Benchmarks Online Table Of Contents  Search Benchmarks  Project 2061  Printer Friendly Version

15. The Research Base

  1. The Role of Research
  2. The Nature of the Research Literature
  3. Research Findings By Chapter and Section
  4. References

The task in preparing this report has been to create clear and specific benchmarks and place them at a reasonable grade level. To allow for differences among students, grade spans of 3 or 4 years are used rather than single years, although even then it is difficult to ascertain when students are able to learn particular concepts and skills effectively. Over-estimation of what students can learn at a given age results in student frustration, lack of confidence, and unproductive learning strategies, such as memorization without understanding. Underestimating what students can learn results in boredom, overconfidence, poor study habits, and a needlessly diluted education. So it is important to make decisions about what to expect of students and when on the basis of as much good information as possible.

The presence of a topic at a grade level in current textbooks or curriculum guides is not reliable evidence that it can be learned meaningfully at that grade. For example, atoms and molecules sometimes appear in a 4th-grade science reader. Yet extensive research on how children learn about these ideas suggests postponement until at least 6th grade and perhaps until 8th grade for most students.

As noted in Chapter 13: The Origin of Benchmarks, the benchmark decisions have been based on inputs from three sources: the Project 2061 school-district teams, learning specialists, and published research. Although each of those sources has its own particular limitations, each contributed significantly to the final product.

The single most important source of knowledge on student learning comes from thoughtful teachers. They have firsthand experience in helping students acquire science, mathematics, and technology knowledge and skills. Their input is limited, however, by the realities of the usual teaching situation. Teachers have little time to conduct careful assessments of student learning, lack instruments for assessing richly connected learnings and higher-order thinking skills, and rarely have opportunities to compare their experiences with others who teach the same concepts and skills. The Project 2061 teachers were able to overcome some of these limitations as a result of having ready access to learning specialists and time for exploring questions of benchmark placement with other teachers. Whenever benchmark-writing conferences were held with Project 2061 teams, several research consultants participated as well. They helped by elaborating on their knowledge of important topics on which they had already done research, extrapolating that knowledge to other topics and helping the teachers extend and interpret their own experience.

Researchers have the advantage of being able to work out a careful design, having time and other resources (including special training in research methods) that teachers seldom have, and undergoing systematic peer review. Research studies into whether students at a certain age can understand a certain idea fall into four categories:

  1. Students do understand the idea after traditional instruction—or even without any instruction at all. So placement at the given grade level is all right and might possibly be earlier.
  2. Students do not understand the idea after traditional instruction. (This is the most common category of published research.) So the idea may be intrinsically too sophisticated for the grade level or may not have been adequately prepared for. The research may provide clues about how to design more effective instruction.
  3. Students do understand the idea after special instruction. So placement at the given grade level is possible, depending on whether adequate additional resources are likely to be available.
  4. Students still do not understand even after special instruction. (This is fairly common.) The idea has to be simplified, better prepared for, or postponed until students are more ready.

As those categories imply, research findings have led to several different kinds of adjustments in writing benchmarks:

  1. Stating less-sophisticated precursors of an idea. For example, research suggests that the notion of a "fair comparison" can be understood in lower grades as a preliminary form of the later concept of a controlled experiment.
  2. Adding prerequisite components for learning outcomes. For example, research draws attention to the need for understanding how people see things by reflected light as a prerequisite to a benchmark for understanding the phases of the moon.
  3. Changing benchmarks to different grade levels. For example, research shows that natural selection is still a difficult idea for many college students—even after special instruction. So the benchmark for natural selection was moved from 8th grade (where some teachers thought it could be taught) to 12th grade.

But research, too, has its limitations. For decades, the chief tool for investigations of what children know was multiple-choice testing, but researchers eventually found such tests often inflate estimates of students' understanding and disguise their misunderstandings. Studies of students' general cognitive development over the decades have usually proven to be too general to be of much help in making decisions about particular concepts and grade placements. Nonetheless, developmental psychology does provide some guidance where there are no specific research findings. (For example, the late acquisition of the concept of proportionality implies that it is very unlikely that students could understand probability in a quantitative way before middle school.)

Beginning in the late 1970s, cognitive studies in many different countries began to focus on the learning of particular science and mathematics concepts, often using in-depth interviews of students in place of conventional testing. Interviews have revealed that students usually have ideas about how the world works even before instruction. Some of the students' ideas work fairly well in familiar contexts and are highly resistant to change. Moreover, students' lack of understanding or misunderstanding of ideas in science is often masked by their ability to memorize the right words. In-depth interviews with students have often shown consistent deficits or peculiarities in student understanding that had not been identified by conventional testing. Still, the total number of concepts or skills investigated is very small and is unevenly distributed across fields.

Evidence on learning from both teacher experience and research ought to be interpreted cautiously because it necessarily refers to today's students taught in today's schools by today's teachers. There are so many variables operating in the learning process—teacher and parent expectations, the learning environment, the methods and materials used, the previous knowledge and experience of individual learners, and more—that the failure of students to learn something currently leaves open the question of whether they could have done so if they had had ideal learning conditions from the beginning.

Some learning difficulties students have may not be modifiable at all until brain maturation allows them to use higher levels of abstract thinking, whereas some current learning difficulties might be readily ameliorated by improved resources such as better books, more hands-on work, or more time for teachers to plan.

So benchmark statements and placements, including estimates of what might reasonably be possible in the future, have been attempted with the evidence from various sources—teachers, general principles of developmental psychology, research on specific topics, and prominent researchers.

