1

What’s Really Different?

A group of second-grade students were engaged in an animated discussion. Their teacher had brought in some colorful ears of corn she wanted to use to decorate their classroom for Thanksgiving.¹ But, she explained, she had left the corn outside and it had gotten wet in the rain. She wondered what might happen to it, and whether she could still use it to decorate the classroom. As they discussed the problem, the students asked questions and made assertions.

Some students thought the corn was ruined and might “rot and get stinky.” Others thought it would “get darker because it’s going to die” or “turn black because water makes stains.” One suggested that “it might grow because it has water,” but another doubted it would grow and wondered, “Is it fake corn?” A student asked whether “people do something differently or grow it differently [so that] real corn could became all different colors?” The class decided they should “observe it every day” to see “did the colors change” and “draw sci-agrams [science diagrams] to show what’s happening” and then “draw models at every stage so we can teach people what happened.” Figure 1-1 shows how the teacher captured the students’ discussion and what they decided to do to solve their problem.

Over the next few days, the students noticed something happening to the corn (see Figures 1-2 and 1-3) and became even more excited. They noticed that some green things were growing out of the corn and thought these looked like leaves. They also observed that brown things were growing out of the bottom of the corn, toward the water, and some students thought these must be roots.

___________________

¹ This description is a summary of classroom experiences in the second-grade classrooms of Lori Farkash, Nancy Jo Michael, and Ruth Purdie-Dyer, who worked with Brian Reiser and Michael Novak to develop a three-dimensional storyline intended to involve their students in science practices to get at key elements of the disciplinary core ideas about plant growth. Their storyline, with examples of student work, can be downloaded at http://www.nextgenstorylines.org/why-is-our-corn-changing [April 2016].

Each day the students rushed to see what was happening with the corn, and their teacher encouraged them to try to explain what they were seeing. One day a new finding led to a vigorous debate. The students had agreed that somehow the corn was “growing” and must in some way be “alive.” But they disagreed about what part of the corn was actually growing. Some students thought that each leaf was coming from an individual piece of corn (which the class learned are called “kernels”). Other students, however, responded that this meant that each kernel was like a seed, and it didn’t make sense for a plant to have so many seeds. Still other students pointed out the kernels falling off the cob, and they reasoned that something inside the corncob was growing and maybe pushing off the kernels as it grew. The class decided they needed to do a “fair test” to figure out where the leaves and roots were growing from.

The teacher encouraged them to brainstorm ways to conduct a fair test that would answer their questions. Students wrote and drew their ideas in their journals and then discussed them as a group. After some discussion the class came up with the idea of separately planting some kernels in one pot and pieces of the corncob in another pot to see which would grow. But then they discovered even more things they were not sure about. Should they plant these in soil? Would light be important? They were pretty sure water was important, given the evidence so far. From talking through these possibilities, they realized they needed even more fair tests. They decided to get some “regular” seeds—ones they knew for certain were seeds—and see what they needed in order to grow. Then they could set up the same conditions with the kernels and the corncob to see which part of the corn is the seed.

What these second-grade students were doing may not look radically different from what happens in many lively science classrooms. They were describing experiments that are common in many elementary classrooms: exploring whether plants need soil, water, or light to grow. What is important about this scenario, though, is that the students were engaged in this investigation in an attempt to make sense of something they had observed. They were comparing different growth conditions not because they were told to or were given the opportunity to pick a variable they could manipulate; rather, the students themselves identified the need to resolve a question, and the teacher supported them in designing an experiment that would resolve it.

Figure 1-1 shows part of the record the teacher made of the class discussion in which the students decided how to explore their questions about the corn.

Where does assessment fit into this picture? This teacher is using what the students say and do in this activity as part of an assessment of how well they can actually do science for themselves. The students have identified a question they wanted to answer because they observed something puzzling, and they have figured out a way to tackle it. They used what they know about investigation, argumentation, and explanation. They connected it to what they already know and are learning about the needs of living things. As the students engage in these various practices, the teacher is able to assess how they connect claims to evidence, how they connect their explanation of the new evidence to what they have previously figured out about plants, and how well they can develop a plan for an experiment. In other words, the work the students produce as part of their

learning also provides valuable information the teacher can use to assess their learning.

This way of embedding the assessment within the learning tasks reflects new ways to think about what a classroom science assessment can be:

• a tool for you as a teacher to collect information about what and how your students are learning;
• a way to help your students see what they’ve learned;
• a way to help your class figure out where they are in an investigation; and
• a tool for you to use in deciding on next steps for instruction and identifying the supports individual students need.

In other words, assessment can be an integral part of your teaching practice, rather than an interruption.

This approach to assessment reflects the thinking about science learning and instruction described in A Framework for K–12 Science Education (National Research Council, 2012) and new science standards that many states and districts are adopting—in many cases, the Next Generation Science Standards (NGSS).² In guiding her students to explore what they observed about the corn, the teacher in the example has actively involved students in the practices of science in the way those two documents describe: the students design experiments, collect data, make inferences from the data, and so on. She is also using these tasks to assess learning that cannot be measured by traditional assessments that rely on multiple-choice and other types of selected-response questions.

This chapter describes what is different about this kind of learning and why it means that different thinking about science assessment is needed. It explores the characteristics that assessments need to have if they are to measure this kind of learning, and how this kind of assessment supports instruction. This chapter also describes some basic principles to guide assessment of how students’ learning develops, and it looks at how those principles fit with the existing principles of good testing practice.

___________________

² The Next Generation Science Standards (NGSS) are available online at http://www.nextgenscience.org [March 2016].

