at SciVerse ScienceDirect

Teaching and Teacher Education 32 (2013) 31e42

Contents lists available

Teaching and Teacher Education

journal homepage: www.elsevier .com/locate/ tate

Leveraging cultural resources through teacher pedagogical reasoning:Elementary grade teachers analyze second language learners’ scienceproblem solving

Cory A. Buxton a,*, Alejandra Salinas b, Margarette Mahotiere c, Okhee Lee d, Walter G. Secada c

aUniversity of Georgia, College of Education, 630 Aderhold Hall, Athens, GA 30602, United StatesbBoston University, United StatescUniversity of Miami, United StatesdNew York University, United States

h i g h l i g h t s

< Teachers of second language learners analyzed students’ science problem solving.< Significant differences were seen across levels of pedagogical reasoning complexity.< Teachers provided generalizations rather than explanations based on evidence.< Teachers overlooked SLL students’ strengths, focusing on perceived student deficits.

a r t i c l e i n f o

Article history:Received 4 May 2012Received in revised form28 December 2012Accepted 4 January 2013

Keywords:Second language learnersEnglish language learnersTeacher reasoning

* Corresponding author. Tel.: þ1 706 424 4924.E-mail address: buxton@uga.edu (C.A. Buxton).

0742-051X/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.tate.2013.01.003

a b s t r a c t

Grounded in teacher professional development addressing the intersection of student diversity andcontent area instruction, this study examined school teachers’ pedagogical reasoning complexity as theyreflected on their second language learners’ science problem solving abilities using both home andschool contexts. Teachers responded to interview questions after watching a video of one of their stu-dents engaged in a science problem solving task. Over a 5-year period, 206 teacher interviews wereconducted with a total of 133 teachers. Results indicated significant differences across the dimensions ofpedagogical reasoning complexity as teachers expressed both deficit and resource oriented thinking.

� 2013 Elsevier Ltd. All rights reserved.

Industrialized nations across the globe have seen increasedlevels of immigration over the past few decades, as citizens of lesseconomically developed nations migrate in search of enhancedeconomic opportunity (Brown, 2012; Lipsmeyer & Zhu, 2011).While there are numerous sociocultural implications of thesemigration patterns, the youthfulness of immigrant populationsoften results in greater shifts in the demographics of school-agedchildren than in the overall demographics of the nation(Leyendecker, 2011; Passel, 2011). At the same time, the populationof teachers in compulsory-grades education in industrialized na-tions continues to be dominated by socioeconomically middle-classfemales from privileged ethnic and linguistic groups (Blomeke,Suhl, Kaiser, & Dohrmann, 2012).

All rights reserved.

For example, the teacher pool in the United States continues todraw largely from White, female, middle-class demographics,whereas the student population reflects a steadily expandingracial/ethnic, cultural, and linguistic diversity (Jorgenson, 2000;Sable & Plotts, 2010). Certainly, teachers need not share theirstudents’ backgrounds in order to teach effectively (Ladson-Billings, 1994), and the diversity within individual classroomsoften makes this impossible in any case. Still, many teachers,regardless of background, have only limited awareness of thecultural and linguistic knowledge that their nonmainstream stu-dents bring to the classroom (Garcia, Arias, Murri, & Serna, 2010;Gay, 2002; Villegas & Lucas, 2002). Too few teachers receive sub-stantive professional development opportunities focused on howstudents’ ethnic, cultural or linguistic backgrounds may affecteducational experiences (Gere, Buehler, Dallavis, & Haviland, 2009;Morrier & Irving, 2007; Selwin, 2007). The teacher professionaldevelopment literature also makes it clear that facilitating changes

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e4232

in teachers’ knowledge, beliefs, and practices is a long anddemanding process, not conducive to quick fixes or one-time in-terventions (Wayne, Yoon, Zhu, Cronen, & Garet, 2008; Wilson &Berne, 2004).

Given the shifting student demographics, professional learn-ing opportunities where teachers have ample time to exploretheir beliefs and practices regarding student diversity have animportant part to play in helping teachers to develop effectiveclassroom practices (Dilworth & Brown, 2001; Seidl, 2007). Callsfor such professional development with a multicultural andmultilingual focus have been heard in various nations withchanging student demographics (Molto, Florian, Rouse, & Stough,2010; Ricucci, 2008). Given the dominance of culturally and lin-guistically mainstream teachers in these settings, however, it isnot surprising that much of the research on professional devel-opment focusing on cultural and linguistic diversity has involveda preponderance of teachers who are not demographically rep-resentative of the students they now teach. Thus, there is a uniquevalue in studying teacher professional development where theparticipants are more culturally and linguistically representativeof evolving student demographics. The present study took placein urban elementary schools in the United States where theteacher demographics were closely aligned with the student de-mographics e predominantly Latino, African American andHaitian.

Further, while much of the teacher professional developmentliterature on supporting culturally and linguistically diverselearners has been built on broad ideas such as fostering culturallyrelevant pedagogy (Ladson-Billings, 1995) or funds of knowledge(González, Moll, & Amanti, 2005), there is more limited research onhow to better connect students’ cultural and linguistic experiencesto specific content area learning (Lee & Luykx, 2005). One poten-tially fruitful way to make these connections is through havingteachers explore their students’ thinking about content-specificproblems in both school and home contexts (Buxton & Lee, 2010).

The purpose of this study was to examine the relationshipsbetween teachers’ pedagogical reasoning about their students’problem solving and those teachers’ abilities to make connec-tions to their students’ cultural and linguistic strengths. Wehypothesized that professional learning that supported teachers’pedagogical reasoning would translate into an improvedawareness of how students express cultural funds of knowledgeduring problem solving tasks. To test this hypothesis, we askedteachers to respond to interview questions after watchinga video of one of their students engaged in a science problemsolving task.

1. Conceptual framework

This study was based on two distinct but complementary areasof research guiding the design and implementation of our teacherprofessional development that was aimed at promoting problemsolving of second language learners (SLLs): (a) teachers’ integrationof students’ linguistic and cultural resources into science instruc-tion and (b) teachers’ pedagogical reasoning about students’problem solving in science.

1.1. Students’ linguistic and cultural resources in science classrooms

In order for teachers of science to take up the challenge ofexplicitly integrating students’ linguistic and cultural resources intotheir instruction, teachers must believe that such efforts willenhance their students’ academic performance (Lee & Buxton,2010; Rodriguez & Kitchen, 2005). The literature, however,indicates numerous obstacles to implementing professional

development that simultaneously supports culture and content. Forexample, many teachers are unaware of linguistic and culturalinfluences on student learning, do not consider “teaching for di-versity” as their responsibility, overlook cultural or racial differ-ences, accept inequities as a given condition, or resist multiculturalviews of learning (Luykx et al., 2007; Zhao, 2010). Additionally,teachers often work in policy contexts that assume that SLLs mustacquire the dominant language of instruction before engaging insubject matter learning, despite research indicating that thisapproach typically leads SLLs to fall behind their dominant-language-speaking peers (Lyster & Ballinger, 2011; Teachers ofEnglish to Speakers of Other Languages, 2006). While these find-ings paint a rather bleak picture of thewaysmany teachers perceivetheir nonmainstream students, it is again worth noting that muchof this research base has primarily involved teachers who are,themselves, members of the mainstream ethnic, cultural and lin-guistic group.

In our previous research with urban elementary school teacherswho were predominantly the same ethnicities as their students(Latino and Haitian), we found that even at the beginning of theintervention, most participants acknowledged the importance ofrecognizing students’ first cultures and languages when planningfor science instruction (Lee, Luykx, Buxton, & Shaver, 2007). How-ever, these same participants showed little change in their actualclassroom practices regarding teaching for diversity, despite twoyears of involvement in a professional development interventionthat included this focus. This discrepancy between teachers’ posi-tive beliefs about the value of students’ diverse backgrounds andtheir failure to act on those beliefs in concrete ways was a deviationfrom the outcomes of other well-known studies of culturally andlinguistically diverse students’ funds of knowledge (González et al.,2005; Warren & Rosebery, 1995). These studies demonstratedchanges in teacher practices that our project failed to realize. Thisunexpected outcome led us to consider a different approach inhopes of better facilitating teachers’ integration of students’ cul-tural and linguistic resources into their science instruction withSLLs.

