What every teacher should know about cognitive science
1. What every teacher should know about
cognitive research
Or
How People Learn
Dr. Stephanie Chasteen
Physics Department
University of Colorado at Boulder
Stephanie.Chasteen@Colorado.EDU
2. This presentation is copyrighted under the Creative Commons License
Attribution Non-Commercial Share-Alike
That means: Please watch it, share it, and use it in your presentations.
Just give us credit, don’t make money from it, and use the same kind of
license on the works that you create from it.
More information about Creative Commons licenses here:
http://creativecommons.org/licenses/
Credit should be given to: Stephanie Chasteen and the Science Education
Initiative at the University of Colorado,
http://colorado.edu/sei
3. PER at Colorado
Faculty: Grad Students:
Melissa Dancy Stephanie Barr
Michael Dubson Kara Gray
Noah Finkelstein Lauren Kost-Smith (PhD May 11)
Valerie Otero May Lee
Kathy Perkins Mike Ross
Steven Pollock Ben Spike
Carl Wieman (on leave) Ben Van Dusen
Postdocs/ Scientists: Bethany Wilcox
Charles Baily Teachers / Partners / Staff:
Danny Caballero Shelly Belleau
Stephanie Chasteen Jackie Elser,
Julia Chamberlain Trish Loeblein
Kelly Lancaster Susan M. Nicholson-Dykstra
Laurel Mayhew Sara Severence
Emily Moore Emily Quinty
Ariel Paul Mindy Gratny, Kate Kidder
Rachel Pepper John Blanco, Sam Reid
Noah Podolefsky Chris Malley, Jon Olson
Benjamin Zwickl Oliver Nix, Nina Zabolotnaya
4. Major advances past 1-2 decades
Consistent picture ⇒ Achieving learning
classroom brain
studies research
cognitive
psychology
5. Some big outcomes:
• Learning is constructing understanding
• People organize their experiences into
patterns or mental models
6. Outline
• What people know affects what they
learn (context is important)
• Preparing your students to learn
• What we remember is affected by how
our brain works (the limits of retention)
If time
• Motivation is important
• Feedback is important
7. Outline
• What people know affects what they
learn (context is important)
• Preparing your students to learn
• What we remember is affected by how
our brain works (the limits of retention)
If time
• Motivation is important
• Feedback is important
11. Strong indication:
Prior knowledge matters
Sometimes prior knowledge gets in the way of learning
12. Tools allow thought
A Story of Galileo: 6 theorems of a genius
Theorem: If a moving particle, carried uniformly at
constant speed traverses two distances, then the
algebra
time interval required are to each other in the ratio of
their distances.
(followed by 2 page geometric proof).
d1 = r * t1 t1 d1
=
d2 = r * t 2 t 2 d2
From diSessa (2000) Changing Minds
13. THE MONTILLATION AND USES OF TRAXOLINE
It is very important to
learn about traxoline. QUIZ:
Traxoline is a new form of 1. What is traxoline?
zionter. It is montilled in
Ceristanna. The 2. Where is it montilled?
Ceristannians found that 3. How is traxoline
they could gristerlate large quaseled?
amounts of fervon and then
4. Why is traxoline
bracter it to quasel
traxoline. This new, more important?
efficient bracterillation
process has the potential to
make traxoline one of the
most useful products within
the molecular family of
lukizes snezlaus.
14. So, lack of context inhibits students from
building productive mental models. It
encourages memorization of facts and
pattern-matching.
But… students are not always aware of
the context of what they’re learning!
Why?
16. The “dead leaves” model
(a) Write down every equation or law the teacher writes
down that is also in the book
(b) Memorize these, together with end-of-chapter formula
(c) Do enough homework and end-of-chapter problems to
recognize which formula is applied to which problem
(d) Pass the exam by selecting the correct formulas for the
problems on the exam
(e) Erase all information from your brain after the exam to
make room for the next set of material.
Redish, Implications of cognitive studies for teaching physics. Am. J.
Phys. (1994).
17. Discussion
• How have you seen this apply in your
classroom?
• Where can/do we take into account
students’ prior knowledge?
