Fourth-Grade Emergent Bilingual Learnersí Scientific Reasoning Complexity, Controlled Experiment Practices, and Content Knowledge When Discussing School, Home, and Play Contexts


by Cory A. Buxton , Ale Salinas, Margarette Mahotiere, Okhee Lee & Walter G. Secada - 2015

Background: In exploring how emergent bilingual learnersí prior knowledge from home and play contexts might influence their scientific reasoning, this study drew upon two distinct research traditions: (a) experimental research from the developmental and cognitive psychology tradition, and (b) research on culturally and linguistically diverse learners from the sociocultural tradition.

Purpose: As part of a larger research project to improve science teaching and learning in culturally and linguistically diverse elementary schools, we explored the knowledge that fourth-grade emergent bilingual learners brought to the classroom from home and play contexts, as well as the knowledge that was developed in the classroom. We considered how this out-of-school and in-school knowledge related to studentsí academic abilities to reason scientifically, to follow controlled experiment practices, and to demonstrate knowledge of core science concepts in school.

Setting: The research was conducted in elementary schools in a large urban school district in the southeastern United States with a linguistically and culturally diverse student population.

Participants: A total of 81 fourth-grade students from 27 teachersí classes across six schools were interviewed during a three-year period. These students were selected for equal distribution across four ESOL levels, two home languages (Spanish and Haitian Creole), and two genders.

Intervention: After being taught by their classroom teacher using a project-developed curriculum unit on the topic of energy that was developed to specifically support bilingual learners, selected students participated in an interactive interview with a member of the research team.

Research Design: The design can be described as analytic interview, in which (a) neither control group nor pre/post comparisons were used, (b) students were selected purposefully from classrooms based on demographic criteria, (c) student responses were coded both qualitatively and quantitatively, and (d) a sufficiently large sample size was used to allow for statistical analysis of student responses.

Findings: Studentsí English proficiency level correlated with their ability to express scientific reasoning (in English), but not their ability to engage in controlled experiment practices. The home, school, or play context of the interview questions correlated with studentsí ability to express science content knowledge about energy.

Conclusions: The uneasy tension of applying both cognitive and sociocultural theoretical traditions enriches and also complicates our understanding of how students learn to reason scientifically, how they engage in controlled experiment practices, and how they express science content knowledge. Curriculum materials, student assessments, and teacher professional development can all benefit from a better understanding of how emergent bilingual learners leverage their prior knowledge and epistemologies from both home and school contexts, as they engage in science learning.

Most classroom teachers fundamentally believe that their students are capable of learning grade-appropriate material (Klassen & Chiu, 2010). With a new wave of more rigorous standards and assessments, however, many teachers are worried that schools are setting expectations out of reach for their students (Haager & Vaughn, 2013). Teachers are particularly concerned about students with additional needs in the classroom, including emergent bilingual learners (Santos, Darling-Hammond, & Cheuk, 2012). In this study, we set out to explore the range of experiences that fourth-grade emergent bilingual learners draw upon when challenged with rigorous inquiry-driven science instruction. We believe it is critical for teachers not to underestimate the abilities of their emergent bilingual learners as capable science thinkers, and that more research is needed to better guide teachers in leveraging their students’ abilities and experiences.


Bringing together two distinct research traditions that have been used to study students’ scientific reasoning, we examine the strengths and limitations of fourth-grade emergent bilingual learners’ reasoning as they engaged in a problem-solving task on the topic of transfer of energy. Although some students may demonstrate only limited proficiency in English language use, such limitations should not diminish teachers’ perceptions of these students’ competence in content area understanding or problem solving. Too often, educators make decisions about the opportunities offered to emergent bilingual learners based on an implicit or explicit deficit perspective (Buxton & Lee, 2014; Reeves, 2004). It is important to recognize that labels, such as English language learner, limited English proficient student (the federal term, often used in schools), or emergent bilingual learner, are one-dimensional categories used to identify students in schools and in research studies, and mask a wide range of prior academic knowledge and rich linguistic and cultural experiences. While highlighting the potential strengths that emergent bilingual learners bring to the science classroom, we must also realize that the increased cognitive and linguistic demands of the Next Generation Science Standards (NGSS) will create new challenges for all students, but will be especially demanding for students who must meet these challenges in a language that they are still mastering (Lee, Quinn, & Valdés, 2013).


Thus, as part of a larger research project to promote improved science teaching and learning in culturally and linguistically diverse elementary schools in a large urban district, we explored the prior knowledge that fourth-grade emergent bilingual learners brought to the classroom from home and play contexts. We considered how this out-of-school knowledge might be related to students’ academic abilities to reason scientifically, to follow sound scientific inquiry practices, and to demonstrate knowledge of core science concepts in school—all skills that will be more important under the NGSS. Better understanding of the connections that emergent bilingual learners make between in-school and out-of-school experiences may lead to both increased teacher awareness of the academic resources that students bring to their classroom and improved teaching strategies that enhance emergent bilingual learners’ academic success. In short, a focus on scientific reasoning may help teachers, students, and researchers to better understand the connections that students make between in-school and out-of-school science learning.


Scientific reasoning indicates the logical thought processes used for problem solving during science tasks (Kuhn, 2005). Scientific reasoning constitutes a core element of scientific practice, yet teachers in elementary science classrooms often give little explicit attention to such reasoning (Osborne, 2010). At home and in other non-school contexts, children experience phenomena closely related to the elementary science curriculum, but may not be asked by family members, caregivers, or peers to explore or discuss those connections (Bell, Lewenstein, Shouse, & Feder, 2009). By examining emergent bilingual learners’ reasoning about science tasks through discussion of both in-school and out-of-school contexts, science educators can develop a better understanding of the cultural resources and experiences that students bring to the science classroom, while also clarifying the challenges faced by emergent bilingual learners when attempting to explain science concepts and engage in scientific inquiry practices in a second language.


CONCEPTUAL FRAMEWORK


In exploring how emergent bilingual learners’ prior knowledge from home and play contexts might influence their scientific reasoning, this study built a new conceptual framework that draws upon two distinct research traditions to examine scientific reasoning with emergent bilingual learners: (a) experimental research in developmental and cognitive psychology, and (b) research on culturally and linguistically diverse learners from sociocultural perspectives. As these research traditions have developed out of separate fields of study, bridging the two traditions produces an uneasy pairing that may pull the research, and the researcher, in quite different directions. We believe, however, that neither research tradition alone can fully explain both the strengths and challenges that emergent bilingual learners face when engaging in the scientific reasoning that will be so important to the goals of the NGSS. Pushing beyond our epistemological and methodological comfort zones may allow us to raise important new questions and develop new models that would not seem pertinent when viewed from one research tradition alone.


We wish to note from the outset that in the present study, we did not find it possible to give equal emphasis to both research traditions. In this study, the developmental and cognitive psychology tradition took precedence, at least methodologically, while the sociocultural tradition provided important interpretations of results that would likely be missed using the developmental and cognitive psychology tradition alone. Thus, our methods were guided more by the developmental and cognitive psychology tradition (e.g., one interview per student occurring in the school setting rather than ethnographic data collection; large numbers of students interviewed in order to better discern patterns, rather than fewer students studied in greater detail; and a focus on students’ scientific reasoning, content knowledge, and controlled experiment practices over a focus on language development); yet we also intentionally used the sociocultural tradition to offer alternative interpretations of the findings.


DEVELOPMENTAL AND COGNITIVE PSYCHOLOGY TRADITION AND STUDENTS’ SCIENTIFIC REASONING


Research on scientific reasoning has its foundations in the highly controlled experimental studies of classical developmental and cognitive psychology. Since the work of Piaget (Inhelder & Piaget, 1964; Piaget, 1973), reasoning has been construed as the required thinking skills for doing scientific inquiry through experimentation (Keil & Wilson, 2000), evaluating evidence (Klahr, 2000), and engaging in argumentation in the service of promoting scientific understanding (Kuhn, 1991). From this perspective, reasoning has been viewed primarily as an individual cognitive process.


Cognitive scientists have long argued for the powerful role that scientific reasoning plays in promoting robust conceptual understanding. Brown and Campione (1986) claimed that understanding is more likely to occur “when a student is required to explain, elaborate, or defend his or her position to others; the burden of explanation is often the push needed to make him or her evaluate, integrate, and elaborate knowledge in new ways” (p. 1066). Similarly, Bereiter and Scardamalia (2003) warned about the danger of shallow constructivism, in which students do activities but are not asked to think deeply about the phenomena they observe. Thus, one of the strengths of considering scientific reasoning from the developmental and cognitive psychology perspective is that it helps to maintain a focus on moving students (and teachers) toward rich knowledge construction that is built on both theory and experience, and strengthened through explanation and elaboration.


