Engagement in high-leverage science teaching practices among novice elementary teachers

BY: Elizabeth A. Davis and Annemarie S. Palincsar

First published: 14 September 2022 https://doi.org/10.1002/sce.21766

Abstract

As part of learning to teach, teachers must learn to use a range of teaching practices. In this longitudinal study the authors explore how novice elementary teachers learn to engage in a set of high-leverage science teaching practices, such as leading a science sensemaking discussion and setting up and managing small-group investigations. Drawing primarily on videorecords from 33 lesson enactments, supplemented by transcripts of interviews with the participants, the authors follow five participants from their second year of a teacher education program through their first year of teaching. Findings suggest that high-leverage science teaching practices can be untangled from one another and that teachers mostly demonstrated some important strengths in the dimensions of the practices. Furthermore, the teachers were able to use the practices synergistically, though in different ways, toward their goal of supporting students' sensemaking. The findings also suggest that teachers' engagement in the practices varied considerably as a function of lesson and school context. The findings have implications for the design of practice-based teacher education experiences and for scholarship on teacher development.

1 INTRODUCTION

Learning to teach is challenging. The challenges are particularly daunting in elementary science, because, despite many strengths, elementary teachers may not envision themselves as science teachers and may lack confidence and extensive knowledge in the science they will be expected to teach (ACARA, 2006; Davis et al., 2006; Banilower et al., 2018). Perhaps as a result, science is taught infrequently in the elementary grades (Banilower et al., 2018), and students infrequently engage in investigation, data representation, argumentation, and other science practices (Plumley, 2019). Elementary science is also more poorly resourced, as measured by funding per pupil, in comparison to both high school science and elementary math (Banilower et al., 2018). This is all compounded by new visions of science teaching and learning, which prioritize integrating content and disciplinary practices to understand phenomena, setting high expectations for students (and thus teachers) from kindergarten onward (National Academies of Sciences Engineering and Medicine, 2022; NRC, 2012).

Practice-based teacher education is a vehicle for supporting new teachers in meeting these challenges. Drawing on the literature in practice-based teacher education (e.g., McDonald et al., 2013) and science education (e.g., Kloser, 2014; Windschitl et al., 2012), we focus on six high-leverage teaching practices (e.g., eliciting students' thinking) that support students' sensemaking in elementary science. We ask, How do novice elementary teachers engage in key high-leverage science teaching practices? This study describes how participants enacted, across 2 years, key elements of the practices, characterizing similarities, and variations. It also provides “images of the possible,” illustrating, for each practice, some of these novice teachers' strengths.

2 LITERATURE REVIEW AND PURPOSE

We draw on literature related to practice, practice-based teacher education, and sensemaking in science, making connections to our teacher education program context.

2.1 Meanings of “practice”

From a sociocultural perspective, practice “involves the orchestration of understanding, skill, relationship, and identity to accomplish particular activities with others in specific environments” (Grossman, 2018; p. 4; see also Kavanagh & Danielson, 2019). Practices are “a set of sensible actions that are performed by members of a community and that evolve over time” (Berland, 2011; p. 627). Extending and building on these definitions, we use “practice” in both science and teaching to refer to (a) the recurrent work of a field or profession, (b) the work of learning to do the practice, and (c) those who work together enacting shared practices (Lampert, 2010; see also Arias & Davis, 2017; Kang & Windschitl, 2018). Definition (a) is most salient for this paper—a community's recurrent work or “sensible actions.” Science practices, such as developing models or constructing arguments, are elements of the work of doing science. Science practices work together to allow individuals to move toward collective understanding of natural phenomena (Berland, 2011). Teaching practices, similarly, are elements of the work of teaching that function together to support equitable learning.

Teaching practices draw on teachers' knowledge and constitute one aspect of what teachers bring to bear as they support student learning. Like science practices, teaching practices can be distinguished from one another (e.g., we can distinguish when a teacher is eliciting ideas, leading a discussion, or establishing norms for discourse, even though these practices are related to one another, share some common sub-practices, and may be used together). A teaching practice (again like a science practice) can be decomposed (Grossman et al., 2009) into its constituent parts, to be learned, but then must be recomposed (Janssen et al., 2015) to be useful. Sets of teaching practices must be employed—orchestrated—synergistically toward a greater endeavor (Ball & Forzani, 2009; Grossman et al., 2009), such as promoting students' sensemaking about a natural phenomenon—a communal goal (Berland, 2011).

2.2 Three-dimensional science learning

Current visions of science learning (and thus science teaching) emphasize understanding phenomena through the integrated application of science practices, disciplinary core ideas (such as the particulate nature of matter or predator-prey relationships), and crosscutting concepts (such as cause and effect or structure and function); this three-dimensional learning is reflected in the performance expectations of the Next Generation Science Standards (NGSS Lead States, 2013) in the United States, based on the Framework for K-12 Science Education (NRC, 2012). Similar integrations of content and practice are reflected in the standards documents of other countries (e.g., ACARA, 2014; NAE, 2018; United Kingdom Department for Education, 2014).

Instruction that prioritizes the integration of content and practice fosters students' sensemaking in science (e.g., Manz, 2015b; McNeill & Krajcik, 2009; Windschitl & Thompson, 2006; Zembal-Saul, 2009; see Fitzgerald & Palincsar, 2019; for a review). With support, even young children are able to engage in sophisticated scientific sensemaking (National Academies of Sciences Engineering and Medicine, 2022). This kind of instruction sets high expectations for every student and works toward more equitable schooling experiences across all schools and communities. Thus, elementary teachers need to learn how to foster such sensemaking through their teaching practice.

2.3 Practice-based elementary teacher education and the development of practice

Practice-based teacher education supports novices learning to do the work of teaching. Evolving from decades of work in teacher education, the current focus on practice in teacher education sees teaching as “a series of moment-to-moment judgments calling on knowledge about instructional goals, students, and the integrity of the discipline” (McDonald et al., 2013; p. 379) to work toward justice and toward disciplinary engagement (Kang, 2021; Kang & Zinger, 2019; Kavanagh & Danielson, 2019; Kavanagh & Rainey, 2017), and reflecting the complex orchestration highlighted above (Grossman, 2018). Developing one's science teaching practice involves deliberately working to improve it, with guidance and feedback from others, to work toward a communal goal of supporting the equitable science learning of one's students.

In our practice-based elementary teacher education program, we use pedagogies of practice (Grossman et al., 2009) to support novices in learning to teach. For example, preservice teachers examine and reflect on videorecords of teaching—their own, their colleagues', and that of experienced teachers. We provide, as decompositions of practice, analytic tools to help novices see the component parts of given elements of teaching and key in on salient features. We also use approximations of practice such as rehearsals. Rehearsals provide lower-stakes opportunities for deliberate practice (Ericsson et al., 1993), or opportunities for repeated, purposeful engagement in the activity to be learned, with focused feedback.

2.4 High-leverage science teaching practices

High-leverage teaching practices are practices that teachers—across grade level, content, and context—use regularly to support students' learning. These are fundamental to teaching and are selected to be generalizable, useful, and teachable (Ball & Sleep, Boerst, et al., 2009; see also Davis & Boerst, 2014; Grossman, 2018; Kloser, 2014; Windschitl et al., 2012). High-leverage practices “include tasks and activities that are essential for skillful beginning teachers to understand, take responsibility for, and be prepared to carry out to enact their core instructional responsibilities” (Ball & Forzani, 2009; p. 504). Kloser (2014) and Windschitl et al. (2012) proposed core or high-leverage practices for secondary science teaching—practices like supporting students' evidence-based explanations (Windschitl & Barton, 2016) or facilitating classroom discourse (Kloser, 2014). There has been little work on high-leverage teaching practices in elementary science, and it is an open empirical question as to whether the same set of teaching practices would be “high leverage” across elementary and secondary, given the differences in both children's development and the standards themselves.

Drawing on literature in science education and teacher education (e.g., Ball & Sleep, Boerst, et al., 2009; Kloser, 2014; Windschitl et al., 2012), we focus on six high-leverage practices that we see as crucial when enacting science teaching that supports scientific sensemaking among elementary-aged learners. These high-leverage science teaching practices include: (1) Supporting students to construct scientific explanations and arguments; (2) Choosing and using representations, examples, and models of science content and practices; (3) Leading science sensemaking discussions; (4) Eliciting and probing students' thinking about science; (5) Setting up and managing small-group investigations; and (6) Developing norms for discourse and work that reflect the discipline of science. Many of these align with explorations of high-leverage practices outside of science education (e.g., eliciting students' thinking in math, Shaughnessy & Boerst, 2018; leading a discussion in English Language Arts, Alston et al., 2018), as well as with Kloser's (2014) and/or Windschitl et al. (2012) sets of practices. Differences are, in part, differences in how practices are organized or parsed. For example, the set explored here and the two science education sets mentioned include practices related to eliciting student thinking and to evidence-based explanations or models. The two practices we focus on that are not quite in either of the other sets are practices (5) and (6) above. While Kloser includes “engaging students in investigations,” that version of the practice does not focus as much on the management issues related to hands-on investigation. Neither set explicitly calls out developing norms for discourse in the discipline, though scientific discourse and reflecting the discipline of science are threaded throughout both sets. These practices might be particularly salient in elementary grades.

