Journal of Science Teacher Education

, Volume 24, Issue 2, pp 297–322

Best Practice in Middle-School Science

Authors

    • Educational Theory and Practice DepartmentState University of New York
  • Kristen C. Wilcox
    • Educational Theory and Practice DepartmentState University of New York
  • Janet Angelis
    • Educational Theory and Practice DepartmentState University of New York
  • Arthur N. Applebee
    • Educational Theory and Practice DepartmentState University of New York
  • Vincent Amodeo
    • Educational Theory and Practice DepartmentState University of New York
  • Michele A. Snyder
    • Educational Theory and Practice DepartmentState University of New York
Article

DOI: 10.1007/s10972-012-9293-0

Cite this article as:
Oliveira, A.W., Wilcox, K.C., Angelis, J. et al. J Sci Teacher Educ (2013) 24: 297. doi:10.1007/s10972-012-9293-0
  • 935 Views

Abstract

Using socio-ecological theory, this study explores best practice (educational practices correlated with higher student performance) in middle-school science. Seven schools with consistently higher student performance were compared with three demographically similar, average-performing schools. Best practice included instructional approaches (relevance and engagement, inquiry, differentiated instruction, collaborative work, moderate amounts of homework, and integration of language literacy and science) and administrative practices (nurturing a climate of opportunity to succeed in science, offering professional development based on data and dialogue, engaging teachers in standards-based curriculum revision and alignment, and recruiting the right fit of teacher). It is argued that best practice entails multiple levels of teaching and administrative praxis that together form a school-wide socio-ecological system conducive to higher performance.

Keywords

Best practiceScienceMiddle schoolStudent performanceInquiry teachingSocio-ecological theory

Introduction

A growing number of science educators have focused on the identification, documentation, and sharing of best practices, that is, instructional strategies conducive to higher student performance in school science. While seeking to support educational research and professional development, these educators have relied on written, audio, and video recordings of classroom practice such as snapshots of practice—Internet videos of inquiry-based instruction (Van Zee and Roberts 2006), interactive videodisc cases of conceptual change lessons (Abell et al. 1998), written classroom cases (Koballa and Tippins 2004), and practitioner accounts of best practice techniques of middle-school science (Swango and Steward 2003). This literature underscores a widespread interest in best practice among science educators.

Despite this increased interest, science educators have yet to develop a sound theoretical foundation and a solid scholarly basis with empirical evidence such as higher student performance on which to support and justify their identification and sharing of best practice in school science. Furthermore, best practice in science has been examined mostly at the classroom level, without taking into account how school-level practices located outside classroom boundaries (e.g., administrative practices) interact with classroom-level practices to shape students’ science learning experiences. The present study attends to this issue by expanding beyond the boundaries of the science classroom and embracing the school at large with the intended purpose of painting a more complex picture of what best practice in science entails.

In this study, we conceive of best practice as a set of practices correlated with consistently higher student achievement, a term operationalized by taking into account socio-economic conditions and student performance on standardized science assessments. The central question guiding this study is What constitutes best practice in New York State (NYS) middle-school science? To answer it, we comparatively examine two sets of NYS middle schools with similar demographics: (1) higher-performing schools where students consistently have performed relatively better on state science tests and (2) average-performing schools where students have consistently demonstrated mediocre performance.

Best Practice in Science Education

In this section, we describe a set of instructional and administrative practices previously associated with higher student achievement in the science education literature.

Inquiry-Based Teaching

In this study, we use the terms inquiry and inquiry-based teaching in reference to constructivist strategies and techniques that teachers use to engage students in the design and implementation of investigative tasks (Colburn 2000). There are different variations of inquiry. Open inquiry is more student-centered in the sense that students are allowed to pose their own questions and receive little direction from the teacher (NRC 1996). Guided scientific inquiry (Furtak 2006) is a more structured pedagogical variation wherein students are provided with research questions and then guided through investigations whose answers are known by the teacher. Although inquiry is still widely used, more recent national standards have favored the term scientific and engineering practices (NRC 2011). In sharp contrast, the expression traditional instruction is used in this paper to refer to learning environments where primary emphasis is placed on passive acquisition of collections of static scientific facts through direct instruction (teacher lectures) and rote learning (textbook study).

Students taught through inquiry-oriented pedagogies have been shown to demonstrate higher levels of science performance than pupils who receive traditional textbook-based instruction (e.g., lectures), including statistically significant higher levels of achievement (content, process skills, scientific reasoning, and argumentation) (Geier et al. 2008; Lynch et al. 2005; Wilson et al. 2010); more positive attitudes toward science and more interest in scientific careers (Gibson and Chase 2002); and higher levels of engagement and motivation (Lynch et al. 2005). Similarly, using evidence to evaluate claims and explaining phenomena scientifically—“essential features of classroom inquiry” (NRC 2000, p. 25)—have both been identified as instructional strategies most effective in achieving positive student learning outcomes (Banilower et al. 2010). These studies provide evidence that inquiry-based science teaching generally leads to higher student performance on achievement tests.

Collaborative Instruction

Collaborative instruction or work—teaching practices structured around small groups wherein students are engaged in a common task and learn by means of mutual interactions (Oliveira and Sadler 2008)—has also been shown to successfully promote better science learning outcomes than more traditional instructional methods, including greater feeling of competence and increased motivation and improved academic performance (Hanze and Berger 2007), as well as statistically significant higher levels of achievement (Bowen 2000; Chang and Mao 1999; Lou et al. 1996; Shachar and Fischer 2004; Stamovlasis et al. 2006). These studies provide evidence that students generally perform better and are more motivated to learn science when working collaboratively in small groups.