Historically, there have been two main approaches used to study and interpret scientific reasoning within the developmental and cognitive psychology tradition, sometimes referred to as experimentation strategy and conceptual change (Schauble, 1996) and other times as domain-general and domain-specific reasoning. These approaches have often been viewed as contradictory, with experimentation strategy research attempting to develop domain-general reasoning with the focus on skill transfer (e.g., Kuhn, 2005), whereas conceptual change research attempts to develop domain-specific reasoning that directly connects reasoning and content knowledge (e.g., Keil & Wilson, 2000). Thus, in the first approach, reasoning skills can be developed separate from content knowledge, whereas in the second approach, reasoning and content knowledge are inextricably linked.


The more recent syntheses of how and where children learn science and how this knowledge can be leveraged to improve science teaching (National Research Council, 2007, 2011) have brought together the domain-general and domain-specific approaches to reasoning by supporting the idea that useful knowledge of science includes a blend of science and engineering practices, crosscutting concepts, and disciplinary core ideas that are developed in a broad range of life-wide learning contexts. When viewed in this way, a learner benefits from repeated opportunities to reason about complex ideas in both domain-general and domain-specific ways and across a range of formal and informal learning settings. Thus, the lessons learned about scientific reasoning from the developmental and cognitive psychology tradition are critical for understanding the challenges that are embedded in the goals of the NGSS. When considering the science learning needs of emergent bilingual learners, however, these lessons prove necessary but not sufficient. Additional insights about students’ diverse cultural and linguistic resources from the sociocultural tradition are also necessary for a more complete understanding of how emergent bilingual learners learn to reason about science (Janzen, 2008; Rosebery & Warren, 2008).


SOCIOCULTURAL TRADITION AND STUDENTS’ SCIENTIFIC REASONING


Scientific reasoning has seldom been the primary focus of research on science teaching and learning within the sociocultural tradition. Instead, much of this research has centered on issues of equity and social justice, such as ensuring equitable learning opportunities (Calabrese Barton, Tan, & O’Neill, 2014; Tate, 2001), closing the achievement gap (Lee, Deaktor, Enders, & Lambert, 2008), or understanding why students of color may not readily embrace a science learner identity (Brown, 2006; Parsons, 2014). In addition, there are two strands of sociocultural research in science education that are directly relevant to understanding emergent bilingual learners’ scientific reasoning: research on epistemological and linguistic continuity and research on contextualized content.


The first strand of sociocultural research involves epistemological and linguistic continuity: the notion that learners perceive a sense of continuity or discontinuity between (1) the academic tasks that they are asked to perform in formal science learning contexts, and (2) the cultural ways of viewing and communicating about the natural world that are prevalent in their home and community. For example, Aikenhead’s (1997, 2001) research on integrating Western and Aboriginal views of human and nature relationships has shown that students engage more fully in academic learning when they do not perceive a fundamental conflict between how humans’ place in the natural world is portrayed in school science and the lessons they have learned about human and nature interactions at home. Similarly, research by Hudicourt-Barnes (2003) in a Haitian community and Parsons (2008) in an African-American community has shown that linguistic continuity may either naturally exist or be intentionally fostered between the inquiry and discourse practices used in the scientific community and the discursive styles common in certain linguistic minority communities. Complicating these findings somewhat, Brian Brown and colleagues (Brown, 2006; Brown & Spang, 2008) studied the ways in which marginalized students negotiate discursive identities in high school science classrooms. By deconstructing the cultural and linguistic moves that define and grant access to the classroom culture of science, Brown was able to help marginalized students, including emergent bilingual learners, become comfortable in adopting both scientific worldviews (epistemology) and science inquiry practices. However, many of Brown’s students expressed a great deal of resistance toward appropriating the discursive norms of science, including the style of arguing from evidence that is a central marker of scientific reasoning within the developmental and cognitive psychology tradition.


The second strand of sociocultural research involves contextualized content, which focuses on linking specific classroom science content to be learned (as opposed to the epistemological and discursive norms discussed above) with the relevant knowledge and skills that marginalized students develop within families and communities. Much of this work has been built upon the funds of knowledge model (González, Moll, & Amanti, 2005) and has shown how students bring relevant (and often untapped) prior knowledge to the science classroom on a wide range of topics including botany (Rahm, 2002), simple machines (Hammond, 2001), and robotics (Basu & Barton, 2007).


The research on science learning from the sociocultural tradition has typically adhered to a resource pedagogies orientation (Ladson-Billings, 1995), in which the strengths and resources that marginalized students, including emergent bilingual learners, bring to the science classroom are highlighted as resources to be built upon. Some research within this tradition, however, has also pointed to the multiple epistemological, linguistic, and cognitive complexities facing emergent bilingual learners in the science classroom, as they are asked to engage in rigorous science learning in a new language (Buxton & Lee, 2014).


PURPOSE OF THE STUDY AND RESEARCH QUESTIONS


Our reading of the research on science learning from developmental and cognitive psychology, as well as sociocultural traditions points to the value of bringing these distinctive traditions together to help teachers and researchers better understand and support emergent bilingual learners in learning to reason scientifically, to engage in scientific inquiry practices, and to make connections between lived experiences and canonical science content knowledge. As we noted above, however, any individual study is likely to place one of these traditions in a central position and the other in a supportive role. Specifically, we used aspects of both traditions to consider the following three research questions:


1.

How does the scientific reasoning of fourth-grade emergent bilingual learners vary across four dimensions of reasoning complexity (generativity, elaboration, justification, and explanation) on our reasoning task?

2.

What differences can be seen in student reasoning, controlled experiment practices, and science content knowledge across English proficiency, home language, and gender on our reasoning task? How can those differences be explained?

3.

What differences can be seen in student reasoning, controlled experiment practices, and science content knowledge across the discussions of home, school, and play contexts on our reasoning task? How can those differences be explained?


METHOD


This study reports on the findings from three years of data collected from reasoning interviews with fourth-grade emergent bilingual learners on the topic of energy. Participating students were taught a project-developed curriculum that covered the topics of energy, force and motion, and processes of life using a controlled experiment approach that was aligned with state expectations for science learning and our own ideas about supporting students’ scientific reasoning. The energy unit was completed during the first third of the school year and consisted of lessons on forms of energy, energy transformations, electricity and magnetism, heat transfer, and renewable and non-renewable energy. Lessons included activities to promote understanding of each energy concept and practice collecting and analyzing energy data. One of the lessons involved a ball-and-ramp experiment, and this task was modified to create the hands-on portion of our reasoning interview about energy. Thus, students who participated in the interviews had completed a similar investigation in class with a different variable being explored (height of ramp instead of mass of ball). There is a rich body of research on elementary students’ developing an understanding of energy through force and motion activities (e.g., Lee & Lu, 2010; Toth, Klahr, & Chen, 2000), and both our curriculum unit and our student elicitation task built on this literature.


STUDY SETTING AND PARTICIPANT SELECTION


The research was conducted in a large urban school district in the southeastern United States with a linguistically and culturally diverse student population. During the year we began the study (2005–2006), the ethnic makeup of the student population in the school district was 62% Latino/a, 26% Black (including Haitian and Caribbean Islanders), 9% White Non-Latino/a, 1% Asian or Native American, and 2% other. Across the school district, 72% of elementary students participated in free or reduced price lunch programs, and 24% were designated by the district as Limited English Proficient and were actively receiving English for Speakers of Other Languages (ESOL) services. An additional 37% of elementary students in the district were classified as ESOL exited or ESOL monitored, meaning that they had been classified as English Language Learners earlier in their schooling but were no longer receiving ESOL services.


Six elementary schools participated in the project and were selected based on three criteria: (a) percentage of district-designated Limited English Proficient students above the district average of 24% at the elementary school level, (b) percentage of students on free or reduced price lunch programs above the district average of 72% at the elementary school level, and (c) school grade of C or D according to the state’s accountability plan. Schools with grades of C and D were selected because these schools met our broader project goal to work with schools where students were struggling academically. Schools with a grade of F were omitted because these schools were placed under direct control of district administration, and their teachers had little to no flexibility to engage in a research project such as ours. Three of the six participating schools enrolled emergent bilingual learners who were predominantly Spanish speaking and the other three enrolled emergent bilingual learners who were predominantly Haitian Creole speaking.