Thus, in order to help teachers better support their culturallyand linguistically diverse students’ academic performance, wedecided to focus on student problem solving. Student problemsolving has been shown to engage and refocus teachers in otherprofessional development efforts, particularly in mathematics ed-ucation (Carpenter, Fennema, Franke, Levi, & Empson, 2000;Steinberg, Empson, & Carpenter, 2004). We noted that teacheranalysis of student problem solving has not been a significant focusof professional development in science education (Lehrer &Schauble, 2000). We considered that encouraging teachers tothink less about what they themselves were doing and to thinkmore about what their students were thinking, saying and doingduring problem solving tasksmight serve to refocus teachers on thequestion of where students get their ideas. Consideration of theorigins of students’ ideas would, in turn, provide a natural bridge tothinking about their first cultures and languages since elementarygrade students’ ideas often have clear connections to their homeand community contexts.

1.2. Pedagogical reasoning complexity

Our approach to understanding how teachers reason abouttheir students’ problem solving during science tasks focuses on thenotion of pedagogical reasoning complexity, or the quality of anindividual’s reasoning about another person’s engagement ina learning task. Duschl and Grandy (2008) have argued thatworthwhile science learning tasks make student thinking visiblefor the purpose of helping teachers give feedback to improve

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e42 33

students’ scientific ideas. We view teacher pedagogical reasoningas an underdeveloped link in the research on reasoning skills.Pedagogical reasoning complexity may be seen as related to butdistinct from pedagogical content knowledge (PCK; Shulman,1987) and disciplinary content knowledge (DCK). In science edu-cation, research on PCK has largely ignored the need for anawareness of sociocultural and sociolinguistic considerations ofstudent learning (Park & Oliver, 2008). In this paper, teachers’pedagogical reasoning complexity explicitly includes an emphasison linguistic and cultural awareness in assessing students’ problemsolving.

To conceptualize teachers’ pedagogical reasoning complexityabout students’ problem solving, we began with the science rea-soning complexity frameworks developed by Hogan, Nastasi, andPressley (2000), and Resnick, Salmon, and Zeitz (1993). Theseframeworks were originally designed to evaluate students’ rea-soning about science content during inquiry activities. We adaptedthese reasoning typologies to create a new framework designed toassess teachers’ pedagogical reasoning about their students’ prob-lem solving in science. Specifically, we initially borrowed the cat-egories of the reasoning framework from the Hogan et al. (2000)study but determined that these categories needed to be modi-fied for our context of one-on-one interviews because the Hoganstudy focused on classroom discourse. The Resnick et al. (1993)study was helpful in that it considered individual interviews.Other frameworks for assessing reasoning in conversation havebeenwidely used in science education in recent years, most notablythe framework proposed by Toulmin (1969). That framework hasgenerally been applied to the analysis of conversations amonggroups of students debating a topic (e.g., Osborne, Erduran, &Simon, 2004). We found that this argumentation framework didnot fit well for our design, as it is best suited to small groups ofstudents critiquing and supporting each other’s ideas. The frame-work we developed was more conducive to our analysis of indi-vidual teachers reasoning about one of their students engaged ina problem solving task.

In our framework, we consider four key dimensions of reason-ing: (a) the generativity of assertions, (b) the elaboration of asser-tions with supporting examples, (c) the justification of assertionswith evidence, and (d) the explanation of assertions through linksto underlying structures, mechanisms or theories. The degree towhich each of these dimensions of reasoning is present in an in-dividual’s argument can be used to assess the overall strength ofthe reasoning. In this way, the dimensions of reasoning becomeevaluative criteria. The first two dimensions (generativity andelaboration) assess the variety and richness of ideas raised by thespeaker, and the last two dimensions (justification and explan-ation) assess the logical power of the individual’s reasoning abouta given topic.

While the four dimensions of pedagogical reasoning complexityin our framework can be evaluated individually, we also view themas a connected set of skills that may be developed together. The firstdimension, generativity or the generation of simple statements andassertions, can be offered without any clear examples, justifica-tions, or rationales, and thus can be considered a fairly simple typeof reasoning. For example, a teacher might claim, “My studentshave more trouble conceptualizing how to measure volume thanthey do length.” As a stand-alone statement, this would be anexample of a generic assertion that lacks examples or justification;it has little logical power.

The second dimension, elaboration, involves providing sup-porting details or examples to enrich a statement or assertion.Following Hogan et al. (2000), we consider such elaborations tobe a more complex form of reasoning that builds directly ongenerativity. Thus, if the teacher who made the statement above

continued to say, “Most of them get fooled by different sizes andshapes of containers, even if the containers hold the same volumeof water. Length doesn’t have that complication.” These state-ments are elaborations that serve as examples to support theoriginal assertion and, as such, they indicate more complexreasoning.

The third dimension, justification, is the process of linkinga claim to some rationale or inference. While the Hogan et al.framework considered justification to be a more complex form ofreasoning than elaboration based on the higher cognitive de-mand of providing evidence for a claim, we found in our studythat justifications were often given arbitrarily in ways that werenot warranted by the evidence or supported by examples. Thus,our teacher might claim that, “The kids who come from Haititoday don’t have the same academic skills with measurement asthe kids who came ten years ago because the kids coming nowdidn’t go to school consistently in Haiti.” This statement includesa justification (interrupted schooling in Haiti) linked to anassertion (the students have weak prior academic skills withmeasurement), but no actual evidence or concrete example isprovided to support the claim e that is, a justification can rep-resent a fairly arbitrary belief statement. Thus, in our pedagogicalreasoning model, we did not necessarily consider the use ofjustifications to imply more complex reasoning than the use ofelaborations.

The final dimension of our model of pedagogical reasoningcomplexity, explanation, involves grounding a claim in an under-lying structure, mechanism or theory. That is, we view explanationas requiring the application of theory to practice. Thus, ourexample teacher might claim, “Some of my third graders at thestart of the measurement unit hadn’t yet grasped the concept thatjust because one container is taller than another doesn’t neces-sarily mean it has a larger volume. After they worked through theunit, though, almost everyone was able to talk about volume asa three dimensional concept.” In this case, the teacher is providingan explanation for her assertion, grounded in cognitive develop-ment theory. We consider explanations that make such connec-tions to be the strongest type of pedagogical reasoning in ourframework.

1.3. Problem solving tasks

The literature in cognitive and developmental psychology in-cludes a long history of experimental research on problem solvingwith both children and adults (Klahr, 2000; Kuhn, 1989;Zimmerman, 2000). One key finding of this body of work is that anindividual’s degree of expertise in a given problem solving domainhas a greater influence on one’s approaches to problem solvingwithin that domain than the individual’s age or general devel-opmental level. Thus, children are capable of more advanced levelsof problem solving if they have some degree of expertise or con-crete prior knowledge about the task in question. Such expertisemay come from home or community experiences as well as fromprior schooling.

The student problem solving tasks that we created for third,fourth and fifth grade students addressed the topics of measure-ment, energy, and the changing seasons, respectively. The taskswere designed to draw out any student expertise about these topicsfrom both home and school contexts and to make that expertisevisible to the teacher in a way that might not take place duringnormal classroom instruction. We selected topics at each gradelevel that were covered in the classroom curriculum but would alsobe topics with which most students would have informal experi-ence in their homes and communities, including their countries oforigin.

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e4234

1.4. Purpose of study

We focus here on teachers’ pedagogical reasoning about studentproblem solving during their participation in our professionaldevelopment project.1 We wished to see if an explicit focus onstudent problem solving might lead teachers to better leveragetheir students’ cultural and linguistic resources to supportimproved understanding of how students form and develop theirideas about the natural world. Specifically, we addressed the fol-lowing two research questions:

1. How did teacher pedagogical reasoning about student problemsolving vary across the four dimensions of pedagogical rea-soning complexity?

2. What connections did teachers make between a student’sperformance on the problem solving task and the student’s firstculture and language?

