From Tradition to Enlightenment:
The Evolution of Intro Physics
at the University of Illinois
8/30/2012 Texas A&M
Mats Selen
UIUC Physics
Talk Outline
• Ini%al Illinois course revisions (1996-­‐2001)
– Why this was a big deal
– Why everyone should do it
• Two cool projects enabled by this – Prelectures (2005 -­‐ ...)
– Interac%ve Online Labs (brand new)
2
How I know Jairo & George
If you are a young scientist you need to check out their
Cottrell Scholars program: http://www.rescorp.org
3
Faculty:
Gary Gladding
Jose Mestre
Mats Selen
Tim Stelzer
Affiliates:
Michel
Herquet
Morten
Lundsgaard
Michael
Scott
Grad Students:
Zhongzhou
Chen
Witat
Sara Rose
Katie
Crimmins Fakcharoenphol
Noah
Schroeder
Intro Physics at Illinois
Enrollment
(as of yesterday)
PHYS 100: 536
PHYS 101: 516
PHYS 102: 278
PHYS 110: 158
PHYS 140: 196
PHYS 150: 56
PHYS 211: 966
PHYS 212: 1203
PHYS 213: 553
PHYS 214: 600
Total:
5062
The spring
is bigger
Parallel Parking an Aircraft Carrier
Forum on Education of the Am. Phys. Soc, Summer 1997
Intro Calculus Based Physics at Illinois
Physics 211 (4 hrs, mechanics)
Physics 212 (4 hrs, E&M)
Physics 213/214 (2+2 hrs, SM, QM)
About 4400 students/year in these 2 classes
Mostly Engineering & Physics majors
Pretty Typical Class Structure:
Lecture, Lab, Discussion…
Big changes in these courses
happened over 15 years ago…
1300 in 211
960 in 211
940 in 212 1200 in 212
Fa
g
ll
in
r
Sp
How it worked in 1993:
(sometimes good, often terrible)
• Tradi%on, Tradi%on, Tradi%on
– Lecturer “owns” the course and is free to “reinvent the flat %re” every semester.
– Discussion TAs preQy much on their own. – Labs intellectually disconnected from rest of course.
• Results: Nobody is happy !!
– Lecturer dislikes it since it’s a monster teaching assignment.
– Students hate it because they see the lecturer dislikes it, (and its oYen not very good)
– Engineering dislikes it because the product is inconsistent (and because their students hate it).
How we* changed it around 1996:
– Permanent Infrastructure
* Lead by Gary Gladding &
David Campbell
• Course material is basically fixed
• Recycled & tuned from semester to semester.
• People don’t need to re-­‐invent the whole stew, but can focus on the spices!
– Typically 3-­‐4 faculty share the load:
• Lecturer(s)
• Discussion Director • Lab Director
– Course administra%on is shared responsibility:
• Directors meet weekly with their TA’s.
• Overall co-­‐ordina%on is very %ght.
• Everybody works on crea%ng & giving exams.
Advantages of this approach:
– Exis%ng infrastructure lowers the bar for par%cipa%on.
• 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).
•This approach enables innova%on
– Pain & Gain are shared
• No burnout & No heroes.
• Makes it possible to keep quality high and material consistent even though instructors are changing.
•This approach is sustainable.
Summary of the iniKal renovaKon:
OLD (Spring 1995)
NEW (Spring 2001)
Lectures
Blackboard, Transparencies
& Demos
75 min, PowerPoint, Demos, Peer InstrucKon, Just in Time Teaching
Discussion
TA’s worked problems on the board
CollaboraKve Learning, materials based on PER
Labs
Intellectually disconnected, focus on measurement
Predict-­‐Observe-­‐Explain,
Focus on concepts
Homework
End of Chapter problems, (or NovaNet)
Web Based (TYCHO),
InteracKve Examples
Exams
MulK part problems, free response, hand graded
MulKple Choice,
Created by faculty team
TA Instruc%on
None
OrientaKon & training,
Mentor TA observaKon
Overview: Discussion SecKons (2 hrs/week)
TA will guide, not lecture
A Question!!
No HW problems here
Key Idea: Collabora%ve Learning
Students work in groups of 4 on problems prepared by the senior staff. TAs act as facilitators, not lecturers.
 TA prepara%on very important (extensive training program).
Orienta%on, Weekly Mee%ngs, Mentor TAs, Observa%on
 Content of prepared materials very important

Example discussion problem
Two astronauts, each having a mass 75 kg, are connected by a light rope 10 meters long. They are isolated in space, and are ini%ally revolving around their common center of mass once every 6 seconds. They now pull in on the rope, halving the distance between them. How much work is done by the astronauts as they shorten the rope?
Overview: Lab SecKons (2 hrs most weeks)
PREDICT
OBSERVE
EXPLAIN