Students were categorized by English proficiency (or ESOL) level based on the state English language assessment during the year they were interviewed, and were considered to be in one of four ESOL categories: (a) ESOL levels 1 and 2—referred to as beginning English speakers, usually new arrivals who had been in U.S. schools for less than one year; (b) ESOL levels 3 and 4—referred to as approaching English proficient, usually students who had been in U.S. schools for between one and three years; (c) ESOL level 5—referred to as English proficient, ESOL exited students who no longer received daily ESOL support but continued to be monitored and tested by ESOL staff for two additional years; and (d) Non-ESOL—referred to as fluent English speakers, primarily those students who had exited from ESOL programs more than two years ago; although they were no longer distinguished from native English speakers, they still needed language support in English in academic settings. We consider all of these students to be emergent bilingual learners, as they had all learned a language other than English as their mother tongue and then learned (or were learning) English as a new language.


During each of the three years of the study, one student was selected from each fourth-grade teacher’s class to participate in the scientific reasoning interview. There were a total of 27 fourth-grade classrooms across the six participating schools, and all fourth-grade classrooms were equally represented in the sample. The classrooms were distributed in a fairly uniform way across the six schools, with smaller schools having three fourth-grade science classes and larger schools having five. Each teacher was assigned a particular student demographic of ESOL level and gender (e.g., beginning English speaker, female) and asked to give a research consent form to all students who met those criteria. This ensured that the student participants would be evenly distributed across the demographic variables of ESOL level and gender.1 Thus, the student sample was representative of the racial/ethnic and gender demographics of each school as a whole but was over-representative of emergent bilingual learners as these students were the focus of our research. From the pool of students who returned signed consent forms in each class, one student was selected at random (see Table 1 for student demographics).


Each student was interviewed one time, following completion of the energy curriculum unit in class, for a total of 81 student interviews over the three years. There were no pre-instruction data collected, only performance after instruction. This design choice was partially a result of limited project resources and partially a result of our broader project goal of helping teachers to see the connections between students’ in-school and out-of-school experiences. Thus, the relevant in-school experiences were a necessary part of the interview protocol and a pre-instruction interview could not have included questions regarding in-school experiences learning about energy.


Additionally, while we were interested in students’ experiences with energy across home, school, and play contexts, the scope of the project limited our data collection to the school setting. Thus, we explored participating students’ home and play experiences with energy through asking them specific questions about energy use in all three contexts, but did not conduct observations or other data collection in home or play settings. This can be seen as another example of how the developmental and cognitive psychology tradition was more central to our conceptual framework in this particular study, with the sociocultural tradition playing a supporting role. As we describe in our conclusions, we see this study as part of a broader project to compare and contrast what can be learned about emergent bilingual learners’ scientific reasoning across multiple study settings and distinct research traditions.


DATA COLLECTION


Each student participated in the reasoning about energy interview task with a member of the research team. The energy task consisted of three parts: (a) discussing lived experiences with forms of energy at home (home context), (b) designing and performing an experiment to measure the force of rolling balls of different masses down a ramp with follow-up discussion (school context), and (c) discussing lived experiences with transfer of energy during play (play context). Sample questions from each part of the interview are included in the appendix. The school context portion of the interview was closely related to a particular controlled experiment investigation that students engaged in during the energy curriculum unit, while the home and play context portions of the interview addressed topics that were used as discussion examples in the curriculum but were not part of any controlled experiment investigations.


All interviews were conducted in English, a design decision that was made for two reasons. First, because all classroom science instruction was in English (the state in which this research was conducted has an “English only” instructional policy), we conducted the interviews in the language in which the students had received the relevant science instruction. While many of the students were more conversationally fluent in a language other than English (Haitian Creole or Spanish), they had not studied the academic material in their home language and thus were not necessarily better prepared to discuss their science ideas in that language (Solano-Flores & Trumbull, 2003). Second, the other half of this study involved showing the videotapes of the student interviews to each student’s teacher and then interviewing the teacher about the student’s reasoning skills (Buxton, et al, 2013). Thus, the student interviews needed to be in a language that was comprehensible to the teachers. While some participating teachers were bilingual in the home language of their emergent bilingual learners, many were not. Resources did not allow us to conduct, translate, and transcribe each student interview in two languages. However, all interviewers were trained ESOL specialists and skilled at supporting students in expressing their ideas in the language of instruction (English). Each interview took approximately 40 minutes and was videotaped for transcription and analysis.


CODING


Three rubrics were constructed for assessing student performance during the interviews: a scientific reasoning rubric, a controlled experiment practices rubric, and a science content knowledge rubric.


Scientific Reasoning Complexity Rubric


This rubric assesses students’ abilities using four levels of reasoning (generativity, elaboration, justification, and explanation) when describing their thinking about energy. This rubric was adapted from the reasoning analysis developed by Hogan, Nastasi, and Pressley (2000). In our modifications, specific attention was given to linguistic scaffolding in our questioning (i.e., the questions modeled language that the students could then use to frame their responses) and to the relationships between language and reasoning in our scoring (i.e., our scoring system allowed relatively simple language to result in scores representing more advanced levels of reasoning complexity). Such modifications were explicit attempts to more accurately assess emergent bilingual learners’ scientific reasoning in English while reducing the confounding of reasoning ability with English language proficiency. In other words, we attempted to minimize the role of language proficiency (whether L1 or L2) in our assessment of students’ scientific reasoning. While acknowledging the connections between language proficiency and reasoning, as discussed in the findings section below, we leave the detailed study of these relationships to a subsequent study that foregrounds the sociolinguistic tradition by using ethnographic methodology, bilingual data collection and analysis, and exploration in home and play settings.


In the present study, we focused on categorizing students’ scientific reasoning complexity by developing a two-step analytical and coding scheme. In the first step of the analysis, a semantic map of verbal and procedural responses and actions was constructed for each of the 81 student interviews. Figure 1 shows an example of one student’s semantic response map. These semantic maps allowed for identification of students’ conceptual propositions, or statements that made evidence-based claims. In the second step, these conceptual propositions were scored using the scientific reasoning complexity rubric (see Table 2) based on four reasoning criteria: (a) generativity (the amount and types of topics and assertions generated by the student), (b) elaboration (supporting details added to ideas within a topic), (c) justification (how well ideas were warranted with evidence and inference), and (d) explanation (the use of underlying concepts or mechanisms to explain an assertion).


The first two criteria of generativity and elaboration assess the variety and richness of ideas raised by students, while the final two criteria of justification and elaboration assess the structure and depth of students’ scientific reasoning. Each of the four criteria was scored on a 5-point scale (0-4), yielding four subscores plus a composite score (0-16). The following excerpts, coded from one student’s transcript (female, Haitian, approaching English proficient), provide examples of the four categories of reasoning addressed in the reasoning complexity rubric:


When I slide down the sliding board, the weight I have counts. (example of generativity—the student is making a relevant assertion)

Like if I’m really skinny I’ll go slow and not go fast, but if I’m really big I could slide down fast. (example of elaboration—the student is adding supporting details to the assertion)

The batteries make the flashlight light because if the batteries die then the light go out. (example of justification—the student is making a claim that is warranted with evidence or inference)

The batteries are chemical energy and then changes to electric energy in the flashlight when I turn it on and energy flows, and then changes to light energy when it light up, it’s a circuit. (example of explanation—the student explains an assertion with reference to underlying concepts or mechanisms)


Because the scores on the rubric are summative, the reasoning complexity score for a complete portion of the interview (home connections, school connections, play connections) cannot be inferred from any given excerpt from the interview. Rather, the scores are arrived at by first coding the entire relevant portion of the semantic map for instances of generativity, elaboration, justification, and explanation, and then applying the rubric to arrive at the total score for each reasoning complexity category.


Controlled Experiment Practices Rubric


This rubric assesses how emergent bilingual learners adhered to and explained a set of controlled experiment practices they had been taught as part of the project curriculum. These practices were aligned with the state science standards that all students were expected to master. The controlled experiment practices rubric could be applied only to the school connections section of the interview, as that was the only section in which students conducted an actual science investigation. This rubric included scales for three practices—controlling variables, taking and recording accurate measurements, and using multiple trials. The first two practices were scored 0-3 and the third practice was scored 0-2, yielding one controlled experiment score with a possible 11 total points.


