunlight, to germinate. Other students still believed that seeds need either sunlight or darkness to germinate. Students that had already understood the scientifically correct concept tried to explain to students that did not yet understand. For example, Kate said, “It's [sunlight] not even a need;” Neal said, “It doesn't matter, it just needs warmth.” However, these explanations did not seem to help. For example, Kalie repeatedly voiced her uncertainty throughout the conversation, “Oh man, now I'm confused again!” Kalie explained her uncertainty in this way, “If it's not a requirement to put it in sunlight, and it's not a requirement to put it in darkness, and shady is just like, in between, and then … you don't know where to put it, so that's why I'm confused!”
Thus, the conversation became circular. Mr. Meyer mentioned an everyday example to help students connect their life experience to the discussed concept (Table 8).
| Turn | Person | Dialogue |
|---|---|---|
| 1 | Mr. Meyer: | So, if the sun's out in the winter time, why don't they [farmers] plant the seeds? |
| 2 | Kalie: | I think they plant it during the summer because it gives off heat, so it grows better. |
| 3 | Mr. Meyer: | So, the sun doesn't give off heat in the wintertime? |
| 4 | Kalie: | It does but it's just warmer in the summer because the sun's out longer and it's hotter because the snow's not on the ground, and the snow is like really cold water. |
| 5 | Kaleb: | The ground is frozen during the winter. |
| 6 | Mr. Meyer: | So, do they need sunlight or do they need warmth? |
| 7 | Group: | They need warmth. |
| 8 | Kalie: | They just don't need sunlight. You could give something warmth without getting it sunlight. |
| 9 | Class: | [Noise. Students are discussing the ideas] |
| 9 | Mr. Meyer: | I've actually seen something around Apple Valley. Not very often, but sometimes in the fall after farmers take out their crops, they have this machine and a drill that plants all these seeds. It's very, very cool – later than this time of year, germinate, snows over them, they take off in the spring, they harvest and they're able to plant another crop for the summer crop. So, it was very cool, right? So, do you think all seeds need the same warm temperature, or does the right temperature depend on the seed? |
| 10 | Group: | Depends on the seed. |
| 11 | Mr. Meyer: | So, do they need a sunny area or dark area, or does that not affect it? |
| 12 | Group: | Doesn't affect it. |
| 13 | Mr. Meyer: | How about the temperature deal? |
| 14 | Group: | Yes! |
In this excerpt, the everyday example of farmers not planting seeds during winter connected students' everyday experience to the concept being discussed (Turns 1 and 2). Students, such as Kalie, were stimulated by Mr. Meyer's example, and were able to reduce their uncertainty. Kalie elaborated on her own experience (Turns 4 and 8) and made the connection to her uncertainty. Thus, through the connection between a familiar phenomenon and an uncertain concept, uncertainty was reduced and resolved.
Several researchers have advocated that when introduced to a new concept, students' everyday experience, prior knowledge, and familiar phenomena can be used to initiate a conversation and make students uncertain or unsatisfied with their current understanding (e.g., Hmelo-Silver, 2004; Tsai & Chang, 2005). For example, this study suggests using students' everyday phenomenon to raise uncertainty through problematizing a phenomenon that is meaningful and authentic for the students. However, this excerpt suggests that utilizing student experience, prior knowledge, and familiar phenomena can be a resource to reduce uncertainty “to reflect on how the new knowledge fits into their existing schema, engaging in some metacognitive thought” (Richards et al., 2020, p. 1138). In this case, the phenomenon did not function as a catalyst to raise student uncertainty, but it served as a connection between student existing knowledge system and the target concepts. Therefore, the target concepts are not discrete entries isolated from students' existing knowledge “lacking structural connections to lend coherence” (Fries et al., 2020). Instead, the target phenomenon helps students to build a coherent knowledge system and understand how the concept can be applied to their everyday life.
6 DISCUSSION AND IMPLICATIONS
We identified varied and nuanced tactics teachers employed to manage student uncertainty to develop collective knowledge through whole-class discussion. The following discussion focuses on how our findings are related to the current research, and how they bridge gaps in the literature related to uncertainty management in science teaching. Pedagogical implementations are also discussed. Our analysis resonates with recent studies that support the role of uncertainty in students' learning in science (e.g., Hartner-Tiefenthaler et al., 2018; Kirch, 2010; Manz & Suárez, 2018; Phillips et al., 2018; Sezen-Barrie et al., 2020; Tiberghien et al., 2014).