Engage the students in the learning process and promote mastery of concepts by manipula%on of experimental apparatus.
Prelab assignments; Lab reports finished within class period. (and much easier to grade than the old style)
Effect of iniKal structural changes:
AIer (2001)
Before (1995)
bad
good
Before (Spring 95)
Total Physics TAs = 77
# “Excellent” = 15
19 ± 5 %
good
bad
AIer (Spring 01)
Total Physics TAs = 75
# “Excellent” = 58
77 ± 6 %
The Challenge (part 1)
Faculty Buy-­‐in
“The Course” is no-longer just “My Lectures”
Question: How do lectures fit right now?
Time (Students)
Time (Professor)
70%
15%
15%
Lecture
Discussion
Lab
Homework
Exams & Quizzes
15%
18%
35%
16%
16%
15
The Challenge (part 1)
Faculty Buy-­‐in
“The Course” is no-longer just “My Lectures”
Question: How do lectures fit right now?
Fraction of Grade
5%
10%
15%
60%
10%
Lecture
Discussion
Lab
Homework
Exams
16
The Challenge (part 1)
Need a paradigm shift:
Lab
Homework
& exams
Lecture
Discussion
17
Time to do other things.
Paradigm shift
Reluctant Prof:
Gary Gladding:
It won’t be my own
course anymore!
This is good. You won’t
have to do it all by yourself.
But I like to do
stuff my way
You can still personalize
it, you just can’t wreck it.
Whats in it for me
Time to do other things.
Infrastructure is the key !
18
The Challenge (part 2)
Sustainability:
“People who create reform are usually not the same
people who enjoy making the trains run on time”
Again - Infrastructure is the key !
(i.e. Departmental commitment)
Cost per student is no higher than the
old approach (and it scales better).
19
Effect of iniKal structural changes:
AIer (2001)
Before (1995)
bad
good
Before (Spring 95)
Total Physics TAs = 77
# “Excellent” = 15
19 ± 5 %
good
bad
AIer (Spring 01)
Total Physics TAs = 75
# “Excellent” = 58
77 ± 6 %
IncenKve for further innovaKon...
What’s the next big problem?
Students are not reading the text and aren’t prepared for class

Lecturer has to assume that students know nothing coming
into the classroom.

We spend (waste) a lot of time going over very basic material.