A NEW WAY TO THINK ABOUT SCIENCE LEARNING

The purpose of new standards for science education—whether the NGSS or similar ones developed by states—is not just to rearrange the order in which topics are taught across the grades. These standards are based on current understanding of how kids learn—and how science teaching can reflect the way scientists and engineers do their work. They are designed so that students will do science themselves, not just learn about how other people have done it or memorize facts. Good teachers have always known that learning doesn’t happen in a tidy, straight line, but now research has given us ways to describe science learning more accurately. A key idea from that research is that in order for learning to really “stick,” students need continuous opportunities to engage in scientific thinking and practices and to gradually build their understanding of how new knowledge fits with what they already know.³

Scientists and engineers have a lot of specialized knowledge and skills. But what makes them experts is not that they have command of a lot of facts or are especially skillful at using technical equipment or performing experiments. These experts have had years of study and experience that give them a broad understanding of how the science ideas and practices they have learned fit together. They have developed the capacity to use their understanding and expertise to form and investigate hypotheses, solve problems, and develop new knowledge.

To be science literate is to be able to see how and why science and engineering really matter, to know how to reason from evidence, and to have a sense of how scientists and engineers do what they do. This understanding of science is not only important for students as they progress in secondary and postsecondary science study. When they are adults, today’s students will need to apply their capacity to think scientifically about important global challenges—such as climate change, the production and distribution of food, the supply of water, or pandemic diseases—even if they are not scientists or engineers themselves. The capacity to see connections across disciplines and contexts and to understand how scientists think will help students at each stage grasp scientific ideas that will be useful to them in their everyday lives long after they finish school.

___________________

³ Two National Research Council reports provide more information about research on learning: How People Learn: Brain, Mind, Experience, and School: Expanded Edition (National Research Council, 2000), available at http://www.nap.edu/catalog/9853 [May 2016] and How Students Learn: Science in the Classroom (National Research Council, 2005), available at http://www.nap.edu/catalog/11102 [May 2016]. An update of How People Learn is forthcoming in 2017.

Science Learning Is Three-Dimensional

These insights about learning are reflected in a key idea laid out in the 2012 framework: namely, that science learning should be three-dimensional. Instruction that develops scientific thinking and learning will integrate three dimensions: (1) the practices through which scientists and engineers do their work; (2) the crosscutting concepts that apply across science disciplines; and (3) the core ideas of the disciplines.⁴ What are these three dimensions?

Scientific and Engineering Practices

Scientists and engineers rely on eight key practices, such as asking questions and defining problems, planning and carrying out investigations, and analyzing and interpreting data. These are called practices—not skills—because when and why a practice is needed is just as important as how it is done. As students engage in these practices for themselves, they come to understand that science and engineering are creative processes of developing explanations and solutions. These practices are not isolated from core ideas; they are the means by which scientists investigate and build models and theories.

The eight key practices are:

1. Asking questions (science) and defining problems (engineering)
2. Developing and using models
3. Planning and carrying out investigations
4. Analyzing and interpreting data
5. Using mathematics and computational thinking
6. Constructing explanations (science) and designing solutions (engineering)
7. Engaging in argument from evidence
8. Obtaining, evaluating, and communicating information

Students cannot master these practices without opportunities to directly experience them. When students have the opportunity to “do” science, they don’t

___________________

⁴ The 2012 framework describes the three dimensions in detail. The NGSS describe grade-level performance expectations associated with these three dimensions. The easiest way to see how the three dimensions work together is to explore the NGSS website: http://www.nextgenscience.org [July 2016]. The description of these NGSS ideas in this book closely follows the text of the report on which this book is based, Developing Assessments for the Next Generation Science Standards.

just learn facts and ideas; they learn to engage in complex scientific reasoning. By “doing” the practices students also demonstrate how they are thinking through a challenge and provide opportunities for educators to assess what they are learning. As the quick look at student scientists at work shown in Box 1-1 demonstrates, teaching students to do science for themselves can be messy but can help them develop a love for science.

Crosscutting Concepts

Some scientific concepts are important within and across disciplines. Crosscutting concepts are important tools for making sense of phenomena that can be observed. They help students structure their thinking about new observations and information they encounter and provide a scaffold upon which they can build understanding. These ideas—such as “cause and effect,” “systems and system models,” and “energy and matter”—help students make connections across contexts and over time. The crosscutting concepts are:

• Patterns. Observed patterns of forms and events guide organization and classification and prompt questions about relationships and the factors that influence them.

• Cause and effect: mechanism and explanation. Events have causes, sometimes simple, sometimes multifaceted. A major activity of science is investigating and explaining causal relationships and the mechanisms by which they are mediated. Such mechanisms can then be tested across given context and used to predict and explain events in new contexts.
• Scale, proportion, and quantity. In considering phenomena, it is critical to recognize what is relevant at different measures of size, time, and energy and to recognize how changes in scale, proportion, or quantity affect a system’s structure or performance.
• Systems and system models. Defining the system under study—specifying its boundaries and making explicit a model of that system—provides tools for understanding and testing ideas that are applicable throughout science and engineering.
• Energy and matter: flows, cycles, and conservation. Tracking fluxes of energy and matter into, out of, and within systems helps one understand the systems’ possibilities and limitations.
• Structure and function. The way in which an object or living thing is shaped and its substructure determine many of its properties and functions.
• Stability and change. For natural and built systems alike, conditions of stability and determinants of rates of change or evolution of a system are critical elements of study.

Disciplinary Core Ideas

There are core ideas to learn in each of the main disciplines of science—the physical, life, and Earth and space sciences—and in engineering, technology, and applications of science. Students could never learn all the important knowledge in these fields, but learning the core ideas will prepare them to understand and evaluate new information they learn on their own. A few examples of core ideas are that all living things are made up of cells, that plants depend on water and light to grow, and that wind and water can change the shape of land.⁵

___________________

⁵ Chapters 5–8 of A Framework for K–12 Science Education describe the core ideas in each field (National Research Council, 2012).

Core ideas for K–12 science:

• have broad importance across science and engineering fields or are key organizing ideas in a field;
• provide a key tool for understanding or investigating more complex ideas and solving problems;
• relate to students’ life experiences and personal concerns; and
• can be taught and learned over multiple grades at increasing levels of depth and sophistication.