6.1 Beyond ideas and asking questions
Over the last few decades, several studies advocated that science discussion and inquiry should be driven by student ideas (e.g., Luna, 2018) or the asking of questions (i.e., higher order questions vs. lower order questions, different types of questions) to engage students in reflective thinking and help students construct knowledge (e.g., Chen et al., 2017; Chin, 2007; Oliveira, 2010; Roth, 1996; Wei et al., 2018). However, managing uncertainty in scientific discourse involves more than just having ideas and asking questions. It involves a larger scope and a coherently sequenced storyline using dialogic pathways to thoughtfully navigate specific uncertainties. Implementing our findings, teachers must horizontally and vertically construct a storyline. The storyline is dynamic and coherent, and its nature depends on which tactics teachers plan to use to involve the students. Importantly, the dynamics and coherence of the storyline are derived from student uncertainty, not the teacher's.
Several studies related to argumentation and classroom discussion have reported that although teachers may raise student uncertainty in the beginning of discussion or activity, they reduce it immediately to avoid unexpected challenges from students (e.g., Chen et al., 2019; McNeill et al., 2017). Typically, teachers lecture about correct answers to reduce uncertainty. Tactics identified in this study guide teachers to maintain and reduce uncertainty to help students build a coherent and meaningful conceptual understanding.
The three stages (Raise, Maintain, Reduce) identified in this study reflect specific functions of and purposes for uncertainty management. To productively manage uncertainty to develop a robust understanding of scientific concepts, teachers should not just ask questions, they should also frame and focus more squarely on the uncertainty along the dialogic pathways (Chen, 2020). Uncertainty is a pedagogical resource to drive the dialogue toward establishing acceptable knowledge.
6.2 Beyond problematizing phenomena
Several studies advocated problematizing phenomena “as the intellectual work” (Phillips et al., 2018, p. 982) to identify and articulate a gap in students' or a community's current understanding (e.g., Engle & Conant, 2002). Problematizing phenomena echoes the first stage of managing uncertainty, Raising. However, based on the findings of this study, we extend the idea that managing uncertainty is more than problematizing. It also includes Maintaining and Reducing. We consider the three stages of uncertainty management to be sequential and interdependent.
Maintaining students' uncertainty is about keeping students motivated about the targeted concepts throughout the activity and discussion. Teachers should sustain or increase students uncertain and engage them in navigating the uncertainty through which teachers create learning opportunities for students to articulate their confusion and incoherent arguments. The tactics for maintaining uncertainty identified in this study not only helped students understand different ideas and explanations, but also provides a “struggle” experience to explore what caused their uncertainty, and to find a better solution. Sinha et al. (2021) suggested “persisting through this uncertainty” and engaging in generative activities “will give students a clearer picture of the problem and solution spaces” (p. 19). Therefore, students can truly engage in deep learning in science classrooms as generative and sensemaking processes in which they have opportunities to explore and navigate their uncertainties on their own or with teachers' scaffolds.
Reducing students' uncertainty is about helping students construct a coherent knowledge system through resolving the identified knowledge gap. It is not about lecturing to students about correct concepts or removing students' uncertainty from their “brain.” Rather, it helps students to find solution to fill the gap (Wickman & Östman, 2002), and it guides them to move forward to a more coherent knowledge system as they synthesize new and existing knowledge.
6.3 Response to NGSS
While NGSS (2013) emphasizes eight essential practices that teachers are encouraged to implement in science classrooms, NGSS does not clarify and address the “needs” to engage in those essential practices. Student needs can refer to the notion that the student has an inherent, internal desire to understand the world that leads them to do certain things to attain a deeper understanding. Teachers do not necessarily see or understand the need for or value of students engaging in argumentation, modeling, or engineering practice (Overman et al., 2019). It is important to point out that scientists implement the essential practices not just because they want to, but because they have uncertainty about encountered phenomenon that drives them to use the practices to develop knowledge (Kampourakis & McCain, 2019). Studies supporting NGSS-aligned instruction reported that teachers do not significantly shift their practical teaching approaches because they do not understand the needs of students as they acquire a working understanding of science, nor the value of implementing the eight practices (Allen & Penuel, 2015; Miller et al., 2018; Moore et al., 2015; Osborne, 2014; Overman et al., 2019). Researchers and teachers designing learning environments should carefully consider building in the needs for students to understand the world and uncertainty in it. This study suggests that student uncertainty has a significant role in science education, to be embedded in curricular considerations. Student uncertainty is a pedagogical resource that should drive lesson planning and instructional design (Tekkumru-Kisa et al., 2020). Students can then truly become epistemic agents centered in the learning process when the lesson is centered on and driven by their uncertainty (Chen & Qiao, 2020). In this way, students may come to see the need to engage in NGSS-aligned practices though raising, maintaining, and reducing their own uncertainty. The tactics identified in this study for different functions and stages of uncertainty management can be a guide for classroom planning, and for teacher education and professional development.