Science Content Knowledge Rubric


The rubric assesses how well students give canonically correct responses to the questions about energy in each section of the interview. Some responses were scored as either correct or incorrect (e.g., What kind of energy does a flashlight need to work?), while other responses were scored on a 2- or 3-point scale (e.g., Can you explain the transfer of energy that happens when you turn the flashlight on?). Separate science content knowledge scores were calculated for the home connections section (15 total points possible), the school connections section (11 total points possible), and the play connections section (16 total points possible) of the interview. As with the controlled experiment practices, measures of acceptable science content knowledge were aligned with the state science standards for fourth grade.


Three members of the research team carried out the coding of the semantic maps each year. Each member scored the same 20% of the semantic maps using all three rubrics (scientific reasoning complexity, controlled experiment practices, and science content knowledge), and scores were compared and differences were reconciled through negotiation until consensus was reached. The same three scorers then scored additional sets of three semantic maps until at least 90% between-scorer agreement was reached. From that point on, one member of the team became the primary scorer and took responsibility for coding all remaining student semantic maps for that year. The primary scorer would bring questions to the other two scorers for discussion on an as-needed basis. The same process was followed in the second and third years of the study.


While we did not share the scoring rubrics with the students, considering the rubrics too abstract for most fourth graders, we did share our rubrics with the teachers as part of their professional development. Particularly in the case of scientific reasoning complexity, this was a significant emphasis in our professional development workshops, and we encouraged the teachers to push their students to give more details and examples and to explain their thinking in their science classes. Thus, while students participating in the interviews did not know exactly how their responses were being evaluated, we did make it clear that we were looking for them to explain their thinking in as much detail as possible, repeatedly prompting them to do so.


DATA ANALYSIS


Once all data were coded, tables were constructed to look for cross-case patterns in students’ scientific reasoning complexity, controlled experiment practices, and science content knowledge. In each case, analysis of variances (ANOVAs) was conducted with regard to three student demographic variables (ESOL level, home language, and gender) to address each of the three research questions. Effect size magnitudes are reported as partial eta22): 0 < ή2 < 0.6 is considered a small effect size, 0.6 < ή2 < .13 is considered medium, and .13 < ή2 is considered large (Cohen, 1988).


To enrich this quantitative analysis of students’ scientific reasoning about energy, a member of the research team other than the primary coder randomly selected half of the coded transcripts (40 transcripts) for additional qualitative analysis. This coding involved categorical analysis (Saldaña, 2012) of the different ways in which students connected their home, school, and play experiences to the topic of energy. Exemplars from these transcripts were then selected on the basis of being representative of students’ descriptions, rather than for being outliers or exceptional examples. In the results section below, selected exemplars are presented using the voices of the participating students to elaborate on statistical patterns.


When taken together, the three rubrics for scientific reasoning complexity, controlled experiment practices, and science content knowledge, along with the subsequent categorical analysis, provide a fairly complete picture of the participating students’ abilities to make sense of, describe (in English), and demonstrate how energy is transformed and used across a range of contexts both in and out of school. While this methodological design follows more of the developmental and cognitive psychology tradition, these results are interpreted through both psychological and sociocultural lenses described in our conceptual framework.

RESULTS


We begin by presenting the results for students’ scientific reasoning complexity, controlled experiment practices, and science content knowledge. In each case, we present the data for the complete student sample and then for subgroups disaggregated by ESOL level, home language, and gender. We also present exemplars to give student voice to patterns in the statistical results.


SCIENTIFIC REASONING COMPLEXITY


Student Sample


The means for the four categories of scientific reasoning complexity for the 81 fourth graders in the sample are shown in Table 3. Recall that the scientific reasoning complexity rubric assessed the students’ responses, starting with basic assertions and then looking to identify successively more complex levels of reasoning, including elaborations, justifications, and explanations. Results from a repeated measures ANOVA indicated significant differences across the categories of scientific reasoning complexity; F(2.389) = 195.646, p < .001. Results yielded a large effect magnitude; ή2 = 0.726. Mauchly’s test indicated that the assumption of sphericity was not met; Mauchly’s W = .671, χ2(5) = 29.009. As such, the Greenhouse-Geisser correction was applied (ε = .796).


A Bonferonni pair-wise comparison test indicated that the combinations of pairs were statistically different from one another (p ≤ .01), with the exception of elaboration and justification. In other words, students’ mean scores for each of the scientific reasoning complexity categories (generativity, elaboration, justification, and explanation) were significantly different from each other, with the exception that elaboration and justification did not differ from each other.


The highest mean score was for generativity (making relevant assertions; 3.84 [.41]), with a mean that was quite close to the maximum possible score of 4.0.


(female, Haitian, English proficient) The heavier the ball was, the farther it pushed the cup. (generativity)


The scores for elaboration (supporting details added to assertions within a topic; 2.23 [1.04]) and justification of assertions (warranting assertions with evidence or inference; 2.19 [.77]) were both intermediate among the four categories and close to the midpoint score of 2.0.


(female, Latina, fluent English speaker) When you turn on the light bulb, it feels warm and then a few minutes later it feels hot. (elaboration)

(male, Haitian, beginning English speaker) No, they don’t have same energy cause none of them don’t have same weight. (justification)


The lowest mean score was for explanation (explaining assertions with reference to underlying concepts; 0.32 [.60]), with a mean that was quite close to the minimum possible score of 0.


(male, Latino, English proficient) It depends if they are side by side or one behind the other. If they have more momentum, they’ll have more kinetic energy. (explanation)


Subgroups


Students’ ESOL level, home language (Haitian Creole or Spanish), and gender were used, respectively, as between-subject factors to determine whether any of these had a significant effect on students’ scientific reasoning complexity. Neither students’ home language (F(1) = 1.056, p > .05) nor gender (F(1) = .036, p > .05) resulted in variability in scientific reasoning complexity. Students’ reasoning complexity, however, varied by students’ ESOL level; F(3) = 6.554, p = .001 (see Table 4). A Games-Howell comparison indicated that non-ESOL students (native English speakers and students who had exited ESOL programs over two years prior to the reasoning interviews) and ESOL level 5 students (exited from ESOL programs within two years) scored significantly higher on scientific reasoning complexity than students who were still actively receiving ESOL services (ESOL levels 1/2 and ESOL levels 3/4).


(female, Latina, fluent English speaker) The washing machine, it uses the electric energy from the wall to turn the metal pole. It mixes the clothes and that’s physical. (justification)


As compared to:


(female, Latina, approaching English proficient) The washing machine does electrical for power . . . [followed by long pause but no additional answer] (assertion)


As we address in greater detail in the discussion section, this relationship between lower ESOL levels and lower scores in reasoning complexity on our interview should not be interpreted to mean that students who were less proficient in English were less capable of scientific reasoning. Rather, the relationship might suggest that they were less comfortable expressing their reasoning about the topic of energy in English on this particular assessment. While all students struggled with the more advanced levels of scientific reasoning complexity, and especially with the highest level (explanation), this struggle was exacerbated for students at the lower ESOL levels.


CONTROLLED EXPERIMENT PRACTICES


Student Sample


The mean of controlled experiment practices scores for the 81 fourth graders in the sample is shown in Table 5. Recall that the controlled experiment practices rubric assessed how well students adhered to and explained the specific practices of controlling variables, taking and recording accurate measurements, and using multiple trials, which they had been taught as being necessary to create a fair test as part of the project curriculum. The results indicate that students were able to demonstrate a moderate ability at using these practices in appropriate ways during the energy activity section of the reasoning task (56.5% accurate).


(female, Haitian, beginning English speaker) Put it here. On top. Right here, then let go. Don’t push.

(male, Haitian, English proficient) You could put the cup anywhere, but then you need to put it back to the same place each time.

(female, Latina, approaching English proficient) You have to roll it three times because you have to see each time so one time isn’t enough.


The first of the above exemplars is typical of how students, regardless of ESOL level, considered the need to control variables, in this case, the need to release each ball from the same place and in the same way each time. The second exemplar addresses how students typically thought about recording accurate measurements, such as accurately positioning the cup before the ball pushed it. The final exemplar shows how students discussed the need for multiple trials, typically stating that they were necessary without clearly elaborating why.


Subgroups


None of the subgroups of ESOL level (F(3) = 2.493, p > .05), home language (F(1) = 0.409, p > .05), or gender (F(1) = 0.501, p > .05) resulted in variability in controlled experiment practices scores. In other words, a student’s ESOL level, home language, and gender did not influence how well he or she used controlled experiment practices during the energy activity section of the reasoning task. This is an encouraging finding in that emergent bilingual learners demonstrated these practices in comparable ways to their more English-proficient peers.