Difficult material is often rushed and student only see it once.
New Approach to Lecture (2008 - ...)
Pre Lectures
Checkpoints (JiTT)
Peer Instruction
(old: 75 min, new: 50 min)
Pre Lectures
Viewed prior to each lecture
(usually the night before)
Students do this instead
of reading a textbook
Introduces all concepts
for the coming lecture
and provides feedback to
both students and
professor
Our students watch the prelectures
Textbook
Prelectures
Next: Checkpoints
(aka Just in Time Teaching)
Online knowledge check of
prelecture concepts
Completed after Prelecture
but before Lecture.
Increases student buy-in
for upcoming lecture
Feedback to professor
helps lecture prep.
(we’ve been doing this for 15 years)
Checkpoint Example
Grade for participation,
not for correctness
Useful information…
wrong
right
Student explanations for “it spills over”…
CheckPoint
When the ice melts, the level of the water in the glass will:
A) Go up, causing the water to spill out of the glass. B) Go down. C) Stay right at the brim.
A) The water level will rise. I learned that from Al Gore in
"An Inconvenient Truth"
B) volume of ice is greater than volume of water
In Class
C) The melted water has exactly the same mass as the
ice cube, and the volume of water displaced is equal to
the mass of the ice cube.
“The ice melting in cup of water question confuses me. So if the
water level doesn't change, why worry about global warming?”
From subsequent lecture
29
Lectures = Peer Instruction
• Lectures are very interactive
– We know students are prepared (Prelectures)
– We know their misconceptions (Checkpoints)
• Typically ask 6-10 clicker
questions per lecture
• Valuable feedback for both
teacher and students
15 years ago, class questions made me realize my
lectures were not as great as I though they were.
That’s what started all this for me.
How does all this impact our students?
Am. J. Phys. 78, 755-759, 2010
Phys. Rev. ST Phys. Educ. Res. 6, 1-5, 2010
Checkpoint Study
A measurement of students concept knowledge
Spring 06
56%
Spring 07
57%
Checkpoint Study
Overall Results
Checkpoint Study
Overall Results
Viewers vs. Non-Viewers
Non-Viewers
Viewers
Audio narration time of Prelecture
35
Viewers vs. Non-Viewers
36
Checkpoint Study
Significant improvement seen for all students
Changes Made Learning Easier!
Before
After
Before
After
Before
After
What’s Next?
Present Status
Lecture: (50 min) JiTT & Peer Instruc%on, smartPhysics (good)
Discussion: Peer instruc%on, trained & mentored TA’s (good)
Homework: Online, including Interac%ve Examples (good)
Labs: Group work (tradi%onal) (not so good)
Exams: Mul%ple Choice (good)
Hmmm…
Labs are typically “underappreciated”
• Students don’t rate them highly
– Hard to see connec%on to rest of class
– Group work leaves some people inac%ve
– Focus is to get finished & leave
• OYen don’t quite “fit in” with the course
– Timing rarely op%mum
• They are expensive to run
– TA’s, equipment, space…
Interactive Online Labs
Hands-on activities delivered & graded online.
The Big Idea:
Each student has their own device (buy cheap).
They are guided through each ac%vity by interac%ve soYware.
Timing of ac%vi%es driven by pedagogy, not space/budget.
Not just a simulation…
Lets look inside
•
MSP430 processor
– 8 bit RISC, 12 A/D channels, 8 I/O,
– CC2500 2.4GHz wireless transceiver sends data to PC (similar unit plugged into PC)
•
Peripherals
– 3D accelerometer
– 3D Hall probe
– 40kHz transducer (range)
– A/C coupled amp input (x40,000)
– V1, V2 “scope” inputs
– Pulser output
Enabled by cheap consumer electronics
Next version smaller & more capable
(force probe, gyro, etc… more later)
Demo – friction, SHM
We built 60 prototypes
What we have tested so far:
(using prototype hardware & soYware)
•
Mechanics (using accelerometer):
– KinemaKcs
– RelaKve MoKon
– FricKon
– Free-­‐fall & air resistance
– Pendulum
– Centripetal acceleraKon
•
E&M (using voltage inputs)
– MagneKc inducKon & Faraday’s law
Modern Physics (using RF signal)
– ProperKes of microwaves (2.4GHz)
– ReflecKon of electromagneKc waves by a metal surface
– Speed of light (~2% accuracy!)
– OpKcal density and index of refracKon
•
Circuits (using voltage In/Out)
– InvesKgate Ohms & Kirchhoff's laws
– InvesKgate RC circuit, exponenKal decay & measure Kme constant
E&M (using Hall sensors)
– Earth magneKc field
– Field from current carrying wire (strength ~ 1/r)
– Field from a dipole (strength ~ 1/r3)
•
•
•
Other
– Wireless heart monitor (ECG)
– VibraKonal Spectroscopy
– Temperature monitoring
It really works (tested by > 500 UofI students)
History & Plan:
• 2010: Created prototype hardware & firmware & soYware. Preliminary tes%ng with volunteer students. • Jan/2011: Wrote NSF/TUES proposal to study IOLab pedagogy.
– Funded for 4 years (Stelzer & Selen).
• Spring/2011: Developed IOLab PC applica%on.
– Applica%on walks students through hands-­‐on ac%vi%es • Fall/2011: Tested in Physics 100 (500 students)
– Used IOLab in discussion to explore kinema%cs & rela%ve mo%on. • March/2012: Tested in Physics 212 (30 students)
– Used IOLab to independently explore magnets, coils & Faraday’s Law. • Summer/2012: Version 2 prototypes
• Fall/2012: Develop V2 soYware. Develop content.
A Peek at Version 2 Hardware
• Features
– 3D accelerometer
– 3D magnetometer (.01 BE)
–
–
–
–
–
–
–
–
–
–
–
Force probe (± 10 N)
Ultrasound ranger (dual mode).
3D gyro (gives orienta%on)
Posi%on encoder wheel for x, v, a rela%ve to desk
1 DC & 3 AC coupled high gain differen%al amplifiers w/ external inputs
Light intensity sensor
Atmospheric pressure sensor
Temperature sensor
Buzzer & microphone
Much higher sample rate & data transfer rate
Extensive expansion port including ADC in, PIO & DAC out, I2C, SPI
Demo !
Amusing Distraction #1:
Cheap ECG
Amusing Distraction #2:
Diagnostic Spectroscopy
Idling
Car
IOLab
Sensor
Time Domain
FFT
Idling
FFT
Idling + A/C
Frequency
Domain
Aside: Speed of light
Pre-prototype hardware shown
2.4
Rx
GHz
Tx
Signal Strength
X
6cm = λ/2
c = f λ = (2.48x109 Hz)(.012 m)
= 3x108 m/s
X (cm)