We are not arguing that this is all an elementary teacher needs to be able to do, nor do we claim this to be a definitive set of high-leverage elementary science teaching practices (or even that there is such a set). We do, however, argue that a teacher does need to be able to do these things to be effective. Furthermore, this set is consistent with the literature on teachers' practices that support student sensemaking—practices such as engaging in discussion, teacher questioning, making connections, and increasing challenge (Fitzgerald & Palincsar, 2019). How are these practices high-leverage in science teaching?

Given the ways in which the development of scientific explanations and arguments is emphasized in current literature and reforms (e.g., Chen et al., 2017; C. Schwarz et al., 2017; Manz, 2015a; McNeill, 2009; NRC, 2012; Varelas et al., 2008; Zembal-Saul, 2009), the teaching practices to support this development become paramount. For example, teachers take on a range of roles to facilitate discourse (e.g., moderator, coach; Chen et al., 2017) to help students make sense of material activity and press them to make evidence-based explanations (Windschitl et al., 2012) and to generalize across cases (Manz, 2015b), sometimes supported by a claim-evidence-reasoning framework (e.g., McNeill, 2009). To support the development of explanations and arguments, one important element entails choosing and using representations—for example, representations of students' ideas and/or data (Windschitl et al., 2012) and instructional representations that are accurate, comprehensible, appropriate, and contribute to learning (McDiarmid et al., 1989).

Teachers also must be able to skillfully facilitate a discussion in science—encouraging students to take up one another's ideas and selecting particular students to present their ideas (e.g., B. A. Brown & Spang, 2008; Cartier et al., 2013; Kloser, 2014; Windschitl et al., 2012). At a smaller grain size, teachers must be able to elicit students' ideas to learn more about their thinking, often through using carefully chosen questions or tasks (Driver et al., 1994; Kloser, 2014; Vosniadou & Brewer, 1992; Windschitl et al., 2012), to check interpretations (Shaughnessy & Boerst, 2018), and to hear the science in children's thinking (Suárez, 2020). Pragmatically, science teachers must be able to manage students engaging in investigations, as emphasized in the Framework (NRC, 2012), in a manner that is productive for sensemaking (Manz, 2015b; National Academies of Sciences Engineering and Medicine, 2022; Patterson, 2019). This requires management skills including, for example, clear directions as well as purposeful circulating and monitoring (e.g., Fantilli & McDougall, 2009).

Throughout their work, teachers should be able to develop norms for students' discourse and work that reflect the discipline of science (C. Schwarz et al., 2017). For example, Berland and Reiser (2009) discuss the importance of teachers engaging students in scientific discourse, Zembal-Saul et al. (2013) present models of teacher talk for supporting students in argumentation (e.g., requesting evidence to support claims), and Herrenkohl and Guerra (1998) show how routines can be put in place to support students to internalize such norms.

Table 1 identifies key dimensions or sub-practices that may be visible in enactments of each of these six teaching practices, based on literature reviewed above and summarized in the table (Table 1 thus also provides the basis for our coding rubrics, as discussed in the Section 3).

Teachers must enact these practices synergistically and equitably in their classroom science teaching to support each student in sensemaking. There are interrelationships among the dimensions of practices, as well. For example, to develop evidence-based explanations through discourse, one needs to be able to support students in responding to their classmates' ideas—also part of leading a discussion (e.g., Alston et al., 2018)—and to request evidence to support claims—also part of developing norms for discourse that reflect science disciplines (Arias & Davis, 2017; Windschitl et al., 2012). Such overlaps are inherent in the work of teaching. We identify the key sub-practices of any given teaching practice while acknowledging these overlaps and synergies. Decomposing these practices (Grossman et al., 2009) provides analytic clarity but is not intended to suggest that any practice is constituted by a unique set of sub-practices.

2.5 Purpose and contribution

In sum, teaching science rigorously and equitably entails teaching practices, including the six featured in the present study, that teachers use together to support student sensemaking. While the field has made progress identifying the features of teachers' practice that are consistent with contemporary reform efforts (e.g., Kang & Windschitl, 2018; Thompson et al., 2013), and is beginning to develop methods for analyzing teachers' emergent practices (e.g., Kang & Zinger, 2019; Shaughnessy & Boerst, 2018; Sleep, 2012), we know little about how beginning elementary science teachers engage in these practices. This gap motivates the present study. We focus on describing how beginning elementary teachers engage in high-leverage science teaching practices, identifying similarities and variations, and illustrating “images of the possible,” thus highlighting some of these teachers' many strengths.

2.6 Theoretical framework

From a situated perspective (J. S. Brown et al., 1989; Putnam & Borko, 2000), concepts, events, and practices are given meaning in the contexts in which they are encountered and learning is accomplished by being situated in specific phenomena, rather than in an abstract, general, or context-free manner. Over time, with repeated opportunity, what is learned in a situated manner can be generalized across contexts. This perspective undergirds practice-based teacher education, demonstrating interconnections among activity, knowledge, and practice.

Furthermore, as noted, from a sociocultural perspective, practice entails the orchestration of skill, knowledge, and identity in working toward a communal goal (Berland, 2011; Grossman, 2018). To develop a set of high-leverage teaching practices, beginning teachers can be supported to engage in deliberate practice (Ericsson et al., 1993). By engaging with representations, decompositions, and approximations of practice (Grossman et al., 2009), using tools and frameworks to support them (Ghousseini et al., 2015; Patterson Williams et al., 2020), novice teachers begin to develop their capacity with a set of high-leverage teaching practices, as well as content knowledge for teaching (Ball et al., 2008). We trace their engagement in specific high-leverage practices, acknowledging that the enactment of these practices is likely to reflect variation based on a teacher's classroom context (e.g., school setting, grade level, student demographics), the lesson's focus, and the supports in place.

3 METHODS

To explore how novice elementary teachers engage in key high-leverage science teaching practices (HLSTPs) over time, we follow them longitudinally across their teacher education program and through their first year as classroom teachers. This study is one slice of a larger longitudinal study that looks at these same beginning teachers' knowledge and practice related to elementary science teaching, following them through 2 years of teacher education and into 1 or 2 years of classroom practice. Here, we zoom in on the teachers' engagement in a set of HLSTPs. Other work from the larger project reports on the teachers' content knowledge for teaching science and knowledge of science practices (Bennion et al., 2022b; Bismack et al., 2022), opportunities to learn (Bismack, 2019), and the role of context in shaping the teachers' instruction (Bennion et al., 2022a).

3.1 Teacher education program as initial context

The first component of this study took place within a coherent, practice-based elementary teacher education program in the United States (Davis & Boerst, 2014). The program enrolls undergraduate students in their third (i.e., junior) and fourth (senior) years at a large, public, Midwestern, research-intensive university. During the first and second semesters of the program, interns (our program's name for preservice teachers) took classes (e.g., multicultural education focused on culturally responsive pedagogy) and were placed in elementary classrooms for 6 h/week. During the third semester (the beginning of the timeline of this paper), interns took science methods (designed and taught by the first author) as well as other courses (e.g., mathematics methods, multicultural education focused on working with families), and had a third field placement. Interns taught in the same field placement during the third semester (for 9 h/week) and fourth semester (full time student teaching). Throughout the program, they also experienced less typical classes or sequences, including a course focused on children's sensemaking (designed and taught by the second author). Several of these courses, including the sensemaking class and elementary science methods (ESM), incorporated rehearsals and other pedagogies of practice. Throughout their coursework, interns gained exposure to—and used and learned to use with children—a wide range of forms of scaffolding or cognitive tools, such as sentence starters, instructional frameworks, and resources about children's ideas (Kademian & Davis, 2020).

3.2 Participants and teaching contexts

We purposively selected five participants during their second year in the program.1 They were selected to capture interest in science and feature some diversity (e.g., science concentrators and nonconcentrators) among a population that, for the most part, was not highly diverse and represented the elementary teaching force in the United States (i.e., mostly white, middle-class women; National Academies of Sciences Engineering and Medicine, 2022). For example, Diana had some previous science teaching experience in a camp setting, Ginny had camp counselor experience, and Sandra was a science concentrator.2 Claudia had a strong science score on a college admissions test and was a particularly strong participant in the science methods class (as was Harry). Harry identified as a white man, and the other four participants as white women. All five were enthusiastic about science teaching and about participating in the study and anticipated being able to teach science during student teaching.

In addressing our research question, we were interested in both similarities and differences in how the teachers engaged in the high-leverage practices, and we suspected that their teaching contexts (see Table 2) might shape that engagement. In general, Allium School District is a relatively highly-resourced district and York School District is a relatively poorly-resourced district with greater demographic diversity and lower socioeconomic status than Allium; all interns complete student teaching in one of these two towns. The participants went on to different public school contexts for their first year of teaching, the specifics of which, of course, we could not anticipate at the time of their selection as focal participants. Their experiences differed. For example, Claudia continued in an international baccalaureate (IB) school context, and occasionally linked science concepts or practices to school-wide IB value statements. Ginny continued in Allium, as did Diana, though Diana worked as a long-term substitute. Harry moved to a western state, where his school served many recent immigrants. Finally, Sandra went to a poorly-resourced district and shared, in an email to the first author during her first week teaching: “My school unfortunately does not have science on the schedule until third grade. We are supposed to integrate science into teacher directed reading time as well as guided reading.” Sandra nonetheless wanted to continue to participate in the study, in part because she valued the teaching of science and wanted to teach science despite the school-based limitations on how she could do so. In both her student teaching and first-year classrooms, Sandra faced the biggest challenges in her school contexts, among all of our participants, with limited support for her science teaching and more classroom management difficulties; her engagement in science teaching was itself a success given her schools' explicit focus on reading.