We note that we originally set out to also investigate howteachers’ pedagogical reasoning about student problem solvingvaried with years of participation in the project. However, due tohigh rates of teacher attrition and mobility, as discussed below, wewere unable to adequately answer this question.

2. Method

2.1. Study context

This study was part of a larger research project that had the goalof improving the science achievement of SLLs, especially in thecontext of high-stakes testing and accountability policy in science.Over a 5-year period, the research used a longitudinal design toimplement a professional development intervention in U.S. urbanelementary schools.

2.2. Professional development

The professional development component of our project con-sisted of curriculum materials for students and teachers and a se-ries of teacher workshops throughout the school year. Thecurriculum materials and workshops were designed to comple-ment and reinforce each other in improving teachers’ knowledge,beliefs, and practices in science instruction and English languagedevelopment for SLLs (Buxton, Lee, & Santau, 2008). In addition tothese primary goals, secondary goals of the professional develop-ment included improving teachers’ pedagogical reasoning andstudents’ science problem solving, and capitalizing on students’first languages and cultures.

Professional development workshops were differentiated fornew and returning teachers at all three grade levels. Each year,teachers who were new to the research project participated in sixdays of workshops with a central focus on implementing theinquiry-based project science curriculum (with a focus on meas-urement, water cycle and weather in 3rd grade; energy, force andmotion, and processes of life in 4th grade; and nature of matter andEarth systems in 5th grade). Second and third year teachers par-ticipated in four days of workshops that reinforced elements of theproject curriculum while placing a greater emphasis on the addi-tional project goals, including teachers’ pedagogical reasoning andstudents’ science problem solving. For example, teachers brought

1 Elsewhere (Santau, Secada, Maerten-Rivera, Cone, & Lee, 2010) we have writtenabout teachers’ classroom practices during participation in the project, includinghow those practices were connected to students’ cultural and linguistic resources.

students’ work samples to discuss student problem solving relatedto science concepts or skills. Project personnel presented samplesof students’ problem solving focused on experimental design andengaged teachers in analyzing students’ capabilities and diffi-culties. We shared the emergent results from our research,including data that highlighted students’ out-of-school cultural andlinguistic experiences that could serve as intellectual resources forlearning school science.

2.3. Study setting and participant selection

The research was conducted in a school district located in thestate of Florida in the southeastern United States. The school districtwas in a large urban area with a culturally and linguistically diversestudent population. During the 2004e2005 school year, the firstyear of the project, the ethnic makeup of the student population inthe school district was 60% Latino, 28% Black (including Haitian,Caribbean Islanders and African American), 10% White non-Latino,and 2% Asian and Native American. Across the school district, 72% ofelementary students participated in free or reduced price lunchprograms, and 24% were designated as limited English proficientaccording to the state’s term for SLLs.

Of the 205 elementary schools in the district, the projectinvolved 15 elementary schools that were selected based on threecriteria: (a) percentage of SLLs (predominantly Spanish or HaitianCreole speaking) above the district average at the elementaryschool level (24%), (b) percentage of students on free or reducedprice lunch programs above the district average at the elementaryschool level (72%), and (c) school grades of predominantly “C” or“D” (on an “A” through “F” scale) for the previous 5 years accordingto the state’s accountability plan. Teachers from six of these schoolswere involved in the pedagogical reasoning study, in which threeschools had SLLs who were predominantly Spanish speaking andthe other three had SLLs who were predominantly Haitian Creolespeaking.

As noted earlier, the teacher demographics in this study wereuncommon in that they approximated the student demographics interms of ethnic, racial, and linguistic background. Of the 133 totalteacher participants in our project, 39% identified themselves asLatino, 37% as Black non-Latino, 12% as White non-Latino, 8% asHaitian, and 4% as French. Linguistically, 74% of the teachersreported that they spoke English as their first language, 26%Spanish, 8% Haitian Creole, and 4% French.2 In terms of gender, theparticipants were more typical of elementary teachers, as 90% ofthe teachers were female and 10% were male. In terms of profes-sional background, 67% of the teachers reported having a bachelor’sdegree as their highest degree earned, 29% a master’s degree, and4% did not respond. None reported having a specialist or doctoraldegree. Teaching experience ranged from 1 to 36 years, with anaverage of 8 years. Participants had been teaching at their currentschool for an average of 6 years.

During each year of the project, one student was selected fromeach teacher’s class to participate in a student problem solving task.Each teacher participated in the project for a maximum of threeyears; however, as can be seen in Table 1, high teacher attrition andmobility rates resulted in only a small percentage of teachers par-ticipating at the same grade level for the full three years. A total of221 student problem solving interviews were conducted froma total of 133 teachers’ classes over the course of the project.3

2 Some participants identified more than one first language.3 Each student interview was meant to be paired with a corresponding teacher

interview. However, a small number of the teacher interviews were not completed,resulting in the 206 teacher interviews discussed below.

Table 2Teacher grade levels at time of interview.

Teacher grade level Number of interviews Percent of interviews

3rd 72 354th 77 375th 57 28Total 206 100

Table 1Years of teacher participation at time of interview.a

Years of teacherparticipation

Number ofinterviews

Percent ofinterviews

1 133 652 59 283 14 7Total 206 100

a Teachers who participated for more than one year are counted more than onetime in the table. That is, they were counted as a first year teacher the first time theywere interviewed, a second year teacher the second time they were interviewed,and again as a third year teacher if they remained in the project and were inter-viewed all three years.

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e42 35

Students were purposefully selected to create balanced groupsbased on four SSL categories and two gender groups. The four SLLcategories included: (a) new arrivals and emergent speakers ofEnglish (English levels 1 and 2), (b) conversational competence inEnglish but still being enrolled in SLL programs (English levels 3and 4), (c) exited from SLL programs but still being monitored(English level 5), and (d) first language speakers of English. Withineach SLL category, male and female students were equally repre-sented. Each teacher was randomly assigned a particular studentdemographic (e.g., SLL 5 female) and that teacher was asked to giveconsent forms to all students in the class who met the given de-mographic criteria. From the pool of students in the class whoreturned signed consents, one student was selected at random forthe interview.

2.4. Data collection

Selected students participated in the grade-appropriate prob-lem solving task with a member of the research team. Each tasktook approximately 30e45min, with the third grade task averagingthe shortest time to complete and the fourth grade task averagingthe longest due to the science investigation component of theinterview. These interviews were video recorded and the re-cordings were then shared with each student’s teacher. Eachteacher was asked to watch the recording of her or his student onthe evening before participating in a semistructured interviewwitha member of the research team. The student video was cued andavailable for reference throughout the interview. The interviewquestions focused the teacher on specific ways that his or herstudent viewed the given science topic as well as where the studentobtained his or her ideas relevant to the topic. The interviews lastedfrom as short as 10 min to over an hour and were audiotaped andthen transcribed. The teacher pedagogical reasoning interviewprotocol can be seen as Appendix A, and the student interviewprotocols are available as supplementary material accompanyingthe online article.

2.4.1. Third grade taskThe problem solving task for 3rd grade students was about

measurement and addressed four topical areas covered in themeasurement curriculum unit: length, weight, volume, and tem-perature. The first part asked students to discuss experiences theyhad with measurement in the home context. The second part askedstudents to perform a measurement task using a ruler, kitchenscale, graduated cylinder, measuring cup, and thermometer. Thethird part asked students to make estimates about each of the fourtopical areas of measurement. The final part asked students todiscuss experiences they had with measurement in the context ofplaying with their friends. Together, parts one and four were takento represent the home and community context and parts two andthree to represent the school context.

2.4.2. Fourth grade taskThe task for 4th grade students was about energy and addressed

topics central to the 4th grade energy curriculum unit. First, stu-dents were asked to discuss their experiences with forms of energyat home, both in Florida and in their country of birth if applicable(home context). Second, students were asked to design, performand discuss an experiment to measure the energy of balls of dif-ferent masses rolling down a ramp (school context). Third, studentswere asked to discuss their experiences with transfer of energyduring play (play context). The design of the energy task wasmeantto provide students with ample opportunity to talk about relevantprior knowledge, both in school and out of school, and then to applythat knowledge to a hands-on problem solving task.