The core ideas fall under four broad areas of science:

Integrating the Three Dimensions

The idea that these three elements of learning—practices, crosscutting concepts, and disciplinary core ideas—need to be woven through every aspect of science education, not taught as separate entities, is called three-dimensional learning. It is important to learn science this way because doing science requires multilayered thinking. The practices that scientists and engineers use are tools for collecting or making sense of data, testing a hypothesis, or in some way answering a question

or solving a problem. Identifying which tools are needed and applying them is at the heart of doing science. Often it is the creativity of an experimental design or a way to collect information, or the originality of an approach to data analysis, that constitutes a scientific breakthrough. Scientists and engineers see how to apply the practices they have mastered because of what they understand about crosscutting concepts and core ideas.

Rather than simply mastering what is on a list of important science skills, students need to learn to use the practices the ways scientists do: that is, applying them as they are needed to solve a problem. At the same time, students cannot fully understand scientific information and conclusions without some understanding of the methods scientists use to investigate, analyze, communicate, and draw conclusions about the subject. Students need to gradually build their understanding of crosscutting concepts and begin to see how they apply in different contexts. And students need to gradually build an understanding of how core ideas from one branch of science relate to others: for instance, how ideas from chemistry can help them understand what happens when the body digests food. Three-dimensional instruction teaches students to think like scientists and provides the foundation for them to develop sophisticated science reasoning skills.

Key Idea

Three-dimensional learning is not just important for the education of future scientists and engineers; it is essential to help all students understand how science works and how to use that understanding to learn, make decisions, and understand their world throughout their lives and in their careers as adults. It’s a key aspect of the new vision for science learning.

Science Understanding Develops Gradually

When we see the purpose of science education as helping students learn to reason, ask questions, and test their ideas the way scientists and engineers do, it’s a reminder that learning develops gradually over time. This is not a new idea: teachers know that students learn as they are ready to do so and that they need to continuously build on what they have already mastered. For decades researchers have also been exploring the way learning develops and have provided insights that are already strong influences on education in science and other subjects. These insights were a foundation for the development of the 2012 framework, the NGSS, and other standards that were based on the framework’s vision for science learning.

Students have many ideas of their own that are not supported by scientific evidence. These ideas are generally grounded in experience and are starting points that teachers can work with to help students build more accurate and complete understanding. This is important on a daily basis in the classroom: Students may begin the study of, say, the solar system, with their own ideas about the orbits of the planets and other topics they are learning about. Teachers can uncover those ideas and use them to draw students into the ideas and activities they are teaching.

Understanding how students’ understanding develops is also important in the longer term: Expectations for students in early elementary school are naturally quite different from those for high school students. What students can readily understand in elementary school will be simpler than what they are able to understand later. Recognizing the steps along the learning pathway, and having strategies to address them, allows teachers to use and build on them and, as students are ready, to address ideas that are not scientifically accurate.

Ideally, students will encounter crosscutting concepts and disciplinary core ideas in different ways in each of their science classes, in many different contexts. The sophistication of their understanding will grow as they use and apply them and gain facility with science practices in the course of countless science experiences. This kind of learning requires coordination, or what educators call coherence, across disciplines and across students’ years of schooling.

Coherence

Science education that helps students develop requires careful coordination of all its elements, or coherence. For a science education program to be coherent:

• instruction, curriculum, and assessments, as well as professional development and other key elements, must all be aligned to the same learning objectives and work together to support student learning; and

• learning goals (performance expectations) and curriculum must be aligned across all the years of science education so that the instruction in each year builds on—rather than repeats—what came before and prepares students well for what comes after.

One key to coherence is the way goals for learning, or performance expectations, are defined. These are descriptions of what students are expected to be able to do and understand. The performance expectations in the NGSS are targets for assessment, and they describe how students can demonstrate their ability to understand and use their knowledge (crosscutting concepts and disciplinary core ideas) through engaging in science and engineering practices. Other standards that are using the ideas in the framework also tie these aspects of learning together.

In a coherent science education system, performance expectations help guide each element so it supports students as they broaden their understanding of crosscutting concepts and core ideas and expand their facility with the practices of science and engineering. This is a collective responsibility, but you contribute to the coherence of your students’ science education in many ways, as we discuss in Chapter 5.

Learning Progressions

Coherence is a general principle, but at its root are the expectations for student learning. The path toward mastery or expertise in a particular area can be called a learning progression. Researchers have examined learning for particular kinds of understanding in science and other subjects. For some topics they have described specific learning progressions that are typical for most students and thus can be used by educators to shape instruction. By explaining what students are expected to know and to be able to do, standards describe performance expectations for each level and each topic or area. Standards that reflect three-dimensional learning include explicit descriptions of how the understanding of crosscutting concepts and core ideas, as well as facility with practices, develops over time. Students will develop an understanding of knowledge and concepts by engaging in practices that help them question and explain what they observe.

Newer standards such as the NGSS reflect the importance of the learning process. They make clear both that students should be encouraged to use what they know in reasoning about what they observe and that struggling is critical to learning, just as it is a critical part of the way science is done.

The example in Box 1-2 shows a learning progression from the NGSS. Researchers have developed descriptions like these for some science topics, but not all are based on empirical research; others are hypothetical.

This NGSS learning progression describes stages in understanding that are natural and expected: what students learn in the earlier grades is not wrong, but it is less complete than what they learn later. A big difference across the stages is seen in how crosscutting concepts and core disciplinary ideas are woven in. The phrases in italics in the learning progression in Box 1-2 show how ideas become more sophisticated over time—each building on the strong foundation established in earlier years. This happens when the curriculum and instruction are coherent across time, and teachers are aware both of the ways their students have encountered these ideas before and of how they will continue to develop later.