Along with the others, we maintained Diana and Sandra in our study, despite the differences in their first-year teaching (1YT) contexts (i.e., long-term sub and lack of school support for science), because we suspected that we could continue to learn about beginning elementary teachers' practice from studying their teaching. Indeed, we see the variability among the five participants' 1YT classroom contexts as a strength that allows us to explore how these teaching practices can be enacted in a range of settings and situations, including—like Sandra's and like many in the United States (Banilower et al., 2018; National Academies of Sciences, Engineering, and Medicine, 2022)—elementary classrooms where there is little explicit administrative support for science.

3.3 Data sources

Because of our interest in the participants' science teaching, we drew primarily on videorecords, including records from the participants' science lessons taught during the science methods course, student teaching, and their first year of teaching. We collected between five and 11 lessons for each participant (see Table 3). These lessons ranged from roughly 25 min to 2 h each, with some lessons stretched over multiple days. For the science methods course, interns developed, enacted, videorecorded, and reflected on an investigation-based science lesson that entailed student sensemaking. We asked them to record a similar lesson during student teaching. Then, at the start of their first year of teaching, we asked each participant to provide us—over the course of the school year—with videorecords from three science lessons; we emphasized our goal of seeing investigation-based lessons that included a sensemaking component, but recognized that there would be variation in what teachers provided. In the most extreme case, as noted above, Sandra was told by her school administrators that she was not to teach science except as a vehicle for teaching literacy; two of her three lessons from her first year were literacy lessons with a science content focus. Claudia and Harry both chose to provide more first-year lessons than we requested. We draw on all of the lessons each participant provided.

We used interview data and written artifacts to triangulate and elaborate our claims. We conducted a semistructured interview after each lesson or sequence of lessons provided. The interviews during the first year of classroom teaching included a stimulated recall portion in which we asked directly about events in the video that were flagged as of interest within the larger project. Each interview lasted approximately 60 min and was conducted face-to-face or via videoconferencing, and was audio- or video-recorded. Each was then transcribed verbatim. We coded these interviews along many dimensions across the project as a whole, including where participants were discussing specific teaching practices. We also asked participants for written artifacts (e.g., lesson plans), which they sometimes provided. These were not coded, because of their idiosyncratic nature. While interview transcripts provided important insights into participants' ideas about the high-leverage teaching practices (e.g., what they entailed and why participants employed them), and the written artifacts provided additional insights for some participants (e.g., their intentions for a lesson), in this study we drew mainly on the video data as a way to answer our fundamental question about the teachers' practice.

3.4 Data coding

To code the video data, we developed detailed rubrics that described key characteristics of each HLSTP and identified indicators of different levels of performance. Then, through three phases, we coded the data. In Phase 1, we reviewed each videorecord, jotting notes about what was happening moment-by-moment and indicating when specific HLSTPs were occurring. In Phase 2, we watched each videorecord again, this time using the qualitative data analysis software Atlas.ti to more precisely code specific time segments that involved a high-leverage teaching practice, and sharpening the identification of the practices in play. In Phase 3, we watched each videorecord again, this time consolidating our viewing to focus on all segments related to a particular HLSTP. We used the coding rubrics to characterize the participant's engagement in the dimensions of the HLSTP. Table 1, above, summarizes these rubrics, providing references for the literature on which we drew as we developed each dimension.3 Drawing on the literature in this way helped us to work toward both construct validity and face validity (Creswell, 2007) thus ensuring the credibility and, more generally, trustworthiness of our findings (Lincoln & Guba, 1985) (How we addressed credibility and trustworthiness in our coding itself is discussed below).

As an example, Figure 1 presents the coding rubric used for supporting students to construct scientific explanations and arguments. To meet the standard for this HLSTP as a novice elementary teacher, we determined that the teacher must support students in making a claim that answers the investigation question (aligning with questions labeled “dimension 1” in Table 1, for supporting students to construct scientific explanations and arguments) and is supported by appropriate and sufficient evidence (“dimension 2” questions). If the claim was also supported by reasoning (“dimension 3”), the reasoning might elaborate on how a mechanism functioned, but not the underlying why (which would be “exceeds expectations”). The teacher would need to demonstrate understanding of the components of an argument (i.e., what is meant by claim, evidence, and reasoning (“dimension 4”); and engage in some form of discourse to codevelop an argument with their students (“dimension 5”). We gave one score for each dimension or sub-practice per lesson. After watching all of the segments associated with the HLSTP for a particular participant and lesson, we scored them using the relevant rubric, giving one overall score per lesson. These scores ranged from 1 (did not meet expectations), 2 (approached expectations), 3 (met expectations), to 4 (exceeded expectations).4

To further bolster trustworthiness and credibility, we also wanted to ensure interrater reliability in our coding. A second rater scored approximately 20% of the full data corpus from all five participants to check for interrater reliability. Cohen's κ for five of the practices ranged from 0.81 to 0.92, indicating substantial agreement. One practice, leading a science sensemaking discussion, was more challenging to reach agreement on, particularly in relation to whether the teacher was making purposeful contributions and how they oriented students to one another. We reached moderate agreement (Cohen's κ 0.60) for this practice, resolved differences through discussion, and clarified the coding scheme.

3.5 Data analysis

To guide our analyses of the videos, we first explored how each focal participant engaged in each HLSTP at the different time points, developing matrices and descriptions to characterize the performances. Next, we used those analyses to describe performance of each practice across time, using color-coded matrices, descriptive statistics, and graphs. Then, we used descriptive statistics to characterize the scores in multiple ways (e.g., mean score across all instances of a practice, mean score across practices for a participant, etc.). We used these analyses to identify patterns and deviations from patterns, rather than to make strong claims about general performance.

Then, we looked at dimensions of each practice, counting how often each score (from “does not meet expectations” to “exceeds expectations,” or 1−4) was applied at the level of dimension (i.e., a row from a scoring rubric, rather than for the rubric as a whole). That is, given the coded data as described above, we counted the scores for each dimension, across the lessons. For the supporting students to construct scientific explanations and arguments practice depicted in Figure 1, for example, the mode for dimension 1 (about making a claim that answers the investigation question) was a score of 4, with n = 11 of the 23 lessons scoring at that level (and n = 5 scoring 2, and n = 7 scoring 3.) This means that the most typical way of engaging in this sub-practice was to exceed expectations: to help students make an accurate claim, supported by evidence and reasoning, that addresses the investigation question. Using modal scores helped us to characterize what typical practice looked like, as well as how variable the engagement was.

Finally, we used all of these analyses to develop exemplar cases for each HLSTP. For each, we chose instances of performance that “exceeded expectations” and organized case descriptions by the specific dimensions identified in the scoring rubrics. The Appendix summarizes these exemplars and why we selected them; one for each practice is presented in the Results. The exemplars are intended to reflect images of the possible, not typical engagement.

3.6 Researcher positionality

Though there are differences in our backgrounds, both authors are white, cisgender women with settler heritage, with privilege in our current professional and social contexts. This positionality is in some ways similar to that of our participants, four of whom were white women and all of whom attended a top public university. Further, we share an interest in learning about the successes, and challenges, of our interns, with the goal of enhancing our program. However, our backgrounds may limit what we see in these teachers' practice, by shaping our expectations around students' participation and classroom norms. We have attempted to guard against such limitations (e.g., by discussing the data and our interpretations with our research team members).

4 RESULTS: EARLY ENGAGEMENT IN HIGH-LEVERAGE SCIENCE TEACHING PRACTICES

In presenting the results, we first provide an overview of how the participants engaged in the practices, which was variable. The mean scores indicate that across the practices, the participants demonstrated the strongest performance with eliciting students' thinking (with an overall mean of 3.23, n = 30 instances) and the least strong performance with choosing and using representations (with a mean of 2.75, n = 16 instances) and establishing norms for discourse and work that reflect the discipline of science (with a mean of 2.73, n = 33 instances). Participants varied in the strength with which they engaged in the practices. For example, Diana and Harry showed the greatest strength with supporting students in constructing explanations and arguments, whereas Claudia showed the greatest strength with several other practices. Claudia and Ginny both demonstrated particular strength with eliciting students' thinking. Table 4 presents the means for each practice as demonstrated by each participant and overall.

Table 4 also shows the number of instances of each practice seen across the lessons, providing evidence that, on average, some practices appeared in teachers' lessons more frequently than others. For example, choosing and using representations was the practice least likely to appear, with just over half of the lessons (17 out of 33 that we received total) receiving no score for this practice; the practice was also used less frequently during the first year of teaching (in 10/24 lessons) than it was during the program (in 6/9 lessons). Supporting students in constructing scientific explanations and arguments was required during the program, but during the first year of teaching, 10 lessons (of 24) received no score for this practice. Not surprisingly, given the parameters of her classroom contexts and their emphasis on English Language Arts, Sandra's lessons were less likely than the other teachers' lessons to include some practices; for example, none of Sandra's lessons from her first year of teaching included supporting students in constructing scientific explanations and arguments or leading a science sensemaking discussion. Otherwise, participants' lessons often included the practices.