2.4.3. Fifth grade taskThe task for 5th grade students focused on the changing seasons

and was aligned with one of the central activities of the 5th gradeEarth systems curriculum unit. The first part asked students todiscuss common activities that they engaged in at home, again bothin Florida and in their country of birth if applicable, that wereunique to particular seasons of the year (home context). The secondpart asked students to usemanipulatives (Styrofoam balls, markers,tooth picks and a light bulb) to demonstrate a model of how andwhy the seasons change (school context). The third part askedstudents to discuss games and sports that they played at differenttimes of the year and how this was related to the changing seasons(play context). Again, the task was designed to highlight and thenapply prior knowledge developed across multiple contexts.

The teacher interviews focused on student task performancethat served as the primary data source for the study of teachers’pedagogical reasoning (research question 1), particularly as itrelated to students’ first culture and language (research question 2).According to the original study design, each teacher should haveparticipated in the pedagogical reasoning interview each year, fora total of three interviews over three years. However, as notedabove, due to high rates of teacher mobility and attrition, fewteachers actually remained at the same grade level for the fullduration of the project. As new teachers replaced prior participants,these additional teachers joined our project and were consideredfirst-year teachers for purposes of both their professional devel-opment and this research (see Table 1). Thus, over the three years,a total of 133 teachers participated in a total of 206 interviews.Additionally, while the total number of interviews conducted ateach of the three grade levels was designed to be equal, fluctuationsin the number of classrooms led to differences in the total numberof interviews conducted at each grade level over the course of thestudy (see Table 2).

2.5. Data analysis

An analytical coding scheme was developed to categorizeteachers’ pedagogical reasoning complexity as they discussed theirstudents’ engagement with the problem solving tasks. As describedin the pedagogical reasoning complexity section above, ourapproach was adapted from earlier science reasoning complexityframeworks. The core of our analytical framework was our

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e4236

pedagogical reasoning complexity rubric (see Table 3). This rubricwas used to code and score the teacher interview transcripts. Eachof the four dimensions of pedagogical reasoning complexity (i.e.,generativity, elaboration, justification, and explanation) was scoredon a 5-point scale, with a score from 0 to 4 assigned to each teacherinterview transcript for each of the four dimensions. As can be seenfrom the rubric, scoring for each pedagogical reasoning constructwas based on a combination of the number of occurrences ofa given type of reasoning and the occurrence of multiple exampleswithin the same (or closely related) response(s).

In applying the rubric, some practical definitions and distinc-tions emerged. First, within the dimension of generativity, we dif-ferentiated between restatements and assertions, in whichrestatements are rewordings of the question that has been askedwith no new meaningful information given in the response, whileassertions are claims that go beyond the information given in thequestion. Second, the key feature in identifying an elaboration wasthat it built on an assertion through the use of one or more ex-amples, but did not provide a clearly stated reason for why theassertion might be true. Third, a justification was identified whenthere was a clear attempt to answer the “why” question in supportof an assertion or an elaboration, usually through the use of anexplicit or implicit “because” statement. Finally, an explanationcould be distinguished from a justification by the presence of anexplicit link(s) to underlying structures, mechanisms, or theoriesthat included but could be generalized beyond the specific casebeing discussed.

Three members of the research team initially carried out thecoding of the teacher interview transcripts. Each member scoredthe same 20% of the teacher interview transcripts from the first yeardata, and scores were compared and differences reconciled throughnegotiation until consensus was reached. The same three scorersthen scored additional sets of three transcripts until greater than90% between-scorer agreement was reached. From that point on,one member of the team became the primary scorer and took re-sponsibility for coding all teacher interview transcripts for the year.The primary scorer would bring scoring questions to the rest of thegroup for discussion on an as-needed basis. The same process wasfollowed in each successive year of the study.

Once all teacher interviews were coded and scored using thepedagogical reasoning complexity rubric, basic descriptive (meanand standard deviation) and inferential (ANOVA) statistics werecalculated to compare teacher reasoning across the four di-mensions of pedagogical reasoning complexity (research question1). To consider the connections that teachers made between thestudents’ performance on the problem solving task and the stu-dents’ first culture and language (research question 2), a member ofthe research team other than the primary coder read through all ofthe coded transcripts and identified all places where a teacher

Table 3Teacher pedagogical reasoning complexity rubric.

Criteria 0 1 2

Generativity Noobservations

Observations limitedto restatements

One, two or three assertionsbeyond information given inthe question

Elaboration Noelaborations

One or two elaborationsof one idea

One or two elaborations eachof two or more distinct ideas

Justification Nojustifications

One or two justificationsof a single assertion

One or two justifications eachof two or more distinctassertions

Explanation Noexplanations

Single example of anunderlying structure,mechanism or theoryto explain one assertion

Single examples of underlyingstructures, mechanisms ortheories to explain two ormore distinct assertions

explicitly mentioned connections to a student’s home experiences,culture, or language. These excerpts were categorized thematically,and exemplars were then selected on the basis of being typicallyrepresentative of those categories, rather than for being outliers orexceptional examples. Representative examples of teachers’ peda-gogical reasoning were then paired with excerpts from the corre-sponding student interview that the teacher was discussing tocreate minivignettes. In the results section below, selected mini-vignettes are presented to elaborate on statistical patterns.

3. Results

3.1. Teacher reasoning across four dimensions of pedagogicalreasoning complexity

We begin by presenting results for the first research question:How did teachers’ pedagogical reasoning about student problemsolving vary across the four dimensions of pedagogical reasoningcomplexity? Statistical analysis using ANOVA is presented first,followed by illustrative minivignettes. While a total of 206 in-terviews were conducted with the 133 teacher participants, for thisanalysis we included only one interview for each teacher; theinterview during their final year of participation. For example, fora teacher who participated for two years, we included the year 2interview in the analysis, but not the year 1 interview. The rationalefor this decision is that if we included all 206 interviews, we wouldbe over-representing the reasoning of teachers who took part in thestudy for more than one year.

The means for pedagogical reasoning complexity for all teachersin the sample are shown in Table 4. Results from a repeated mea-sures ANOVA indicated significant differences across all four di-mensions of reasoning complexity for the entire sample;F(2.277) ¼ 472.561, p < .001. Mauchly’s test of sphericity indicatedthe need for a correction; Mauchly’s w ¼ 0.528, c2(5) ¼ 88.473,p < .001, ε ¼ 0.759. Results yielded a large effect magnitude;h2 ¼ 0.77. A Bonferroni pair-wise comparison test indicated thateach combination of pairs was statistically different from eachother (p � .01). In other words, teachers’ mean scores for eachreasoning complexity category (generativity, elaboration, justifi-cation, and explanation) were significantly different from eachother. The highest mean score was for generativity, followed byjustification, elaboration, and explanation, respectively.

Next, we present minivignettes to serve as exemplars for each ofthe four dimensions of pedagogical reasoning complexity. We notethat the student and teacher utterances did not actually occurtogether as part of the same interview e the two interviews wereconducted at different times, with the teacher responding to thevideo of the student interview. Student SLL level and teacher par-ticipation year are indicated in parentheses.

3 4

Four or more assertions aboutdifferent topics

Three or more assertions buildingon the same idea or topic

Three or more elaborations ofthe same (or similar) idea

Three or more elaborations eachof two or more distinct ideas

Three or more justifications ofthe same (or similar) assertion

Three or more justifications eachof two or more distinct assertions

Multiple/chained examples ofunderlying structures, mechanismsor theories to explain the same(or similar) assertion

Multiple/chained examples of underlyingstructures, mechanismsor theories to explain two or moredistinct assertions

Table 4Teacher pedagogical reasoning complexity results.

n Generativity Elaboration Justification Explanation df F p h2p

M (SD) M (SD) M (SD) M (SD)

133 3.99 (0.12) 1.89 (1.21) 2.77 (0.91) 0.44 (0.72) 2.277 472.56 <0.001 0.77

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e42 37

3.1.1. GenerativityTeachers almost universally made numerous assertions about

various aspects of their students’ understanding of science problemsolving, as seen by the very high mean score and low standarddeviation for generativity in Table 4.