As we will see in later chapters, this idea is the key to assessing three-dimensional learning as it is described in the 2012 framework. The assessments we will explore demonstrate ways of collecting evidence about how students are

progressing along a pathway that leads to the three-dimensional performance expectations, whether for a lesson, a unit, or a grade level.

Key Idea

The idea that student learning develops gradually and cumulatively fits perfectly with a three-dimensional approach to learning: this kind of understanding can come only through repeated opportunities to do science across units, disciplines, and years. The goals for units and stages of growth describe what can be expected of students at a particular stage; they are also stepping stones for the more sophisticated and complex understanding students will be capable of as they integrate crosscutting concepts with disciplinary core ideas, using their developing expertise with science and engineering practices. New science standards and curricula are designed so that students encounter concepts and practices as they are ready for them; the concepts and practices become more elaborate as students grow and gain experience.

CHANGES IN THE CLASSROOM

The idea of three-dimensional science learning can sound abstract, but it translates into immediate practical changes in instruction in the classroom. Teachers who are responding to this vision are shifting their focus. Rather than concentrating on helping students absorb sets of factual knowledge, the teacher can focus on strengthening students’ capacity to think and reason about the ideas and information they are tackling. Students who learn this way gradually expand the breadth of their understanding of crosscutting concepts and core ideas as they gain mastery at posing and investigating scientific questions and analyzing their findings. This shift in focus is expressed in many changes large and small in daily classroom practice, such as the ones listed in Table 1-1.

These differences in classroom practice also reflect the idea that learning develops gradually and cumulatively. Notice that the activities in the right-hand column tend to be ones that take time, may be done in multiple steps, and involve many different types of tasks. These activities also allow the students to direct their own learning—pursuing their ideas and hypotheses within an instructional structure rather than simply following instructions. Many are also activities that can be done with different degrees of sophistication at different grade levels.

TABLE 1-1 Changes in Classroom Practice

SOURCE: Adapted from National Research Council (2015).

Capitalizing on Students’ Natural Curiosity

One way to engage students in active science thinking—and help them to see connections and to understand how and why science ideas are important—is called “anchoring instruction in a phenomenon.” With this approach, the teacher identifies a phenomenon—a puzzling or counterintuitive circumstance, event, or process—that is apparent to her students. A phenomenon might be an unusual

weather pattern, the behavior of animals that students observe on a field trip, or questions students have about how a particular technology works. The teacher challenges the students to explain or resolve it or to design a solution to a problem. Often the teacher is responding to what students notice and ask about, and he or she uses that opportunity to structure instruction, as the second-grade teacher in the example at the beginning of the chapter did when her students noticed changes in the corn.

This teacher identified an opportunity that was not only fun and interesting but also engaged students’ genuine curiosity about something that can be explained scientifically. Furthermore, the science needed to explain this phenomenon was part of the learning objectives for the class, so the teacher could use what the students noticed about the corn to move learning forward in a purposeful way.

Wherever a teacher finds it, a well-chosen phenomenon will focus the students on connections between what they are learning and what they observe in the world. It will provide the students with a shared experience to which they all have equal access. It will require the students to integrate different science or engineering ideas and practices in order to collect the evidence and other information they need to meet the challenge. Students will draw on what they already know and use the practices they have at their command to explore further.

A phenomenon or problem that would work well for instruction will⁶:

• build on everyday experiences and relate to things students do or recognize. It’s especially important that it engages students from varied cultural and language backgrounds.
• relate to performance expectations, engaging students in core ideas and crosscutting concepts and requiring them to use science and engineering skills.
• be too complex for students to solve in a single lesson. The possible solutions should be ones that students could not find online or reach without some teacher guidance.
• be observable to students. The phenomenon should be something students can learn more about by, for example, using scientific procedures or technological devices such as telescopes or microscopes, collecting data outside, or finding patterns in data.

___________________

⁶ These ideas are drawn from material posted by the R+P Collaboratory; see http://stemteachingtools.org/assets/landscapes/STEM-Teaching-Tool-28-Qualities-of-Anchor-Phenomena.pdf [May 2016]. For more ideas about activities that are engaging to students in this way, as well as instructional materials, see other practice briefs at http://stemteachingtools.org/brief/26 [April 2016] and the resources at http://ngss.nsta.org [May 2016].

• be something students can learn more about with data, images, and text that are accessible to them. It should allow them to use science and engineering practices to conduct firsthand or secondhand investigations.
• be a specific set of circumstances, such as a case or a problem (e.g., an infestation of pine beetles), or something that puzzles the students (e.g., why isn’t rainwater salty?), or something students wonder about (e.g., how did the solar system form?).
• is important. The phenomenon should be a problem or question that people care about so the students will clearly see why their findings are important.

In one example of this approach, developed by the Inquiry Hub, the phenomenon to be explained is the planting of trees in cities—an activity that is intended to be beneficial to the environment but, in many cases, actually disrupts the local ecosystem.⁷ The students are asked to figure out what kinds of trees should be planted, and where, to maximize the intended benefits. They can then do a series of investigations to explain such specific questions as How do trees affect an ecosystem and its habitats and food web? and What trade-offs are involved when trees are planted? They are given the challenge of choosing a species of tree to plant in their schoolyard that would be optimal for promoting biodiversity as well as benefits to human beings and other organisms.

Engaging Students from Diverse Backgrounds

Science educators are aware of the importance of respecting students’ diverse experiences and cultural backgrounds and using them in instruction. All students bring valuable life experience and ideas to their classrooms, and their science learning is most successful when instruction draws on and connects with that richness. Because science involves specialized language, as well as the precise use of words and ideas that are understood more loosely outside the science context, English-language learners have an extra challenge in the science classroom. It can also be challenging for educators, who must bear these issues in mind along with the many other goals they have for their teaching practice.