These quantitative results are intended to provide an overview of how often and how skillfully the beginning teachers engaged in these practices, at a gross level. The more interesting story, though, is what it looks like when beginning elementary teachers engage in this study. Next, then, we report results for each practice in turn. For each, we describe the characteristics of the Enacted Practice that we tend to see (and not see), using the modal subscores associated with each dimension of that practice's rubric and drawing on data from our observations and interviews to illustrate these sub-practices. In doing so, we draw on data from all of the participants. Then, for each practice, we present an exemplar intended to reflect high-quality novice engagement in the practice, focusing on a specific lesson enactment. These exemplars are intended to serve as images of the possible—to illustrate what novice teachers are able to do at their best. They thus complement the analysis based on the modal scores, which show typical performance. Throughout, we use quotes from interviews to round out these findings.

4.1 Supporting students in constructing scientific explanations and arguments

Supporting students in constructing scientific explanations and arguments is a key practice for supporting sensemaking. As summarized in Table 1, this practice, ideally, entails making a claim that answers the investigation question, supporting it with appropriate and sufficient evidence and with reasoning that addresses the mechanism underlying a phenomenon (e.g., how energy transfer causes a container of water to heat up). It also requires that the teacher understand the notions of “claim,” “evidence,” and “reasoning,” and—again, ideally—involves developing an argument through rich discourse with children. Table 4 shows that 23 of the 33 total lessons included this practice (9/9 lessons in the program and 14/24 lessons in the first year of teaching). On average, participants “met expectations” with this practice (mean of 3.04).

Teachers addressed in interviews how they valued explanation and argumentation for sensemaking. Claudia, for example, identified the importance of evidence and reasoning, saying, in an interview after science methods, that the end of a lesson would involve “making sense of the data and then synthesizing it into a larger scientific concept” (Claudia, int. post-ESM). We turn next to how this practice typically looked in our data.

4.1.1 Characteristics of the Enacted Practice

We examined modal subscores associated with each dimension of the rubric (summarized above and in Table 1; some elements are combined in the headings here, for readability) for the practice of supporting students in constructing explanations and arguments, and draw largely on video data from participants' enactments to illustrate typical engagement in these sub-practices. Table 5 presents these modal scores to give a flavor of the typical performance. We also examined interview transcripts to learn more about how the teachers thought about this practice and its sub-practices.

Table 5. Descriptive statistics (modes) related to supporting students to construct scientific explanations and arguments

Making a claim, using evidence, and using reasoning

Based on the modal subscores, participants typically were able to support the development of an accurate claim that answered the investigation question, using evidence and reasoning. Furthermore, they typically used appropriate and sufficient evidence and were able to support students to use reasoning that addressed the mechanism underlying a phenomenon, when they attempted to develop reasoning. For example, in an engineering lesson exploring how a tightrope walker stayed balanced, Harry asked his students to consider why their model tightrope walker stayed balanced. He asked them to use the sentence stem, “I think this works because…” (Harry, 1YT winter, Lesson 4). Similarly, in a lesson on batteries and bulbs, Harry used a series of questions (e.g., “Do the wires have the energy? What purpose do the wires serve?”) to help his students recognize the importance of the flow of electric charge (Harry, 1YT spring, Lesson 6). These are examples of Harry supporting his students in mechanistic reasoning. Toward the same end, as a scaffold, Claudia sometimes seeded ideas for students to incorporate into their reasoning. For example, in a force and motion lesson, she asked students to include key words from their investigation, like inertia, force, and speed, as they developed their reasoning (Claudia, 1YT winter, Lesson 2).

Teachers often discussed using the claim-evidence-reasoning framework (a framework introduced in science methods class) to support this practice. In fact, some teachers talked about using the claim-evidence-reasoning framework across subject areas. After science methods, for example, Harry noted that “It's a structure that I've learned from science methods that I've used in all my subjects actually—[the] explicit structure of the claim, evidence, reasoning” (Harry, int. post-ESM), noting that he used sentence starters to help his students use the framework. The teachers also, however, expressed less confidence with the use of reasoning. Ginny, for example, revealed that she was glad her class had run out of time so she didn't have to get to the science principle that would support reasoning (Ginny, int. post-ESM).

Understanding the elements

As shown in Table 5, participants typically met expectations in demonstrating an understanding of the elements of an explanation or argument (i.e., claim, evidence, and reasoning). After using argumentation less during the early part of his first year, by spring of his first year of teaching, Harry was using similar sentence starters (introduced in science methods) for argumentation across all subjects, including science, to give his students “common language” across subjects (Harry, int. spring 1YT). In a lesson right before the interview, Harry walked his students through the meaning of each element, saying, for example, “What's your evidence? How do you know that works?” and “The reasoning is why” (Harry, 1YT spring, Lesson 6). As another example, Claudia commented that a student “jumped ahead” to provide reasoning when Claudia had asked for a claim, and then, shifting to focus on reasoning, asked “Tell me what that teaches you” (Claudia, 1YT winter, Lesson 2). This exchange suggested that Claudia had a sense of the roles that claim, evidence, and reasoning play in building an argument.

Using discourse to build an argument

Participants typically met expectations in constructing the argument with students; these instances reflected a mid-level performance between a rote or algorithmic development of the argument and a rich, discursive construction. Harry's sentence starters reflect a typical way teachers supported this kind of discourse.

Thus, the analysis of the modal subscores demonstrates variability across instances of the practice, but, overall, shows that participants engaged in each of the sub-practices effectively.

4.1.2 Exemplar of the Practice

We use an exemplar to illustrate what the practice of supporting students in constructing explanations and arguments can look like in a novice elementary teacher's classroom when enacted well. In the exemplar in Figure 2, Diana was teaching a lesson on the five senses (using the challenge of identifying unlabeled salt and sugar jars) to the kindergartners in her student teaching classroom (Diana, ST) (Diana and Harry were the most successful with this practice; Table 4). The exemplar is organized around dimensions of the practice, as specified in our coding rubrics. Throughout the lesson, Diana connected to the work of scientists and had students build an age-appropriate explanation and argument. Diana's engagement in supporting explanation and argumentation shared characteristics with that of other teachers who supported their students effectively in explanation and argumentation.

4.2 Choosing and using representations, examples, and models of science content and practices

The practice of choosing and using representations, examples, and models of science content and practices is complex. As summarized in Table 1, it involves using (ideally multiple) representations that are accurate and comprehensible, contribute effectively to students' learning, and are appropriate within the context at hand. Furthermore, the practice involves using precise language in discussing the representation(s), and identifying salient aspects of the representation(s) while sidelining those that are potentially distracting. In addition, the teacher needs to be able to make their own thinking about the representation visible and accessible to students. As noted above, this practice appeared relatively infrequently (n = 16 times, among 33 total lessons) in the data, with quite a drop in the first year of teaching (10/24 lessons, compared to 6/9 lessons during the program). Participants on average scored somewhat below “meets expectations,” with an overall mean score of 2.75 (Table 4).

We chose to use “representations” as a broad term including scientific models, data tables, graphs, and other instructional representations. We wanted to be inclusive to capture the kinds of things that were happening in elementary classrooms around the more general work of representation. Ginny, for example, spoke about the importance of helping students recognize how what she called “models” can be used to support student learning:

[O]ne of the big objectives that was written into the lesson was actually introducing the concept of models to students. And it was a lesson all about bird beaks, bird bills and how their different shapes, different structures for different purposes. So we had different stations set up with different models to represent birds. … So with all these models at the beginning of each lesson we're talking about what do I mean by model, how do these work as models or why are these good models being used, and also talking about how we handle models (Ginny, int. post-ESM).

The teachers reflected on how representations furthered their instructional goals for sensemaking and about the importance of helping students recognize the utility of the representations.

4.2.1 Characteristics of the Enacted Practice

We used our rubric's sub-practices to dig deeper into how participants engaged in the practice of choosing and using representations. Table 6 presents these modes. In interviews, teachers discussed their work with representations.

Table 6. Descriptive statistics (modes) related to choosing and using representations, examples, and models of science content and practices

Number and quality of representations

Based on the analysis of the modal subscores, our results suggest that participants typically were able to use a single representation that correctly and appropriately represented the content, and that was comprehensible to students, contributed to student learning, and was appropriate in context (see McDiarmid et al., 1989). Half of the lessons scored showed this characteristic. For example, Harry spoke about coconstructing a data table and helping students see how the structure of the table would help them to answer their investigation question (Harry, int. post-ESM). Later, he described his use of a food web as a model that allowed for the development and communication of scientific ideas (Harry, int. post-student teaching [ST]). Ginny used light-ray diagrams at multiple points in her lesson (including having students generate them on the board) and also briefly used a physical model (Ginny, ST). However, participants did not typically use multiple representations. Most of the time, when teachers used a representation, they used but one.

Using language carefully

Participants typically used language carefully in discussing representations (see Table 6). For example, Diana corrected and clarified a student's use of language to refer to different caterpillar parts, when the class was looking at a representation of a caterpillar (Diana, 1YT spring, Lesson 2).