• Where does context come into our
instruction?
18. How context can help…
The card game
Rule: If there is a vowel on one side,
there is an even number on the other
In order to verify the rule isn’t broken,
which card(s) do you need to flip over?
E 2 L 5
19. Adapted from Johnson-Laird ‘83
The bartender game
You are a bartender and need to verify
that the following drink orders/ ages
don’t break the law: if you drink alcohol
you must be 21 or older. Whose IDs do
you need to check?
Gin/ Age: Age:
Coke
Tonic 16 52
20. If letter = vowel, then number = even
E 2 L 5
If drink = alcohol, then age>21
Gin/ Age: Age:
Coke
Tonic 16 52
21. So, prior knowledge can be used to help
process information more readily.
Prior knowledge can be accessed by
providing useful, authentic context.
It is easy to learn something that matches or
extends an existing mental model!
(And it is hard to learn something we don’t
almost already know)
Much learning is done by analogy
24. Comparing Activity Design
Make the man start at Sketch what you think the
-5 meter mark, move with graphs will look like for this story
that Jill told:
constant speed to the 2
meter mark and then “Bobby was talking to me on his
accelerates to the 8 meter cell phone standing by his car.
mark. The phone signal was poor, so
he walked toward his house
A. Sketch the position, trying to get a better signal and
velocity and acceleration then stood still so we could talk.”
graphs that you see.
A. Explain why each part of your
B. How do the three graph makes sense.
graphs relate?
B. Test your ideas using the
Which activity are you more simulation
likely to use with students?
B.Green. B. Red.
25. The importance of context
The procedure is quite simple. First arrange items into different
groups. Of course one pile may be sufficient depending on how
much there is to do. If you have to go somewhere else due to
lack of facilities that is the next step; otherwise, you are pretty
well set. It is important not to overdo things. That is, it is better
to do too few things at once than too many. In the short run this
may not seem important but complications can easily arise. A
mistake can be expensive as well. At first, the whole procedure
will seem complicated. Soon, however, it will become just
another facet of life. It is difficult to foresee any end to necessity
for this task in the immediate future, but then, one can never
tell. After the procedure is completed one arranges the material
into different groups again. Then they can be put into their
appropriate places. Eventually they will be used once more and
the whole cycle will then have to be repeated. However, this is
part of life.
* Bransford, & Johnson(1972). Journal of Verbal Learning and Verbal Behavior 11, 717-726
28. Outline
• What people know affects what they
learn (context is important)
• Preparing your students to learn
• What we remember is affected by how
our brain works (the limits of retention)
If time
• Motivation is important
• Feedback is important
29. “hooks” for memory
Hooks for retention-- mental connections
e.g. lesson on fasteners-- here are all the
types and how they are used.
vs.
Here is an interesting job problem, here
are possible types of fasteners for solving
6 kg
problem, and here is how a certain type
of fastener solved it.
31. A study…
• Population: cognitive psychology students.
• Content: cognitive theories of memory
• Question: How well do students
understand theories of memory from…
– reading a textbook about classic
experiments?
– analyzing and graphing simplified data sets
from these classic experiments?
33. Study design
First… Second… Assessment
A Graph Data Graph Data
Factual Test
B Summarize Chapter Lecture on Theory
C Graph Data Lecture on Theory
Which do you think did better on the test?
A B C or D- other
Schwartz, Bransford and Sears, 2005. Efficiency and Innovation in Transfer.
34. Score on factual recall test
A B C
Graph data Reading + Graph Data +
Graph data Lecture Lecture
35. So that means…
• Data analysis activities are useless, lecturing
is key. Right?
“Wouldn’t it just be more efficient to tell them?”
• No… this is the ‘conspiracy theory’.
– Assessments designed to test efficient learning of
facts make fact-based instruction look good.
36. Assessment Design
Activity 1 Activity 2 Assessment
Graph Data Graph Data
Factual Test
Summarize Chapter Lecture on Theory
Transfer Test
Graph Data Lecture on Theory
Add a new “transfer” assessment
Asked to predict outcomes of a novel experiment.
