Transcript Cornell Colloquium on Course Revisions

Educating in Bulk:
The Introductory Physics Course
Revisions at Illinois
Mats Selen, UIUC Department of Physics





Our Clients
 Faculty, Students, College of Engineering, U of I
Our approach
 Infrastructure
 Pedagogy
 Technology
Some Research Results
 The particle physics approach.
“Physics 100”
 Helping under-prepared students
Deep thoughts
 Just Do It
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 1
Our Clients:
Calculus Based: Total enrollment of about 3500/year
Mostly Engineering (& Physics) students
Physics 111 (4 hrs, mechanics)
Physics 112 (4 hrs, E&M)
Physics 113 (2 hrs, thermo/stat-mech)
Physics 114 (2 hrs, waves/quantum)
Algebra Based: Total enrollment of about 1100/year
Mostly pre-med & biology students
Physics 101 (5 hrs, mechanics, heat, fluids, waves)
Physics 102 (5 hrs, E&M, Light, Atoms, Relativity)
www.physics.uiuc.edu
Each week:
Lectures (2x75 min)
Discusison (2 hrs)
Lab (2 hrs)
Each week:
Lectures (2x50 min)
Discusison (2 hrs)
Lab (3 hrs)
Cornell (Mar/10/03): Pg 2
How it used to work:

Tradition, Tradition, Tradition
Lecturer “owns” the course and is free to
“reinvent the flat tire” every semester.
Discussion TAs pretty much on their own.
Labs intellectually disconnected from rest of course.
Typically only quantitative problems on exams.

RESULTS: NOBODY IS HAPPY !!
Lecturer dislikes it since it’s a monster
teaching assignment.
Students dislike it because they see
the lecturer dislikes it and because the
organization is often “uneven” at best.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 3
How we do it now:

Integrate all aspects of a course using active learning
methods in a team teaching environment.
 Typically 3 faculty share the load for a class:
» Lecturer (lectures, ACTs, preflights, exams).
» Discussion Director (TA training, quizzes, exams).
» Lab Director (TA training, web homework, exams).
 Course administration is shared responsibility:
» Faculty meet at least once a week with each-other
and with their TA’s to plan the campaign.
» Overall co-ordination is very tight (web helps this).
» Everybody works on creating exams.
 Course material changes adiabatically:
» Recycled & tuned from semester to semester.
» People don’t need to re-invent the whole stew,
but can focus on the spices!
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 4
Advantages of this approach:
Existing (evolving) infrastructure lowers the bar for
participation.
» This is now seen as a reasonable teaching load.
» Most of our new junior faculty start teaching in
these courses (i.e. not a heavy assignment).
Pain & Gain are shared
» No burnout & No heroes.
» Makes it possible to keep quality high and material
consistent even though instructors are changing.
42 of ~70 faculty have taught
in these courses since 1995 !
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 5
Feedback
(are things better now ?)
THE NEW
THE OLD
Student Attitudes Towards Physics 101 (fall99)
THE OLD
Spring 95
Total Physics TAs = 77
# “Excellent”
= 15
19 ± 5 %
www.physics.uiuc.edu
Before Course
awful
negative
good
neutral
After Course
positive
bad
100
90
80
70
60
50
40
30
20
10
0
enthusiastic
awful
negative
neutral
After Course
positive
good
Before Course
No of Students
35
30
25
20
15
10
5
0
enthusiastic
No of Students
Student Attitudes Towards Physics 102 (fall99)
bad
THE NEW
Spring 01
Total Physics TAs = 75
# “Excellent”
= 58
77 ± 6 %
Cornell (Mar/10/03): Pg 6
Details of some key components:
WEB-centric organization
Peer instruction in Discussion & Lab sections
ACTs & Preflights in Lecture
Homework & Interactive Examples
Exams
Standard stuff these days
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 7
WEB-centric organization

All course materials available on-line.
 Lectures, discussion & lab materials, exams…
 Makes our job easier (copy spring01  fall01).