Identifying core features

As shown in Table 6, participants typically identified some of the salient aspects of a given representation. For example, Ginny briefly highlighted for students that a cross-section (of a stem) could let one “see what is going on inside” and “zoom in—see those details” (Ginny, 1YT spring). Similarly, Sandra, in her lesson on the life cycle of a butterfly, helped students notice that the diagram showed that all the elements in the diagram were connected by arrows to indicate the cyclicality (Sandra, 1YT spring, Lesson 1). Typically, though, the participants neglected to identify all the salient features of a representation.

Sidelining distracting features

Claudia was the only participant who occasionally focused students' attention away from potentially distracting ideas about representations (a dimension of practice we considered to exceed expectations). For example, in working with her students on graphing their data, Claudia stated, “We always start with the x-axis” when a student said they would begin plotting a point by “going up to” the y-value (Claudia, 1YT fall, Lesson 4). While, of course, one can go up and then over when graphing, this was a reasonable teaching move within the context of a lesson focused not on graphing but on pendulum motion. The students might need a reminder that the x coordinate is written first in the x, y pair.

Making thinking visible

We found that only Claudia and Diana made their own thinking visible about representations (another dimension we deemed as exceeding expectations). For example, Diana made her thinking visible by noting, “I made a mistake” (in identifying a caterpillar's parts) and then going on to correct the mistake through modeling what she observed as she looked at an image of a caterpillar (Diana, 1YT spring, Lesson 1).

4.2.2 Exemplar of the Practice

Claudia taught a lesson in the fall of her first year of teaching that we use as an exemplar of choosing and using representations, examples, and models of science content and practices (see Figure 3), providing an image of the possible for this practice. She used a representation that allowed students to see patterns in data, building a physical graph out of pendulum strings to show the relationship between a pendulum's length and its period (Claudia, 1YT fall, Lesson 3). In an interview, Claudia noted the graph's importance for supporting student sensemaking. She and her students also codeveloped a data table.

4.3 Leading a science sensemaking discussion

Another practice to support students' sensemaking is leading a science sensemaking discussion. As summarized in Table 1, leading such a discussion involves using students' ideas as resources to build toward collective knowledge in the classroom and making purposeful contributions oneself. It also involves supporting multiple students in contributing orally, listening actively, and responding to (and learning from) other students' contributions. Ideally, it also involves purposefully selecting ideas to infuse into the discussion (e.g., by listening to small group conversations and enlisting students to share their ideas or calling on specific students at specific moments in the discussion). Table 4 shows that lessons often included this practice (26 of 33 lessons, with all 9 lessons during the program and 17/24 lessons in the first year), but participants typically scored short of meeting expectations (mean of 2.77).

The teachers often described the purpose of discussion as supporting student sensemaking. For example, after student teaching, Claudia said,

[D]iscussion is really important because I think that's where all of the construction of knowledge comes. They've seen all these things. They've made some inferences. They have this data … from this lesson, but once we have the discussions I can kind of guide the student[s] to where they need to go (Claudia, int. post-ST).

When asked how she tries to support student sensemaking, during her first year of teaching, Claudia said straightforwardly, “I really want to do it through discussion” (Claudia, int. 1YT fall). The teachers often used discussion to support student sensemaking and used a range of approaches to help them reach this goal.

4.3.1 Characteristics of the Enacted Practice

We examined the participants' practice using the dimensions of our rubric. Table 7 presents the modal scores. In interviews, the teachers also described some characteristics of their discussions, noting the importance of asking probing questions, using multiple participation structures, and giving ownership to students.

Table 7. Descriptive statistics (modes) related to leading science sensemaking discussions

Using student ideas as resources

Based on the modal subscores, participants almost always used students' ideas as resources to build toward collective knowledge. For example, Claudia described attributing ideas to specific students. She said, “I really like to give ownership to the students. So if Ayesha comes up with a really great idea then I'll refer back to it as, “Ayesha told us or predicted” or “Ayesha's theory was this”” (Claudia, int. post-ST). We saw Claudia do this consistently throughout her lessons, and other participants made similar moves.

Making purposeful contributions

Participants almost always made purposeful contributions of their own. For example, in his tightrope walker lesson, Harry led a discussion of the students' models. He asked questions like “Was the pole straight, or bent?” (Harry, 1YT winter) and encouraged students to constantly map back and forth between their models and the real-world phenomenon they had observed in a video. Similarly, in a later lesson on magnets, Harry asked “Do magnets stick to every type of metal?” and, through his contributions, eventually led students to recognize the importance of iron in the materials they had tested (Harry, 1YT spring, Lesson 5). Toward a similar end, during her student teaching lesson on light, Ginny provided important insight about a child's light-ray drawing on the board (Ginny, ST).

Supporting student talk and purposefully selecting ideas for discussion

Participants typically supported several students in contributing to a discussion and listening actively. Ginny worked on broadening student participation. After her science methods class lesson, Ginny said:

Every student needs to gain something from it and they all need to be able to understand and access the content. That's why I try and incorporate a lot of discussion into my lessons where I'm getting representatives from each group talking and I'm getting students to do turn and talks, and I'm saying, “How does this compare to this?” … And I'm trying to be intentional about who I call on, who I'm asking to speak, whose ideas we're listening to so we hear a wide variety of perspectives and thoughts. (Ginny, int. post-ESM)

However, although Ginny mentions trying to be intentional, we saw limited evidence of participants tracking students' ideas to purposefully infuse them into the discussion (see Table 7). Diana was one teacher who did engage in this purposeful work. In her first science lesson, from ESM, Diana tracked students' ideas and purposefully selected ideas that she wanted to infuse into the discussion. She taught a kindergarten lesson on buoyancy in her student teaching classroom. At one point in the discussion, she connected back to a specific idea that had been brought up in the initial discussion, asking a student to comment on his initial thinking, now that the class has explored the ideas with the investigation. In an interview after this lesson, Diana, like Ginny, talked about the importance of broadening participation, describing how she determines who to call on: “now I'll highlight a couple names sometimes that I want to—if I hear them say something, like, “Can I call on you?”” (Diana, int. post-ESM).

Furthermore, participants sometimes supported students in participating in the discussion, but not in learning from and responding to one another's ideas; to meet expectations, the teacher needed to orient students to one another. Diana did emphasize collective knowledge-building through talking and listening to each other (Diana, int. post-ST; Diana, int. 1YT fall).

Thus, the analysis by dimension of this practice indicates strong performance related to using students' ideas as resources and making strong contributions oneself. Participants were more variable, though, in how they supported student talk so students would learn from one another. Interviews suggest that they did, however, recognize the importance of trying to do so.

4.3.2 Exemplar of the Practice.

As an exemplar of leading a sensemaking discussion, we turn to a lesson from Claudia's first year of teaching, focused on how mass affects the motion of a vehicle (Claudia, 1YT winter, Lesson 3); the lesson took place after a series of investigations adding masses to carts in different ways to track effects on motion. Figure 4 illustrates Claudia's lesson to provide an image of the possible with regard to leading a discussion (Claudia and Harry were most successful, among their colleagues, with this practice). Claudia's enactment reflected strength in several dimensions of the leading a science sensemaking discussion practice. This exemplar reflects the kinds of work that teachers were able to do when they successfully led a sensemaking discussion.

Using effective questions and using tasks

Based on the modal subscores, participants typically used effective questions and tasks to elicit students' ideas. For example, Claudia and Ginny both spoke of how they elicited students' observations, with Ginny saying:

While they're in their exploration investigation or whatever, we're asking questions like, “Well, what are your observations so far? What are you noticing? Why do you think that's happening?” (Ginny, int. post-ST)

Ginny commented regularly in interviews on how she wanted to become better at eliciting and responding to students' thinking. After her science methods class, she stated, “Well, with my peer teaching [rehearsals] I didn't do very well at it. That was something I got a lot of feedback on, “You should probably have some more questions to elicit student thinking and prior knowledge”” (Ginny, int. post-ESM). Ginny seems to have taken this feedback seriously; she consistently showed effective skill with this study.

Following up on student language

As shown in Table 8, participants typically followed up on the specific language students used, asking questions like “What does that mean?” (e.g., Harry, 1YT fall, Lesson 2) or pushing for detail (e.g., Harry, 1YT spring, Lesson 5).

Checking interpretations of student ideas and making reasonable interpretations

Table 8 shows that participants typically checked alternative interpretations of students' ideas. For example, Diana made statements like “I hear you saying…” (Diana, 1YT fall) and Claudia similarly used language like “But what I heard you saying is…” (Claudia, 1YT fall, Lesson 3), or, for Ginny, “Let me know if I've got this right” (Ginny, ESM). They also typically made reasonable interpretations of students' ideas. For example, Diana connected a student contribution about bristles on a caterpillar to “sensitive hairs” (Diana, 1YT spring, Lesson 2).

Thus, in sum, participants, with the exception of Sandra, effectively engaged in most dimensions of eliciting students' ideas, yet there was still variability across dimensions.

