Midterm marks spread

[Subbarao Kambhampati <rao@asu.edu>]

Folks:
 The midterm marks spread was 

Max: 80/80
Min: 6/80
Avg: ~40ish

The full xcel spreadsheet is with Romeo (which for some reason he cannot
email to me!)--Romeo--can you send the Avg/st dev directly?

I was somewhat surprised at the low marks as I thought it was a fairly
straightforward exam that might bore you.. 

Which brings us to your perceptions. As per the numbers you gave today
(listed below; summarized here)

Raw data for first qn "for most people, the midtem probably fairly
tests their understanding of material covered in the semester" (-5
to +5 scale)

 The mean: 2.17  median: 2   Max: 5 Min: -4 Stdev:
2.03

Raw data for second qn "My performance in the midterm can be taken
to be indicative of my understanding of what has been covered in the
class until now"

 mean: 0.7  median: 1  Max: 5 Min: -5  Stdev:
3.35

[After adding the "incumbent" factor (+1.5), the
"turnout" factor (+0.57) and the "cell phone" factor
(+4), the mean and median for both questions come to 6.78. 

Also, the predicted effectiveness of the overall exam--using the Gibson's
poisson equation (Cosh(Mean(q1))+log(|Stdev(q1)-stdev(q2)|)+ 1/log(1)
)--
works out to +infinity (+/- 5 percentage points). 

]

Rao

 "It's not a good idea to change horses on the
way to the Apocalypse."


3 3 
2 4 
1 0 
3 -1 
2 -1 
2 5 
4 4 
4 4 
1 1 
2 4 
0 -1 
5 -5 
2 2 
3 -3 
3 -4 
4 4 
-4 -4