Differentiated Instruction

Differentiated instruction has also been shown to enhance science learning for diverse learners and promote higher student achievement (than traditional science teaching strategies) at the middle-school level through practices such as a peer tutoring and differential curriculum supports (Mastropieri et al. 2006; McDuffie et al. 2009; Schroeder et al. (2007). These studies show that differentiated instruction is an effective way to help students become more successful in science.

Homework

The positive effect of homework on student achievement has been reported in a number of studies (Cooper et al. 1998, 2001, 2006; Cooper and Valentine 2001; Keith et al. 2004, Van Voorhis 2001, 2003). As science teachers engage their students in classroom environments that are supportive and motivating, homework has been positively associated with achievement and shown to be a contributing factor toward student success and academic motivation (Bempechat 2004; Epstein and Van Voorhis 2001; Katz et al. 2010; Trautwein and Ludtke 2007; Trautwein et al. 2006). These studies provide evidence that homework in moderation and directed toward specific content with clear objectives in a teacher supportive environment improves student achievement.

Administrative Practices

Administrative practices related to the provision of support, professional development, and leadership to science teachers have been shown to be an integral part of high-performing schools. This research shows that teachers’ ability to improve student achievement is contingent upon the availability of support from colleagues with knowledge about reform documents (Johnson 2006), purposeful integration of district and administrative policies combined with the availability of multiple interdisciplinary teacher teams (Trimble and Peterson 2000), effective principal leadership (Browne-Ferrigno and Fusarelli 2005), and the existence of administrative support systems and structures (City et al. 2009; Elmore 2004). Similarly, teachers’ use of investigative science curriculum materials has been shown to be positively correlated with teachers’ perceptions of principal support (Banilower et al. 2007). These studies provide evidence of the importance of administrative practices for student performance in science.

Theoretical Framework

In this study, we explore student performance in middle-school science from a socio-ecological perspective based on the theoretical model proposed by Bronfenbrenner (1979). Central to this perspective is the notion that student performance in science is related to multiple contextual influences or processes situated both inside and outside the classroom. Rather than simply assuming a direct causal relationship between classroom practices and student achievement, we adopt a contextual and holistic view of student performance in science as a set of cognitive abilities and knowledge bases that pupils develop as a result of their participation in activities situated within a set of nested social structures or systems. Each of these systems resides inside another as a set of Russian dolls: a science classroom nested in a school nested in a school district nested in a state and national policy context. Contextual processes within all these different systems interact and influence students’ academic performance in science.

Ecological theory, as described by Bronfenbrenner (1994), situates individuals within a set of systems: microsystems, exosystems, and macrosystems. Microsystems or the immediate classroom settings are those in which students experience proximal processes (e.g., instructional practices) that directly influence their performance in science. Exosystems are those settings in which students do not usually participate directly (e.g., teachers’ lounges, school administrative offices, state education department meetings) but within which distal processes may occur that can indirectly influence students’ performance in science (e.g., administrative practices, high stakes assessment, and standards decisions). Finally, macrosystems are the wider societal and institutional contexts in which particular schools are located (e.g., school district’s policies and student population’s socio-economic status and cultural background) that can also distally influence students’ performance in science.

The present study uses the lens of socio-ecological theory to explore processes and practices in microsystems (i.e., classrooms), exosystems (i.e., middle schools), and macrosystems (i.e., districts) that are effective in fostering higher student performance in science. In doing so, we seek to identify what constitutes best practice in NY State middle-school science, that is, teaching and administrative practices that enable educators to promote consistently higher science performance among middle-school students. For a diagrammatic representation of our socio-ecological lens, see Fig. 1 below.
https://static-content.springer.com/image/art%3A10.1007%2Fs10972-012-9293-0/MediaObjects/10972_2012_9293_Fig1_HTML.gif
Fig. 1

Socio-ecological framework

Methodology

This mixed-method study is part of a larger research project on schools achieving consistently higher student performance on standardized tests than other schools in the state of New York (Wilcox and Angelis 2006, 2007; Wilcox 2008). The science tests are described below.

NY State Science Exams

In New York, middle-school students generally take a state science assessment at the completion of eighth grade. The Grade 8 science exam is designed to measure the content and skills contained in the Intermediate-Level Science Core Curriculum grades, 5–8, which is based on the NYS Learning Standards for Mathematics, Science, and Technology (NYSED 1996). According to these standards, middle-school students will “use scientific inquiry to pose questions, seek answers, and develop solutions” (Standard 1); “understand and apply scientific concepts, principles, and theories pertaining to the physical setting and living environment and recognize the historical development of ideas in science” (Standard 4); and “apply the knowledge and thinking skills of science to address real-life problems and make informed decisions” (Standard 7). At the time of this study, the grade 8 science exam consisted of two components: a written test and a performance test; the latter is in the form of a set of laboratory experiments requiring short answer responses. The written test included 45 multiple-choice and 34 open-ended questions, and the performance test consisted of hands-on tasks set up at three stations (NYSED 1996). Each open-ended question on the test was evaluated with its own rubric based on the above standards.