A key idea in the 2012 framework is that science is critical for all students. The varied science activities that allow students to develop three-dimensional

___________________

⁷ The example, developed by Sam Severance, can be found at http://learndbir.org/resources/2-NSELA2015-Tools-for-NGSS-aligned-Unit-Development.pdf [May 2016]. See http://www.inquiryhub.org [May 2016] for more information about Inquiry Hub.

understanding provide many entry points for students as well as opportunities for educators to elicit and connect with their experiences. Students learning this way can demonstrate their understanding in a variety of ways, and teachers working this way can take advantage of phenomena that capture their students’ imaginations.

Key Idea

Teachers who are teaching in a three-dimensional way will structure student-centered instruction that

weaves together a wide variety of science practices with learning about important crosscutting concepts and disciplinary core ideas;

is flexible, allowing students to explore as they pursue learning objectives;

works cumulatively, helping students develop their understanding over time, and provides continuous support at all stages of the learning process;

engages students in investigating phenomena from everyday life;

recognizes that learning requires repeated engagement with important ideas, guidance, and opportunities for reflection; and

provides all students with avenues to science learning.

A NEW WAY TO THINK ABOUT ASSESSMENT

What exactly do all these changes mean for assessment? All assessments are intended to support teaching and learning by collecting information about what students know and can do. As teachers move to teaching in a three-dimensional way, focusing on learning over time, they will need to collect different kinds of information.

To measure three-dimensional learning that develops over time, assessments need to:

• examine how students use science and engineering practices in the context of crosscutting concepts and disciplinary core ideas;

• use a variety of tasks and challenges to give students multiple opportunities and ways to demonstrate what they have learned;
• provide diverse and specific information that shows teachers where students are struggling in their learning and helps them decide on next steps; it also helps students understand the progress they have made and where they need to go next; and
• focus on students’ progress along a learning pathway rather than what is correct or incorrect at a particular time.

Teachers will want to use assessments that accomplish these goals to support their own teaching, but it will be just as important that other kinds of science assessments be developed with the same goals in mind. Ideally, next-generation classroom assessments will need to be part of assessment systems in which all the components are also designed to measure three-dimensional learning that develops over time. This will not happen right away, of course, and teachers are sometimes caught in the middle as large changes such as these develop. Regardless of how quickly districts and states develop such assessment systems, the change will begin with classroom assessment, and there is much that an individual educator can do.

Assessment in the Classroom

Performance expectations that reflect the 2012 framework, whether for curriculum units or for grade levels, are not descriptions of isolated concepts or skills students are expected to master by a particular point in time. Instead, they describe how students are expected to make connections among practices, crosscutting concepts, and disciplinary core ideas and to show how they have built on the understanding they developed in an earlier unit or grade. This means that assessments should not focus on isolated ideas or skills either. They should provide evidence of students’ developing capabilities and insight into their partially correct or incomplete understanding—they should help teachers “see” their students’ learning. Assessments that accomplish this will have multiple components and be constructed so they can provide information about this kind of multidimensional learning.

Without a doubt, this is a big change, but the three-dimensional assessment you do in your classroom can be an important contribution to the process. What you do in the classroom starts with the performance expectations for the grade and the subject you are teaching. Whether your school is using the NGSS or other standards that support three-dimensional science learning, you have learning goals

for the year as well as more specific objectives. These objectives also drive when and how you assess your students’ learning. You most likely already use many kinds of assessments to check on your students’ progress because doing so is essential to good instruction.

As you adapt your instruction to the new vision of science learning, you will be finding many more ways to engage your students in doing science for themselves. The second-grade students who were investigating wet corn in the case at the beginning of the chapter were engaging in science practices from the start of their discussions. They raised scientific questions based on their observations. They collected evidence and attempted to explain the evidence, and they engaged in argumentation to compare competing ideas and reach consensus. This class investigation demonstrates how different types of activities can provide natural opportunities for assessment. As you gain experience with new approaches to assessment, you will be able to use classroom activities to collect new kinds of information about what your students are learning and to use new strategies to gain insight from your assessments.

For example, when your students are developing and using models, you can observe how they explain and discuss them with classmates. Their discussion can give you a window into their thinking and help you make instructional decisions. To assess your students’ progress toward learning objectives, you may use material they produce as a part of such classroom activities, such as lab reports or data displays, or choices they make using a computer-based activity rather than tests or quizzes.

These and other activities are all ways students can show you what they understand and give you clues about where they need help while they are actively engaged in doing science. You may not necessarily have thought of all these sorts of activities as assessment opportunities, but the information they provide can become an integral part of your instruction.

Assessment as a System

The assessments you use in the classroom—whether they are tests, quizzes, or other activities, and whether you developed them or are using assessments that are part of your instructional materials—are closely linked to the instructional activities you are doing. Your students most likely also take tests developed by your district or state that address the content you are covering but may not reflect the specific content you were teaching before they were given.

Ideally, the assessments you use in your classroom will be reinforced by the other kinds of assessments used in your district and state. While this would be the hope with all assessments, districts and states can develop systems for science assessment that are purposefully coordinated. In this kind of system⁸:

• all assessments are truly linked to instruction and curriculum so that each one is measuring—and reinforcing—what and how students are really taught; and
• all assessments work together—so that regardless of their purpose they reflect the same vision of how students learn science.

The idea of an assessment system begins with a commonsense point: no one assessment—or assessment occasion—can meet all the needs for information about what students know and can do in science. The purpose of collecting assessment information is to use it in some way to benefit students, but the different people involved in science education will use information in different ways. For example, parents need to have an accurate understanding of what teachers are doing in the classroom, and of what students are expected to do to graduate, so they can provide effective support for their children. Administrators use assessment data to evaluate programs and monitor how the students in their school are progressing. Policy makers use it to monitor larger groups, and the public uses it to hold the school system accountable for the education it is providing.