4.4.2 Exemplar of the practice.

We use an example from a lesson Ginny taught in her first year of teaching to serve as an exemplar of how novices could elicit students' thinking in science (Ginny and Claudia were most successful with this practice, across the participants see Table 4). Ginny taught a lesson on the purpose of stems for transporting water as a part of a unit on plants, getting at structure-function relationships (Ginny, 1YT, spring, Lesson 1). In this lesson, a celery stem is placed in water with red food dye, and students notice red coloring in the “strings” (xylem) and leaves, illustrating the role of stems in water movement. Ginny's strong approach for eliciting ideas here was fairly consistent with what she and other participants did throughout most of their lessons. Figure 5 illustrates Ginny's lesson and how she elicited ideas.

4.5 Setting up and managing small group investigations

To support students' sensemaking about natural phenomena, teachers must be able to set up and manage small group investigations. Table 1 summarizes that this practice involves using small groups at appropriate moments, and giving clear directions for students' small-group work. It also involves developing mechanisms for holding groups and individuals accountable for small-group work and, at the same time, supporting students in collaboratively engaging in the investigative work. This practice also involves circulating purposefully and monitoring students' work. This practice was prominent in participants' lessons (27 of 33; 7 of 9 and 20 of 24 lessons in the program and first year of teaching, respectively). Participants scored, on average, just short of meeting expectations (mean of 2.93) see Table 4.

Claudia commented on the importance of having students work in small groups during investigations, noting that different children observe different things (Claudia, int. 1YT winter), while Sandra expressed a preference for whole-group work (Sandra, int. post-ST). In interviews, the teachers identified characteristics of effective small group management, including giving directions for the investigation procedure, identifying safety issues, conceptualizing and assigning roles, and circulating and monitoring.

4.5.1 Characteristics of the Enacted Practice

We used participants' scores on the dimensions of the rubric associated with setting up and managing small group investigations, to examine their practice. Table 9 presents these modes. We also examined interview transcripts to explore how they discussed the sub-practices associated with this practice.

Table 9. Descriptive statistics (modes) related to setting up and managing small-group investigations

Appropriate use of small group work and giving clear directions

Participants used small group work when appropriate in the context of the lesson, based on the modal subscores in Table 9. One exception revealed the challenges of limited resources: Sandra, teaching a lesson on life cycles, brought a single jar of caterpillars to each group for quick observations, rather than having students work in small groups to observe the caterpillars (this is in contrast to the successful use of small groups by Diana for a similar lesson, described in the exemplar below).

Based on the analysis of the modal subscores, participants almost always gave clear directions to allow groups to work independently. Claudia talked about the importance of clearly explaining the procedure for the investigation and noted that she used interactive modeling (a strategy she learned about in a teacher education course immediately before science methods) to support giving these directions (Claudia, int. post-ESM). All participants were able to do so.

Holding students accountable

Participants almost always implemented mechanisms that would hold students accountable, as Table 9 indicates. Usually, this involved having an individual notebook, packet, or worksheet. They never, however, implemented mechanisms that would hold students accountable for both collective and individual learning. We did not, for example, see teachers have students work together on written work.

Supporting development of collaboration strategies

The modes in Table 9 show that participants almost always supported students to develop strategies for collaborating. Several teachers discussed the importance of assigning roles to group members. For example, in describing a lesson on heat energy transfer, Claudia said:

I had a time keeper, a recorder, and then a hot thermometer reader and a cold thermometer reader and talking about checking up on each other and each person having a role. And so we were really working on being accurate but also making sure that each person had a role within the experiment because I have kids that will just do the whole thing and the other kids will just sit back. So I really wanted to provide a way for each kid to be invested. (Claudia, int. post-ESM)

We saw Claudia use roles consistently throughout her lessons. In her enactments, she clearly described each group member's responsibilities.

In addition to group roles, Harry also used sentence stems to scaffold students' language for groupwork. Harry taught his first year in a context with a high proportion of emergent multilingual learners, so language was a prominent concern for him. Harry told us,

[At the start of the school year] I would probably have sentence frames to actually scaffold the conversations they were having in their groups. So saying there is a partner A and a partner B, partner A is going to do this part of the task and they're going to say—they're going to make an observation about this. Then the other student is going to respond by saying, “Oh, I agree with you.” Or, “I want to build on this.” … [So] we've built that up just being able to have discourse in the classroom. (Harry, int. 1YT spring)

While most teachers used roles, at least sometimes, Harry was the only teacher using language scaffolds for small group conversations.

Monitoring and circulating

Participants almost always monitored groups' work, and some discussed the importance of monitoring in interviews. Participants also almost always circulated purposefully and with consistent questions or comments to drive toward learning goals—but never seemed to deliberately choose which groups to work with, when, and on what (see Table 9). Claudia described “going around and talking to groups [and asking questions like] “what are you noticing”?” (Claudia, int. 1YT fall). We saw her (and others) doing this consistently. For example, while small groups were brainstorming an investigation question to ask, Claudia circulated and asked “Looking at these samples [of questions], how could you rewrite this?” or “How could you write this as a question?” (Claudia, 1YT fall, Lesson 3). Later, when students were building their pendulums, as she circulated, she said “What's the first step?” and “Let me see the one you built. How long is it? Prove to me that that is 45 cm” (to emphasize the importance of using accurate lengths in the investigation) (Claudia, 1YT fall, Lesson 3).

In sum, participants' engagement in the dimensions of supporting small groups' investigations was uniformly strong, with only a few cases where participants did not meet or only approached expectations. In contrast to some of the other practices, they also had numerous specific moves or techniques that they described using in their interviews, suggesting that perhaps they had more nuanced understandings of this management practice than they did of some of the other practices in which they engaged.

4.5.2 Exemplar of the Practice.

We draw on a lesson Diana taught in the spring of her first year of teaching as an exemplar of how a novice can support students in working in small groups on science investigations (see Figure 6). The lesson focused on observing caterpillars to illustrate structure-function relationships (Diana, 1YT spring, Lesson 1). Diana had been in this long-term substitute position long enough to have established rapport with her second-graders and to have set up some of the classroom culture that she wanted. Diana's work here was consistent with that of other participants who were able to effectively support small-group investigation work.

4.6 Establishing norms for discourse and work that reflect the discipline of science

There were two major categories of work within the practice of establishing norms for discourse and work that reflect the discipline of science: supporting the science practices and supporting the development of academic language. Table 1 summarizes what is entailed in this practice, including helping students to learn the norms for discourse and work that are important in science, and helping them know why these norms are important. Another aspect of this is helping students to understand the intent of these norms, and providing opportunities to practice them. One key element here is regularly asking students to provide evidence to support their claims. This practice also involves supporting students in developing key academic language to aid them in communicating about science. This practice was visible in all 33 of the lessons, but, perhaps reflective of the multiple challenging facets of this practice, participants scored worse on this practice, on average, than any other (mean 2.73; Table 4).

4.6.1 Characteristics of the Enacted Practice.

We used participants' scores on the dimensions of the rubric associated with establishing norms for discourse and work that reflect the discipline of science to help us examine their practice carefully, and examined interview transcripts to explore their talk about the sub-practices. Table 10 presents the modes.

Table 10. Descriptive statistics (modes) related to developing norms for discourse and work that reflect the discipline of science, such as asking for evidence to support claims

Supporting learning about the norms for discourse and work in the discipline

Based on the modal subscores, participants typically taught students about the norms of discourse and work in the discipline. Some teachers commented on the importance of engaging students in science practices. For example, regarding her pendulums lesson described above, Claudia said:

[S]cientific investigation and just the variables are [ideas] that stem throughout all of science. … [R]eally, I want [students] to be able to use data, use a graph to draw a conclusion. And so, it's less so just the literal, what is a pendulum? It's more so, like, how do I use an investigation to answer a question? And then, how do I use the evidence to draw a conclusion? (Claudia, int. 1YT fall)

Claudia and Harry consistently named science practices as a key learning goal, in addition to science content; Harry commonly made comments like “They're learning how to do science” (Harry, int. 1YT fall). In these ways, throughout their lessons, they knitted together the science practices with the science content of the lesson. Similarly, Diana discussed how she came to focus a kindergarten unit around the science practices:

I had been talking to the class, like even before I started my unit, and we were talking about scientists. And one kid was like, “Well, I'm not a scientist. I'm not an adult.” So then I was like, “Well, I'm going to switch the focus of my unit to being like, this is what scientists can do and you can do all these things. And you're a scientist.” (Diana, int. post-ST).

Indeed, Diana consistently treated her students as people who construct knowledge in and do science and referred to them as “scientists.” Through her use of a poster focused on “what scientists do,” Diana regularly supported her kindergartners in learning about norms of discourse and work in the discipline of science during her student teaching lesson. Toward the beginning of the lesson, Diana asked her students to put on their (imaginary) scientist outfits, including goggles, gloves, and a lab coat. With her students, she explored the ideas that scientists make predictions, ask questions, observe, record, and test. For example, toward the beginning of the lesson, Diana noted that a student's contribution was a prediction. She said to the class, “What is a prediction?” After taking a student contribution, Diana revoiced the idea as, “It's a guess and you see if you're right.” She went on to say, “Scientists can predict. We're going to add this [picture] to our poster of what scientists can do” (Diana, ST). At the end of the lesson, Diana recapped by narrating what scientific work the students did that day, noting that they observed, tested their predictions, asked questions, and recorded what they observed—all in the context of the content they were working on related to the five senses (Diana, ST).