It should be noted that not all students take the grade 8 science exam. Some students have the option to accelerate and take a high-school-level earth science course. These students then may opt to take the Earth Science Regents Exam, a high-school-level standardized test designed to measure the content and skills contained in the state Physical Setting/Earth Science Core Curriculum. Although this curriculum focuses on the same standards as those for 8th grade science, it places stronger emphasis on process skills and student-centered, problem-solving. At the time of this study, the Earth Science Regents exam was a written test and included a multiple-choice section and an open-ended response section. Like the 8th grade exam, students were expected to identify, discuss, explain, and analyze on this test.

Students’ raw scores on the grade 8 science exam and Earth Science Regents exam are converted into a 1–4 scale. These scale scores indicate different standardized levels of student performance. Students who do not meet the state standards receive a scale score of 1. Students who do not fully meet the standards (i.e., demonstrate only minimal proficiency) receive a scale score of 2. Students who meet the standards receive a scale score of 3. And, students who meet the standards with distinction are assigned a scale score of 4 (NYSED 2010).

School Selection

Publicly available school performance data for the grade 8 science exam and the Earth Science Regents exam for test years 2006, 2007, and 2008 were used for school selection. The number of students earning a passing grade was calculated from those receiving a scale score of 3 or 4 on the grade 8 science exam or the Regents exam. The number of high-performing students was calculated from the number attaining a scale score of 4 on either exam. The percentages of passing students and high-performing students were calculated from these numbers and then divided by the number of tested students in each school.

To identify the sample of schools for this study, for each test year (2006, 2007, and 2008), the percentages of passing and the percentages of high-performing students were regressed on background characteristics of the school and its student population as a means to adjust achievement for demographic features known to be related to overall school achievement. Background variables included in the analysis were those used by the NYS Education Department in equating schools, including school size, the stability of the student population (defined as the percent of students in the highest grade who were present in the school during the previous academic year), the percentage of English language learners, the percentage of African-American students, the percentage of Hispanic students, the percentage of Asian students, and the percentages of students eligible for free or reduced-price lunch. This regression analysis was conducted using the Statistical Package for the Social Sciences (SPSS).

Studentized residuals representing school science performance after adjusting for background variables were calculated for each of these regressions, six in all for each school (percent passing and percent high performing for each of the three years). The mean and standard deviation of the standardized residuals were then calculated, yielding a z score that reflects the science performance of each school relative to other schools in NYS, after adjusting for background factors, and a standard deviation that reflects the stability of achievement across years and across percent passing and percent high achieving. As shown in Table 1 below, seven higher-achieving schools were selected based on z scores greater than 1, with relatively small standard deviations, and three average-performing schools were identified to closely match demographic characteristics of the higher-performing set, with z scores less than + or −0.05. Additionally, all schools included in the study had per pupil expenditures clustering near the state average (to control for factors of community wealth) and open admissions policies.
Table 1

2006–2008 school regression

Schoolsa

#

8 graders tested

%

Free lunch

%

Reduced lunch

%

ELb

%

African-American

%

Hispanic

%

White

%

Other

Z

Combined rank

Higher performing

 Bolivar-Richburg

68

25

13

0

1

0

98

1

1.30912

 Geneseo

77

13

3

2

3

4

91

2

1.43154

 Greene

95

30

8

0

2

0

97

1

1.61084

 Jefferson

89

55

14

1

12

7

79

2

1.71341

 Johnson City

170

36

10

1

13

6

76

5

1.48842

 Oliver W Winch

270

9

5

0

1

1

98

0

1.39108

 Wayne

198

11

8

0

2

2

95

1

1.55713

Average performing

 Bram

83

30

17

0

1

1

97

1

−0.02520

 City

135

82

9

1

89

8

3

0

0.03935

 Vantin

333

26

11

3

18

10

70

2

0.02743

State average

N/A

36

8

7

19

21

52

8

N/A

aNames of the higher-performing schools are actual; leaders of those schools agreed to have their schools identified, and a case report for each school has been published. Names of the average-performing schools are pseudonyms

bEnglish language learner

Our sampling method sought a representative sample of schools from different geographic regions and with different ethnic compositions while also seeking schools with close to or higher SES challenges than the state average. Therefore, for each higher-performing school in the final cut, at least one average-performing school with similar demographic characteristics was included. For example, Jefferson and Johnson City (higher performing) were at or above the state average for poverty as was City Middle school (average performing). “Saturation” and redundancy are criteria typically used to determine the extent of sampling in multiple case study designs (Patton 2002; Merriam 1997). Based on previous studies conducted in this project, a minimum of three school cases was found to be sufficient in reaching saturation around salient themes of best practice; hence, this informed the sampling for average-performing schools. Since this study also intended to provide individual case studies of higher-performing schools as models for school improvement efforts and funding permitted inclusion of a total of 10 site visits, seven higher-performing schools were included. See Table 2 for school descriptions.
Table 2

Higher- and average-performing middle-school descriptions

School

Description

Higher performing

 Bolivar-Richburg

This rural school is located in one of the poorest counties in the State. Its district serves nearly 850 students, a little more than half of whom are in the middle/high school. Despite a nearly 40 % poverty level, grade 8 students have consistently scored well on the state’s science examinations