37. Score on transfer test (predict new experiment)
A B C
Graph data Reading + Graph Data +
Graph data Lecture Lecture
38. Assessment and Instruction
• So, lectures can be an effective tool for
instruction… IF the students are
prepared to learn from them
• What was so special about the
“graphing the data” activity, especially
compared to summarizing the chapter?
39. Creating a “time for telling”
• Data graphing oriented students to key
features
• They needed to account for variation in
the data = contrasting cases
• This struggle towards meaning prepares
them to learn from lecture, enabling
better transfer
40. The importance of contrasts
What is relevant?
Circle Biggish Empty Solid White Line Left Side of Screen
41. But we do learn to perceive…
Despite variations in surface features
42. The importance of contrast
How do you teach Japanese speakers to hear
“L”? How do you teach someone to taste
the difference between Merlot and Cabernet?
Do you give them the purest example of “L” ? Of a Cabernet?
Learning depends on finding structure in
variability.
Need both positive and negative variations.
But you can’t just throw contrasts at people
43. Invention Activities
orient to key features
• Before a lesson on deviation in statistics
• Ask students to develop “reliability index”
for pitching machines
• Students don’t need to discover right
answer. Prepares them to “get it” when
you give them lecture.
* Schwartz, D. L., Bransford, J. D., Sears, D. L. (2005). Efficiency and
innovation in transfer
* Schwartz and Martin (2004), Inventing to Prepare for Future Learning
44. Pitching
machine
example
• Create a reliability
index that
differentiates
between these
different machines
Implemented in high school Algebra 1.
Himmelberger, K., & Schwartz, D. L. (2007). It’s a homerun! Using
mathematical discourse to support the learning of statistics.
Mathematics Teacher, 101(4), 250-256.
45. A. Area covered by
pitches
B. Perimeter using grid
marks
C. Average distances
between pairs of points
D. Average distance from
random point to all
points
E. Frequency of balls in
each of 4 quadrants
F. Average distance
between all pairwise
points
46. Pitching redux
• Wide variety of sophisticated solutions
• Solutions themselves not critical
• Generates discussion about how to
handle aspects of variability
• Prepares to understand formula
• 9th graders after invention did better on
test than college students after a
semester of statistics!
ave deviation =
å x- X
n
49. Worksheet: Invention
activities
A. Crowded Clowns
B. Popcorn
Work through with two partners. 5 minutes.
50. Thinking about it all
Might you use an “invention activity” like
this in your class?
B.Definitely (why?)
C.It depends (on what?)
D.Definitely not (why not?)
Are there other ways you use contrasting
cases in your teaching?
Are there other ways to let students “struggle
towards meaning”?
51. Summary of contrasting cases
• Asking students to invent a description
of different contrasting cases (e.g.,
crowded clown index) helps them learn
the important features
• Prepared them to learn from lecture,
creating a “time for telling”
• Helps them to transfer to new situations
(but not necessarily in factual recall)
52. Outline
• What people know affects what they
learn (context is important)
• Preparing your students to learn
• What we remember is affected by how
our brain works (the limits of retention)
If time
• Motivation is important
• Feedback is important
53. How much do you remember
from this talk already?
Probably 10% of you remember any non-obvious fact
from 15 minutes ago
54. Working Memory Capacity
VERY LIMITED!
every added demand hurts
learning (“cognitive load”)
(remember/process max 4-7
unrelated items)
Without great care,
exceeded in almost
every lecture.
Mr. Anderson, May I be excused?
My brain is full.
55. What does help memory?
Quiz:
What is Traxoline? It’s a new form of…
D.Montillation
E.Quasel
F.Zionter Testing is a learning event!
G.Bracter
57. Some interesting findings on
studying…
• Under time pressure, people study the
easiest items
• People often stop studying before they
have learned the information
• Spaced vs. massed practice is better
• Self-testing is important
• There are benefits to retrieval even if it
fails, especially with corrective feedback
Kornell and Bjork, The promise and perils of self-regulated study
58. Implications
• Provide opportunity for retrieval in
lecture
• Space repetitions across
lecture/homework
• Help students learn how to study
59. Classroom application
• What kinds of things might you do to
help improve students’ memory of facts
and vocabulary?
