Science education reform efforts set clear expectations for K-12 education: to produce scientifically literate adults (American Association for the Advancement of Science [AAAS], 1990, 1993; National Research Council [NRC], 1996). The reforms also universally recognize that, in order for this goal to be achieved, the process of educating a scientifically literate population must start in the earliest grades and proceed in a coordinated fashion throughout the public school years. Science is a basic, as important as acquiring the skills and knowledge needed to read, write, use mathematical ideas, apply technology, understand the social sciences, and appreciate the fine arts (Loucks-Horsley et al., 1990).
 

Despite such goals, 25% of all elementary teachers do not teach science at all and, among those who do, science accounts for less than 2 hours of the instructional time each week (Raizen & Michelsohn, 1994; Tilgner, 1990). These figures are offered in contrast to the recommended 120 minutes of weekly science instruction at the K-3 levels and the 300 minutes at the 4-6 grade levels (Loucks-Horsley, 1990). Why don't elementary teachers teach science? Potential explanations include insufficient science content knowledge; limited resources, academic time, and physical space; an overly crowded elementary curriculum; and low levels of science self-efficacy (Raizen & Michelsohn, 1994; Schwartz, Abd-El-Khalick, & Lederman, 1999; Tilgner, 1990). Regardless of explanation, limited science exposure in the elementary grades results in low levels of science achievement. In addition, poor attitudes toward science developed by students in the elementary years may inhibit future science pursuits (National Center for Educational Statistics, 1995). Both achievement and attitudinal outcomes will inhibit reaching the ultimate goal of scientific literacy.
 

How do we best ensure adequate and appropriate science instruction at the elementary level? This article will explore potential answers to this question. Through an examination of the specific outcomes that constitute scientific literacy, the attributes needed in a teacher for effective science instruction will be outlined. This needed knowledge base will be used to evaluate five delivery models for elementary science instruction, as well as associated financial, institutional, and human resource costs. The article concludes with a call for research to determine the best delivery models for elementary science instruction.
 

Scientific literacy is not a single characteristic, but a compilation of qualities. On the most basic level, scientific literacy includes a basic understanding of key science concepts, often delineated into the categories of physical, life, earth, and space science (NRC, 1996). Hand-in-hand with conceptual understandings, a scientifically literate individual is facile with the methods of science, can use these methods to answer personal questions and judge the answers proffered by others, and can recognize the strengths and limitations of scientific inquiry. More important, a scientifically literate person should have an understanding of the concepts and processes that unify science and recognize the interrelationships and interdependencies between science and other disciplines (AAAS, 1993; NRC, 1996). Finally, the recognition of the personal and social value of science provides the rationale and motivation to study the discipline. These four characteristics (conceptual knowledge, nature of science, integration, and relevance) could then be considered fundamental elements of scientific literacy. Despite variations in wording and organization, these elements are pervasive in the reform documents and have specified outcomes at the lowest grades as precursors to adult science literacy.
 

Teaching science in the manner specified by the Standards "requires integrating knowledge of science, learning, pedagogy, and students; it also requires applying that knowledge to science teaching" (NRC, p. 62). Using this statement and the elements of scientific literacy, I suggest that there are four teacher attributes necessary for quality science instruction: content knowledge and attitudes, pedagogical knowledge, knowledge of students, and knowledge of curriculum. Content knowledge and attitudes are composed, at minimum, of an understanding of the four elements of scientific literacy: conceptual knowledge, nature of science, integration, and relevance. Attitudes that support science teaching include an enthusiasm and a willingness to create time for science instruction and recognize that all students have the right to be engaged in meaningful science instruction. Teachers with positive attitudes toward science will encourage similar attitudes in their students by modeling curiosity, using problem solving approaches when answering questions, relying on data, being skeptical of explanations while being open to new ideas, and respecting reason and honesty.
 

Pedagogical knowledge and skill refer to the ability to plan, implement and assess student engagement in meaningful science instruction. This instruction should be active, relevant, developmentally appropriate, and build on prior knowledge. Activities should be inquiry oriented, support the social construction of accurate science knowledge, and develop classroom community. The range of activities used should promote the science learning of all students and assist in the development of positive attitudes toward science.
 