 Geneseo

In Geneseo, the middle school shares a building with the high school and a campus with the elementary school and district offices. The community features working farms, a college campus, and a growing commercial area. About a quarter of the 1,000 students overall are eligible for free or reduced-price lunch, although the percentage in the 500+ -student middle/high school is less

 Greene

Although the local population is declining as income levels fall and the cost of living rises, the Greene school district still serves about 1,200 students, including approximately 280 in a three-grade middle school. The middle-school building was once the high school, and the science program benefits from having inherited the fairly spacious science laboratory/classroom facilities of the former high school

 Jefferson

Jefferson Middle School is in Jamestown, a small city, whose industrial base is disappearing. The school faces increasing levels of student poverty (more than 50 % overall) and high transiency, the result, in part, of relatively high levels of rental property and relatively low rents

 Johnson City

The Johnson City school district serves more than 2,500 students in four buildings, one of which is the middle school with its 575 students. Committed to high achievement, this middle school has been consistently successful in science

 Oliver W. Winch

Identified as a School to Watch by the National Forum to Accelerate Middle-Grades Reform (an honor recognizing its high academic standards and developmentally appropriate learning opportunities), Oliver Winch serves about a quarter of the rural/suburban district’s 3,300 students, receiving students from four different elementary schools

 Wayne

This school is part of Wayne, a district that serves students from a number of hamlets and rural areas, some 2,615 students total in five buildings, including a 600+ student middle school. Across this large area, socio-economic conditions vary widely and students from all communities come together for the first time in the middle school. Student performance in science has been consistently high

Average performing

 Brama

In a rural setting with a nearly 50 % free and reduced-price lunch rate, Bram serves a relatively ethnically homogeneous population of fewer than 200 students

 Citya

In this urban school, more than three-quarters of students qualify for free lunch, and approximately 10 % more for reduced-price lunch. A majority of the student population is African-American, with some Hispanics and few Whites. Many students attend Bram from elementary through the middle grades

 Vantina

Located in a small city, Vantin serves nearly 1,000 middle-grade students. More than one-third are eligible for free or reduced-price lunch. The student population is ethnically mixed, with a small percentage classified as English language learners

aThese names are pseudonyms. As in Table 1, the higher-performing schools are actual names

Since our goal in using a case study approach was to deliberately take into account contextual conditions believed to be highly pertinent to the phenomenon of study (Yin 2005), using demographic factors that have been correlated in previous research on student performance was an appropriate criterion for school selection. Furthermore, although state assessments represent only one metric of student achievement for which limitations must be acknowledged, these data offer the benefit of a variable that can be used to systematically compare complex data sets. Although our sampling procedure is not without limitations, it nevertheless provides two sets of schools with statistically significant differences in science achievement and is not atypical of a multiple case study research design.

Data Sources

Data collected for this study included transcribed interview recordings, documentary evidence, and field observations. Volunteer teachers and administrators in each higher- and average-performing school were interviewed over 2 days by a research team. These interviews lasted for approximately 40 min. Interviewees for each school typically included two to four administrators and four to six teachers totaling 83 interviews (see “Appendix 1” for a teacher interview protocol). Administrator and teacher interview protocols were similar with the exception of some minor adaptations and a few additional questions related to school- and/or district-level management. Documentary evidence collected before and during site visits included curriculum maps, pacing guides in science; professional development information/materials; teaching evaluation information/forms; staff selection materials; unit and lesson plans; school schedules; district, school, and classroom assessments; and academic intervention services specific to science. Furthermore, field observations were guided by a semi-structured observation protocol (“Appendix 2”). For each observation, one entire science class period (~45 min) was observed and 33 observations in total were conducted (nine in average-performing schools and 24 in higher-performing schools).

Data Analysis

Individual research teams crafted case studies for each school which were shared with study participants to validate findings (Denzin and Lincoln 1995; Maxwell 1996). In addition, using a multiple case study method to build explanations of practice across cases, data were sorted and organized using qualitative database management software (HyperResearch) and the constant-comparison method for the coding and categorization of data (Miles and Huberman 1994; Yin 2005). A matrix was then developed to record whether particular practices were present, absent, or “in process” (i.e., in a planning phase), and memos were taken throughout this analysis regarding developing interpretations. Characteristics of practice in higher-performing schools were compared with those of average-performing schools. Practices that were both salient (i.e., of high importance) and typical in higher-performing schools and either not salient or non-existent in average-performing schools were identified as best practice.

To increase the validity of our study, we used data triangulation (of documentary evidence, interview, observation, and researcher memos), investigator triangulation (of multiple research teams and member checking of both individual case studies and the cross-case analysis of best practice), and triangulation in time and space (of multiple years of student performance data and multiple locations for site study). In particular, the 33 field observations were used to triangulate self-reported and actual proximal processes at the classroom microsystem level (i.e., check the extent to which practices reported in the 83 interviews actually occurred in classrooms).

Findings

In this section, we describe the set of best practices we identified at the classroom microsystem level and at the school exosystem and district/state macrosystem levels (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs10972-012-9293-0/MediaObjects/10972_2012_9293_Fig2_HTML.gif
Fig. 2

Overview of best practices identified

The Classroom Microsystem

Instructional approaches varied widely among higher- and average-performing schools. Nonetheless, a few practices stood out which are described below.