Explaining, modeling, labeling, and providing practice with the norms

Participants typically made some moves that supported students in learning to engage in the norms of discourse and work—but sometimes did not do this at all and only occasionally provided multiple ways of promoting this learning. For example, Harry modeled how to measure the size of a rock he was observing, defining diameter and circumference (Harry, 1YT fall, Lesson 2) but did not explain why one might want both of these measurements. Similarly, Ginny intentionally modeled a poor observational drawing, and then, in explaining how one would make a stronger drawing, she noted that “Scientists are very detailed. That's why I gave you a magnifying glass” (Ginny, 1YT fall)—but like Harry, she did not explain why the details would be important.

Supporting learning about the why and how

Thus, participants rarely helped students know why the norms of the disciplines were important and how to use them, as shown in Table 10. While Harry did not often explain why science practices were important in science, he did consistently describe students as “being scientists” (e.g., Harry, 1YT fall, Lesson 3). Ginny used similar language, noting that student were working “as scientists” (e.g., Ginny, 1YT fall), as did Diana, as noted above—getting at a more superficial level of why the practices mattered. Claudia was the only participant who more fully supported the “why” of the norms of science practices in multiple instances. For example, across her first year of teaching in multiple lessons on force and motion, Claudia supported her students in understanding why students could use a graph to predict a result they hadn't directly measured, why it was important to control variables, why it was important to use measurement units, and why multiple trials might be important.

Asking for evidence for claims

Table 10 shows that participants typically requested evidence for claims and engaged students in argumentation. For example, in a lesson that involved observing isopods, Harry circulated among the small groups and consistently asked students what their observations could tell them about the isopods' preferences. He did this argumentation work with students in a subsequent whole-class discussion, as well (Harry, 1YT fall, Lesson 3). Claudia, similarly, routinely asked her students for evidence from their investigation in a food webs lesson during student teaching (Claudia, ST).

Developing academic vocabulary

Finally, participants typically supported students in developing academic vocabulary, though in less than half of the instances did participants support academic vocabulary by coconstructing definitions with their students, as Table 10 indicates. With regard to academic language, Claudia was explicit about how students needed to have a reason to use the academic language. For example, she said:

I think that language is such an important thing and that you shouldn't dumb it down …. And I think with a lot of scientific language, that's best understood through experiences and not just giving definitions. …. And I think language is something that every kid has the right to and like I don't want to look at an ESL kid and be like, “They're not going to get it so I just won't talk—I'll use the simple language with them,” because that's not supporting their learning. (Claudia, int. post-ESM)

As an example of Claudia's emphasis on using accurate academic language, after the pendulum investigation described above, Claudia commented, “they [the curriculum materials] used the word swingers, which drives me nuts that they don't call it pendulums” (Claudia, int. 1YT fall). We saw Claudia (and others) use appropriate language throughout multiple lessons.

As another example of developing academic language, Diana also supported her young students in learning language beyond the language of the science practices, like descriptive words they could use for describing textures or tastes. For example, she asked, “What are some words we use to describe what we feel?” (Diana, ST). She used similar approaches for developing vocabulary around each sense.

In sum, engagement in the establishing norms for discourse and work high-leverage practice was variable across all of the dimensions of the practice, but we saw some instances of effective engagement in these sub-practices.

4.6.2 Exemplar of the Practice

As an exemplar of the establishing norms for discourse and work in the discipline practice, we turn to a lesson from Claudia, from the winter of her first year of teaching. Claudia scored notably higher than her colleagues with this practice (Table 4). The lesson focused on what happens to a vehicle when a weight attached to it falls (Claudia, 1YT winter, Lesson 2). In this lesson (and others), Claudia taught her students about norms of discourse of science and how knowledge in science is constructed and shared, as the class worked together to understand and engage in science practices in ways that were similar to what Claudia had done during ESM. Figure 7 illustrates Claudia's lesson, providing an image of the possible while revealing the complexity of doing this study.

5 DISCUSSION AND AREAS FOR FURTHER WORK

Our central research question was, How do novice elementary teachers engage in key high-leverage science teaching practices? We hoped to identify similarities and variations in how novices enacted these high-leverage practices and to illustrate if and how these novices might engage in exemplary versions of the practices. In the sections that follow, we first summarize the main results, briefly describing how these participants engaged in the practices. We identify some of the strengths that seemed unexpected given the literature, and consider how teachers' scaffolding for their students seemed to support their practice—another unexpected finding. Second, we conjecture about the role of school context in the teachers' performance, with particular attention to Sandra and Harry. Third, we examine what counts as a high-leverage practice and how empirical findings can inform that question. Finally, we reiterate the importance of seeing how these practices work together toward supporting sensemaking, as part of a teaching system. With each contribution, we also note areas for future research.

5.1 Novice elementary teachers' engagement in high-leverage science teaching practices

The five participants in this study were mostly able to engage in a set of high-leverage science teaching practices. In our analysis, we were able to untangle teaching into some of its component parts, separating an individual teacher's performance on six different teaching practices. Different teachers showed particular strength with different practices. For example, Ginny, who did not experience as much success with supporting students in explanation and argumentation, was able to shine when she elicited students' thinking. Harry, on the other hand, exhibited less strength than his colleagues with eliciting students' thinking. He excelled, however, at supporting students in explanation and argumentation and in leading a science sensemaking discussion. While the teachers did not demonstrate exceptional practice all the time, there were important glimmers of success across these novice teachers.

5.1.1 Surprising strengths of these novice elementary teachers

The literature base focused on novice teachers, and novice elementary teachers of science in particular, often defaults to a deficit framing, focusing on what teachers cannot do or do not know (Gray, McDonald, & Stroupe, 2021; Zembal-Saul, Carlone, & Brown, 2020). Taking, instead, an asset framing, we name here three strengths of these participants' teaching. These are strengths that the literature would lead us to expect to be missing from novice practice, but that we believe may serve as important building blocks for future practice. Each warrants further research. First, these participants were relatively successful in supporting students to use mechanistic reasoning in supporting their claims, such as when Harry encouraged his students to consider the role of the wires in a lesson on electricity. Some teachers, for example, seeded science concepts that had arisen during a discussion as terms that students could use to develop their reasoning, as Claudia did with terms like “inertia” and “speed.” This work on reasoning is challenging for both teachers and students (e.g., McNeill & Knight, 2013; Zangori & Forbes, 2013). Another surprising strength these teachers showed was their work with small groups. Classroom management is known to be a struggle for new teachers (Fantilli & McDougall, 2009). Yet these teachers, for the most part, were able to give clear directions, monitor effectively, and help students learn to collaborate effectively and work together in groups. Finally, a third surprising strength was the degree to which these teachers engaged their students in the science practices, integrating science practices with science content in meaningful ways, throughout the many investigations we observed. Their support for establishing norms for discourse and work in the discipline was not especially sophisticated—but they did not shy away from trying to engage in this integration and support children in science practices as a way to support sensemaking about natural phenomena. Given the prominence of activity-focused elementary science, at the expense of sensemaking (Appleton, 2002; Plumley, 2019), we found this attention to science practice—even though it sometimes missed the mark—to be promising.

5.1.2 The role of scaffolding students in supporting novice teachers

Furthermore, we gain some hints about what might support some of these novices' strengths. One emergent theme that appeared in our video and interview data was that these teachers considered carefully the scaffolding and cognitive tools they provided for their students, integral to sensemaking (Fitzgerald & Palincsar, 2019) in situated and sociocultural perspectives on learning. For example, participants talked about the sentence stems they provided for students for constructing explanations and arguments (e.g., McNeill & Krajcik, 2009; Zembal-Saul et al., 2013); Harry provided similar sentence stems for small-group work (e.g., Herrenkohl & Guerra, 1998); Claudia had a poster of discussion talk moves (e.g., Zembal-Saul et al., 2013) for her students to use; Diana made a poster naming and depicting the science practices her kindergartners were engaging in; Ginny and Claudia both used back-pocket questions (Windschitl et al., 2012) in their interactions with small groups who were making observations—the list goes on. They learned about some of these (though not all) in our teacher education program (e.g., the CER structure for explanation and argumentation, some uses of sentence stems, the importance of talk moves and back pocket questions).

The scaffolding and cognitive tools that these teachers used with children gave them insight into what their students were thinking and were able to do. We hypothesize that this helps teachers experience more success with some of these sophisticated teaching practices, which might, in turn, lead them to continue to work on and thus improve their use of these practices. We see this as an important area for further research, because it suggests that one important focus of teacher education might be to help novices learn specific ways of scaffolding their students to engage in ambitious work. Tools (e.g., Kademian & Davis, 2020; Ghousseini et al., 2015; Windschitl et al., 2012) provide one way to provide this support.

5.2 Context as a shaper of teacher development

We have said relatively little about Sandra, who experienced less success with these practices than did her colleagues. Given her schools' priorities, Sandra's teaching of science at all was in some ways a subversive act, and one to be celebrated. For example, she infused first-hand investigation into her instruction as a first-year teacher in a context where teaching science exclusively as a part of literacy instruction was both the norm and the stated expectation. In both Sandra's student teaching placement and her first year teaching classroom, Sandra's management issues were significant. She had far fewer physical and curricular resources than the other teachers, and science was barely taught—reflecting common circumstances for teachers in lower-resourced schools (Banilower et al., 2018). Ronfeldt (2012) explored the role of high-needs field placements during teacher education and found, overall, that novice teachers may not be well-served by such placements in terms of their long-term success. Sandra's experience seems to fit that pattern. While Sandra's field placement during student teaching had numerous strengths, it did not seem to afford her the opportunities she may have benefited from for growing her science teaching practice—and her first-year school was similar. These findings also may suggest the need for a more expansive set of high-leverage practices that better suit elementary science teaching in contexts with fewer resources.