7 LIMITATIONS AND FUTURE RESEARCH
Although this study identifies several tactics for productive whole-class discussion across different contexts, grades, and subjects, we did not unpack the challenges and difficulties that the teachers encountered when they adopted the orientation of using student epistemic uncertainty as a pedagogical resource. During our interactions with the six teachers, we found some rapidly adjusted to this orientation, but some did not. For instance, we identified more events related to using student epistemic uncertainty as a resource in the classrooms of Mrs. Weber, Mr. Meyer, and Mr. Fischer than in Ms. Chandler's classroom. We suspect one of the reasons is a difference in epistemic orientation related to teaching science (Suh & Park, 2017; Zeidler et al., 2013). Such orientations can refer to how they perceive the ways to construct and teach scientific knowledge. Some teachers, such as Ms. Chandler, more frequently followed the lesson plan, and less frequently used student uncertainty to maintain discussion. A teacher may raise student uncertainties but not maintain them for deep reasoning. In this situation, without maintaining uncertainty, students did not have ample opportunity to explore their uncertainties and understand what they did not know. The discussion was closed prematurely. We conjecture that teachers may be more challenged to maintain student uncertainty than manage the other two stages. Exploring this possibility will require targeted research on the relationship between teachers' epistemic orientation and the ways they manage students' epistemic uncertainty to promote students' conceptual understanding.
This design-based study identified 38 events, however, not all events incorporated all three stages. Some events focused on the first stage, Raise, while some events focus on the others. This could be because we did not clearly identify the three stages and seven tactics during the time of our collaboration with teachers. Our identification of the stages and tactics emerged from the design-based study process. Teachers may differ in how they adapt the storyline-based framework to their teaching. This would be a fruitful question to explore. It would also be interesting to compare situations and the depth of student learning in which only one or two versus all three stages occur in an event. Does the “success” or “productiveness” of managing all three stages in one event depend up the topic, students' preparedness, and ease of getting them interested? It would be worthwhile to unpack the underlying factors that influence the “success” or “productiveness” of managing uncertainty in science classrooms.
In addition, we cannot predict how teachers will respond to uncertainty even when they have knowledge of a storyline-based framework. It is necessary to understand a teacher's perception of uncertainty, and how this perception influences the way they adopt the storyline framework in their lesson plans and instructional tactics to facilitate student learning in science. Future research could focus on teacher perception and difficulties in using the storyline-based framework across the three stages of management. In addition, this study focuses on the context of whole-class discussion, rather than other activities, such as small group or one-on-one discussion. It would be worthwhile to examine how productive the identified tactics are when applied to other forms of discussion.
Finally, a survey-based study conducted by Lee et al. (2014) showed that students that held a higher degree of uncertainty toward a topic tended to construct higher levels of warrants between theory and evidence. In a similar vein, Wu and Wu (2020) conducted a survey with 11th graders, and they found that degree of student uncertainty was positively associated with inquiry ability. These survey studies suggest that future research should unpack the relationship between the students' existing knowledge and experiences, how they were acquired, and the management of uncertainty at the three stages of the storyline. and their learning outcomes (e.g., reasoning skills, critical thinking skills, and representation competency).
8 CONCLUSION
We suggest that student epistemic uncertainty is central to these processes by driving them to make meaning of their experiences to develop deep scientific knowledge. The teaching of a deep understanding of science can be a daunting task. Our work demonstrates seven practical tactics that teachers can employ to help students manage their uncertainty and truly engage in deep learning in the science classroom, as a generative process (e.g., Hand et al., 2020) when they have opportunities to explore and solve their uncertainties on their own or with teachers' scaffolds. We also demonstrate that teachers can facilitate deep learning in science classrooms as a sensemaking process (e.g., Odden & Russ, 2019), as students connect, integrate, and synthesize new information with an existing knowledge base. We suggest that student epistemic uncertainty plays a critical role in their generative and sensemaking process and drives them to make meaning to their collective experience and scientific knowledge. Only when students identify a gap in knowledge and actively resolve that gap are they able to construct their own knowledge meaningfully (Fiorella & Mayer, 2016).
ACKNOWLEDGMENTS
We gratefully acknowledge the feedback of Editor-in-Chief, Dr. Felicia Mensah, Associate Editor, Dr. Audrey Msimanga, and three anonymous reviewers, on different versions of this paper. We wish to thank Dr. Terry Christenson at Knowledge Enterprise and Dr. Michelle E. Jordan at Mary Lou Fulton Teachers College of Arizona State University, for their invaluable feedback and support on this paper.