All students do several on-line assignments every
week:
 Homework, Interactive Examples, Quizzes
 Preflights for lectures, labs & discussion
 Exam preparation & exam results
 All grades & progress throughout the semester
» Students know in advance what everything is
worth and the final thresholds for A,B,C,D,F etc
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 8
Details of some key components:
WEB-centric organization
Peer instruction in Discussion & Lab sections
ACTs & Preflights in Lecture
Homework & Interactive Examples
Exams
This is hard for TA’s to get used to: Training !
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 9
Discussion Sections
TA to the rescue !

A Question
Key Idea: Collaborative Learning
Students work in groups of 4 on problems prepared by the senior
staff. TAs act as facilitators, not lecturers.
TA preparation very important (extensive training program).
» Orientation, Weekly Meetings, Mentor TAs, Observation
Content of prepared materials very important
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 10
Lab Sections
Engage the students in the learning process and promote mastery of
concepts by manipulation of experimental apparatus.
Web-based Prelabs; Lab reports finished within class period.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 11
Details of some key components:
WEB-centric organization
Peer instruction in Discussion & Lab sections
ACTs & Preflights in Lecture
Homework & Interactive Examples
Exams
The most fun you’ll ever have teaching!
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 12
Pre-Flights !!



Students are asked to answer a set of conceptual questions
(on the Web) prior to every lecture (and discussion, and lab).
The main structure is:
Students read about material in text.
Students answer pre-flight questions on material prior to lecture.
» Physics 101 PF’s due at 6am, lecture starts at 1pm.
» Graded on participation, not correctness.
Instructor uses pre-flight responses to guide lecture preparation.
Pre-flights are reviewed during lecture, often presented again as
ACTs, and often capped off with a demo.
» Use their own words, (both right & wrong)
With careful preparation, the pre-flights can form the “backbone” of
the lecture.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 13
What the students see on the web:
www.physics.uiuc.edu
What I typed in
a simple text file:
Cornell (Mar/10/03): Pg 14
The instructors interface to the
student responses (also on web):
Statistics:
Free response:
“NTUPLE” inspired
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 15
Lecture 6, Pre-Flight Questions 7&8
Two identical boxes, each having a weight W, are
tied to the ends of a string hung over a pulley
(see picture).
What is the tension T in the string?
1. T=0
2. T=W
3. T=2W
T
W
W
correct
44%
23%
This is exactly what I
prepare before and
show during the lecture
33%
0%
www.physics.uiuc.edu
20%
40%
60%
Cornell (Mar/10/03): Pg 16
Students see their own answers
Two identical boxes, each having a weight W, are tied
to the ends of a string hung over a pulley (see picture).
What is the tension T in the string?
1. T=0
2. T=W
3. T=2W
T
W
W
Due to Newton's second and third laws, the rope itself is
massless, so any force transmitted across it is done so
without the diminishing of any magnitude. As each box has
an equal weight, the tension T must be zero, as each box's
force cancels the other's out.
The force applied to the rope is transmitted to the
other side. This example would be just like a person
hoisting up a box, pulling on the rope with a force of
W. In this case, the tension would just be W.
The string has the tension of two weights.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 17
Students have fun with answers...
Shown is a yummy doughnut. Where
would you expect the center of mass of
this breakfast of champions to be located?
(Explain your reasoning Homer).
Homer: "mmmm.....donut...(slobbering)...center of mass in tummy...."
Flanders: "why no diddly-o there Homer. The center of mass would be in
the center of the hole."
Homer: "Doh!"
CORRECT
You're not getting my answer unless i get sprinkles.....suckers !
unfortunately, i think the center of mass of this perfectly
symmetrical donut would be the center of the donut which does
not seem to exist; so, i'll just say homer ate it.
I think it would be in a the middle of the dough in a
circular pattern. Kind of like the onion in an onion ring.
UMMMMM..... Onion rings!!!!
www.physics.uiuc.edu
INCORRECT
Cornell (Mar/10/03): Pg 18
Details of some key components:
WEB-centric organization
Peer instruction in Discussion & Lab sections
ACTs & Preflights in Lecture
Homework & Interactive Examples
Exams
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 19
Interactive Examples (Socratic Dialogue)
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 20
Details of some key components:
WEB-centric organization
Peer instruction in Discussion & Lab sections
ACTs & Preflights in Lecture
Homework & Interactive Examples
Exams
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 21
Students are happier…
but are they learning more ??
Physics Education Research
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 22
For Example:
Do BEFORE vs AFTER study to see
if student learning was impacted by the
introduction of Interactive Examples (IEs):
Choose Homework B “on-line quiz”
performance as metric of learning
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 23
“Homework B”
Example