SCIENCE CONTENT KNOWLEDGE


Student Sample


The mean scores for science content knowledge for the 81 fourth graders in the sample are shown in Table 6. Recall that the science content knowledge rubric assessed how well students could give canonically correct responses to questions about transfer of energy asked during all three sections of the interview (home connections, school connections, and play connections). Scores were analyzed based on the percentage correct in each category. Students expressed moderately correct science content knowledge related to the school context (50.8% accurate), less correct knowledge related to the home context (27.9% accurate), and little correct knowledge related to the play context (18.9% accurate). Results from a repeated measures ANOVA indicated significant differences across the three contexts; F(1.631) = 102.62, p < .001 (see Table 6). Results yielded a very large effect magnitude; ή2 = 0.568. Mauchly’s test indicated that the assumption of sphericity was not met; Mauchly’s W = .744, χ2(2) = 19.771. As such, the Greenhouse-Geisser correction was applied (ε = .815).


A Bonferonni pair-wise comparison test indicated that the combinations of pairs were statistically different from one another (p ≤ .05). In other words, students’ mean scores for science content knowledge across the three contexts (school, home, and play) were significantly different from one another. The highest mean score was for the school context, followed by home, and finally play.


(male, Haitian, approaching English proficient) First, the ball has energy and the cup has no energy. Then the ball hits the cup and the ball gives energy to the cup.

(female, Haitian, beginning English speaker) The battery gives electric energy, it goes to the light when I press the button on and off.

(male, Latino, English proficient) When I kick the ball up, it’s gaining energy and then when it’s coming down, it’s losing energy.


The first exemplar above is typical of the science content knowledge that students expressed when discussing the school context questions, in this case how energy is transferred from the ball to the cup. The second exemplar shows how students often showed a mix of emergent correct and incorrect content knowledge when discussing energy in the home context, in this case how energy is transferred in a flashlight. Finally, the third exemplar shows a typical student response to questions about energy in the play context that demonstrates a misconception of energy transfer, in this case when kicking a soccer ball.


Subgroups


None of the subgroups by ESOL level (F(3) = 13.966,  p > .05), home language (F(1) = 2.726, p > .05), or gender (F(1) = 3.677, p > .05) resulted in variability in content scores. Thus, as was the case for controlled experiment practices (but not the case for reasoning complexity), emergent bilingual learners demonstrated a level of content knowledge related to energy transfer that was comparable to their more English-proficient peers.


DISCUSSION AND IMPLICATIONS


The objective of this study was twofold: (1) to bring together two research traditions that typically have little to do with each other but may provide complementary insights into understanding emergent bilingual learners’ scientific reasoning, and (2) to use aspects of these two traditions to analyze and interpret the reasoning abilities of a group of fourth-grade emergent bilingual learners as they engaged in a reasoning task on the topic of transfer of energy. This was one half of a larger study in which the other part involved showing each student’s teacher the video recording of the student’s reasoning task and then interviewing the teacher regarding the student’s ideas about energy in school, home, and play contexts (Buxton, Salinas, Mahotiere, Lee & Secada, 2013). When taken together, the two parts of the study offer new insights into how emergent bilingual learners make sense of and discuss their ideas about a science topic (in this case, energy) commonly encountered in home, school, and play contexts; the resources and challenges these students bring to science learning; and perhaps most importantly, strategies that teachers (and researchers) may use to better support emergent bilingual learners’ scientific reasoning, inquiry practices, and content knowledge. This multifaceted view of science learning seems to be critical as we prepare to help all students meet the cognitive and linguistic challenges of the NGSS.


Further, this study provides important counterevidence to combat the all too prevalent deficit-oriented discourse regarding academic capabilities of emergent bilingual learners as they work to master English for academic purposes. This study highlights key aspects of science learning (e.g., applying controlled experiment practices during an investigation and expressing relevant content knowledge during an oral interview) in which emergent bilingual learners performed on par with their fluent English-speaking peers.


We begin our discussion by interpreting findings regarding emergent bilingual learners’ scientific reasoning complexity, controlled experiment practices, and content knowledge through the lens of the developmental and cognitive psychology tradition, and then follow this by reinterpreting the findings through the lens of the sociocultural tradition. As we have already noted, while our methodology was primarily grounded in the developmental and cognitive psychology tradition, interpreting the findings through a dual lens can challenge basic assumptions of both psychological and sociocultural traditions of how students learn to reason scientifically. Thus, this study highlights the need for new applications of these existing theoretical traditions.


INTERPRETING SCIENTIFIC REASONING COMPLEXITY


Developmental and Cognitive Psychology Perspective


In the developmental and cognitive psychology tradition, scientific reasoning is seen as an essential skill for gaining a conceptual understanding of science processes and principles (Bereiter & Scardamalia, 2003). While there is general agreement that reasoning skills are developed through repeated practice, there are somewhat contradictory views about the role that content knowledge plays in this process. Some research has focused on developing reasoning skills that are meant to be content-independent (Kuhn, 2005), yet research has highlighted the centrality of content knowledge in supporting reasoning abilities (Keil & Wilson, 2000). Our interest in how these contradictory claims play out with emergent bilingual learners led us to analyze our student interviews in terms of science content knowledge and controlled experiment practices as well as scientific reasoning complexity.


In our study, fourth-grade students’ scientific reasoning complexity varied substantially, with the scores being quite high for generativity of assertions, intermediate for elaboration and justification of claims, and quite low for explanation grounded in mechanism. From a developmental and cognitive psychology perspective, this pattern might well be expected. Students who participated in the project did have multiple opportunities to participate in scientific inquiry activities, with a focus on explaining their thinking. However, in order to keep pace with state science standards, no one topic or experiment was explored with the kind of depth or repetition that has typically been used in experimental research to demonstrate improved scientific reasoning complexity in either children or adults (Zimmerman, 2000). Additionally, from the observations we conducted in our project classrooms, we found that our project teachers were inconsistent in their implementation of the project curriculum when compared to the high fidelity of implementation that is common in the more controlled research that typifies the reasoning work in developmental and cognitive psychology (Lee & Buxton, 2010).


Thus, the reasoning complexity results can be explained as follows. Through project participation, students had just enough practice to become comfortable using the less complex levels of reasoning, such as generating their own ideas and assertions. As would be expected, students were largely unable to use the more complex levels of reasoning, such as explaining their assertions with reference to underlying concepts and mechanisms, as they had not received the in-depth scaffolding and repeated practice necessary to develop the content-specific expertise that is prerequisite for these advanced reasoning skills. Thus, to support upper elementary-grade students in reasoning at higher levels, explicit strategies for scaffolding these reasoning skills, including more systematic and repeated practice with key experiments, would be required.


Sociocultural Perspective


While the developmental and cognitive psychology interpretation of the patterns we saw in scientific reasoning may be reasonable for the student sample as a whole, this lens may not prove adequate for interpreting the findings in light of the unique characteristics of emergent bilingual learners. Research from the sociocultural tradition has paid particular attention to how various cultural learning and discourse patterns may at times be highly congruent with the reasoning skills necessary for successful school science learning (Rosebery &Warren, 2008), while at other times may require explicit scaffolding to support students in making (and wanting to make) these connections (Brown & Spang, 2008). Thus, a sociocultural interpretation of students’ scientific reasoning should take into account additional factors such as: (a) the overlap between normative science practices and students’ ways of making sense of and communicating about the natural world, (b) the degree to which students are willing to modify their worldviews and linguistic practices in order to more closely align with normative science practices, and (c) the connections between specific content knowledge to be learned in the science classroom and the prior knowledge and skills that students may have developed outside the classroom.


These sociocultural insights may help to further explain some of the patterns in the scientific reasoning complexity data. For example, the finding that most of the emergent bilingual learners in the study, regardless of ESOL level, expressed scientific reasoning complexity at intermediate levels of elaboration and justification, can be seen as evidence of fairly good epistemological and linguistic continuity between the science being studied in class and the experiences the students had with energy outside of school. Students did make many of these connections as they reasoned about their experiences with energy in their out-of-school lives (e.g., using flashlights or washing machines, or sliding down a sliding board) and linked those experiences to their study of energy in the school context.


However, the challenges of adopting the most complex levels of reasoning, such as explanations based on underlying mechanisms, are both linguistically and cognitively demanding. Thus, it is not surprising that English proficiency influenced students’ ability to express these advanced levels of scientific reasoning more so than the basic levels of reasoning. The finding that non-ESOL and English proficient (ESOL 5) students reasoned at significantly higher levels than beginning (ESOL 1/2) and approaching proficient (ESOL 3/4) English speakers may be interpreted not as indicative of differences in reasoning ability, but rather of differences in educational opportunity, as emergent bilingual learners were not supported in grappling with challenging content in the language in which they were most competent. Thus, if emergent bilingual learners are to be expected to reason at higher levels, science curriculum and instruction must not only make stronger connections to their prior knowledge, but must also pay specific attention to building students’ linguistic competence and comfort in using the academic language of science, including home language support.