Relevance and Engagement

Engaging adolescents who often lose interest or become what one teacher characterized as “downtrodden” was a priority in higher-performing schools. In contrast with average-performing schools, higher performers typically showed evidence of applying multiple efforts to keep students’ motivation high. As emphasized by the Greene middle-school principal, “if it’s [science] not fun, they [students] won’t do it.” These educators recognized that “fun” for adolescents has to do with keeping science relevant and is the result of purposeful planning to ensure that students’ curiosities are stimulated and they can connect their own life experiences to the subject at hand. This shared view of fun was consistent with the notion of “engagement” typically used in inquiry lessons structured as a 5-E learning cycle (Marek 2008; Trowbridge et al. 2004).

Hands-on Activity

Hands-on instruction was a popular and often-used approach to teaching science in higher-performing schools, where teachers viewed it as a means to make science engaging and relevant to students’ lives. As a Geneseo teacher put it, “I try to make it as hands on and as pertinent to their lives as possible.” Nonetheless, these teachers strived for instructional balance by combining lecture with purposeful hands-on science activities as a way to both prepare students for more advanced science and still communicate increasingly complex ideas. As stated by a Bolivar-Richburg teacher, “You’ve always got to have a little bit of lecture… the experiences that they have as far as hands-on in science will keep them attached.” Such preference for hands-on activity was paralleled by an emphasis on inquiry-based learning which was often cited by higher-performing school educators. During our interviews, a Geneseo administrator touched on questioning as an essential feature of inquiry-based activity by describing how “creating the classroom environment where questioning is valued” and, in Jefferson, an administrator pointed to the importance of “looking at data, making predictions and realizing those ah-ha moments.” Similarly, our classroom observations revealed that textbook study paled in comparison with hands-on activity. The textbook was used in a much higher percentage of average-performing school classrooms than in higher-performing ones (33 and 17 %, respectively). Moreover, as shown in Table 3, hands-on activity typically took the form of guided science inquiries.
Table 3

Hands-on science activities observed in higher-performing schools

School

Activity description

Bolivar-Richburg

Students watered plants and recorded their growth in a plant journal

Geneseo

A student’s echocardiogram (measure of electrical impulses in the heart) at rest and after doing 20 push-ups was measured and displayed on a smartboard. Other students speculated on what spikes in the ECG meant

Greene

In the engagement phase of an environmental learning cycle, students played a version of the Survivor television game showa

Jefferson

Students utilized snapshot depictions of Pangaea from 65 million years ago to present to investigate Earth’s changes over time and created a “flip book.”

Johnson City

A teacher set up three stations: Plant, Mapping, and Matching of biome with appropriate animals and plants. At these stations, students learned facts, investigated patterns, made descriptive classifications, and solved puzzle-like problemsa

Oliver W. Winch

Sponges, starfish, mollusks, worms, crustaceans, centipedes, and arachnids were placed at different laboratory stations in addition to a lobster kept moist and cool in a refrigerator. Students had 5 min to identify the characteristics of at least two organisms and fill in their laboratory data table

Wayne

The teacher related a lecture on timelines to an upcoming field trip where students would identify fossils from different eras. The teacher demonstrated how to use metric measurements in creating a timeline by taping a paper timeline to the board. Students in small groups used paper rolls at their desks and then added in different important moments in Earth’s historya

aImplemented as part of larger investigative instructional units

Differentiated Instruction

Overall, the classes observed in higher-performing schools were more advanced in their efforts to integrate differentiated science instruction. As reported by a Wayne teacher: “We’re getting much better at differentiation, changing our lessons in three or four different tiers to teach to the different modalities and the different readiness levels. Especially at this age level, we’ve got kids who are such concrete learners and we’re trying to give them very abstract concepts like density.” Making a concept like density understandable for adolescents with differing background knowledge, learning styles, and intelligences brought in another instructional characteristic common among higher-performing schools, namely the provision of input and support from specialists like English as a second language (ESL) and special education teachers. As reported by a Wayne special education teacher: “Working with [the eighth-grade science teacher] and his students before the state exam was big, because we actually got to do some of the hands-on things that we wouldn’t necessarily be able to do if we were contained in this [special education] classroom. We got to participate with the regular ed and have his expertise as a science teacher and demonstrating a lot of those activities that the students are asked to do on the test, giving them the supports and the modifications that they need to understand the curriculum definitely helps them succeed in science.” Higher-performing schools also showed evidence of differentiating science instruction more frequently by using teaching assistants in 43 % of the observed classes (compared to only 11 % in average-performing schools).

Collaborative Work

Higher-performing school educators also offered students more opportunities for pair and group work than their average-performing school counterparts. Although teacher-led class time was approximately the same, the time allotted for student–student interaction, including pair and small-group activity, was higher in the observed higher-performing school classrooms (29 % of class time) than in average-performing ones (only 12 %).

Homework

Average-performing school teachers were more likely to spend more time reviewing larger amounts of homework than their higher-performing counterparts (18 and 8 % of class time, respectively). In contrast, higher-performing school teachers used more of their instructional time on new material rather than homework.

Summary

Within the microsystem of the science classroom, therefore, we observed the proximal processes directly associated with higher student performance in science. These proximal processes had elements that were both affectively and cognitively appropriate for adolescents. Affective elements included keeping science content and teaching relevant, engaging, hands-on, and collaborative while carefully adapting to students’ cognitive needs through inquiry-based and differentiated instructional strategies.