Given C1, C3, and C4
 Three Qualitative
» Compare Q1 and Q4
» Compare Q1 and Q3
» How does energy change if
dielectric added to C2?
 Two Quantitative
» Given Q1 and E, what is C2?
» Given Q1 and E, what is V3?
• Four different versions
given each week
• Total of 269 different
questions!!
64%
74%
68%
56%
54%
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 24
Homework B Analysis
The Homework B questions have a wide variation in difficulty
# Questions
Homework B Question Difficulty
45
40
35
30
25
20
15
10
5
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Score
Computing the total Homework B average (pre-IE), we see:
Fall Avg
= 63.2  0.2
Spring Avg = 67.8  0.3
We take this difference as a measure of the difference in populations,
on-sequence vs off-sequence.
Until we understand how to correct for this difference, we will
compare only Spring to Spring and Fall to Fall.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 25
Homework B: Pre IEs
Compare performance on Homework B questions (269 total!)
in Physics 112 between Spring99 and Spring00 (pre IEs)
Physics 112 Sp99, Sp00
<Sp00Difference
> - <Sp99>
0.4
0.3
0.2
0.1
0
-0.1
0
50
100
150
200
300
250
-0.2
-0.3
Question
Question
#
Looks pretty similar.. We try to quantify..
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 26
Distribution of Differences
d=0
sp99 - sp00
70
Sp99 – Sp00
60
# Questions
Histogram the points on the previous
plot to get a distribution of the
normalized differences:
d  (<sp99>-<sp00>)/s
50
40
sp
30
20
10
0
-6
-5
-4
-3
-2
-1
0
d
1
2
3
4
5
3
4
5
Distribution centered at 0 (mean = 0.09), with width 1.17
70
Question Difficulty Distribution
Student Performance Distribution
Run “simulation” of two identical
semesters and plot d distribution:
Model Works
www.physics.uiuc.edu
60
# Questions
Create model from the following input:
Sp99 – Sp00
50
40
30
20
10
0
-6
-5
-4
-3
-2
-1
0
d
1
2
Cornell (Mar/10/03): Pg 27
Compare Performance Post IE – Pre IE
70
60
112
Sp02 – Sp00
60
50
50
112
Fa02 – Fa00
40
40
fall02 - fa
sp02 - sp00
30
112Sp00z
30
112Fa99
20
20
10
10
0
0
-6
-5
-4
-3
-2
-1
0
d
1
2
3
4
-6
5
60
50
CONCLUSION:
Systematic Improvement
in all semesters of
both courses!
-5
-4
-3
-2
-1
0
d
111
1
2
3
4
5
Fa02 – Fa01
40
fall02-fa
30
111Fa01
20
10
0
-6
www.physics.uiuc.edu
-5
-4
-3
-2
-1
0
d
1
2
3
4
5
Cornell (Mar/10/03): Pg 28
112comb
More Observations:
On avg, HWB score
has improved by 1.6s
A total of 47
questions have
improvements > 3s !
Look in a little
more detail at
these 47 questions
(18% of total)