Student (SLL1/2): For volume? This (holds up the graduatedcylinder). So I put the water in there (pours until the cylinder isfull and realizes that the cylinder will not hold all the water inthe cup). It’s 100. And this much left in the cup (holds up the cupand looks, then estimates). like 15. So 115. milliliters.

Teacher (Y1): I’m not really worried. I think he gets it. He showshis understanding of measurement with the tools evenwhen hehas trouble saying it.

The above excerpt shows a typical example of an assertion inwhich the teacher claimed that the student demonstrated hisunderstanding with his actions even when the student had diffi-culty demonstrating it through words. This is a general claim thatis not supported or elaborated with specific examples from thestudent interview, such as mentioning how the student showed anunderstanding of volume by summing up the amount of water inthe cylinder and in the cup. Such examples were very commonthroughout the teachers’ pedagogical reasoning interviews,despite ongoing professional development designed to encourageteachers to elaborate on their reasoning about student problemsolving.

3.1.2. ElaborationDespite the large number of assertions, teachers were less likely

to provide elaborated examples to support these assertions. Themajority of teachers’ assertions were either unelaborated or onlyminimally elaborated with general statements rather than concreteexamples taken from the student interviews. The following excerpt,also taken from the 3rd grade measurement task, typifies a case inwhich the teacher did elaborate on an assertion with a specificexample.

Student (SLL 5): From home to know how long something is?Like one time my dad, he was building a table and he needs toknow if the legs, like if they were all the same size, so we had tomeasure how long they are with a tape.

Teacher (Y2): He gets his ideas from class, from the experimentswe did, but he also showed that he gets them from home. Hementioned the time that his dad was building a table andmeasuring the legs, but it was unclear to me if he was reallymeasuring or if he was just watching his dad.

The teacher begins with a basic assertion (that the student’sideas about measurement came from both school and home), fol-lowed by a single elaboration of that statement, drawing upona specific example the student gave (measuring table legs to be surethey are the same length). Across grade levels, we found thatdespite having just watched the student video the evening beforethe interview, most teacher assertions were not well supported byelaborations in the interviews and that the majority of elaborationsthat were given were single elaborations using just one examplefrom the student problem solving task.

3.1.3. JustificationWhile teachers were less likely to support their assertions with

concrete examples taken from the student interviews (elaboration),as can be seen in the Table 4 results, they were more likely toprovide a general rationale to support their claims (justification). Inother words, teachers’ elaborations took the form of specific ex-amples drawn from watching the student videos, while their jus-tifications were typically drawn from awide range of sources, oftenbeyond the scope of the student videos.

Student (SLL 3/4): The weight [of the ball] gives more energy.The heavier ball, it pushed the cup more because it has moreenergy. The energy makes the cup go longer.

Teacher (Y1): I’mpretty happy that he couldmake some sense ofthis. You know we did activities in class that were really similar,with different ramps. And also the kids kick soccer balls all thetime. So I think he can explain energy pretty well because he’smaking these connections to things he’s done.

In this case, the teacher’s assertion that the student could ade-quately explain transfer of energy was based on the justificationthat the student was making connections to both in-school andout-of-school contexts. The teacher drew general connections toother school and play activities, but did not provide any specificexamples (elaborations) from the student interview. In ourframework, broad justifications of this kind are considered to beindicative of more complex reasoning than assertions alone, but arenot considered to be more complex reasoning than elaborationsthat draw examples specifically from the student interview. This isbecause in developing reasoning skills, it is important to beginwithconcrete examples, and then to use those examples as evidence fordrawing justifications and more generalized explanations.

3.1.4. ExplanationOf the four dimensions of pedagogical reasoning complexity in

our framework, explanation was the least commonly observed inour teacher interviews. As can be seen from the results in Table 4,teachers were much less likely to support their assertions withtheory-based explanations than they were to give more anecdotaltypes of justifications.

Student (non-SLL): Summer and winter, well you know it has todo with the Earth going around the Sun and how much Sun weget, and. and. I forget the rest.

Teacher (Y2): I guess that as soon as she learned it, she musthave forgotten it. In class she participated a lot using the modelsand I remember she could explain the seasons. But when it cameto the interview she had trouble pulling it all together. I thinkthat a lot of our Haitian students have trouble with changingseasons because you know, back home the weather just doesn’tchange that much, no matter what time of year it is, so it’s kindof an alien concept for them.

In this case, the teacher supported her assertion (that the studenthad learned but quickly forgotten the reason for the changing sea-sons) with an elaborated example (that the student had demon-strated the knowledge during participation in class). The teacherthenwent on to give a generalizable theoretical explanation forwhy

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e4238

the student failed to learn the knowledgedeeply (that students fromthe tropics, who lack direct experience with large seasonal differ-ences, are likely to struggle to make sense of seasonal change as anabstract concept beyond their lived experience). Although weattempted to support teachers in developing more theory-drivenpedagogical reasoning through our professional development ac-tivities, we saw few examples during the interviews.

3.2. Teacher reasoning related to students’ first culture andlanguage

Next, we turn to our second research question: What connec-tions did teachers make between a student’s performance on theproblem solving task and the student’s first culture and language?The teacher interview protocol (see Appendix A) implicitly probedteachers’ thinking about students’ cultural and linguistic resourcesby asking teachers to discuss where they believed their studentsgot the ideas about the topics of the problem solving tasks. Theconnections between the science topics and home and communitycontexts were quite explicit in the student interview protocols (e.g.,Do you ever go to the food store or market? Have you ever seenanyoneweigh anything there?What did they weigh? How did theyweigh it?). Thus, the teacher heard the student respond to ques-tions that directly prompted for funds of knowledge constructedduring out-of-school experiences related to the science topic.Below, we present additional minivignettes to demonstrate teacherthinking about the role of students’ first culture and language asrelated to the problem solving task for each of the three grade-levelscience topics. We address each of the three grade levels separately,as we noted somewhat different patterns in the connectionsteachers made to students’ linguistic and cultural resources,depending on the topic of the student problem solving task(measurement, energy and changing seasons).

3.2.1. Third grade measurement taskMost 3rd grade students interviewed were able to provide

multiple examples of measurement experiences fromout-of-schoolcontexts and to elaborate on those experiences with supportingdetails. Of course, these students also showed limits of theirknowledge and naïve conceptions about different aspects ofmeasurement in these out-of-school contexts. When the 3rd gradeteachers discussed their students’ home experiences, their com-ments were more likely to highlight these limitations than thepotential resources in prior knowledge.

Student (SLL 3/4): Here we use the English system, but back inCuba we use metric. I like metric better because I know it more.The English system is confusing [to] me.

Teacher (Y1): Well, she wasn’t completely lost, but she wasconfused a lot of the time [during the interview]. She had seensome of these things at home. She talked about them. And thenshe had done measurement in class and I think some of thehome ideas were confusing her. She had a better understandingof the school activities than the home questions. What we haddone in class was helping her get clear about measurement.

In this example, typical of many third grade teacher responses,rather than seeing this student’s home experience (familiarity withthe metric system) as a strength that could be leveraged to supportscience learning, the teacher focused instead on the student’sconfusion that resulted from trying to integrate the measurementsystemmore commonly used in the U.S. (Imperial System) with thesystem she was already familiar with (metric). The teacher’s per-ception was that the student’s school experiences would help her“correct” her confusion that came from home experiences.

3.2.2. Fourth grade energy taskAs was the case with cultural connections to science content,

our teacher interviews typically indicated a failure to view a stu-dent’s first language as a resource for science learning, even thoughthis was an ongoing topic of discussion in our professional devel-opment. For example, our project curriculummade use of trilingual(EnglisheSpanisheHaitian Creole) science vocabulary guides.Many teachers still expressed the belief, however, that language(and particularly vocabulary) served as a barrier more than asa resource for their SLLs. This focus on language was particularlycommon in the 4th grade interviews on the topic of energy, astypified by the following response:

Interviewer (from student interview): During the experiment,did energy get converted from one form to another?

Student (SLL 3/4): The energy? Converted? Like more or less?Here (points to the ramp).

