Transforming Education:

1
Transforming Education using the tools of
science to teach science
and most other subjects
Carl Wieman, Univ. of Colorado
I) Data on effectiveness of traditional science
teaching. II) A better approach. Specific
examples. (based on research, CU tested)
(CU physics chem ed. research, W. Adams, K.
Perkins, Kara Gray, Linda Koch, Jack Barbera,
Sarah McKagan, N. Finkelstein, Steve Pollock,...
NSF, Kavli)
2
Science education more important, different
purpose than in the past.
Not just for scientists

• Survival of world.
• Wise decisions by citizenry on global (technical)
  issues.
• Workforce in High-Tech
• Economy.

Need to make science education effective and
relevant for large fraction of population!
3
Essence of an "effective education".
Transform how think about science-- novice
attitudes and problem solving into expert.
Think about science like a scientist.
4
II. Some data on effectiveness of traditional
approach to science teaching. lecture, textbook
homework problems, exams (most data from physics
but applies to other sciences, etc.) 1.
Retention of information from lecture. 2.
Conceptual understanding. 3. Beliefs about
science.
5
1. Lecturing and retention

• Explain about sound violin.
• Show class a violin
• Tell them that the strings cannot move enough
  air
• Point inside violin to show a sound post
• Tell them strings causes back of violin to move
  and back is what makes the sound

• 15 minutes later in the lecture
• Question to Class The sound you hear from a
  violin is produced by
• mostly by strings, b) mostly by wood in back,
• c) both equally, d) none of the above.
• What fraction gave the correct answer?
• a. 0, b. 10 , c. 30, d. 50, e. 80

6
"Sound you hear from a violin is produced " a.
mostly by strings, b. mostly by wood in back, c.
both equally, d. none of above.
84
ans. B. (students had been told 15 minutes
earlier)
very typical for nonobvious fact (even with profs
and grad students)
responses ()
10
3
3
0
A B C D E
later in talk-- how to get gt90 after 2 days
7
Redish- interviewed students as came out of
lecture. "What was the lecture about?"
unable to say anything but vaguest generalities
Rebello and Zollman- had 18 students answer
six questions, then told them to get answers to
these 6 questions from following 14 minute
lecture. (Commercial video, highly prepared and
polished, "world's most wonderful physics
lecturer")
Most questions, less than one student was able to
get answer to questions from listening to
lecture.
8
2. Conceptual understanding. How well are
physics concepts mastered by students who
complete traditional intro physics course?

• Force Concept Inventory- conceptual content
  survey on
• basic concepts of force and motion

• traditional approach not effective for developing
  conceptual understanding.
• Lecturer quality, class size, institution,...doesn
  't matter!

R. Hake, A six-thousand-student survey AJP
66, 64-74 (98).
9
Conceptual understanding (cont).
Eric Mazur (Paired problems)
10
3. Beliefs about science and problem solving
(measured)
Expert
Novice
Content isolated pieces of information to be
memorized. Handed down by an authority.
Unrelated to world. Problem solving pattern
matching to memorized arcane recipes. (boring,
useless, "antiscience")
Content coherent structure of
concepts. Describes nature, established by
experiment. Prob. Solving Systematic
concept-based strategies. Widely applicable.
nearly all physics courses ? more novice ref.
Redish et al, CU work--Adams, Perkins, MD, NF,
SP, CW
adapted from D. Hammer
11

• Science Education Research Conclusions
• Hard to know what students actually are (and are
  not) learning.
• Most students "learning" rote memorization of
  facts and problem solving recipes, not
  understanding. Useful only to pass class.
• also learning science is uninteresting and
  irrelevant

12
How to improve this situation?
Use tools of science to teach science!

• Practices and principles based on research and
  data, not tradition.
• Effective use of technology.
• Disseminate and build upon proven methods.
• like science research, copy what works!