60
Physics 112
50
d=0
Sp02 + Fa02
40
30
112co
20
10
0
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
ovf
Were improvements made in qualitative or quantitative questions?
 Qualitative/Conceptual: 35 questions
 Quantitative:
12 questions
 Improvement occurs uniformly in both areas
» fraction of all HWB questions that are quantitative = 25%
Very encouraging, but much more work is needed to understand this.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 29
Is Easy to Forget that Physics is Hard
And its even worse if you are not well prepared !
Can we help under-prepared students to “get it” ?
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 30
Physics 100:
”Thinking About Physics” (1 credit hour)

Despite the University of Illinois’ C of E very high admission
standards, nearly 20% of accepted students are inadequately
prepared to pass our introductory mechanics course Physics 111
(i.e. they earn a D or F).
The failure rate is even higher for minority groups.
» As high as 68% for African Americans !!
(About six times higher than the average)

Many students do not realize that they are poorly prepared.

We need to identify inadequately prepared students and help
them gear up for Physics 111 and beyond.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 31


Typical Physics 100 enrollment ~ 100
Getting students is a challenge…next slide.
Course Format
Online “Asynchronous Lectures” (interactive slides with sound)
Online Homework (heavy use of Interactive Examples)
Two hour discussion/tutorial once per week (expert instructors)
Material covered is first 30% of Physics 111
Follow-up prep classes in || with Physics 111 & 112
Under-prepared
Mainstream
Physics 100
Physics 111
Fall(n)

Physics 111 + 1hr prep
Spring(n+1)
Physics 112 + 1hr prep
Physics 112
Fall(n+1)
Physics 100 does not start until ~ 3 weeks into the fall semester.
Gives students time to evaluate their situation
Many decide to take Physics 100 after doing poorly on the first
Physics 111 mid-term exam.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 32
Identify students at the beginning of the semester using
an online Self Evaluation quiz in “Engineering-100”.
Students receiving a score below a certain cutoff are
“invited” to take Physics 100 (1 credit-hour).
Fewer than half of identified students choose to
participate initially. (This should be a placement exam!)
Physics 111 Grade

Self Evaluation Score
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 33
Is Physics 100 helping students ?
Q: Can we reduce the failure rate of under-prepared students
taking Physics 111/112/113/114 ?
A: Probably YES (research by Gladding & Shoaf)
Physics 111 Grade
Average physics 111 students
Average physics 100 students
All students
Self Evaluation Score
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 34
Is Physics 100 helping students…

It seems like we might be helping, however there is a is a
big caveat: Physics 100 students are self selected !
 Are we getting only those
students that were going
to do well anyway ?
 We need more data to
study this.
» A real placement test
would be very helpful !
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 35
Summing Up



We have revised all of our “big” intro classes
(both calculus and algebra based).
So far so good.
Feedback very positive.
» Students like it, the faculty likes it, the University likes it.
Maintaining momentum will be the challenge for the future
Our Physics Education Research Group is growing
This is becoming a real PhD research program
» Graduate students are very interested
We have lots of data and many analysis ideas (like CLEO)!
Other projects
New Inquiry based physics course for elementary-ed students
» K-5 teachers need more science.
» I’m very excited about this…ask me later.
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 36
Concluding Thoughts




Strong departmental support is needed to pull this off:
Vision, leadership, money (faculty release time).
Developing a sustainable infrastructure is the first part
of the battle.
We are eager to give away any/all of the materials & tools we
have developed, and (of course) hope to get new ideas back.
Getting faculty to “buy in” is the second necessary ingredient.
Not everyone likes this approach.
At UIUC, most people have bought
in to the “new” way.
42 of ~70 faculty
“I can do it better
have taught in these
all by myself”
courses since 1995 !
www.physics.uiuc.edu
Cornell (Mar/10/03): Pg 37