Teacher (Y1): I think part of his problem was that [studentname] was having problems with the vocabulary and that dis-tracted him from the question you wanted him to answer. Thevocabulary is a big challenge for our EL students. This is all newvocabulary for them and it gets in the way sometimes of tryingto teach them the material.

While interpreting the vocabulary in the question might indeedhave been a challenge for this student, we were struck by the fre-quency of teacher statements such as “this is all new vocabulary forthem.” In fact, for Spanish speakers, including both the student andthe teacher in the above example, most of the science vocabularywords in the sentence under discussion are close cognates in En-glish and Spanish.We tried to support teachers in coming to see theexplicit use of these cognates as a linguistic resource. Someteachers did begin to express this idea when observing their stu-dents’ problem solving task, as in the following example froma second year participant, but these teachers were in the minority.

Teacher (Y2): Sometimes my class struggles with the vocabu-lary, but the Spanish can actually help the kids whenwe teach it,and probably the same for the Creole. Because, look at thewords, the similarities between many of the science vocabularywords in English and in Spanish. If they learn the word in bothlanguages, then they’re going to be more likely to remember itthe next time they need to use it in either language.

3.2.3. Fifth grade changing seasons taskLastly, we consider how the 5th grade teachers thought about

first language and culture as they reflected on their students’ par-ticipation in the changing seasons task. As with the 3rd and 4thgrade students, most of the 5th grade students were able to providevarious examples of experiences they had at home (both in Floridaand in their home country of origin when applicable) and whileplaying with their friends that were relevant to the science topic ofthe changing seasons. Although many of the 5th grade teachersdiscussed these student home experiences as obstacles to schoolscience learning, the 5th grade teachers were more likely than the3rd or 4th grade teachers to also discuss the potential of homeexperiences as resources to be leveraged.

Student (SLL 5): It’s like why you plant [crops] in the spring. Thedays are getting longer and there’s more Sun and more heat andthe plants like all that Sun and they grow taller.

Teacher (Y2): That answer he gave about why you plant seeds inthe spring. That was a great example. So he learned this growingup on a farm and he pretty much understands the practicalreason for why you do it, but now I can make that connection to

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e42 39

the science. Why are the days getting longer in the spring?What’s going on? He’s more likely to understand it because hehas that background for this very abstract idea.

Here the teacher made use of information the student sharedabout out-of-school experiences with the seasons (crop cycles) andconsidered how this practical knowledge could be connected tomore theoretical science concepts about changing seasons. Theteacher might not have tried to elicit such prior knowledge fromher students in the past, but now that she directly saw the value ofsuch connections through this interview, she might dedicate moretime and attention to student prior knowledge as part of her sci-ence instruction.

4. Discussion, conclusions, and implications

The research on teacher professional development has docu-mented the need for thorough, multifaceted, and ongoing teacherlearning opportunities in order to support teachers’ enhancedknowledge, beliefs, and practices related to content area instruction(Wayne et al., 2008; Wilson & Berne, 2004). Further, professionaldevelopment that connects content area instruction with students’cultural and linguistic experiences presents an additional layer ofchallenges for teachers and teacher educators (Janzen, 2008;Rodriguez & Kitchen, 2005). The limited research in this area,however, has pointed to the importance of fostering just such anintersection in order to support teachers in making academiccontent more accessible and relevant to the increasingly diversestudent populations now found in nearly all industrialized nations(Lee & Buxton, 2010; Leyendecker, 2011; Warren & Rosebery, 1995).Facing this challenge in our own previous research (Lee et al., 2007)caused us to seek out new approaches to support teachers inintegrating inquiry-based science with students’ prior knowledgefrom home and community contexts.

4.1. Discussion

Our resulting professional development model featured bothexplicit and implicit approaches to leveraging students’ culturesand first languages as instructional resources. Explicit approachesto support teachers’ cultural and linguistic awareness includedworkshop sessions on social and academic language registers andculturally congruent communication styles. Our implicit ap-proaches emphasized student problem solving, based on thepremise that if teachers practiced focusing their attention on whattheir students were thinking and saying, they would becomeincreasingly aware of where their students’ ideas were comingfrom. Given that the majority of any student’s experiences comefrom beyond the classroom setting, an increased teacher focus onstudent problem solving can potentially translate into an increasedawareness of students’ cultural and linguistic resources from homeand community. The topics we selected for our three studentproblem solving tasks (measurement, energy transfer, and chang-ing seasons) provided rich contexts for students to make connec-tions to their everyday lives beyond school and for teachers to thenreason about those connections and how to capitalize on them inthe science classroom.

As we have noted, our professional development model calledfor all teachers to participate in the project for three consecutiveyears. However, due to high rates of teacher mobility and attrition,in the end, most teachers participated for only one year. Thus, onesubstantial limitation of this study is that this participation patternlimited the overall effectiveness of both the intervention and theresearch efforts, including the teachers’ pedagogical reasoning in-terviews. Specifically, the first year of our professional development

work focused primarily on supporting teachers in implementingthe inquiry-based science curriculum that was central to the proj-ect. Because many of our teachers had limited prior experienceteaching inquiry-driven experimental science, content-basedpedagogical support took priority for first year participants. Dur-ing the second and third years, a greater emphasis was placed onpromoting student problem solving, analyzing student work, anddevising ways to leverage cultural and linguistic resources in theclassroom. As the majority of our participants never engaged in thesecond and third year professional development, however, theyhad limited opportunities to develop pedagogical reasoning in thecontext of our professional development. A second limitation of thestudy was that we conducted our student problem solving in-terviews in English only, as that was the only language of schoolscience instruction. Conducting the interviews in the SLLs’ firstlanguage of Spanish or Haitian Creole (languages spoken by manyof the teachers) might have provided teachers with additional in-sights into how better to leverage students’ linguistic and culturalresources in their science teaching. Nonetheless, the large data setof 206 interviews with 133 teachers did allow us to discoverimportant patterns in teachers’ pedagogical reasoning complexityas they considered the connections between science problemsolving and students’ first language and culture.

4.2. Conclusions

Our first research question examined the degree to whichteachers’ pedagogical reasoning about student problem solvingvaried across the four dimensions of pedagogical reasoning com-plexity. We found statistically significant differences between eachof the four dimensions of pedagogical reasoning. Specifically, nearlyall teachers generated large numbers of assertions about theirstudents’ science problem solving, but were less likely to elaborateon those assertions with concrete examples taken from watchingtheir students participate in the science problem solving task.Additionally, teachers often justified their assertions with anecdo-tal reasons based on other experiences with the student (or withstudents more broadly) rather than with specific examples takenfrom the student videos. Finally, the teachers were the least likelyto provide explanations for their assertions that were grounded inclearly articulated theories, structures or mechanisms.

This finding raises an important question for future professionaldevelopment work with teachers to support student problemsolving. If some dimensions of reasoning complexity are morechallenging to develop than others, and if teachers do not routinelyreason at the higher levels of pedagogical complexity themselves, isit possible for them to support their students in developing higherlevels of science problem solving? While teachers’ pedagogicalreasoning and student science problem solving are distinct types ofreasoning activities, they are surely related and can be mutuallyfostered.

Our second research question addressed the initial impetus forthis work, namely how teachers made connections between stu-dents’ performance on problem solving tasks and potential re-sources drawn from students’ first cultures and languages. Wequestioned whether a focus on supporting the development ofteachers’ pedagogical reasoning could provide an effective way toimprove their ability to leverage students’ cultural and linguisticfunds of knowledge (González et al., 2005) as resources for scienceteaching and learning. We adopted an implicit approach to makingcultural and linguistic connections to science content as a way toelaborate on the more explicit approach that had only limitedsuccess in our previous project (Lee et al., 2007). The student in-terviews in our current project revealed that the majority of stu-dents were able to make a wide range of connections between

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e4240

school-based science tasks and prior experiences beyond theschool context. For example, the 3rd grade students’ problemsolving about measurement indicated that they actually used morecomplex reasoning and language when discussing measurement inthe out-of-school context than in the school context (Buxton & Lee,2010). Nonetheless, the teachers in our project tended to focusmore on the limitations of their students’ experiences than on thepositive connections that their students were making betweenschool science topics and out-of-school experiences.