In contrast, Harry's schools—like Sandra's, situated in districts with fewer material resources—provided a rigorous academic orientation including for science, and his first-year school provided professional development (particularly around supporting emergent multilingual learners), a cohesive grade-level team, and a supportive principal. Overall, Harry's first-year context was extraordinarily supportive for a novice teacher—probably more supportive than any of our other participants'—and Harry clearly thrived in that professional context.

This was not a study of the effects of context, though we take up this focus more centrally elsewhere in our work (Bennion et al., 2022b), showing how characteristics of context can serve as resources, not just constraints; for example, across our sample, teachers were able to draw on varied contextual resources (at the level of classroom, school, and community), including the standards, the priority given to science teaching in the district, the presence of science specialists, and families' funds of knowledge, as well as curriculum materials and time. But the contrasts between Sandra's experience and Harry's are striking to us, and suggest the importance of supporting novice teachers through multiple means, both within and beyond the school (NASEM, 2015). Future work should take seriously how context can bolster or dampen a novice teacher's work in elementary science. We hypothesize that much more is at play than simply demographics or physical location. For example, school leadership likely plays an important role in setting the stage for elementary science teaching (National Academies of Sciences Engineering and Medicine, 2022; Spillane et al., 2001).

5.3 What counts as “high leverage”?

As we engaged in this study, we came to recognize further conundrums we faced. We take up one such conundrum here: the question of how we determine what counts as “high leverage,” particularly with regard to the consideration of “high frequency” that is often used in identifying a practice as high leverage. Our results lead us to question whether one of our practices, choosing and using representations, was truly high leverage for these teachers. Extensive literature in science education (e.g., Harrison & Treagust, 2000; Kloser, 2014; Magnusson et al., 1999; McDiarmid at al., 1989; Windschitl et al., 2012) suggests that this should be a high-leverage science teaching practice, and Kloser's Delphi study, in particular, suggests that it is in high use in secondary science classrooms. Yet this practice was used in less than half of the lessons we observed. How high leverage is this practice if it does not show up in a majority of science lessons? What explains its relative lack of appearance?6

Based on our small study we cannot definitively answer the question of whether choosing and using representations in science is a high-leverage teaching practice in elementary science. Indeed, Plumley's (2019) analysis of the NSSME + survey data (Banilower et al., 2018) shows that in only 34% of elementary classrooms is “organize and/or represent data using tables, charts, or graphs…” done each week (p. 20)—though this was at the same time one of the more prominent science practices-related instructional activities identified in these classrooms—and that frequency drops to 19% for developing or using models. Further research should explore whether there are certain practices that might, for example, be “high leverage” for experienced teachers that perhaps novice teachers are less likely to employ, that may be more frequent in secondary science classrooms than in elementary classrooms, or that may be more prominently prompted by the available curriculum materials. Regardless, our findings illustrate the tension between practices that should be high-frequency and would be high-leverage if used, and those that in reality may be used in limited ways in classrooms.

5.4 High-leverage practices working to support sensemaking: Practice with a purpose

We close this discussion with an overarching point about the value of studying practice as a coherent endeavor. Practice-based teacher education, grounded in situated and sociocultural perspectives on learning (e.g., Grossman, 2018), is based on the premise that teachers can learn a set of teaching practices that can be used across situations to equitably support students' learning (Ball & Forzani, 2009; Grossman et al., 2009; McDonald et al., 2013). Decomposing teaching into constituent parts risks deprofessionalizing teaching into a set of technical skills. However, teaching is inherently interactive and contingent on students (Ball & Forzani, 2009) and must be responsive to students to promote equitable learning (e.g., Colley & Windschitl, 2016; Kang, 2021; Suárez, 2020). We can see such deliberation and responsiveness throughout these data—for example, in Claudia giving ownership to students' ideas, Harry carefully scaffolding his emergent multilingual learners' small group conversations to help them engage in sensemaking discourses, or Ginny puzzling over what a student's statement about a phenomenon means.

Yet the practices do not, and cannot, work in isolation. Ginny's strength in eliciting students' ideas is important, but she is limited in supporting students' sensemaking if she cannot also support them in argumentation and reasoning. Diana was skillful in helping her young students make careful observations of living and nonliving things, but was still learning to orient her students to one another in discussions. These practices are the recurrent work of the profession of teaching (Lampert, 2010), and it is important to be able to see how novices can engage in specific elements of that work. This analysis allows us to see the assets of novice elementary science teachers (National Academies of Sciences Engineering and Medicine, 2022), when so much of the literature focuses on their deficits. In the end, these teaching practices must work together to support student sensemaking for every student—toward the communal goal (Berland, 2011). And indeed, how these teachers enacted the practices suggest how a relative weakness in one practice was sometimes made up for by strength in another. Future research should continue to explore the synergies among the high-leverage practices (National Academies of Sciences Engineering and Medicine, 2022); such research could help strengthen the work of teacher education and promote equitable and just science teaching.

6 LIMITATIONS OF THE STUDY

Our rubrics allow us to characterize some specifics about each high-leverage teaching practice, and our results lead us to suspect that (e.g.,) eliciting students' thinking may be more straightforward for novice elementary science teachers than (e.g.,) establishing norms for discourse and work. A statement like this has face validity, as well. Yet a larger study would be needed to make strong claims about what practices are “harder” or “easier” for novices. Further, as noted briefly above, we—this study's researchers and authors—may be limited in what we see in these data because our perspective, as white women, largely aligns with historic norms at play in elementary classrooms (Carlone et al., 2010). Finally, other researchers might be interested in different high-leverage practices. Exploring novices' engagement in a range of practices is important for understanding their use.

7 IMPLICATIONS AND CONCLUSIONS

Incorporating improved opportunities to learn for novices related to these practices could help to mitigate some of the challenges associated with the teaching practices, which are amply evidenced in the literature (e.g., Berland & Reiser, 2009; Cartier et al., 2013; McDiarmid et al., 1989; McNeill & Knight, 2013; Windschitl et al., 2012). Our analyses informed our ongoing use of some experiences, frameworks, and cognitive tools in practice-based teacher education, and spurred us to develop new ones. Tools are a crucial form of support for novice teachers (Ghousseini et al., 2015; Patterson Williams et al., 2020; Windschitl et al., 2012). Table 11 summarizes some experiences, frameworks, and cognitive tools we suspect may be helpful for novice elementary teachers of science, based directly on these results; some of these refer to experiences, frameworks, or tools that participants had access to and seemed to reference, while others are ideas based on areas of need identified in the findings.

Table 11. Design implications for practice-based teacher education

This study helps to identify how novices engage in dimensions of specific practices, better positioning teacher educators to be able to support them. As just one example, as we looked at the practice of supporting norms for discourse and work in the discipline, novices in this study were unlikely to help their students understand why they were engaging in science practices. Teacher educators can work with preservice teachers on recognizing why sharing the why with students is important, as well as strategies for doing so.

Furthermore, this study has implications for conceptualizing early teaching practice. The field knows relatively little about how new teachers engage in specific high leverage teaching practices, particularly in elementary science. Our descriptive work begins to fill that gap.

We make three main assertions about teacher development on the basis of this study. First, these novice teachers demonstrated some important strengths in their practice. This was true despite the high expectations of current visions of elementary science teaching (National Academies of Sciences Engineering and Medicine, 2022; NRC, 2012). Identifying these strengths can help scholars to orient toward and build on teachers' assets (Gray et al., 2021; Zembal-Saul et al., 2020). Second, it was possible for us to untangle these teachers' performance on different practices (Arias & Davis, 2017; Kang & Windschitl, 2018; Shaughnessy & Boerst, 2018). We could identify strengths and weaknesses with specific dimensions of a given practice, as well. This illustrates the importance of the decomposition and recomposition of teaching practice (Grossman et al., 2009; Janssen et al., 2015)—of recognizing that the practices can be taken apart to be learned and studied but must be put together to fully support student sensemaking. Finally, the teachers' engagement in these practices was variable, across teachers and practices, supporting and extending other findings in the field (e.g., Arias & Davis, 2017; Kang & Windschitl, 2018).

Beginning teaching is exceptionally complex. Our findings should help the field as we work to gain analytic precision in understanding how novices learn to engage in high-leverage teaching practices to move toward more rigorous and equitable schooling experiences in science.

ACKNOWLEDGMENTS

This research is funded by a grant from the Spencer Foundation awarded to Elizabeth A. Davis and Annemarie Palincsar. We appreciate this generous funding. Any opinions, findings, and recommendations expressed in this paper are those of the authors. We are enormously grateful for the participation of the novice teachers who shared their practice with us; we have learned so much from them and their students. We also appreciate the insights and contributions of team members of our research project, past and present, including Amber Bismack, Adam Bennion, John-Carlos Marino, Ben Tupper, Jacquie Handley, and Rachel Kuck.

CONFLICT OF INTEREST

The authors declare no conflict of interest.