Even though the uses are different, the questions teachers who are integrating the three dimensions of science learning will ask about their students’ progress are the same ones that can provide a meaningful picture of how well the students are mastering the standards. This is exactly the information that policy makers who monitor learning on a larger scale want, even though they need to look across much larger numbers of students. They will also want to know whether students have equitable opportunities to learn and do science—and here, also, the answers are found in the classroom.

___________________

⁸ For a detailed discussion of systems for state science assessment, see National Research Council (2006).

This approach is different from the way most states have traditionally assessed science learning, but many states were thinking about taking a systems approach to their science assessments even before the 2012 framework for science learning was put forward. In the past, the goal has been to align instruction, curriculum, and assessment with standards. That’s still just as important, but in a systems approach, the focus is not only on aligning each piece with a central set of content standards but also on the ways each part of the system contributes to the others. In an assessment system, different types of information are collected throughout the year using a variety of assessment tools, but each type of information contributes to a bigger picture of student learning.

The results of each type of assessment will complement the results of others. This doesn’t mean that all assessments will be explicitly linked, however; it does mean that each assessment is designed to measure learning of the three-dimensional performance objectives being taught in the classroom. Different assessments—components of the system—will provide different kinds of information that can be used for different purposes.

It is important to be clear that this will be a complicated transition for states and districts. Developing and implementing such a system poses practical and political challenges that will take time to solve. Most districts and states are likely to make these changes gradually, and educators will need to adapt as materials, external tests (those developed outside the classroom and the school), and other elements evolve. Changing large-scale accountability tests may be the most challenging piece of the puzzle, but teachers can proceed even while system-wide changes are evolving. Chapter 5 talks further about the parts of an assessment system and the role individual teachers can play in fostering these changes. Here we focus on the specific changes in assessments that enable them to measure the development of three-dimensional science learning.

What Will Be Different?

Adaptation to three-dimensional science learning will mean changes in every type of assessment, guided by a few key ideas:.

Assessment Is Grounded in the Classroom

Three-dimensional assessment for any purpose has to be grounded in what takes place in the classroom: that is, in the curriculum and the way it is taught. The classroom assessments you use are not an intrusion into what you do; they are an integral part of your teaching because you want information about what your

students know and can do. The information you collect this way is essentially the same sort of information any science assessment should collect: evidence of what and how well students are learning and understanding. As the three-dimensional approach gets worked into teachers’ practice, curriculum materials, instructional objectives, and professional development, assessments used for any purpose will need to capture this sort of learning if they are to provide meaningful information.

Assessments Should Be Linked to Everything Else

Instruction, curriculum, and assessment must be tightly linked—coherent—if they are to successfully address learning that builds over time. This sort of linkage has always been important, but aligning each element to a set of written standards is not enough. The three dimensions of science learning—practices, crosscutting concepts, and disciplinary core ideas—need to play a part in everything teachers and students do in science class. And they need to be woven together in a way that builds cumulatively.

Ideally, science education is a seamless system made up of several smaller systems. The systems that guide instruction, curriculum, and assessment will evolve together toward the new vision of learning, though in the real world this is not likely to happen in a seamless way. If you are ready to teach in the multidimensional way, but your school and district are still using traditional curriculum and materials, you will need to adapt gradually. As systems take shape, though, it will be critical to remember that if you give students who’ve been exposed to three-dimensional instruction a traditional assessment it won’t provide you with strong information about their three-dimensional learning. When the systems are all grounded in the same model of how students’ learning will develop across the years of schooling, they will all reinforce one another.

Assessments Should Work Together

Teachers want immediate information about how their own students are doing today. Administrators and district and state staff need information about larger groups of students and often across longer time periods. Each of these three parties needs the same kind of information: what and how well students have learned in the classroom. Just as science education should be a seamless system, so should science assessment. If all assessments are designed with the same vision of how students learn science in mind, then the information they collect will fit together—and can be used more effectively. The information from large-scale assessments will reflect what students have been doing in their classrooms and provide information that teachers can use to make instructional decisions.

And, information collected in the classroom itself can be standardized and used in new ways by administrators and district and state leaders.

In an assessment system designed to collect information about three-dimensional science learning, a range of types of assessments, given at different times and for different purposes, provides a corresponding range of information about students’ learning and thinking. These varied, but related, assessments give the students themselves, as well as teachers, parents, administrators, and others, the information they need about the progress of science learning. They also contribute to a more detailed portrait of student learning than any one type of assessment by itself could provide. As you adapt your use of assessments, you will contribute to the building of that kind of system in your district and state.

In other words, large-scale assessments can provide information that is useful beyond the classroom but still measure the learning that takes place within it. These sorts of assessments may be designed to provide evidence about programs or individual students, but they can be designed to measure the learning in the same way classroom assessments are. Large-scale assessments, particularly the yearly tests used by districts and states, play a key role in shaping both expectations for student learning and public discussion and perceptions of science education. Therefore, it is critical that these tests be adapted along with instruction.

These changes are just beginning. Most of the science assessments that districts and states have been using were developed before the new vision for science learning—especially three-dimensional learning—were put forward. The idea of learning progressions has become well known, but it has rarely been used in the development of science assessments, especially the large-scale ones given to all students in a state. Even though many testing programs have included innovative ways to assess practices and complex concepts, they were not designed with three-dimensional learning in mind. States are just beginning to respond to these ideas, but California’s CA-NGSS Summative Assessment Plan is one example of a large-scale assessment program that is based on the 2012 science framework.⁹

Opportunity to Learn Is Key

Ideally instruction, curriculum, materials, and assessment would all be adapted at the same time, but in the real world that may not happen. As you adapt your own assessment practices, it will be important to think about the opportunities

___________________

⁹ For more information about this assessment program, see http://www.classroomscience.org/cde-reveals-and-state-board-approves-californias-ngss-summative-assessment-design-plan [April 2016] and http://www.cde.ca.gov/be/ag/ag/yr16/documents/mar16item02slides.pdf [April 2016]. The New York state grade 8 intermediate assessments are another example; see http://www.nysedregents.org/grade8/science/home.html [May 2016].

and resources your students have access to. Information about what students have had the opportunity to learn helps you—and other people in the system—do your best to give students equitable opportunities to learn science. But it also provides a check on what conclusions can be drawn from any assessment results. You know, for instance, that quizzing a student on material from a lesson he or she missed doesn’t give you useful information.

