1. Lecture Demonstrations and the Force Concept Inventory
Brian Andersson
Physics Education Seminar
3/25/05

2. Motivation I was curious to see how demos impact student learning.
Is there any reason to believe demos have an impact on FCI scores?
Is there enough data to say one way or the other?

3. Overview
To determine if there is any correlation between Phys 1301 FCI results and demo usage
Start with the large picture
Work down to individual questions

4. Previous Work
Sokoloff and Thornton have been working on Interactive Lecture Demonstrations (ILD)
Gains on the Force and Motion Conceptual Evaluation are reported to be significantly higher for students taking lectures involving ILDs
American Journal of Physics 65:1, 14 (1997)

5. Differences ILDs require an entire class period to perform.
Very few demonstrations are actually used.
I am looking at standard lecture demonstrations (0-10 demos per lecture, duration 1-5 minutes)
Presentation of demos vary from one instructor to the next

6. Analysis - First Pass
I looked at the total number of demonstrations used in a semester versus two results from the FCI
Average Post Test %
Average Relative Gain
Analysis covers 1301 classes from Fall Semester 1999 to the present (33 sections)

11. Initial Impressions Very messy. No correlation.
Try limiting analysis to instructors who have taught 1301 at least 4 times since fall 1999.
There are three: Profs X, Y, and Z.

24. Second Pass Again, very little correlation.
Previous analysis includes demos not relevant to FCI (rotational dynamics, SHM, etc...)
Second analysis involves matching existing demonstrations (FCI demos) to specific FCI questions.
Once matches are made, compare usage to Post Test % and RG

25. FCI Demo List Questions Demo(s) 1 1C20.10, .15, .16 3 1C20.15, 1D60.05
4 1H10.13
D55.10, .11
D50.15
8, 9, 10, N20.40
D60.60, .65
D60.05
D60.10
G20.60
D50.37, .38
25, 26, K20.10, .15, .20, .30
H10.10, .12

26. Examples of FCI Demos
Question 1 involves two balls identical in appearance but differing in mass by a factor of 2
Existing demo involves plastic and steel balls

27. FCI Demos
Some demos do not match to questions exactly, but convey the idea
An example is question 28

32. Third Pass
The previous plots show little, if any, correlation between FCI demos and the FCI over a wide range.
The next step was to focus my analysis on the three instructors who have taught 1301 at least 4 times since 1999
I assumed that their teaching style changed little over the years.
It was hoped that changes in demo usage might be reflected in FCI scores.

47. Fourth Pass Still no convincing data.
The next step was to analyze individual instructors and terms.
Looked at the Post Test % and RG for each question.
Would using FCI demos increase these scores compared to questions not associated with a demo?

60. Final Pass
My final analysis involved the comparison of the three instructors on a question by question basis.
Also, compare results for each term taught by each instructor.
Do FCI demos help?

61. Prof X
Prof Y
Prof Z

62. Prof X
Prof Y
Prof Z

63. Prof X
Prof Y
Prof Z

64. Prof X
Prof Y
Prof Z

65. Prof X
Prof Y
Prof Z

66. Prof X
Prof Y
Prof Z

67. Prof X
Prof Y
Prof Z

68. Prof X
Prof Y
Prof Z

69. Conclusions
Do we have enough data to conclude that demos have an effect (positive or negative) on the FCI?
The charts seem to say one of two things:
There is little, if any, correlation between demos and the FCI
Or, there is not enough data

70. Future Work
Look at data and modify FCI demos that may need improvement
Develop new demonstrations to match other questions on FCI
Expand analysis to 1101 and 1201 classes
Continue collecting data from 1301 classes
Analyze multiple choice and final test scores.

71. AAPT Abstract
Students often comment that they find lecture demonstrations interesting. But, do these “interesting” demonstrations help students understand physics, or are they just something cool to see? Using readily available data, can we determine any connection linking lecture demonstrations to student learning? In this talk, I will discuss my search for relationships between different measures of student understanding and the use of standard lecture demonstrations in large introductory physics classes.