13
A few illustrative specifics
I. Use of research on how people learn.
explain results shown, guide how to do better
a. cognitive load b. importance of attitudes
and beliefs c. developing expert competence
II. Effective use of technology a. student
personal response systems b. interactive
simulations
most applies to all subjects
work from Col. sci. ed. research group
14
examples-- using research on how people learn
a. Cognitive load-- best established, most
ignored.
7 ?2 items max. short term working memory.
Implication for teaching Teacher can present
MUCH more material than students can process.
15
b. Importance of student beliefs about physics
(science) and science problem solving
We developed and tested new beliefs
survey. online at CLASS.colorado.edu

• lt10 minutes, Give online pre- and post-
  instruction (gt5000 stds)
• Score agree ( favorable) or disagree with expert
  view

• Piles of Data!
• Beliefs ?? content learning
• Beliefs ?? choice of major/retention
• Teaching practices ? students beliefs

16
Beliefs and choice of major/retention
Men 62 Women 41

• Large gender gap in Personal Interest/relevance

• "Personal Interest" correlates with choice of
  major

Want more students, including women and
minorities, to go into science? Improve their
beliefs ("personal interest")!

Possible with teaching??
17
Teaching practices and beliefs
Beliefs (favorable)
Overall
Post
Pre
56
57
56
58
60
51
64
66
64
58

• Decline in beliefs all intro phys courses (also
  in chem.)
• Even with interactive engagement, good
  conceptual gains.

BUT can avoid decline if explicitly address
beliefs.
Why is this worth learning? Connection to real
world. Why does this make sense?
18
c. Research on developing expert competence
c. Expert competence fact. knowledge
organizational structure
? effective retrieval and application of facts
Can't just pour facts into passive student.

• Actively construct new way of thinking.
• Built on prior thinking
• Organize and use those facts.

or
Make them think, guide that thinking!
19
Some technology that can help. ( when used
properly)

• Personal electronic response systems--facilitate
  active thinking and useful guidance. Relatively
  cheap.

"Jane Doe picked B"
individual
20
"Sound you hear from a violin is produced " a.
mostly by strings, b. mostly by wood in back, c.
both equally, d. none of above.
84
ans. B. (students had been told 15 minutes
earlier)
responses ()
10
3
3
0
A B C D E
responses
21
clickers- technology not automatically
helpful Only require students to commit to an
answer (accountability peer anonymity)
When used properly transform classroom. Dramatical
ly improved engagement, thinking,
discourse, number and distribution of questions.
The KEY to clicker effectiveness use guided by
how people learn

• Questions and follow-up
• Focus students on processing ideas, organize and
  apply
• Communication and feedback (student-instructor,
  student-student) 3 student consensus groups,
  listen in on
• discussions, as well as see histogram. Respond
  accordingly
• Reflection (why answer, why not other answers)

22
Interactive simulations
phet.colorado.edu
Physics Education Technology Project (PhET) Wide
range of physics topics and some chem., well
tested, free online or download. Run in regular
web-browser. Use in lecture, lab, homework.
(often better than reality!)
supported by Kavli Operating Inst., NSF, Univ.
of Col., and A. Nobel
23

• Summary
• Need new, more effective approach to science ed.
• Traditional lecture, textbooks, homework, exams
  often teach against true understanding and
  interest in science.
• Solution Approach teaching as a science
• Guided by good data/research, use technology.
• Copy what proven to work.