61. Outline
• What people know affects what they
learn (context is important)
• Preparing your students to learn
• What we remember is affected by how
our brain works (the limits of retention)
If time
• Motivation is important
• Feedback is important
62. If you see no reason to learn,
you won’t bother!
• Learning takes effort
• Why spend energy if there’s no reason?
• Motivation is highly malleable!
63. “This class is very hard and many of
you will fail so you need to study
really hard.”
How do you think this affects university
student motivation to learn the material?
a. increases b. decreases
Focus groups and
interviews indicate is
demotivating for
university students.
Psychology studies
support.
64. What does motivate?
•What have you found to be the most motivating to
students?
•What did you think would be motivating but
wasn’t?
c.Subject relevance (meaningful context)
d.Instructor attitude. (respect and challenge)
“Subject hard for everyone, but all can master
with effort, and my goal for course is for all of you
to succeed.”
65. Attitudes and Beliefs*
Assessing the “hidden curriculum” -
beliefs about physics and learning physics
Examples:
• “I study physics to learn knowledge that
will be useful in life.”
• “To learn physics, I only need to memorize
solutions to sample problems”
*Adams et al, (2006). Physical Review: Spec. Topics: PER, 0201010
66. How do you think a single
introductory physics class affects
students beliefs about physics?
A. Not much. Their beliefs are pretty well set
by college.
B. Some students probably come out with a
slightly more positive view of physics
C. It varies by students’ individual learning
styles
D. Something else
67. Can we affect students’ beliefs?
Shift (%) “CLASS” survey of
Expert-like beliefs
Real world connect... -6
Personal interest........ -8 The good news: yes…
Sense making/effort... -12
Conceptual................ -11
Math understanding... -10
Problem Solving........ -7
Worse for
Confidence................ -17 females!
Nature of science....... +5
(All ±2%)
Students come out of introductory classes with more negative
views of physics than they came in with!
70. Expert Tutors *
1. Motivation major focus (context, pique curiosity,...)limited
praise, never for person, all for process
3. Understands what students do and do not know ⇒
timely, specific, interactive feedback
5. Almost never tell students anything-- pose questions.
7. Mostly students answering questions & explaining.
9. Asking right questions so students challenged but can
figure out. Systematic progression.
11. Let students make mistakes, then discover and fix.
13. Require reflection: how solved, explain, generalize,…
*Lepper and Woolverton pg 135 in Improving Academic Achievement
72. Outline
• What people know affects what they
learn (context is important)
• Preparing your students to learn
• What we remember is affected by how
our brain works (the limits of retention)
If time
• Motivation is important
• Feedback is important
74. What makes an expert thinker?
It’s not just that an expert knows more
An expert thinks about a subject in different ways
than a novice does
“New wiring!”
We can see that the brain changes through brain
activation and imaging studies, and in what experts
do
75. Feedback helps with constructing
our own understanding
If we’re to change how we think, we need
feedback on our thinking
What does that mean?
What kind of feedback is most helpful?
How can students get it?
77. Feedback through formative assessment
Compare and contrast what students
experience during two different types of
assessment activities.
2. Does the assessment help students
gauge what they know?
3.Does the assessment build skills in
feedback?
4.How does the assessment motivate
students to learn the material?
Adapted from Handelsman, Miller & Pfund, 2007
78. It’s not about our teaching,
it’s about student learning
79. Conclusions
• Educational practice is a researchable endeavor
– We can make systematic progress
– Imperative to include physicists
• Possible to achieve dramatic repeated results
• CU model strongly couples:
– Reform and research
– Education and physics
• Sustaining & Scaling reforms is possible
– Requires theoretical framing
– Both CONTENT and CONTEXT matter
Many Thanks… - Marty G. for ceding his spot… I look forward to it in the spring I’m excited to be speaking with you all today and will be focusing on the field of PER, the broad reseach lines, and some specifics… Basically a bit of the how, when where going of PER --- with the caveat that this is my take. 15 min intor - through CC (10 intro 5 CC) 20 min reps/ analoogy 15 min tutorials 2 min conclusion.