INTERPRETING CONTROLLED EXPERIMENT PRACTICES


Developmental and Cognitive Psychology Perspective


When asked during the interview to design and carry out their own experiment on transfer of energy using rolling balls, most students showed partial understanding of the conventions of controlled experimentation, and the majority succeeded in designing a viable experiment to answer the problem posed. However, most students also violated one or more of the basic conventions of experimentation that were emphasized in the project curriculum (e.g., controlling variables, taking and recording accurate measurements, using multiple trials). Additionally, many students committed substantive errors in measurement techniques and data recording that were also emphasized in the project curriculum (e.g., lack of consistency in measurement practices across trials, misuse of estimation, misunderstanding of units of measure).


From the developmental and cognitive psychology perspective, the findings for controlled experiment practices can be interpreted in terms of scaffolded practice. Because the fourth-grade students in this study did have the opportunity to participate in a number of structured and semi-structured controlled experiment experiences through the project curriculum, they had developed some understanding of the controlled experiment practices to be used during an experiment. However, because their experiences with the project curriculum did not extend to repeated and scaffolded practice with the same or related experiments, nor to ongoing practice designing their own experiments, the majority of students were not yet fully prepared to conduct self-directed experimental design, even when the experimental task was closely modeled on a task they had already completed in class. Thus, students may need repeated opportunities for practice to gain greater ability with the desired conventions of experimentation.


Sociocultural Perspective


Unlike the assessment of scientific reasoning complexity and science content knowledge, our assessment of controlled experiment practices asked students to design and discuss an experiment related to the school context only, without explicit connection to the home and play contexts. Our rubric focused on basic conventions of experimentation that were relevant to the task in question. As noted above, the majority of students, regardless of ESOL level, demonstrated partial competence in designing and carrying out procedures relevant to the rolling ball task. However, most students, again regardless of ESOL level, also violated at least one of these experimental conventions.


From a sociocultural perspective, the finding can be interpreted to suggest that the underlying epistemological basis for controlled experiment tasks may be at odds with some students’ ways of making sense of and communicating about the natural world. That is, to these fourth-grade students, largely from Latino/a and Haitian backgrounds, experimentation may not necessarily seem more valid than knowledge received from perceived authority figures (Cobern, 2000; Lee, 1999). Thus, although ESOL level, home language (Spanish or Haitian Creole), and gender did not influence these students’ application of controlled experiment practices, their sociocultural backgrounds and epistemological worldviews may have played a factor. If we wish for all students to gain greater ability in carrying out experimental tasks, some students may need more explicit classroom discussion about similarities and differences between controlled experiment practices used in school and problem-solving practices used in their homes and communities. For emergent bilingual learners to benefit from inquiry-based science, teachers should understand the epistemologies and worldviews that underlie their students’ thinking about experimentation.


INTERPRETING SCIENCE CONTENT KNOWLEDGE


Developmental and Cognitive Psychology Perspective


Content scores for knowledge about energy varied across the three interview contexts (e.g., home activities, school science experiment, and informal play with friends) for the whole student sample. Students expressed the greatest degree of canonically correct science content knowledge in the school context, less correct knowledge in the home context, and the least correct knowledge in the play context.


From a developmental and cognitive psychology perspective, this finding can be interpreted in terms of the difficulty of transfer of content knowledge to novel contexts. The ball rolling experiment portion of the interview was quite similar to a task that students had carried out in class, so the content knowledge in this context required the least degree of knowledge transfer. Next, some of the interview questions about energy in the home context related to topics that had been specifically discussed in class as part of the project curriculum (e.g., what happens when you turn on a light switch), thus requiring an intermediate degree of knowledge transfer. Finally, the play context questions, while familiar in terms of most students’ lived experiences, were not explicitly discussed in the project curriculum. Connections between energy and play required the greatest degree of knowledge transfer, resulting in the poorest student performance. This interpretation implies that while prior knowledge may be important for new knowledge construction, having lived experiences related to an academic topic does not imply that those experiences can be readily linked to normative scientific explanations. Thus, if teachers or curriculum developers believe it is important to connect students’ out-of-school experiences with academic content, teachers will need to explicitly build in those connections.


Sociocultural Perspective


The finding from this study that may be most surprising, given the propensity for a deficit view of emergent bilingual learners’ content knowledge, was that students at all English proficiency levels performed comparably in expressing their science content knowledge. This finding is important from a sociocultural perspective, because it supports the claim that when emergent bilingual learners are given opportunities to engage in challenging inquiry-oriented science instruction, their science content knowledge improves even as their English language skills are still developing. Historically, emergent bilingual learners have received limited science instruction based on the assumption that a certain level of English proficiency is prerequisite to learning science content (Buxton & Lee, 2014). The contrary sociocultural perspective is that science learning and language learning can effectively develop in parallel, as long as curricular and instructional accommodations are made to take into account the challenges of learning rigorous content in a second language (Lee et al., 2013; Rosebery & Warren, 2008). This study would seem to support the latter view, in that emergent bilingual learners kept pace with their more English-proficient peers in terms of expressing their content knowledge about energy in our interview.


Prior sociocultural research on contextualized content has highlighted the importance of providing richer and more consistent connections between targeted school content and out-of-school knowledge and experiences (González et al., 2005). While we fully support this goal, our finding that students expressed lower science content knowledge connected to out-of-school (home and play) contexts than to in-school contexts highlights the challenge of making these connections in a sufficiently robust manner that allows students to meaningfully integrate their in-school and out-of-school experiences.


LIMITATIONS


In bringing together two theoretical traditions that typically pull the researchers’ attention in different directions, we found that methodological decisions needed to be made that served to highlight one tradition over the other. As we have noted, the broader research project of which this study was a part necessitated that we focus on the school setting, interview data after instruction, and English language data collection, all of which stood as limitations to our goals in this work.


First, we acknowledge that a study wishing to connect students’ home and play contexts to the context of school science would be well served by ethnographic data beyond the school setting, including observations and interactions involving students, families, and friends in various contexts. Such data collection was beyond the scope of this project. Findings from the study, however, provide guidance for a subsequent study to explore the same basic topics, but foregrounding the sociocultural tradition and methodology. Combining such a study with this one may provide a new range of insights for how best to simultaneously support emergent bilingual learners’ scientific reasoning, controlled experiment practices, and content knowledge.


Second, the lack of baseline data prior to instruction is a limitation of our design. A pre-post comparison of student performance on the reasoning interview would provide additional insights about the role of classroom instruction in influencing students’ abilities to express their knowledge of reasoning, controlled experiment practices, and science knowledge. As we have noted, such a design did not align well with the other half of this study in which we showed the video of each student interview to that student’s teacher, and then interviewed the teacher regarding where the student obtained her or his knowledge about energy.


Finally, conducting the student interviews solely in English limited our ability to tease apart students’ scientific reasoning skills from their English language proficiency. Interviewing students from similar backgrounds but from varying ESOL levels, including non-ESOL and ESOL-exited students, did give us some frame of reference for interpreting the role of language in emergent bilingual learners’ scientific reasoning, inquiry, and content abilities. These understandings can be deepened by subsequent research, in which students are interviewed bilingually and linguistic analysis is conducted in addition to analysis of scientific reasoning.


IMPLICATIONS FOR RESEARCH AND PRACTICE


We conclude by considering the implications of this study for future research and practice to support science learning for emergent bilingual learners. First, despite the challenges already discussed, this study points to how the tension of applying multiple theoretical traditions can both enrich and complicate our understanding of how students learn to reason scientifically, engage in science inquiry practices, and express science content knowledge. For example, from the developmental and cognitive psychology perspective, our findings point to the need for students to have repeated opportunities to practice experimentation on the same topics over time. However, when seen from the sociocultural perspective, such repeated exposure seems necessary but not sufficient to support scientific reasoning, controlled experiment practices, and content learning for emergent bilingual learners. Attention must simultaneously be paid to the epistemological and linguistic discontinuity that may cause some students to disengage from science learning due to perceived cultural conflict.