Beyond the Classroom—Exosystems and Macrosytems

Higher-performing school educators enacted a variety of administrative practices that indirectly influenced students’ experiences in science classrooms.

Promoting a Climate of Opportunity

The most consistent exosystem factor among schools whose students consistently achieved well in science was a climate of opportunity. In higher-performing schools, that climate was defined by a focus on ensuring that all students had an opportunity to succeed in science regardless of their socio-economic status, language background, ethnicity, or any other factor. Educators in higher-performing schools typically expressed a shared belief that it was incumbent upon school personnel to keep higher-level science within reach of all students.

Creating a climate in which adolescents felt capable of grasping the more abstract concepts of science was a multifaceted accomplishment. For many of the higher-performing school educators, it began with debunking the commonly held myth that some just cannot grasp higher-level science. As one teacher explained, perceptions of capability are extremely important in science education because they directly impact ways of adapting instruction that can either keep the natural curiosity many children have about science alive or not. In Bolivar-Richburg, a teacher noted the principal’s high expectations for the special needs students and described how a climate of opportunity was manifested daily and supported by the school principal:

The principal is very involved. On a daily basis he’s in the room, observing what’s going on with the students. He knows a little bit about each kid. That’s very important because they know he expects a lot out of each student. They’re held to the same standards as the kids in the general education environment. They may have different modifications to their curriculum, but they’re still expected to produce the same kind of quality work.

Focusing Science Education Efforts

In higher-performing schools, teachers and administrators poured over state test score data to inform their work and continually engaged in a dialogue about impacts of science instruction throughout the school year. This dialogue often involved analyzing school or district benchmark tests to inform revisions to the school curriculum and innovations in instruction. It also included discussions about data from elementary and high schools within a district and in some cases led to changes in policy, staff, and resource allocation. In the words of a Jefferson Middle School administrator “[We] use data … to help us to identify strengths and target weaknesses, particularly, areas in need of staff development and areas in need of curricular attention.” It was in this context that rich discussions around impacts of science instruction, science curriculum organization, and assessment took place.

Making Time for Professional Development

In higher-performing schools, administrators provided time to focus on curriculum development as well as direction on how to engage in the process. In Geneseo, an administrator voiced a sentiment shared among many of the higher-performing educators about the importance of preserving staff development time for teachers to engage in focused curriculum work in their subject area: “We, as a district, made a conscious decision this year that our four superintendent’s days [i.e., professional development days] were going to be utilized for curriculum work, and the principals and superintendents said departments can take curriculum release time, but the curriculum release time is very focused, it’s very purposeful; they meet with me ahead of time [to summarize] what we accomplished, our next steps, and the direction we are headed.”

Aligning Curriculum and Standards

Educators in all schools described the monumental task of aligning the science state standards and assessments with the school curriculum and organizing activities and materials across grades. However, in higher-performing schools, disciplined, reflective, and iterative processes for analyzing assessments to inform curricular revision were in place to facilitate this work. The ultimate aim of this collaborative culture and continuous revision of what constituted foundational knowledge in middle-school science was to afford students opportunities and a feeling of ease to explore scientific topics in more depth. As one Geneseo administrator put it, “Identifying what is fundamental knowledge to any particular scientific field, whether Earth Science or Biology, is a good challenge.”

Making Time for Science through Integration

In all schools, science often received limited instructional time partly due to the prevalence of an accountability structure that emphasized mathematics and English language arts. However, rather than simply allowing science to get “squeezed out,” educators in higher-performing schools set goals to incorporate interdisciplinary connections and supported field trips and extracurricular activities.

Students in higher-performing schools were typically surrounded by educators who were keenly aware of the importance of integrating good linguistic and mathematical literacy practices into science to improve students’ fluency in expressing ideas in science. Two commonly integrated linguistic literacy practices were reading comprehension and word attack skills. A special education teacher said: “Literacy [has] got to be number one. I mean… so much [in science] has got to do with literacy.” Her co-teacher expanded on the importance of literacy-building activities, “It gets back to the vocabulary that we do. It’s so content-specific that if they don’t have that basic understanding of the words that we’re speaking, they’re not going to be getting any of it. I’m constantly pushing literacy and, really, the vocabulary aspect.” A Jefferson teacher provided an example of how she taught vocabulary with her “root of the week” 5-min mini-lessons, explaining that these focused “on vocabulary and etymology, and how it can be used to gain insight into scientific knowledge and tasks.”

Educators in higher-performing schools showed abundant evidence of instruction incorporating other subjects such as math, social studies, health, art, and physical education. As for math connections specifically, as one Johnson City teacher noted, science can make math, a subject some kids “don’t like,” more engaging and comprehensible. In Bolivar-Richburg, a teacher noted the importance of basic math skills in being able to “go further in science.”

Recruiting the Right Fit of Teacher

Higher-performing school administrators strived to recruit the right fit of teacher for middle-school science. Qualities that characterized the right fit of science educator included the ability and willingness to relate to adolescents, knowledge, strong interest, state certification in science, and alignment with the school and district’s educational philosophy and culture. Another important dimension of fit was with regard to the school and district focus of instruction and broader philosophy of education. Alignment to the school’s instructional focus, as one Johnson City teacher asserted, was “huge.” If a teacher could not or did not want to adapt to the schools’ focal emphases in his or her lessons, the friction would be palpable and come with negative consequences. In Oliver Winch, for example, the focus on hands-on work and differentiated instruction was expected and counted as a measure of a teacher’s success. As underscored by the administrator below, the right fit in terms of philosophy and school culture also meant embracing the use of data to inform decision making:

[In Bolivar-Richburg] you need the right people. I mean in any subject, science, English, you name it. You have to have the right people who are willing to… look at the data and analyze that and be willing to make adjustments to your teaching style based upon that, especially once you see a pattern starting to develop.