Knowledge of students includes both a general knowledge of student development and specific knowledge of the students in one's classroom, allowing the teacher to capitalize on student interests and motivations to create a relevant science curriculum. This category also includes knowledge of student misconceptions for commonly taught topics.
 

Knowledge of curriculum allows a teacher to select, adapt, or create instructional materials to meet student needs and recognize how these materials combine to create a coordinated program of science both across grade levels (as specified in state or district guidelines) and across the curriculum (through integration or co-development of knowledge and skills in other content areas). Curricular knowledge synthesizes the other knowledge bases through the selection of developmentally appropriate content and activities based in student interests and experiences, and through the extension of science beyond school boundaries.
 

All elementary teachers possess some degree of each of the attributes needed for quality science instruction, but few possess high levels of knowledge and skills in each of these areas. And, in addition to science, elementary teachers are expected to have this level of knowledge for multiple disciplines. Is it realistic to expect all elementary teachers to possess these attributes in all content areas? If not, what type of elementary science delivery model would provide instructors who possess these attributes?
 

In thinking about the models for the delivery of science instruction, there are a multitude of alternatives. At one end of the spectrum, each elementary teacher is responsible for science instruction; at the other end, only science specialists are adequately prepared to handle such a task. The relative merits of these two positions have been argued in the literature for nearly 20 years (Abell, 1990; Hounshell & Swartz, 1987; Neuman, 1981; Olson, 1992; Williams, 1990). Between these two extremes, other delivery models exist. In this article I will describe five delivery models for elementary science instruction. This delineation of models is not intended to be exhaustive, but illustrative of the options that exist. Each model will be described in terms of the assumed characteristics of classroom practice, teacher preparation, and potential advantages and disadvantages. Following the example set by Abell (1990), similar delivery models in other content areas will be identified.
 

Classroom generalists

Elementary teacher preparation typically follows a generalist model, with each teacher taking content and methods courses in each of the areas typically included in the elementary curriculum: reading, writing, mathematics, science, social studies, health, physical education, and the fine arts. As a result, it is assumed that the teacher has sufficient knowledge and preparation to design and deliver a curriculum that adequately covers each content in a self-contained classroom. Once employed, each teacher has the flexibility to organize and allocate class time to the various content objectives as they see fit.
 

Model advantages include a deep understanding of student interests and development, curricular flexibility in terms of planning for thematic, interdisciplinary or integrated instruction, and no need for additional personnel. Disadvantages include limited content knowledge, limited science-specific pedagogical and curricular knowledge, and dispersed material resources. The most devastating disadvantage of this model may be the lack of time dedicated to science instruction resulting from multiple teaching responsibilities, pervasive accountability measures for reading and mathematics (but not science), and low levels of interest or self-confidence in science teaching.
 

Classroom science specialists

Most classroom teachers find that they have a preference for one or more of the content areas that they teach. Allowing teachers to identify a specialty area--either through a content major or minor, additional course work or workshops, or other forms of formally or informally recognized advanced preparation--would create classroom-based content resources within schools. For instance, a classroom science specialist would still be responsible for a self-contained classroom, but would take leadership or offer assistance to other teachers in the area of science instruction. The science specialist could be given the primary responsibility for previewing and selecting science curricular materials and ordering and maintaining science equipment, thus creating a division of labor for the benefit of all. This model is similar to the one proposed by the National Commission on Teaching and America's Future (1996) and would flatten school-based hierarchies and result in the maximum number of specialists in contact with students.
 

Advantages for the students of the classroom science specialists include increased teacher content and curricular knowledge, and time dedicated to science instruction. While other teachers have access, close proximity, and the support of a science specialist, unless there is an infrastructure of support within the school, few benefits may extend beyond the boundaries of the specialist's classroom. For instance, unless there is time planned into the school day for teachers to collaborate, the presence of the specialist will be underutilized. In addition, unless differential hiring patterns for specialists are introduced, there would be limited incentive for a teacher to become a science specialist. Most elementary teachers, if required to select a speciality area, would gravitate toward the language arts (Tilgner, 1990). Only if a per-school number of science specialists were required would the market demands encourage teachers to focus in this area.
 