Second, a large part of the sociocultural research over the last two decades that has considered the connections between academic content and home context has been built on the funds of knowledge approach (González et al., 2005) in which attempts are made to bring out-of-school experiences into the school curriculum. Findings from this study, while certainly not minimizing the importance of teachers learning about community funds of science knowledge, imply that science educators should also do more to explicitly help students transfer the robust science content knowledge and skills learned in the classroom in the opposite direction, and help students better understand the science taking place in their home and play experiences. That is, simply looking for and pointing out life-wide connections to academic topics will not (as is sometimes implied in the sociocultural research) lead students to understand how challenging academic concepts, such as the principles of energy transfer, apply to those out-of-school experiences. More explicitly teaching about such connections may serve to reduce the transfer distance of academic knowledge and could result in a deeper bidirectional model of funds of knowledge to support all students’ science learning both in and out of school.


Third, some of the findings from this study may serve to problematize the achievement and testing gap data that often paint a deficit view of emergent bilingual learners in science. For example, beginning English speakers and approaching English proficient students (ESOL levels 1 through 4) performed comparably to non-ESOL and ESOL-exited students in both controlled experiment practices and content knowledge during our interview. On the other hand, non-ESOL and ESOL-exited students did demonstrate greater reasoning complexity than beginning English speakers and approaching English proficient students. This finding points to the double challenge of learning to reason scientifically in a language in which one is still gaining mastery, while also implying that learning to reason scientifically may be more highly language dependent than is learning science content or inquiry practices. Additional study, including bilingual interviews, should help to further clarify the role of language in learning to reason scientifically.


Thus, while science testing gaps for emergent bilingual learners do exist and should be a real concern, we must become more sophisticated both in terms of how we measure science learning and how we disentangle the role of language from other aspects of science reasoning and learning (Solano-Flores, 2008; Solano-Flores & Trumbull, 2003). In other words, linguistic challenges for emergent bilingual learners may unfairly underrepresent their true understanding of science concepts when measured by assessments that are also unintentionally measuring English language proficiency, creating testing gaps that are not truly indicative of science abilities. In our most recent work (Buxton, Allexsaht-Snider, Aghasaleh, Kayumova, Kim, Choi & Cohen, 2014), we are exploring the role of bilingual, constructed-response science assessments to support teachers in more thoughtfully examining how assessments can tease apart students’ understanding of science content knowledge, science inquiry practices, and their use of the language of science.


In conclusion, while only partially realized in this study, combining multiple research traditions, such as psychological and sociocultural perspectives, can challenge researchers to develop better models to support emergent bilingual learners in reasoning scientifically, in making use of scientific reasoning skills across in-school and out-of-school contexts, and in demonstrating their reasoning no matter how it is assessed. As we noted in the introduction, a desire among researchers to push ourselves beyond our epistemological and methodological comfort zones may allow us to raise and explore new questions that might otherwise seem irrelevant. Developing new multifaceted frameworks for studying education issues seems especially important at the present historical juncture between the increasing cultural and linguistic diversity in the U.S. student population and new cognitively and linguistically demanding standards and assessments in all content areas. Existing research has differentially highlighted both challenges facing and resources available to emergent bilingual learners in and beyond science; these challenges and resources are language-based, culture-based, and content-based. Researchers, teachers, and students can all benefit from more multidimensional interpretations of how language, culture, and content understanding intersect across life-wide learning contexts.


Notes


1. The exception being ESOL level 1 and 2 students who are underrepresented in the sample because they tended to be grouped in only a few teachers’ classes, thus limiting our access.


References


Aikenhead, G. S. (1997). Toward a First Nations cross-cultural science and technology curriculum. Science Education, 81(2), 217–238.

Aikenhead, G. S. (2001). Integrating Western and aboriginal sciences: Cross-cultural science teaching. Research in Science Education, 31(3), 337–355.


Buxton, C. & Allexsaht-Snider, M., Aghasaleh, R., Kayumova, S., Kim, S., Choi, Y. & Cohen, A. (2014). Potential benefits of bilingual constructed responses science assessments for emergent bilingual learners. Double Helix, (2), http://www.qudoublehelixjournal.org/index.php/dh/article/view/31/113


Buxton, C. & Lee, O. (2014). English language learners in science education. In N. Lederman & S. Abell (Eds.). Handbook of Research on Science Education, Vol. 2. Routledge, 204-222.


Buxton, C., Salinas, A., Mahotiere, M., Lee, O. & Secada, W. G. (2013). Leveraging cultural resources through teacher reasoning: Teachers analyze second language learners’ problem solving in science. Teaching and Teacher Education, 32, 31–42.


Lee, O. & Buxton, C. (2010). Diversity and equity in science education: Theory, research, and practice. New York: Teachers College Press.


Basu, S., & Barton, A. C. (2007). Developing a sustained interest in science among urban minority youth. Journal of Research in Science Teaching, 44(3), 466–489.


Bell, P., Lewenstein, B., Shouse, A., & Feder, M. (Eds.). (2009). Learning science in informal environments: People, places, and pursuits. Washington, DC: National Academies Press.


Bereiter, C., & Scardamalia, M. (2003). Learning to work creatively with knowledge. In E. De Corte, L. Verschaffel, N. Entwistle, & J. van Merriënboer (Eds.), Powerful learning environments: Unraveling basic components and dimensions (pp. 55–68). Oxford, UK: Elsevier Science.


Brown, A. L., & Campione, J. C. (1986). Psychological theory and the study of learning disabilities. American Psychologist, 41(10), 1059–1068.


Brown, B. A. (2006). “It isn’t slang that can be said about this stuff”: Language, identity, and appropriating science discourse. Journal of Research in Science Teaching, 43(1), 96–126.


Brown, B. A., & Spang, E. (2008). Double talk: Synthesizing everyday and science language in the classroom. Science Education, 92, 708–732.


Calabrese Barton, A., Tan, E., & O’Neill, T. (2014). Science education in urban contexts: New conceptual tools and stories of possibilities. In N. Lederman & S. Abel (Eds.), Handbook of research on science teaching (2nd ed.) (pp.246–265). London, England: Taylor & Francis.


Cobern, W. (2000). Everyday thoughts about nature. Dordrecht, Netherlands: Kluwer Academic Press.


Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Mahwah, NJ: Lawrence Erlbaum.


González, N., Moll, L. C., & Amanti, C. (2005). Funds of knowledge: Theorizing practices in households, communities, and classrooms. Mahwah, NJ: Lawrence Erlbaum.


Haager, D., & Vaughn, S. (2013). Common Core State Standards and students with learning disabilities. Learning Disabilities Research and Practice, 28(1), 1–4.


Hammond, L. (2001). An anthropological approach to urban science education for language minority families. Journal of Research in Science Teaching, 38(9), 983–999.


Hogan, K., Nastasi, B., & Pressley, M. (2000). Discourse patterns and collaborative scientific reasoning in peer and teacher-guided discussions. Cognition and Instruction, 17, 379–432.


Hudicourt-Barnes, J. (2003). The use of argumentation in Haitian Creole science classrooms. Harvard Educational Review, 73(10), 73–93.


Inhelder, B., & Piaget, J. (1964). The early growth of logic in the child, classification and seriation. New York, NY: Harper & Row.


Janzen, J. (2008). Teaching English language learners in the content areas. Review of Educational Research, 78(4), 1010–1038.


Keil, F. C., & Wilson, R. A. (2000). Explanation and cognition. Cambridge, MA: MIT Press.


Klahr, D. (2000). Exploring science: The cognition and development of discovery processes. Cambridge, MA: MIT Press.


Klassen, R., & Chiu, M. (2010). Effects on teachers’ self-efficacy and job satisfaction: Teacher gender, years of experience and job stress. Journal of Educational Psychology 102(3), 741–756.


Kuhn, D. (1991). The skills of argument. Cambridge, England: Cambridge University Press.


Kuhn, D. (2005). Education for thinking. Cambridge, MA: Harvard University Press.

Ladson-Billings, G. (1995). Toward a theory of culturally relevant pedagogy. American Educational Research Journal, 32(3), 465–491.


Lee, H., & Lu, O. (2010). Assessing learning progression of energy concepts across middle school grades: The knowledge integration perspective. Science Education, 94(4), 665–688.


Lee, O. (1999). Science knowledge, world views, and information sources in social and cultural contexts: Making sense after a natural disaster. American Educational Research Journal, 36(2), 187–219.


Lee, O., Deaktor, R., Enders, C., & Lambert, J. (2008). Impact of a multi-year professional development intervention on science achievement of culturally and linguistically diverse elementary students. Journal of Research in Science Teaching, 45(6), 726–747.


