Research Matters - to the Science Teacher
No. 9704 November,
When are science projects learning
by Marcia C. Linn and Helen C. Clark, Graduate
School of Education, Education in Math, Science and Technology
University of California, Berkeley
Science projects have played a central role in schools at least
since turn of the century, when they were championed by John Dewey
(1901). How can we ensure that projects are efficient, effective
learning experiences that promote knowledge integration and lifelong
science learning? For answers we draw on more than a decade of
research by the Computer as Learning Partner project (Linn &
Songer, 1991; Linn, Songer, & Eylon, 1996), and specifically on
dissertation research by Clark (1996 May).
Why include science projects in the classroom? Science projects
can engage students in authentic science
experiences--essentially the work of experts. Projects can encourage
sustained reasoning, connect classroom to personal problems, make
science relevant to everyday life, and prepare students for lifelong
learning. Projects give students a window into the complexities and
uncertainties of science. Professional scientists engage in projects
in a supportive community of mentors, peers, and skilled technicians.
They benefit from shared methodologies, standards, and criteria for
success. They follow sanctioned critiquing practices in reviewing
each other's work and in participating in scientific meetings. How
can we make these supports a part of classroom science? Our research
in designing middle-school science projects in the Computer as
Learning Partner project results in four recommendations.
Recommendation 1: Start with small, accessible
First, start small. Experts-in-training (such as graduate students)
often replicate the work of others or apply established procedures
before designing their own projects. In the Computer as Learning
Partner eighth grade classroom, students start with projects that are
slightly more complicated than the most demanding class assignment.
These projects, nevertheless, require the sustained reasoning
necessary to link and connect ideas, reflect on progress, and
We found three types of projects that succeed most of the time. In
a critique project students evaluate an experiment or
conclusion reached by another student or an account of a scientific
result reported in the media. A design project engages
students in building a solution to a project such as designing a
house for the desert. An explain project asks students
to use science principles to account for an observation such as the
"dog dish" project in Figure 1, where students explain why the water
in the dog dish gets warmer than the water in the swimming pool.
These three types of projects allow students to generate several
different ideas about the scientific question, distinguish among
their ideas using evidence from class or other experiences, and draw
Projects succeed when students can connect class experiences to
their project questions (e.g., see Linn & Muilenburg, 1996). This
requires the alignment of class scientific principles with both class
projects and student prior knowledge. Instruction can then promote
knowledge integration across all problems.
Recommendation 2: Develop class criteria for
Second, help students develop shared standards and criteria for the
intellectual work of carrying out a project. We designed "composite"
projects based on several students' work and engaged the class in
critiquing them: encouraging students to identify improvements and
describe weak or contradictory links among ideas or between
explanations and evidence. Often students' intuitive criteria
reflected superficial, schoolish standards like neatness and grammar.
As a group, students developed criteria such as "back up assertions
with evidence from class experiments or personal observations," and
"identify confusing observations and seek additional information."
These group criteria were posted and used regularly in class and
Recommendation 3: Provide support and coaching
Third, provide support and coaching to encourage linking and
connecting of ideas and use of shared criteria. Experts carry out
projects with extensive support and guidance from mentors, peers, and
experts in other fields. They revise their plans based on this
support. Citizens need to learn how to locate dependable coaches or
experts, ways to make sense of the views of others, and strategies
for incorporating useful views into revisions of their projects. To
emulate this process in the science classroom, we used peer coaching,
trained graduate student coaches, and teacher coaches.
We found that coaching helps students learn to monitor their
progress and to make improvements to their projects. When students
revise their projects, they add connections using scientific ideas,
personal experience, and conclusions they have drawn. Fairly specific
coaching comments, such as "what happens to light when it hits
water?" or "describe a class experiment that supports your view" were
more effective than general encouragement to think about related
information, such as "what other variables might be influencing how
warm the water gets." Students tended to add links and resolve
inconsistencies in response to specific coaching comments.
In the Computer as Learning Partner classroom we synthesized the
best individual coaching comments into a cybercoaching system (Clark,
1996 May). Once a project had been coached, we identified frequent
student responses and we selected the most useful coaching comments
for the system. The "cybercoach" allows coaches to match frequent
student responses with appropriate coaching statements, and to send a
message to the student. Cybercoaching took 90% less time and was
almost as effective as individual coaching.
We also designed prompts based on coaching experience. These prompts
raise issues, ask students to analyze their own work, and encourage
them to reflect on their own progress. Using our Computer as Learning
Partner software, students could access these hints or prompts for
some projects. Examples of the kinds of prompts that have proven
successful in our research include, "what do you need to know to
carry out this project?," "what is still confusing about your
results?," and "connect your conclusions to a class experiment."
Recommendation 4: Make project assessment part of
Fourth, make project assessment both efficient and a part of the
learning process. To help students develop shared criteria for
arguments and learn to critique the work of others, we structure oral
project presentations. We require each student or group to
prepare a project report, and to write a
question in response to each project presentation. We randomly select
groups to present their projects and individuals to ask questions so
that each student participates at least once in the project
discourse. In addition, we grade the written questions and written
reports using a straightforward holistic system that rewards
knowledge integration (see Figure 2).