Change Labels (Ack, Fac, etc..)
Why physicists is implicit… need to make more explicit? Include APS backing etc? #’s stats… etc
Why physicists is implicit… need to make more explicit? Include APS backing etc? #’s stats… etc
Time for 30 seconds and see how far we get
Time for 30 seconds and see how far we get
Time for 30 seconds and see how far we get
Galileo - Dialogues Concerning Two New Sciences (early part of 17th C) This is thrm 1 of 6
In communication, new information should always be presented in familiar context = “given new principle”. As scientists, we carry context around with us, we have deep mental models of our discipline.
Fact memorization is what a lot of students see as their job in science class. This is what they fall back on, without context. Physics is a collection of equations on dead leaves. Flip through equations like a set of leaves until one finds the right equation. Want a living tree instead. But how to promote creation of useful mental models of science?
Need to flip over the E to see if there is an even number on other side Need to flip over 5 to see if there is mistakenly a vowel on the other side. Give out DVDs or pencils for giving an answer. Make sure to make it clear what the right answer is and why.
But it’s hard to change existing mental models. You have to offer an alternative that better matches reality than the old model. This is why having students predict outcome of demos is good idea. Gives opportunity for conflict. And, not all students come in with well-defined mental models – sometimes you can just hook onto a vaguely formed idea.
Bring up states of matter. Show real world stuff, and also scaffolding from simple to complex.
Demonstrate “waves” to show scaffolding Show a chemistry sim – Gas Properties?
Talk in your group about these two . Share out. Specific learning goals –both lessons have the same learning goals But this one meets our guidelines for example: First it, Connects to students’ experiences - cell phone Next there is a Connection to students’ knowledge – prediction in A. There are Minimal directions- B just says test, no specifics about sim features given Students are asked to reason and make sense Students self-check understanding- B To get the most out of this lesson, students would be working in collaborative groups.
Analogy to lecturing… students can’t make meaning without the context of laundry… A problem with powerpoint?
What do you see? Trace out the spiral Where is the spiral -- can it exist “decontextualized” I.e. w/o th ebackground?
Show sim. Attend to real life Make visual constrain
A. Analyze and graph data sets from classic memory experiements to find patterns. Requires analyzing contrasting cases. Then graph it again to see if missed patterns. Read book chapter and then write 1-2 page summary of ideas in chapter. Then hear lecture explaining experiments, results, and theories. Or, analyze and graph data, and then hear lecture on theory. Fact based test = “do people tend to remember the first thing they read?” To test this hypothesis, students in one condition analyzed the contrast-ing cases of data. In another condition, students read a modified book chapter that described the same studies and results (in words and graphs), and provided their theoretical significance. This latter group’s task was to write a one- to two-page summary of the important ideas in the chapter. A few days after students completed these tasks, both groups heard a com- mon lecture that explained the experiments, the results, and the theories that were designed to accommodate the results. The question was whether both groups of students had been equally prepared to learn from the lec- ture. We also included a third group that did not hear the lecture. This group also completed the data analysis activity, but instead of hearing the lecture, they analyzed the data a second time looking for any patterns they may have missed. All told, there were three conditions: Data Analysis + Lec- ture, Summarize Chapter + Lecture, Double Data Analysis.
From Efficiency and Innovation in Transfer As an approach to solving this problem, we asked students to analyzeand graph simplified data sets from classic memory experiments to find the “interesting” patterns. Table 1.1 provides a sample of the data sets the students analyzed. Afterward, we asked them questions about what they had studied and compared their performance to other students who had not seen the data but had read summaries of the studies. For example, given the true–false question, “Do people tend to remember the first thing they read?”, students who had graphed the data did not do well compared to students who had written a summary of a chapter on memory. Thus, by a standard (replicative) assessment of knowledge, our method of instruction fared poorly. To assess whether the students learned from the lecture, we employedtwo assessments about a week later as part of a class exercise. The first assessment measured transfer by asking students to read the description of a novel experiment. The students’ task was to predict as many of the out- comes from the experiment as possible. Eight possible predictions were covered in the previous lessons (e.g., primacy). The second assessment used a recognition test that included factual assertions from the lecture. For example, “When people understand something they have read, they tend to remember it verbatim. True or false?”