Good Refs. NAS Press How people learn , "How
students learn" Redish, Teaching Physics
(Phys. Ed. Res.) Mayer, Learning and
Instruction (cog. sci. applied) CLASS belief
survey http//cosmos.colorado.edu/phet/survey/CL
ASS/ phet simulations phet.colorado.edu
24
Nobel prize in physics expert in sci. ed.???
CEW- large science education research
group, Chair--new Board on Science Education
National Academy of Sciences
How Students Learn, eds Donovan and BransfordNAS
Press 20053 well-established principles of
effective teaching. 1. Engage prior
understandings/thinking. 2. Essential role of
both factual knowledge and
conceptual frameworks in understanding. 3.
Importance of student self monitoring and being
mentally active.
Tested case studies implementing in variety of
subjects (history, math, science etc) various
grade levels
25

• For effective science teaching need
• Know subject
• Know student thinking
• Know how students learn subject

26
Interactive simulations
Physics Education Technology Project (PhET) Wide
range of physics topics and some chem., well
tested, free online or download
Carl Wieman Wendy Adams Noah Finkelstein Krista
Beck Ron LeMaster Kathy Perkins Sam Reid Mike
Dubson Noah Podolefsky Linda Frueh
moving man wave on string freq 33, tn max,
damping .01 circuit const.
phet.colorado.edu
supported by Kavli Operating Inst., NSF, Univ.
of Col., and A. Nobel
27
Summary

• research based teaching can achieve much
  better learning.
• Carefully decide what students should learn.
• Make objective measurements of starting point and
  results.
• Conclusions and guiding principles based on data.
• Change practices to improve results.
• Save and build on past materials and results.

• Good References
• How People Learn Brain, mind, experience, and
  School, NAS press
• Learning and Instruction, Mayer (educ.
  psych.-cog. sci.)
• Learning and Understanding, How Students Learn,
  both NAS press
• Teaching Physics with the Physics Suite, Redish

physics simulations--phet.colorado.edu beliefs
survey--CLASS.colorado.edu
28
2. Research on sim as replacement. premed
class ½ did real DC circuits lab, half CCK sim.
Relevant final exam questions 2 months later
Time to build and evaluate real circuit

Mean score on Q1-Q3 CCK 0.59 , TRAD
0.48 Statistically different, plt0.001
29

• CW results- using clickers heavily. Transform
  classroom.
• Much higher retention of information ( gt90 after
  2 days).
• Much better critical thinking and scientific
  discussion.
• More questions (4? 15/class), broader
  distribution.

Student opinion
60
Usefulness of lecture to your learning?
double attendance
clickers and consensus groups
50
40
traditional lecture
30
20
textbook
10
0
great deal
fair amount
some
a little
none
30
Bunch of research on design and effectiveness of
sims. (Wendy Adams, Kathy Perkins, Noah F., et
al) 1. Substantial improvement on concept
questions when used in lecture vs real demos or
static images.
standing wave on string
100
sim demo
right
comparisons of sims vs. static images with
explanation, also big gains
real demo
Q1 Q2
31
How Students Learn, eds Donovan and BransfordNAS
Press 20053 fundamental and well-established
guiding principles of effective teaching. 1.
Engaging prior understandings. 2. The essential
role of factual knowledge and
conceptual frameworks in understanding. 3. The
importance of self monitoring.
Tested case studies on implementation in math
science various grade levels
32
But what about content coverage, don't you have
to give up a lot? Some-- typically perhaps about
1/3-1/4 of material, but if students are not
learning it, what is the point of teaching
it? also-- if students just memorizing facts and
recipes, do about as well with less time spent on
it than usual. Can also have them cover some
material without going over it in class.
33

• Student perspective on talk.
• How does this match my experience?
• Questions for your teachers
• Why should I learn this?
• Why are you teaching like this?
• ("Wieman told me to ask.")

34
Beliefs and choice of major/retention

• Personal Interest correlates with choice of major
• possibly, the most important factor

35
(FCI test) Ref. R. Haake
1 semester intro. physics
48 classes (various approaches to get students
actively thinking more)
14 classes traditional lecture
Fraction of unknown basic concepts learned
traditional lectures and homework just dont
work for conceptual understanding. Lecturer
quality, class size, institution, ... doesn't
matter!)
36
Beliefs and learning

• Calc-based Phys I, Sp03 416 students
• Content Learning FMCE ? normalized learning gain