Lee, O., Quinn, H., & Valdés, G. (2013). Science and language for English language learners in relation to Next Generation Science Standards and with implications for Common Core State Standards for English language arts and mathematics. Educational Researcher, 42(4), 223–233.

National Research Council. (2007). Taking science to school: Learning and teaching science in grades K-8. Washington, DC: National Academies Press.


National Research Council. (2011). A framework for K-12 science education: Practices, crosscutting themes, and core ideas. Washington, DC: National Academies Press.

Osborne, J. (2010). Arguing to learn in science: The role of collaborative, critical discourse. Science, 328, 463–466.


Parsons, E. C. (2008). Learning contexts, Black cultural ethos, and the science achievement of African American students in an urban middle school. Journal of Research in Science Teaching, 45(6), 665-683.


Parsons, E. C. (2014). Unpacking and critically synthesizing the literature on race and ethnicity in science education. In N. Lederman & S. Abel (Eds.), Handbook of research on science teaching (2nd ed.) (pp. 167–186). London, England: Taylor & Francis.


Piaget, J. (1973). The child and reality: Problems of genetic psychology. New York, NY: Grossman Publishers.


Rahm, J. (2002). Emergent learning opportunities in an inner-city youth gardening program. Journal of Research in Science Teaching, 39(2), 164–184.


Reeves, J. (2004). ‘Like everybody else’: Equalizing educational opportunity for English learners. TESOL Quarterly, 38(1), 43–66.


Rosebery, A. S., & Warren, B. (Eds.). (2008). Teaching science to English learners: Building on students’ strengths. Arlington, VA: National Science Teachers Association.


Saldaña, J. (2012). The coding manual for qualitative researchers. Los Angeles, CA: Sage.


Santos, M., Darling-Hammond, L., & Cheuk, T. (2012). Teacher development to support English language learners in the context of Common Core State Standards. Stanford, CA: Understanding Language Initiative.


Solano-Flores, G. (2008). Who is given tests in what language by whom, when, and where? The need for probabilistic views of language in the testing of English learners. Educational Researcher, 37(4), 189–199.


Solano-Flores, G., & Trumbull, E. (2003). Examining language in context: The need for new research and practice paradigms in the testing of English-language learners. Educational Researcher, 32(2), 3–13.


Schauble, L. (1996). The development of scientific reasoning in knowledge-rich contexts. Developmental Psychology, 32(1), 102–119.


Tate, W. F. (2001). Science education as civil right: Urban schools and opportunity-to-learn considerations. Journal of Research in Science Teaching, 38(9), 1015–1028.


Toth, E., Klahr, D., & Chen, Z. (2000). Bridging research and practice: A cognitively based classroom intervention for teaching experimentation skills to elementary school children. Cognition and Instruction, 18(4), 423–459.


Zimmerman, C. (2000). The development of reasoning skills. Developmental Review, 20(1), 99–149.




APPENDIX


Sample Questions from Grade 4 Students’ Scientific Reasoning Task – Forms of Energy

Excerpts from Section 1 – Home Context

[The purpose of the first section of the interview is to get the student to reflect on ways that he or she has experienced various forms of energy in the home context.]

1.

Can you tell me all the things that use energy in your house? Can you give me any more examples?

2.

Sometimes energy is converted from one form to another. Think about a light bulb. What kind of energy does it need to work? What kind of energy is the result? Can you describe what is happening with the light bulb?

3.

Think about a washing machine. What kind of energy does it need to work? What kind of energy is the result? Can you describe what is happening with the washing machine?

Excerpts from Section 2 – School Context: Potential and Kinetic Energy

Tell the student, Now I’m going to ask you to do a science experiment. This may be like the kind of activity you are doing with your teacher during science here in school. Today, we are going to use the controlled experiment framework to think about forms of energy.

Suppose you roll a ball down a ramp. How does changing the weight of a ball change the amount of energy it has?

1.

You have a ramp to roll the balls down and a cup to roll the balls into. You also have three balls of different weights. Remember what you want to find out. (How does changing the weight of the ball change the amount of energy it has?) Tell me your plan for what you will do.

2.

What measurements can you make to help you answer this question?

3.

Compare what you thought would happen with what actually happened. Did the results match your hypothesis? How can you explain this?

Excerpts from Section 3 – Connection to Play

1.

The last thing I want to ask you about is forms of energy when you are playing with your friends.

2.

First, tell me all the things you like to play that use energy.

3.

When you slide down a sliding board on a playground, what kind of energy do you need to slide? What kind of energy is the result? Are there similarities between you sliding down the sliding board and the balls rolling down the ramp? Tell me about it.

4.

Suppose two children are different weights – one is heavy and one is light. When both slide down the sliding board, do they have the same amount of energy? Tell me about it. Do they have the same forms of energy? Tell me about it.


Table 1. Student Demographics (across all three years)


Characteristic

N

Ethnicity

Haitian

Latino/a


44

37

Gender

Male

Female


42

39

ESOL Level

ESOL Level 1 or 2

ESOL Level 3 or 4

ESOL Level 5 (Exited within two years ago)

Non-ESOL (Never in ESOL or exited over

two years ago)


  8

16

25

32



Table 2. Students’ Scientific Reasoning Complexity Rubric


Criteria

0

1

2

3

4

Generativity

No observations

Observations limited to restatements

One, two, or three assertions beyond information given in the question

Four or more assertions about different topics

Three or more assertions building on the same idea or topic

Elaboration

No elaborations

One or two elaborations of one idea

One or two elaborations each of two or more distinct ideas

Three or more elaborations of the same (or similar) idea

Three or more elaborations each of two or more distinct ideas

Justification

No justifications

One or two justifications of a single assertion

One or two justifications each of two or more distinct assertions

Three or more justifications of the same (or similar) assertion

Three or more justifications each of two or more distinct assertions

Explanation

No explanations

Single example of an underlying structure, mechanism, or theory to explain one assertion

Single examples of underlying structures, mechanisms, or theories to explain two or more distinct assertions

Multiple/ chained examples of underlying structures, mechanisms, or theories to explain the same (or similar) assertion

Multiple/ chained examples of underlying structures, mechanisms, or theories to explain two or more distinct assertions



Table 3. Students’ Scientific Reasoning Complexity (maximum possible score of 4)


N

Generativity

Elaboration

Justification

Explanation

df

F

p

ηp2

 

M (SD)

M (SD)

M (SD)

M (SD)

    

81

3.84 (0.06)

2.25 (0.15)

2.14 (0.11)

0.28 (0.09)

2.389

195.646

<.001

.726



Table 4. Students’ Scientific Reasoning Complexity by ESOL Level (maximum possible score of 4)


 

ESOL Levels

M

SD

N

Generativity

ESOL Level 1 or 2

3.75

.71

8

ESOL Level 3 or 4

3.44

.53

16

ESOL Level 5 (Exited within two years ago)

3.88

.33

25

Non-ESOL (Never in ESOL or exited over two years ago)

3.94

.25

32

Total

3.84

.41

81

Elaboration

ESOL Level 1 or 2

1.63

1.30

8

ESOL Level 3 or 4

2.00

.00

16

ESOL Level 5

2.36

.99

25

Non-ESOL

2.34

1.12

32

Total

2.23

1.04

81

Justification

ESOL Level 1 or 2

1.50

.76

8

ESOL Level 3 or 4

1.89

1.05

16

ESOL Level 5

2.28

.54

25

Non-ESOL

2.37

.75

32

Total

2.19

.77

81

Explanations

ESOL Level 1 or 2

.00

.00

8

ESOL Level 3 or 4

.22

.67

16

ESOL Level 5

.40

.65

25

Non-ESOL

.38

.61

32

Total

.32

.60

81



Table 5. Students’ Scientific Controlled Experiment Practices: Percentage Correct


N

Minimum

Maximum

M

SD

81

18%

91%

56.5%

16.8%



Table 6. Student Science Content Knowledge: Percentage Correct


N

School

Home

Play

df

F

p

ηp2

 

M (SD)

M (SD)

M (SD)

    

81

50.8 (16.3)

27.9 (19.5)

18.9 (14.0)

1.631

102.62

<.001

.568



Figure 1. Semantic Response Map


[39_17780.htm_g/00001.jpg]





Cite This Article as: Teachers College Record Volume 117 Number 2, 2015, p. 1-36
https://www.tcrecord.org ID Number: 17780, Date Accessed: 10/26/2021 5:44:02 AM

Purchase Reprint Rights for this article or review