We also assess student learning from projects through written,
in-class tests of knowledge integration. In these tests, students
critique, design, or explain a novel event and give the main reasons
for their choice (see example in Figure 3). We score responses using
the same holistic criteria found in Figure 2.
Figure 1: Example of a CLP project
On a hot summer day Shawn's little sister notices that the water
in the dogs' dish, which is sitting in the sun, feels fairly hot but
the water in the swimming pool is still very cool.
If you were Shawn what would you say to your little sister to help
her understand her observations?
Dog Dish versus Swimming
Figure 2: Holistic scoring for projects and
classroom knowledge integration test items
- No science principle mentioned--descriptive only (bowl is
- Mentions principle but inaccurate or incomplete (small things
- Accurately restates principle without elaboration or
connections (if same heat is added than smaller object reaches
- Clear and accurate understanding of single principle and adds
elaboration and or context. (e.g., if same heat added to two
objects then the smaller object has less space, so heat is more
dense like in the lab where we heated the small and large
beaker... so it reaches a higher temperature).
- Clear and accurate understanding of principle and also ties in
one or more additional principles from the same or related topic
area. (e.g., the light from the sun hits the water and changes to
heat energy which warms both the bowl and the pool, but since the
pool has more water and surface area it doesn't reach as warm a
Figure 3: Classroom knowledge integration
It is a hot summer day and Mac has invited some friends over. Mac
takes two identical pitchers of lemonade out of the refrigerator and
puts one on the counter in the 20 C air-conditioned kitchen and one
on the picnic table outside on the covered porch where the
temperature is 40 C.
a. Which lemonade will warm at a faster rate
_____The lemonade on the kitchen counter
_____The lemonade on the picnic table
_____Both lemonades will warm at the same rate
b. Fill in the blank to make a principle that applies to these
Heat energy flows __________________________when the temperature
faster / slower / at the same rate difference between an object and
its environment is greater.
c. Give the main reasons for your answer.
In general, projects give students a chance to be creative in
science class. Most students become engaged and carry their projects
to completion, providing authentic examples of their thinking for
teachers. The satisfaction of finishing a project is sufficient
reward for some. Since students vary in resources for completing
projects, we place more course evaluation emphasis on knowledge
integration tests, oral presentations, and written questions than on
the completed project.
In conclusion, classroom projects can prepare students to carry
out future personally-relevant science projects,. Projects succeed
when they build on what students know, starting small. Furthermore,
projects are most successful when students have developed shared
criteria for scientific arguments that they can apply to their own
and others' work. In addition, instruction that includes coaching to
stimulate reflection and revision results in more sophisticated
projects. Finally, instructors can best evaluate students using
projects and multiple forms of assessment. Under these circumstances,
projects can engage students in sustained scientific thinking,
prepare them to seek and use feedback from peers or experts, and help
them systematically analyze experiments and claims they encounter in
Clark, H. C. (1996 May). Design of Performance Based
Assessments as Contributors to Student Knowledge Integration .
[Unpublished dissertation]. University of California at
Berkeley, Berkeley, CA.
Dewey, J. (1901). Psychology and social practice,
(Contributions to education). Chicago, IL: University of Chicago
Linn, M. C., & Muilenburg, L. (1996). Creating lifelong
science learners: What models form a firm foundation? Educational
Researcher, 25 (5), 18-24.
Linn, M. C., & Songer, N. B. (1991). Teaching thermodynamics
to middle school students: What are appropriate cognitive demands?
Journal of Research in Science Teaching, 28 (10),
Linn, M. C., Songer, N. B., & Eylon, B. S. (1996). Shifts and
convergences in science learning and instruction. In R. Calfee &
D. Berliner (Ed.), Handbook of educational psychology (pp.
438-490). Riverside, NJ: Macmillan.
This material is based upon research supported by the National
Science Foundation under grants MDR-8954753, MDR-9155744 and
MDR-9453861. Any opinions, findings, conclusions or recommendations
expressed in this publication are those of the author and do not
necessarily reflect the views of the National Science Foundation.
The authors give special thanks to members of the Knowledge
Integration Environment project and the Computer as Learning Partner
project. We appreciate the contributions of the other group members
including Steve Adams, Ben Berman, Julia Claeys, Doug Clark, Alex
Cuthbert, Jeff Morrow, LaShunda Prescott, Linda Shear, Jim Slotta,
Bridgette Sparks, Eunice Yi, Doug Kirkpatrick, Eileen Lewis, Nancy
Songer, Jacquie Madhok, Philip Bell, Helen Clark, Betsy Davis, Brian
Foley, Oliver Grillmeyer, Chris Hoadley, Sherry Hsi, Lawrence
Muilenburg, Staci Richard, Jim Slotta, Erika Whitney, and Judith
Thanks also to Dawn Davidson, Liana Seneriches, Erica Peck, and Mio
Sekine for assistance with the production of this manuscript.
Research Matters - to the Science
is a publication of the National Association
for Research in Science Teaching