There is a vast literature on education, a fraction of which reports research findings. That research literature is spread across different fields of study, grade levels, kinds of schools, and aspects of education. Studies dealing with the learning of specific science, mathematics, or technology content, while small compared to the whole body of educational literature, are growing in number and sophistication.

Research progress in a field can be inferred from an examination of review articles, published bibliographies, and handbooks. They are also good starting places for beginning a literature search. With regard to the topics covered in this report, the following were found to be very helpful:

Elementary School Social Studies: Research as a Guide to Practice, edited by Virginia Atwood. Washington, DC: National Council for the Social Studies, 1986. 176 pp.

This book examines the research base on when and how elementary-school students develop concepts, skills, and attitudes associated with social sciences. Chapters discuss and summarize research findings related to topics in citizenship and law-related education, intercultural and multicultural education, geography, history, economics, anthropology, and sociology.

Children's Ideas in Science, edited by Rosalind Driver, Edith Guesne, and André Tiberghien. Milton Keynes, UK: Open University Press, 1985. 208 pp.

This book documents and explores conceptions that students aged between 10 and 16 hold about topics related to light, heat and temperature, force and motion, the structure of matter, and the earth as a cosmic body. It also examines how students' conceptions develop with teaching.

Handbook of Research on Mathematics Teaching and Learning, edited by Douglas Grouws. A project of the National Council of Teachers of Mathematics. New York: Macmillan Publishing Company, 1992. 771 pp.

The Handbook presents a comprehensive view and analysis of available research on mathematics teachers, teaching mathematics, and the learning process for mathematics content. Areas for further research needed are identified, and, where appropriate, implications of research for classroom practice are provided. Several chapters concern students' learning in content areas that have received considerable research attention: addition and subtraction, multiplication and division, rational numbers, estimation, geometry and spacial reasoning, and probability and statistics.

Bibliography. Students' Alternative Frameworks and Science Education, by Helga Pfundt and Reinders Duit. Kiel, Germany: Institute for Science Education at the University of Kiel, 1991. 270 pp.

The bibliography documents and categorizes research into students' conceptions in science. It contains about 2000 citations to journal articles, research reports, conference papers and whole books on students' learning in physics, biology, chemistry, and earth science. Entries are classified by type of issue. Issues include general considerations concerning research on students' conceptions, relations between students' conceptions and scientific conceptions; relations between the development of student conceptions and the development of notions in the history of science, relations between everyday language and students' conceptions, methods of investigation, investigations of students' conceptions, instruction taking students' conceptions into account, investigations of teachers' conceptions, and consequences of students' conceptions research on teacher training. Investigations of students' and teachers' conceptions are further classified by content area.

Handbook of Research on Social Studies Teaching and Learning, edited by James Shaver. A project of the National Council for the Social Studies. New York: Macmillan Publishing Company, 1991. 661 pp.

The Handbook presents a comprehensive view and analysis of available research on social-studies teachers, teaching social studies, and the learning process for social-studies content. Needed future research efforts are identified and methodological issues for research on social studies education are raised. Several chapters in the Handbook examine student characteristics and development as relevant to social studies. Research findings are summarized related to topics in government, civics, and law; multicultural education; history; economics; geography; and anthropology, sociology, and psychology.

In searching for papers relevant to our benchmarks, Project 2061 also turned to the research literature from other countries and to proceedings from international research conferences. The references below cite the work of researchers from 10 countries in addition to the United States. With few exceptions, English-language versions were found and cited. The bibliography currently contains citations to over 40 professional journals in several categories. Major journals carrying research papers in science, mathematics, and technology education include

There are also research articles relevant to learning concepts in science and mathematics in a variety of psychology journals including

There are a number of journals that occasionally carry research articles and provide helpful commentary on general principles of when and how students learn ideas, including The Science Teacher, Science and Children, The Physics Teacher, The American Biology Teacher, Mathematics Teacher, and The Technology Teacher. Also, education associations sometimes publish monographs on research, an example being the series "What Research Says to the Teacher," published by the National Science Teachers Association from 1978 (edited by Mary Budd Rowe) to the present (edited by Robert Yager). Generally informing Project 2061's interpretation of research findings was long familiarity with the theoretical work of educational psychologists such as Jerome Bruner, Robert Gagné, and David Ausubel and with the successes and failures of the national curriculum projects that were undertaken from the late 1950s to the early 1980s.

The references that follow are organized to match chapters and sections of Benchmarks, which in turn mostly match those of Science for All Americans. The list is very selective and includes only those references that met two criteria. One was relevance—some excellent papers were not included because they did not bear on one of the Benchmarks topics. The other criterion was quality—papers, however relevant, were bypassed if they were seen to have design flaws or their evidence or argument was weak. Even then, however, not all relevant and good papers are included. In many cases, a single paper has been used as representative of a number of similar reports.

It will immediately be clear that mathematics and the physical sciences have had the benefit of many more studies than have other fields. Perhaps that is because the subject matter lends itself to research more easily; in the next few years, though, perhaps the attention to cognitive research will increase in all fields.

  1. The Nature of Science
  2. The Nature of Mathematics
  3. The Nature of Technology
  4. The Physical Setting
  5. The Living Environment
  6. The Human Organism
  7. Human Society
  8. The Designed World
  9. The Mathematical World
  10. Historical Perspectives
  11. Common Themes
  12. Habits of Mind

Browse the reference list: A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All

Keywords can be the author last name, year in which a reference was published, or words from the title of the reference.

Copyright © 1993,2009 by American Association for the Advancement of Science