Opportunity to learn doesn’t just mean that the materials and curriculum were available; it means that all of the students had genuine access to the kind of instruction and learning the new standards envision. It is not fair to judge a student, a teacher, or a school for poor test performance if the students did not have true access to the learning described in the standards being assessed. Many factors could limit students’ access to learning: Some are practical, such as a lack of resources, teachers who have not received necessary training and professional development, or insufficient time allotted for science learning. Other factors are more subtle, such as unrecognized assumptions about what will be familiar to students, or a reliance on communication styles that students are not comfortable with, which might impede their learning.

Teachers know their students best. Your students likely bring a mix of prior educational successes as well as challenges. Each one brings a cultural background and experiences that could influence their responses and contributions to the activities in your classroom. These experiences can be resources for learning. As you get to know a new group of students, you pay attention to students whose first language is not English to see whether their language skills will affect their understanding. You consider whether prior educational experiences, disabilities, or other factors will influence any of your students’ learning.

Assessing students’ opportunity to learn is also especially important while districts and states are gradually adapting their science education systems. This is because the jurisdictions cannot be sure how successfully a program or curriculum is performing without knowing about students’ real access to the new approaches. District and state assessment and accountability systems should include ways to assess opportunity to learn, including documentation of instructional materials and instructional time, questionnaires for students and teachers, and classroom observations. But teachers are in the best position to report on changes they are making in their own classrooms because they know firsthand the experiences their students are having.

Key Idea

Assessments that measure three-dimensional science learning are integral to instruction because they allow the teacher to see how students are progressing. As districts and states adapt the NGSS-based approach to science education, all science assessments will need to be designed to measure three-dimensional learning that develops over time, and to work together as a system. Each kind of assessment will complement the others to build a composite picture of students’ science learning—the system will also track students’ access to three-dimensional science learning.

BUILDING ON ASSESSMENT BASICS—A QUICK PRIMER

These principles for designing new kinds of assessments build on good assessment practice: that is, the ground rules of psychometrics established over decades still apply. Each of the examples in this book illustrates a way to answer new kinds of questions about students’ learning while still making sure the results support valid

and reliable inferences about what students know and can do. In an assessment context, validity and reliability have very specific meanings¹⁰:

• People often describe tests as valid or not valid, but it is the interpretation of a test’s results that must be shown to be valid. The process of validation involves collecting related evidence to support each particular interpretation or use of a test. For example, districts use scores on end-of-year subject matter tests (e.g., large-scale accountability tests) to report the percentage of students who are “proficient.” Validity evidence to support this interpretation should include evidence that the test scores relate to other measures of achievement from that same school year, such as performance on end-of-course exams, or course grades.
• Reliability refers to the degree to which scores are consistent when a test is given at different times or under different conditions. For example, a reliability measure might indicate the extent to which student performance on two different administrations of a test will produce similar scores or the extent to which ratings of student work given by different judges are consistent. In other words, a rating of test reliability is an estimate of the extent to which the test scores are precise, free of random measurement error, and reproducible.

Assessment of any sort is a way of gathering and evaluating information, and in that way, it is like a scientific investigation. As in a science investigation, you have to identify the precise questions you want to answer, design a way to collect data to answer these questions, and design a structure for interpreting your results. To be useful, the individual tasks need to generate the information you want and need—not some other, irrelevant information. They need to generate results that you can interpret accurately and fairly, and you need to have confidence that they can reliably answer your questions even when the assessment is given on different days with different groups of students.

Here are some basic considerations to keep in mind as you adapt your approach to assessment. Our focus is on assessments you use in the classroom, but these principles apply to any type of assessment.

___________________

¹⁰ Professional standards for the ethical and technical requirements of tests, including reliability and validity, are described in the Standards for Educational and Psychological Testing (2014). The National Council for Measurement in Education provides an online glossary of important testing-related terms at http://www.ncme.org/ncme/NCME/Resource_Center/Glossary/NCME/Resource_Center/Glossary1.aspx?hkey=4bb87415-44dc-4088-9ed9-e8515326a061 [October 2016].

The Purpose for Assessing Drives the Design

Science assessment systems are needed because people want information about student learning for many different reasons. Some assessments—such as statewide tests usually given once per year—are used to monitor large groups of students. Chapter 5 talks more about this type of assessment, but there are other reasons to assess in the classroom as well. The terms formative and summative assessment are commonly used to describe two primary purposes for assessing in the classroom. These terms refer to the way the results of an assessment are used—not to its design or characteristics.

Teachers use formative assessments to collect information they need to guide their instruction and that students need to improve their learning. A formative assessment might be as simple as a quick comprehension check, a pop quiz, a conversation with a student, or a classroom discussion. These kinds of assessments help teachers measure their students’ progress and figure out what steps are needed to support them. They also help students understand explicitly what the criteria are for high-quality work and what they need to accomplish to meet expectations.

So, formative assessments might be used to:

• check on individual students’ understanding of what you are teaching,
• get insight into students’ thinking about science concepts and identify misunderstandings,
• make decisions about reteaching material or pacing the instruction, or
• help students evaluate and revise their own work.

Summative assessments are used at the end of a unit or a course to provide evidence of learning that can be used to make decisions such as assigning grades. They may be tests and exams designed by teachers, the results of which are used by them and their students. The large-scale assessments given by the district or state are also summative. Parents, school administrators, district or state officials, or others responsible for making sure that students progress as they should, use the results of these kinds of tests in making decisions.