What if we add a transfer measure? A week later students received description of a novel experiment (no data). Had to predict outcomes, which were derivable from lecture. Very hard task, because the novel experiment was unlike what they had analyzed or read about.
So: Data analysis group learned from lecture, because did better than graphing data only group And graphing the data adds something on top of just summarizing a chapter (= unguided active engagement)
Without contrasts, it’s hard to know what information is relevant. what one notices about the circle depends on the contrast. For example, the fact that it is not filled only becomes apparent when contrasted to the circle that is filled. The information in the circle is infinite... it is on the projected screen, it is on earth, etc., etc., etc. Without contrasts, it is hard to know what information is relevant. Do not assume students already know what is relevant, given that is what you are trying to teach. Give them contrasts so they can figure it out.
The example with chinese character and letter A. For people who do not know chinese, it is very hard to see that it is the same character in all the instances. However, with expertise (in english), it is easy to see that they are all the letter A. The point is that experts can see the underlying structure despite variation in surface features.
If students ask what reliability means, teacher encourages them to create definitely based upon characteristics a basebal coach would look for in purchasing a pitching machine. Uses contrasting cases. Different # of pitches so students notice solution has to handle different sample sizes. Prepares to understand why variability divide by n - Pitching machine with tight cluster of pitches, notice that variability is not the same thing as inaccuracy, which is a common confusion - There is more than one way to measure reliability, so they can generate many feasible solutions This is a form of exploratory behavior. Productive activities to get them to notice and account for contrasts. teracting with each group. We do not encourage teachers to guide students to the conventional solution, because this can shortcut the students’ opportunity to de- velop the prior knowledge that will help them understand the conventional solution at a later time. Instead, we suggest three primary moves for the teachers, which am- plify the three benefits of production previously stated. One move is to ask the stu- dents to explain what they are doing. This places a premium on clarity and consis- tency. A second move is to ask students whether the results of their mathematical procedures correspond to their “common sense.” This ensures that students pay at- tention to specific symbol-referent mappings, instead of simply computing arbi- trary values. The third move is to push students towards more general solutions. The teacher encourages the students to find solutions that generalize across differ- ent legitimate configurations of quantity.
College students solve transfer problem 12% of time; HS students 34%
: The blue circles show that the examples differ on the surface, but the have the same ratio. The red circles make a nice example of a contrasting case at work. Most students start this by just counting the number of clowns. But, by looking at the clowns in the red circles they discover that 2 can't be the right answer for both of them. Using contrasting cases. Like tasting glasses of wine side by side, helps people notice aspects of a situation they might otherwise overlook
Why physicists is implicit… need to make more explicit? Include APS backing etc? #’s stats… etc
Spaced vs massed study. 1 hr for 5 days not 5 hrs for one day We learn by being tested, by retrieval. Explaining is a form of retrieval
Need to flip over the E to see if there is an even number on other side Need to flip over 5 to see if there is mistakenly a vowel on the other side. Give out DVDs or pencils for giving an answer.
Had students study a text, and then study it again. How much do they remember?
When no time pressure, spend time on more difficult items, which will take longer. With time pressure, study those things more proximal to learning. Often drop flashcards that they don’t really know. Will drop after one successful recall, but they are close to knowing those. Spaced study (e.g., pictures by an artist are interleaved) rather than massed (all shown at once) produces better learning. Self-testing is important, but should make retrieval difficult. But if fail to retrieve, doesn’t help recall. Overall, students do whatever is more urgent, and don’t study strategically. Learning by triage rather than what is most effecitve.
Why physicists is implicit… need to make more explicit? Include APS backing etc? #’s stats… etc
Study showed that trait of experts is that they work hard Survival trail
NOT AFFECT
Do we have any chaance of affecting student attitudes /beliefs: good news: yes Bad news worse
Demotivating, and gives the wrong message about what science is about.
Why physicists is implicit… need to make more explicit? Include APS backing etc? #’s stats… etc
Timely specific feedback Tests, homework, peers, clicker questions