Educators in higher-performing schools also managed to avoid the “stepping stone effect,” a phenomenon whereby teachers take middle-school positions as a stepping stone to a high-school science job. Administrators shared a common concern about ensuring that their middle-school science teachers, first and foremost, were willing to teach middle-school children. Many of the science educators in this study explained that middle-school students were quite savvy at picking up when a teacher did not like them, or cared about adapting the content to their interests, or felt comfortable teaching science altogether. Administrators in higher-performing schools looked for teachers who could relate to their students not only in-class, but after class, in hallways, through field trips, and in extracurricular club activities and were capable instructionally.

In sum, at the exo- and macrosystem levels, therefore, we saw distal processes that occurred outside the classroom itself impacting teacher beliefs about student potential, sharing of information and instructional strategies among teachers across content area and grade levels, and collaboration and ongoing professional development of science teachers. Another important distal process was the ability of those in higher-performing schools to successfully mitigate the “squeezing out” of science influenced by state-level policies that emphasized math and English.

Discussion

The above findings provide evidence that student performance in science is indeed the result of many instructional and administrative practices that expand beyond the boundaries of the science classroom. Furthermore, what constitutes best practice in science is not discrete teaching strategies adopted in individual classrooms isolated from the rest of the school, but rather multiple proximal and distal educational processes that together form a school-wide “ecological” system (i.e., a system of interdependent social and institutional factors) that is conducive to higher performance in science. In this section, we discuss the significance of these findings and situate them in the current science education literature.

Proximal Processes

At the classroom microsystem level, best practice in middle-school science included instructional approaches that emphasized relevance and engagement, were hands-on and inquiry-based (rather than textbook-based), frequently offered differentiated instruction, relied heavily on collaborative work, and limited the amount of time devoted to homework and review (as opposed to new material). Another key instructional practice correlated with higher performance in science was the integration of literacy-building approaches and interdisciplinary connections between science and other subjects. These findings indicate that higher student performance in middle-school science is impacted directly by the affective and cognitive supports provided within the classroom, being for the most part consistent with what is currently identified as best practice by science educators (see introduction) as well as by mathematics educators (Ma 1999; Stigler and Hiebert 1997). The only exception is the practice of homework assignment which has been repeatedly reported as conducive to higher student performance (Bempechat 2004; Cooper et al. 1998, 2001, 2006; Cooper and Valentine 2001; Epstein and Van Voorhis 2001; Katz et al. 2010; Keith et al. 2004; Trautwein and Ludtke 2007; Trautwein et al. 2006). In particular, Van Voorhis (2001) reports that students who complete more science homework (which requires a longer period of review time) demonstrate higher levels of achievement. However, the findings of the present study show that classroom time spent reviewing science homework is not correlated with increased student science performance. So, rather simply spending more time reviewing larger amounts of homework (as in the average-performing schools), homework assignment must be appropriate both quantitatively (i.e., given in suitable amounts) and qualitatively (i.e., engaging, clearly linked to content, consistently aligned with instructional objectives).

Despite the growing numbers of advocates, calls for inquiry-based teaching practices have often been met with resistance due to a lack of clear supporting evidence and implementation problems such as teachers’ difficulties in interacting with students (Furtak 2006; Keys and Kennedy 1999; Roehrig and Luft 2004) and grasping their new teaching role (Friedrichsen et al. 2006; Hayes 2002; Lotter 2004; Oliveira 2010a, b, c, d; Oliveira 2009; Oliveira et al. 2007). There is also a concern that integration of inquiry into science classrooms may have a negative impact on student achievement (Bianchini et al. 2003; Crawford 2007; Davis et al. 2006; Kuhn 2007; Roehrig and Luft 2004) and lower student performance on standardized tests (Jones et al. 2003). The present study suggests that (at least at the middle-school level) such concern is unwarranted and, despite its pedagogical complexities, integration of inquiry into science instruction is well worth the effort. As reported above, classroom implementation of inquiry was actually found to be correlated with higher student performance across the higher-performing middle schools of our study, hence consistent with our definition of best practice.

Distal Processes

The reported findings also show that distal processes such as school administration practices, although removed from individual science classrooms, can penetrate students’ immediate learning contexts and impact student performance. In higher-performing schools, administrators pay close attention to purposefully nurturing a climate of opportunity to succeed in more advanced science. Higher-performing school administrators also show evidence of efficiency in directing efforts to improve science education through the use of data and dialogue in professional development efforts, curriculum revision work, and support of recommended instructional approaches. Another key administrative best practice is ensuring a good fit among teachers, students, and administrators. These findings underscore professional development, educational leadership, and creation of shared vision and responsibility as essential administrative components of best practice in middle-school science, thus being consistent with previous studies (Browne-Ferrigno and Fusarelli 2005; Johnson 2006; Osisioma 2007; Rorrer and Skrla 2004; Rutherford and Broughton 2000; Trimble and Peterson 2000).