What was most surprising to us about this finding was that ourteacher participants themselves came from linguistic and culturalbackgrounds that were largely similar to their students, a circum-stance that is quite rare in schools serving large numbers ofimmigrant students in any industrialized country (Blomeke et al.,2012). Thus, our research context might be considered a best-casescenario for classroom environments in which teachers can bothunderstand and attempt to leverage their students’ cultural andlinguistic backgrounds to make connections with science content.Yet in our teacher interviews, participants often related to theirstudents’ experiences in home and community contexts as deficitsrather than as resources. There seemed to be multiple factorscontributing to this finding.

First, teachers in our study often referred to generational dif-ferences rather than cultural or linguistic similarities when com-paring their own life experiences with those of their students. Forexample, our teacher interviews were filled with stories of theteachers discussing experiences they had in their own childhoods,such as cooking, shopping, sewing, and playing games outside,where they felt they had learned about measurement, energy andother topics in the student problem-solving tasks. They juxtaposedthose experiences with the experiences they attributed to theircurrent students, which they portrayed as comprised mostly ofplaying video games, which failed to provide the relevant experi-ences related to science. In other words, while we saw our Haitianand Hispanic teachers and their Haitian and Hispanic students ashavingmuch in common, the teachers tended to focus on how theircommunities, cultures, and languages were changing in ways thatmade each successive generation of students less and less like them.

Second, the teachers were under a great deal of pressure fromaccountability policies and mandates related to student perfor-mance on standardized assessments. Becauseweworked in schoolsthat were struggling according to these accountability measures,the teachers were continually being told that their students werefailing to meet the required academic standards. Thus, there wasa broader deficit-based discourse that was applied to these stu-dents in their schools, which might also have influenced theteachers’ perceptions of their students.

These findings complicate the existing literature indicating thatteachers who share the language and culture of their students aremore effective at promoting student learning than those teacherswho do not share the same or similar backgrounds as their students(Okagaki, 2006). The results of our study imply that simply sharingcultural and linguistic background with students is insufficient toensure that teachers actively build upon those shared experiencesas academic resources; rather, rigorous and ongoing professionaldevelopment to this end would seem to be required.

4.3. Implications

Despite the less than fully encouraging results of this study, wecontinue to believe that attempts to improve teachers’ pedagogicalreasoning provide a viable way to connect content area learninggoals with students’ cultural and linguistic backgrounds. A focus onteachers’ thinking about students’ problem solving has been shownto enhance students’ skills and reasoning inmath at the elementary

grade level (Carpenter et al., 2000) and social studies at the middleschool level (Kuhn, 2005), and has been called for in scienceteaching at all levels (NRC, 2007, 2011). Our results show largedifferences across the dimensions of teachers’ pedagogical rea-soning that indicate limitations on the part of most of our partic-ipating teachers to build upon the concrete examples of studentthinking. To construct well-reasoned explanations of studentproblem solving, teachers first need access to a clear picture of theirstudents’ thinking. The types of tasks designed for this study pro-vide examples of how this can be done. Simply providing this ac-cess is not sufficient, however, as teachers also need practice andsupport to draw well-reasoned conclusions about the resourcesand limitations expressed in their students’ thinking that can beleveraged for pedagogical purposes.

While the clear patterns of teachers’ pedagogical reasoningoutlined in this study may help teacher educators develop newstrategies to support improved student problem solving, there arespecific details and challenges of such an approach that requirefurther research. First, we wonder how the topics and design ofstudent inquiry tasks could be improved to make students’ lin-guistic and cultural resources even more explicit and visible forteachers. Other research (e.g., González et al., 2005; Warren &Rosebery, 1995) has achieved better results in supportingteachers in making these connections, but not at the scale we un-dertook for this project. Thus, design of tasks for large-scale in-terventions would seem to require new conceptualization.

Second, we wonder how professional development workshopscan better provide the scaffolding that teachers need to engage inthe cognitively challenging work of learning from their students’problem solving. Our attempts to balance professional develop-ment that supported the teaching of science content through in-quiry, the development of academic language, and thestrengthening of teachers’ pedagogical reasoning gave participat-ing teachers many complex topics to consider simultaneously.Would a more singular focus on students’ science problem solvingresult in greater improvements in teachers’ pedagogical reasoningand at what cost to the other goals of a multifaceted professionaldevelopment effort?

Finally, wewonder if more can be done to connect teachers’ owncultural and linguistic resources to those of their students. Even ina seemingly best-case situation in which the ethnic and linguisticbackgrounds of the teachers in our study were largely similar tothose of their students, teachers often focused on cultural andgenerational changes that made their students’ experiences dif-ferent from their own. What professional development work couldbetter help teachers, regardless of their own backgrounds andthose of their students, to more closely consider the potential valueof connecting students’ out-of-school experiences with the aca-demic topics in the school curriculum? For example, could askingteachers to participate in the same science problem solving tasksthat we used with students help the teachers to better see culturaland linguistic connections?

Clearer answers to these questions and challenges are needed ifwe are to create successfulmodels of professional development thathelp teachers connect robust content area learning with students’cultural and linguistic backgrounds. Supporting teachers in buildingupon their students’ lived experiences remains a critical goal if wewish to foster enhanced problem solving abilities for all students.

Appendix A

Teacher pedagogical reasoning interview protocol

I’d like to ask you some questions about [student name] fromyour class. You know we interviewed [student name] recently and

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e42 41

asked [him/her] to do some science problem solving activities onthe topic of [grade level topic]. Now I want you to talk to me aboutwhat [student name] was thinking as [s/he] was doing the activitiesand why [s/he] was thinking those things. There are nine questionswe are going to talk about:

1. As you were watching the tape what general impressions didyou have about what [student name] knows about [grade leveltopic]?

2. What specific parts of the tape led you to think that?

PROMPT: Can you be more specific or would you like to look atthat segment of the tape together?

3. Where do you think [student name]’s ideas about [grade leveltopic] come from?

4. What specific parts of the tape led you to think that?

PROMPT: Can you be more specific or would you like to look atthat segment of the tape together?

5. Is there anything interesting or surprising to you in what[student name] said about [grade level topic]?

6. Is there anything that worries you about what [student name]said about [grade level topic]?

7. Would you say that [student name] is typical of the otherstudents in your class when it comes to [grade level topic]?Why do you think that?

8. (focus question e this is a question based on one particularlyrevealing part of each specific student interview e raise thisquestion if it has not already been discussed)

I was especially interested in [describe salient episode]. What doyou think about that?

9. As you think about teaching science to your class, are there anyways in which watching [student name] do this activity mayinfluence how you teach?

10. Do you have any questions for me about anything we havetalked about?

Appendix B. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.tate.2013.01.003.

References

Blomeke, S., Suhl, U., Kaiser, G., & Dohrmann, M. (2012). Family background, entryselectivity and opportunities to learn: what matters in primary teacher edu-cation? An international comparison of fifteen countries. Teaching and TeacherEducation, 28(1), 44e55.

Brown, D. (2012). Migration and rural population change: comparative views inmore developed nations. International Handbook of Rural Demography, 3, 35e48.

Buxton, C., & Lee, O. (2010). Fostering scientific reasoning as a strategy to supportscience learning for ELLs. In D. Senal, C. Senal, & E. Wright (Eds.), Teachingscience with Hispanic ELLs in K-16 8 (pp. 11e36). Charlotte, NC: Information AgePublishing.

Buxton, C., Lee, O., & Santau, A. (2008). Promoting science among English languagelearners: professional development for today’s culturally and linguisticallydiverse classrooms. Journal of Science Teacher Education, 19(5), 495e511.

Carpenter, T. P., Fennema, E., Franke, M., Levi, L., & Empson, S. B. (2000). Cognitivelyguided instruction: A research-based teacher professional development program formathematics. Madison, WI: Wisconsin Center for Education Research.