Science support teams

With the addition of personnel, increased support in the area of science instruction can be achieved. A science specialist, a scientist, or a paraprofessional with science expertise could be assigned to a self-contained classroom for some part of the school day to promote science instruction. Three levels of support could be provided. In the first, the science specialist would take primary responsibility for science instruction with only minimal assistance from the classroom teacher. For instance, the classroom teacher would help select science content or activities that the specialist would then plan and implement. The teacher would assist with classroom and material management. A second scenario would include the teacher and specialist co-planning and implementing the lesson. In the third, the teacher would co-plan the lesson with the specialist and would then take the lead role in lesson implementation. The specialist would help locate curricular resources, collect and manage science materials, and assist in lesson delivery. Versions of the model may be equated with the role played by the school media center specialist (Abell, 1990).
 

Depending on the support level used, the advantages and disadvantages of the science support team model vary. Advantages include the shared expertise of two adults with differing specializations. The classroom teacher would specialize in knowledge of students, pedagogy, and the general curriculum, and the specialist would contribute knowledge of science content and curriculum. Depending on the level of support, teaching generalists would have a structured and supported opportunity to increase their knowledge and confidence in science instruction. Time for science instruction would be guaranteed and opportunities to integrate science into the rest of the curriculum would be maintained. Potential disadvantages are increased personnel costs, administrative structure, and time for collaborative planning.
 

Departmentalization within grade levels

Following a model most often seen in secondary schools, some elementary grades have elected to departmentalize. This arrangement is often informally organized by a group of teachers as opposed to mandated by the administration. In this model, each teacher is responsible for the majority of the academic content taught within a self-contained classroom. During specified times each week, however, the teachers "rotate" classes, teaching a specialized content. Science, along with social studies and health, is often taught in this fashion.
 

Departmentalized models guarantee time for science instruction and allow for science resources to be centralized with each grade level. While instruction from a teacher who has science content and curricular specialization would appear a likely benefit, conversations with teachers indicate that departmentalized arrangements are rarely induced with student learning in mind. More often, efficient use of teacher planning time is cited as the reason for departmentalization and content assignments are made by convenience or seniority rather than by specialized content knowledge. Disadvantages include decreased knowledge of students and minimal opportunities to integrate science with other curricular topics. Most important, departmentalization may result in relegating all the responsibility for science content to a single team member.
 

Science specialists

The delivery model most often depicted in the literature employs a science specialist who maintains a science laboratory/classroom. Hired as a science specialist, this teacher is solely responsible for science instruction and therefore has a higher degree of science interest, enthusiasm, and content expertise. As a recipient of specialized preparation, the specialist is skilled in the implementation of reform-based science instruction, assessment techniques, and knowledgeable of curricular guidelines and resources. In this model, classrooms have regularly scheduled times in the science lab. Science instruction is the primary responsibility of the science specialist who may interact to varying degrees with the classroom teacher. Similar to specialist programs in the fine arts or physical education (Abell, 1990), the classroom teacher generally "drops off" his or her students and uses the much-needed release time for planning rather than assisting in science instruction.
 

Advantages to the specialist model include guaranteed time for science instruction by a highly qualified individual and the centralization of science materials. Disadvantages include limited knowledge of individual student development and interests, decreased opportunities for content integration, the potential creation of elite images of science, and increased costs for personnel and administration.
 

The models offered in this article are not new. Advocates for science specialists argue for the benefits of increased time, interest, and expertise in the teaching of science, as well as the ability to centralize science materials (Abell, 1990; Hounshell & Swartz, 1987; Neuman, 1981; Williams, 1990). In support of this contention, research has shown that content knowledge is often the limiting factor to effective science instruction (Dobey & Schafer, 1984; Gess-Newsome, in press), and that teachers' level of content knowledge positively correlates with student outcomes on standardized tests of science (Schwartz et al., 1999). Specialist preparation programs would concentrate efforts on those individuals who show enthusiasm for the science rather than dilute their impact across all teachers. Proponents of the classroom generalist model worried about the uncertain financial resources for specialists as well as decreased opportunities for content integration, decreased knowledge of individual students, and the creation of the view that science is the providence of only a privileged few (Hounshell & Swartz, 1987; Olson, 1992). Advocates for science specialists counter that integration and knowledge of students can be fostered by communication and planning with the classroom teacher.
 