Pre
Post
37
If lectures are so bad, why are you giving us
a lecture?
Lectures are not inherently bad. It is an issue
of what happens in the lecture. passive
listening vs cognitively engaged Does this
make sense? How is this related to
my experience. Could I use this? How?
Requires suitable match to audience background
and preparation.
Also, lecture and slides incorporate many
features based on cognitive research.
38
Ways in which we use types of questions. Build
class around clicker questions. 1. Start of
class-- 3 question quizzes on reading. 2. Quick
surveys on backgrounds, course issues, 3.
Students predict results for all
demonstrations. 4. Check understanding of
material covered. 5. Reveal prevailing
misconception to confront/get attention leading
into coverage of material.
assigned seats and groups, consensus answers
8-10 substantial questions in 75 minute class
39
Does it work? 2. Student assessment.
60
Usefulness of lecture to your learning?
50
clickers
40
colored cards
30
20
10
0
great deal
fair amount
some
a little
none
line--1010 (fall 01) lots of demos, colored
cards feedback, no groups (text a bit lower than
lect.) column--1020 (spr 03) used clickers,
assigned seats and groups
40
III. Combining research based innovations

• The context
• 1020 Intro algebra-based physics for
  nonscientists.
• 2nd term of 1010-1020 sequence. 1010 has 200
  students,
• 1020 has 55 students (full 1010 grade spectrum
  except Fs)
• (enrollments have increased x 2-3 over 4 years)
• The challenge
• Traditionally unpopular, Challenge to teach.
• 2 x 1.25 hour lectures, no recitations.
• The advantages
• Largely overlooked by rest of dept.
• No constraints on curriculum or methods.

Innovations (and success) based on general
principles. Likely not class specific. Hard to
do, easy to copy.
41
Examples of research-based teaching that works.
Not trying to pour knowledge into passive
student. Student actively engaged in constructing
new way of thinking and solving problems.

• Few topics, explore in depth.
• Collaborative problem solving/scientific
  discourse.
• long hard homework requiring explanation of
  process and reasoning, not just simple answer.
  selling students on collaboration
• Explicit focus on novice/expert attitudes and
  creative problem solving. Start with tie to real
  world, make reasoning explicit focus. Ask
  students questions in class, elicit multiple
  solution approaches e. g. 1) compare with lab
  results, 2) use equations,
• 3) compare with real world observation, 4)
  application of basic concept, 5) reason from
  previous class discussions.

Discuss how all work, advantages of combining
multiple approaches and viewpoints.
42
Testing the approach-transforming CU classes.

• The context
• 1010/1020 Intro algebra-based physics for
  nonscientists.
• The challenge
• Traditionally unpopular, 2 x 1.25 hour lectures,
  no recitations.
• The results
• Time on homework and enrollment both up.
• Better expert-novice physics attitude results.
• Conceptual exam questions-- improve C ? A
• Big increase in questions and comments (
  20/class)
• Scientific reasoning and problem solving,
  enormous change!

Innovations (and success) based on general
principles. Likely not class specific. Recently
seeing comparable results with 2nd intro course.
43
Lesson built around clicker question.
Lightning rods a. attract lightning to tip,
prevent from hitting rest of building. b. prevent
lightning from occurring. c. make it strike
somewhere else. d. dont actually do anything,
are superstition.
lightning rods
---------------------- --------------------------
-----------------------
- -



first asked-- 8 correct. Discuss reasoning,
relate to concepts. Two days later, asked
again. gt90 correct!!





44
Measures of success.
1. Classroom atmosphere- BIG CHANGE in 1020
(35-40 students) 20 questions and comments/75
min class from 1/3 of students. Visitors thought
were physics majors. 2. Homework- work longer,
do better. 3. Enrollments and attendance
up. 4. Exam performance-- 1 sigma increase
2 letter grades on conceptual understanding/transf
er questions. 5. Student assessments of value
toward learning. 6. Problem solving session
environment. 7. CLASS student attitude survey,
even or .