Dilworth, M. E., & Brown, C. E. (2001). Consider the difference: teaching andlearning in culturally rich schools. In D. V. Richardson (Ed.), Handbook ofresearch on teaching (4th ed.). (pp. 905e950) Washington, DC: American Edu-cational Research Association.

Duschl, R., & Grandy, R. (2008). Teaching science inquiry: Recommendations forresearch and implementation. Rotterdam, Netherlands: Sense Publishers.

Garcia, E., Arias, M., Murri, N., & Serna, C. (2010). Developing responsive teachers:a challenge for a demographic reality. Journal of Teacher Education, 61(1e2),132e142.

Gay, G. (2002). Preparing for culturally responsive teaching. Journal of Teacher Ed-ucation, 53(2), 106e116.

Gere, A. R., Buehler, J., Dallavis, C., & Haviland, V. (2009). A visibility project:learning to see how preservice teachers take up culturally responsive pedagogy.American Educational Research Journal, 46(3), 816e852.

González, N., Moll, L. C., & Amanti, C. (2005). Funds of knowledge: Theorizing prac-tices in households, communities, and classrooms. Mahwah, NJ: Lawrence Erl-baum Associates.

Hogan, K., Nastasi, B. K., & Pressley, M. (2000). Discourse patterns and collaborativescience reasoning in peer and teacher-guided discussions. Cognition and In-struction, 17(4), 379e432.

Janzen, J. (2008). Teaching English language learners in the content areas. Review ofEducational Research, 78(4), 1010e1038.

Jorgenson, O. (2000). The need for more ethnic teachers: Addressing the criticalshortage in American public schools. Teachers College Record, On-line formatonly. http://www.tcrecord.org ID Number: 10551.

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

Kuhn, D. (1989). Children and adults as intuitive scientists. Psychological Review,96(4), 674e689.

Kuhn, D. (2005). Education for thinking. Cambridge, MA: Harvard University Press.Ladson-Billings, G. (1994). The dreamkeepers: Successful teachers of African American

children. San Francisco: Jossey-Bass Publishers.Ladson-Billings, G. (1995). Toward a theory of culturally relevant pedagogy. Amer-

ican Educational Research Journal, 32(3), 465e491.Lee, O., & Buxton, C. (2010). Diversity and equity in science education: Theory,

research, and practice. New York: Teachers College Press.Lee, O., & Luykx, A. (2005). Dilemmas in scaling up innovations in science in-

struction with nonmainstream elementary students. American EducationalResearch Journal, 42(5), 411e438.

Lee, O., Luykx, A., Buxton, C., & Shaver, A. (2007). The challenge of altering ele-mentary school teachers’ beliefs and practices regarding linguistic and culturaldiversity in science education. Journal of Research in Science Teaching, 44(9),1269e1291.

Lehrer, R., & Schauble, L. (2000). Modeling in mathematics and science. In R. Glaser(Ed.), Advances in instructional psychology (pp. 101e160). Mahwah, NJ: LawrenceErlbaum Associates.

Leyendecker, B. (2011). Children from immigrant families: adaptation, develop-ment, and resilience. Current trends in the study of migration in Europe. In-ternational Journal of Developmental Science, 5(1e2), 3e9.

Lipsmeyer, C., & Zhu, L. (2011). Immigration, globalization and unemploymentbenefits in developed EU states. American Journal of Political Science, 55(3),647e664.

Luykx, A., Lee, O., Mahotiere, M., Lester, B., Hart, J., & Deaktor, R. (2007). Cultural andhome language influences on children’s responses to science assessments.Teachers College Record, 109(4), 897e926.

Lyster, R., & Ballinger, S. (2011). Content-based language teaching: convergentconcerns across divergent contexts. Language Teaching Research, 15(3),279e288.

Molto, M., Florian, L., Rouse, M., & Stough, L. (2010). Attitudes to diversity: a cross-cultural study of education students in Spain, England and the United States.European Journal of Teacher Education, 33(3), 245e264.

Morrier, M., & Irving, M. (2007). Teaching and learning within and across cultures: edu-cator requirements across the United States.Multicultural Education, 14(3), 32e40.

National Research Council. (2007). Taking science to school. Washington, DC: Na-tional Academy Press.

National Research Council. (2011). Framework for K-12 science education. Wash-ington, DC: National Academy Press.

Okagaki, L. (2006). Ethnicity and learning. In P. Alexander, & P. Winne (Eds.),Handbook of educational psychology (pp. 615e634). New York: Lawrence Erl-baum Associates.

Osborne, J., Erduran, S., & Simon, S. (2004). Enhancing the quality of argument inschool science. Journal of Research in Science Teaching, 41(10), 994e1020.

Park, S., & Oliver, J. S. (2008). Revisiting the conceptualization of pedagogical con-tent knowledge (PCK): PCK as a conceptual tool to understand teachers asprofessionals. Research in Science Education, 38(3), 261e284.

Passel, J. (2011). Demography of immigrant youth: past, present and future. TheFuture of Children, 21(1), 19e41.

Resnick, L. B., Salmon, M. H., & Zeitz, C. M. (1993). Reasoning in conversation.Cognition and Instruction, 11, 347e364.

Ricucci, R. (2008). Educating immigrant children in a ‘newcomer’ immigrationcountry: a case study. Intercultural Education, 19(5), 449e460.

Rodriguez, A. J., & Kitchen, R. S. (2005). Preparing mathematics and science teachersfor diverse classrooms: Promising strategies for transformative pedagogy. Mahwah,NJ: Lawrence Erlbaum Associates.

Sable, J., & Plotts, C. (2010). Public elementary and secondary school student enroll-ment and staff from the common core of data: School year 2008e09 (NCES 2010-347). Washington, DC: National Center for Education Statistics.

Santau, A., Secada, W. G., Maerten-Rivera, J., Cone, N., & Lee, O. (2010). US urbanelementary teachers’ knowledge and practices in teaching science to Englishlanguage learners: results from the first year of a professional developmentintervention. International Journal of Science Education, 32(15), 2007e2032.

C.A. Buxton et al. / Teaching and Teacher Education 32 (2013) 31e4242

Seidl, B. (2007). Working with communities to explore and personalize culturallyrelevant pedagogies: push, double images, and raced talk. Journal of TeacherEducation, 58(2), 168e183.

Selwin, D. (2007). Highly quantified teachers: NCLB and teacher education. Journalof Teacher Education, 58(2), 124e137.

Shulman, L. (1987). Knowledge and teaching: foundations of the new reform.Harvard Educational Review, 57(1), 1e22.

Steinberg, R., Empson, S., & Carpenter, T. (2004). Inquiry into children’s mathe-matical thinking as a means to teacher change. Journal of Mathematics TeacherEducation, 7(3), 237e267.

Teachers of English to Speakers of Other Languages. (2006). PreK-12 English lan-guage proficiency standards. Alexandria, VA: Author.

Toulmin, S. (1969). The uses of argument. Cambridge, England: Cambridge UniversityPress.

Villegas, A. M., & Lucas, T. (2002). Preparing culturally responsive teachers:rethinking the curriculum. Journal of Teacher Education, 53(1), 20e32.

Warren, B., & Rosebery, A. S. (1995). Equity in the future tense: redefining relation-ships among teachers, students, and science in linguistic minority classrooms. InW. G. Secada, E. Fennema, & L. B. Adajian (Eds.), New directions for equity inmathematics education (pp. 298e328). New York: Cambridge University Press.

Wayne, A. J., Yoon, K. S., Zhu, P., Cronen, S., & Garet, M. S. (2008). Experimentingwith teacher professional development: motives and methods. EducationalResearcher, 37, 469e479.

Wilson, S. M., & Berne, J. (2004). Teacher learning and the acquisition of profes-sional knowledge: an examination of research on contemporary professionaldevelopment. In A. Iran-Nejad, & P. D. Pearson (Eds.). Review of research ineducation, Vol. 24 (pp. 173e209). Washington, DC: American EducationalResearch Association.

Zhao, Y. (2010). Preparing globally competent teachers: a new imperative forteacher education. Journal of Teacher Education, 61(5), 422e431.

Zimmerman, C. (2000). The development of reasoning skills. Developmental Review,20(1), 99e149.