Table 1 characterizes the five delivery models of elementary science instruction in terms of the four teacher attributes needed to teach for scientific literacy. Each category has been assigned a ranking of low, medium, or high based on the perceived level of expertise held by the model's science instructor(s). In addition, categories related to the amount of time spent in science instruction, personnel needs (in-house versus extra staffing), and the centralized or dispersed nature of science-based equipment and materials are included. Finally, infrastructure requirements (relating to the time needed for planning, coordination, and collaboration) and teacher preparation needs (as academic generalists, science specialists, or a combination) are considered.
 

Interestingly, this 20-year-old debate continues with little empirical data to support the contentions of either side. With the exception of a recent study by Schwartz et al. (1999), I was unable to find any research that examines the student outcomes of classroom generalists versus science specialists. In the Schwartz et al. study, science specialists were found to have more sophisticated understandings of inquiry, the nature of science, and the national reforms. Their lessons plans clearly reflected these understandings and resulted in increased levels of higher order thinking skills in students when compared to academic generalists. This study also revealed that, in schools that employ science specialists, the regular teachers participated in fewer professional development activities and graduate level courses related to science. These teachers offered unsolicited constraints to teaching science (i.e., time, content, experience, equipment, and space) and cited the need for a science specialist. In the generalist district, the teachers had higher levels of science content preparation and offered no constraints to science teaching.
 

Clearly, research about the use of science specialists in the elementary schools is needed to transform this debate into reasoned action. First, the various delivery models must be evaluated as to the cognitive and attitudinal outcomes they produce in students and teachers. Which models result in increased student learning? Does attention to science content integration and process skills decrease in the presence of a science specialist? Second, why do so many teachers leave their preservice programs uninterested or unwilling to teach science? What program structures produce improved content knowledge and attitudes toward science teaching? A third area of research relates to policy: In the crowded curricula of the elementary school, can science become a basic? If schools hire science specialists, can time be found in the school day for collaboration? Can differential hiring practices increase market demands for teachers specializing in science? Without this knowledge base, academics will continue to argue the value of various models based on opinion rather than empirical evidence, and elementary schools will select models based on tradition, opportunity, or perceived need.
 

    Abell, S. K. (1990). A case for the elementary science specialist. School Science and Mathematics, 90, 291-301.

    American Association for the Advancement of Science (AAAS). (1990). Science for all Americans: Project 2061. New York: Oxford University Press.

    AAAS. (1993). Benchmarks for science literacy: Project 2061. New York: Oxford University Press.

    Dobey, D. C., & Schafer, L. E. (1984). The effects of knowledge on elementary science inquiry teaching. Science Education, 68, 39-51.

    Gess-Newsome, J. (in press). Secondary teachers' knowledge and beliefs about subject matter and its impact on instruction. In J. Gess-Newsome & N. G. Lederman (Eds.), Examining pedagogical content knowledge: The construct and its implications for science education. Dordrecht, The Netherlands: Kluwer Publishing.

    Hounshell, P. B., & Swartz, C. E. (1987). Elementary science specialists? Definitely!/We know better. Science and Children, 24 (4), 20-21, 157.

    Loucks-Horsley, S., Kapitan, R., Carlson, M. D., Kuerbis, P. J., Clark, R. C., Melle, G. M., Sachse, T. P., & Walton, E. (1990). Elementary school science for the '90s. Alexandria, VA: Association for Supervision and Curriculum Development.

    National Center for Educational Statistics. (1995). Understanding racial-ethnic differences in secondary school science and mathematics achievement. Washington, DC: US Department of Education.

    National Commission on Teaching & America's Future. (1996). What matters most?: Teaching for America's future. New York, NY: Author.

    National Research Council. (1996). National science education standards. Washington, DC: National Academy Press.

    Neuman, D. B. (1981). Elementary science for all children: An impossible dream or a reachable goal? Science and Children, 18(6), 4-6.

    Olson, A. K. (1992). In praise of the classroom teacher. Science and Children, 29(1), 16-17.

    Raizen, S. A., & Michelsohn, A. M. (1994). The future of science in elementary schools: Educating prospective teachers. San Francisco, CA: Jossey-Bass.

    Schwartz, R. S., Abd-El-Khalick, F., & Lederman, N. G. (1999, January). An explanatory study of the "effectiveness" of elementary science specialists. Paper presented at the annual meeting of the Association for the Education of Teachers in Science, Austin, TX.

    Tilgner, P. J. (1990). Avoiding science in the elementary school. Science Education, 74, 421-4