The reported administrative best practices are consistent with reform-based strategies previously shown or argued to be effective in improving school learning. Fishman et al. (2003) showed that professional development that promoted reflective practice improved student learning outcomes through sustained evaluation and subsequent change of instructional practice in the classroom. Fullan (2007) argues that collaborative professional learning (an ongoing type of teacher professional development based on the recognition that teachers need to learn every day in and out of the classroom) is required to achieve educational practice that fosters improved student learning outcomes. Similarly, in the higher-performing middle schools of our study, administrative practices were more effective in promoting reflective practice and professional learning.

Best Practice as an Ecological Construct

Our findings underscore that best practice in science education is a dynamic, complex, and organic phenomenon, as many factors can influence student performance in science. Thus, understanding what constitutes best practice requires the development and employment of a powerful analytical framework that can capture proximal and distal processes at the micro-, exo-, and macrolevels of science education. Our socio-ecological approach to understanding middle-school science constitutes an initial step in this direction. While most previous research has been limited to the simple examination of correlation between student science performance and isolated practices (mostly at the classroom level), we utilize a framework that allows us to simultaneously focus on a myriad of instructional as well as administrative best practices and to take into account factors beyond the science classroom such as internal administrative structures. Such holistic approach is closely aligned with Stigler and Hiebert’s (1997) argument that previous efforts aimed at engendering educational improvement have been excessively focused on discrete features of teaching such as materials, questioning type, and small-group instruction rather than on the process as a whole that manifests itself as “a system of tightly connected elements” (p. 17).

As a socio-ecological construct, we use the term “best practice” in reference to a set of ecological conditions (i.e., instructional and administrative practices) that co-occur with higher student performance in science. What we identify as best practice in this study is not generic science teaching practices ‘‘for anywhere’’ (Aikenhead et al. 2006) or “best under all circumstances” (Lynch et al. 2005), but rather “what works”—socio-ecological conditions that are conducive, as evidence through co-occurrence, to higher student science achievement specifically in the context of NYS middle-school ecosystems. It is important to note, then, that a different or modified set of socio-ecological conditions may be “best” in middle-school environments with different science curricula and assessments and geographic, economic, and socio-cultural characteristics. Nonetheless, the best practices we identify in this study shed some light on ways that science educators can begin to systematically (re)shape their school ecosystems in order to make them more conducive to higher student achievement.

Our definition of “best” entails a process of identification of significant patterns of what holds true in participants’ perceptions and what is observed in classrooms across contexts. What we recommend is not mere imitation or wholesale transfer of “best practices” without regard for context, but rather thoughtful and appropriate adaptations of best practices. Further, the reported findings should be used to inform close analysis of strengths and weaknesses across district, school, and classroom levels as they compare to the identified “best practices.” In this analysis, we recommend that the social and cultural contexts of the school figure prominently in discussions of what practices are most appropriate to adopt and in what ways to integrate them.

Significance and Conclusion

This study has important limitations that must be acknowledged. One noteworthy limitation is the unavailability of data on the observed teachers’ certification and years of teaching experience, both previously shown to constitute important socio-ecological factors in urban middle schools where typically large numbers of early career teachers often teach “out-of-field” due to high attrition rates (Darling-Hammond 2000; Ingersoll 1999). Another important limitation was that even though our selection method sought schools with high SES challenges and representative of NYS different geographic regions and ethnic compositions, our final sample of higher-performing schools was predominantly White and contained only a few schools with higher than the state average poverty level. Lastly, our study was also limited by our use of state-based standardized testing as the sole measure to gauge student learning outcomes. Such state assessment represents only one type of metric of student achievement.

Nevertheless, we believe that our findings have important implications for the field of science teacher education. In particular, our identification of inquiry as a best practice at the classroom microsystem level deserves further consideration. The practice of inquiry-based science teaching has prompted critical reactions, including questions such as “who is teaching whom?” (Lewis and Wagner 2002) and accusations of teacher invisibility in the learning-teaching process (Baines and Stanley 2000) as well as arguments in favor of the superiority of instructional modes that emphasize expert-novice differences (Kirschner et al. 2006). While our findings do not provide evidence of a direct causal relationship between inquiry and higher student achievement in science, they do point to inquiry as one of several socio-ecological conditions found in school ecosystems conducive to or supportive of improved student performance in standardized science tests. This correlation, we believe, constitutes reasonable grounds for arguments in favor of informing science teachers (both prospective and practicing) about inquiry-based pedagogical practices as well as encouraging their classroom adoption.

This study also contributes to the existing literature on middle-school science by providing an explanatory theoretical lens to the various factors at the microsystem, exosystem, and macrosystem levels that simultaneously impact student performance. Best practice in middle-school science as defined in this study shows evidence of moving in the direction of breaking down disciplinary lines and promoting students’ capacities to inquire, nurtured through affectively and cognitively appropriate instructional strategies such as hands-on, differentiated, collaborative, and literacy-based work. We can also glean from these findings a raised awareness of administrative practices that seem to have the greatest impact on student performance in science, including the promotion of a school climate of opportunity to succeed in science, accessibility to professional development, engagement in standards-based curriculum revision and alignment, and promotion of a good fit among administrators, teachers, and students. It is our hope that the reported findings can help ensure that an increasing number of middle-school students will have access to a state-of-the-art science education.

Copyright information

© The Association for Science Teacher Education, USA 2012