Summative assessments provide results that can be used for purposes such as:

• assigning grades to individual students;
• informing parents about their children’s progress;

• determining whether an individual student is ready for the next grade, or has performed well enough to graduate from high school;
• providing evidence of how students within a group have performed with respect to general levels of mastery, perhaps identified through research on learning progressions—such as “basic,” “proficient,” or “advanced”;
• providing evidence of how well a group of students has performed in comparison with other groups—for example, fourth-grade students in a school or district as compared with those across their state; or
• providing evidence of how well a change—such as a revised curriculum, a new approach to professional development, or some other policy—has worked.

Just like formative assessments, summative ones should focus on three-dimensional learning and be closely linked both to the curriculum being taught and to specific performance expectations. What makes an assessment formative or summative is what you plan to do with the results. This means that planning for how results will be interpreted and how they will be communicated to the people who need them should be part of the design of the assessment from the beginning.

One Assessment Cannot Serve All Purposes

The purpose for which you need information should drive the design of the assessment you will use to collect that information. If you use an assessment for a purpose other than the one for which it was designed, the results won’t support valid answers to the questions you are asking about student learning. For example, a test that is designed at the district or state level to assess the understanding of a large group of students, such as all fourth graders, may not also provide information about the specific areas that an individual student needs to work on. Different people have different reasons for asking questions about student learning, and you also will have different purposes for testing in your classroom depending on what aspects of your students’ learning you are focused on. You may want a spot check of comprehension or a picture of how well students have learned the material in a complex multi-week unit.

Many of the examples in this book do serve multiple purposes because they are made up of multiple different tasks, each of which is designed for a specific purpose. A few of the examples show how you might design a task so that it can be used for a formative purpose while the students are learning and then used

again, perhaps in a modified form, when summative information is needed. But it is critical to think precisely about the purpose of each use of an assessment task.

The Assessment Should Measure What You Intend It to Measure

Imagine a task that is intended to measure students’ understanding of, say, the water cycle and asks students to write an extended answer. Students who are actually well able to do what the task is supposed to measure—for example marshaling evidence to explain something they observed—may not be able to demonstrate that ability on this assessment because of limited writing skills. There are a few ways to avoid this problem. You can be very precise about what it is you want to measure and how students can demonstrate that they have the understanding and knowledge you are interested in. You can carefully review an assessment task to see whether it poses unnecessary challenges that are not relevant to what you are measuring. You can also develop alternate ways for students to respond to the task and allow them some choice. For example, they might create a table or graph showing their evidence, or draw and label a picture.

It Is Critical to Be Sure Students Understand What They Are Being Asked to Do

For your students to demonstrate what they understand and can do, they need to be sure what the expectations are. If the assessment tasks you use are grounded in your instruction, and resemble the sorts of activities your students have been doing in class, it’s likely that your students will know what is expected. But it’s important to think carefully about the guidance your students will need both about what they are expected to do when they are being assessed and about the kinds of results they are working toward. Your students can show you what they have learned when the tasks you use to measure learning are examples of the same sorts of tasks you’ve used in teaching.

The Assessment Tasks and Context Should Be Consistent If Groups of Students Are to Be Compared

Developers of large-scale assessments are very careful about standardizing testing conditions (such as time allowed or the use of calculators) so that the results are comparable and fair across schools or districts. You also might want to compare results across students or groups of students within your classroom or across classrooms in your school. To do that fairly, you need to be sure that all

students are doing the same assessment task, even if they do it on different days. Likewise, you need to be sure that all of them have been given the same instructions. Many three-dimensional assessments will have students engaged in doing science—and won’t just rely on paper-and pencil tasks—but it is still possible to make these tasks consistent and even to standardize their administration for large-scale accountability purposes. You’ll want to ensure that the conditions are consistent and that the guidance, resources, and support the students have access to as they do the activity are consistent. For example, in an assessment in which your students will work with living organisms it is important that all the students encounter comparable challenges. If the organisms have not been exposed to the same conditions or are not at the same stage of development at the start of the activity, the results may be affected by factors that have nothing to do with how much the student knows or can do.

The Assessment Situation Should Give Every Student a Fair Opportunity to Demonstrate What He or She Has Learned

Teachers may be concerned that these new kinds of tasks are more challenging than traditional ones. If you are wondering whether your students will do well, remember that diverse kinds of evidence can show capacities that may have remained hidden with traditional assessments. You can give them multiple modes for expressing what they know and can do. As you adapt your instruction, your students will also adapt. Ideas for making instruction more inclusive and accessible will be especially applicable as you adapt your assessments.

CHAPTER HIGHLIGHTS

• Three-dimensional science standards—the NGSS and others—are based on the idea that in order to become science literate, students need continuous opportunities to do science. Engaging in science for themselves will teach students how and why science and engineering really matter and how scientists and engineers do what they do.
• Instruction that develops scientific thinking and learning will integrate three dimensions: (1) the practices through which scientists and engineers do their work; (2) the crosscutting concepts that apply across science disciplines; and (3) the core ideas of the disciplines. Coherent instruction and curriculum will allow students to build increasingly complex understanding across their years of schooling.
• Assessment is an integral aspect of this new kind of instruction because teachers need regular information about students’ developing capacity to integrate the three dimensions of science learning: the practices of science and engineering, crosscutting concepts, and disciplinary core ideas.
• Assessments can be designed to measure three-dimensional learning while still following established measurement principles. Assessments of this kind of learning will often look a great deal like classroom activities. Teachers can structure many kinds of science activities to collect information about their students’ learning as it develops.
• Three-dimensional assessments provide more information about students’ developing understanding than many traditional assessments can. This information will help guide your teaching and your students’ learning.
• These new kinds of assessments may be used for formative purposes (to collect information a teacher needs to guide instruction and help students improve their learning) or summative purposes (to provide evidence of learning after, an instructional unit, a course, or a grade is complete).

The specific examples we explore in the following chapters illustrate how this new approach to assessment works.