Be sure to share your comments in the Class Participation section below -- that's the best part! You can turn pages with the arrows on your keyboard ← →.
Join the conversation! There are now 8 comments... what are your thoughts?
Choosing 3 is spot on. The old adage, “pick C” wasn’t a myth. Before we had computers to randomize multiple-choice exams, most instructor-constructed exams had a bias towards C. If you are in a room with a fairly (i.e., n > 25) large audience, ask them to pick a number between 1 and 4 inclusive (if it is a room full of mathematicians, you have to say “pick a natural number…”). The results are very interesting.
But I doubt instructors would use dice to randomize the position of their answers, that’s the point. When people are asked to pick a random number on their own(no software, no tools, no external randomness), they usually fall near the middle.
Yes, but the most important aspect is that humans can’t be random. For example, if you try to win at Rock, Paper, or Scissors with a “plan,” then you will probably lose against a machine. But if you use a tool for randomization, you can win: http://www.nytimes.com/interactive/science/rock-paper-scissors.html?_r=0
The first page they see controls for absolute (red in this comic) vs. relative (yellow) witnesses, since the pictures are all from the control group. If the witness rejects that lineup, then they’re good to go on a lineup with the actual suspect (as well as the rest of the control group).
Choosing 3 is spot on. The old adage, “pick C” wasn’t a myth. Before we had computers to randomize multiple-choice exams, most instructor-constructed exams had a bias towards C. If you are in a room with a fairly (i.e., n > 25) large audience, ask them to pick a number between 1 and 4 inclusive (if it is a room full of mathematicians, you have to say “pick a natural number…”). The results are very interesting.
Before we had computers, we had dice and also playing cards. So I don’t see what computers have to do with it.
But I doubt instructors would use dice to randomize the position of their answers, that’s the point. When people are asked to pick a random number on their own(no software, no tools, no external randomness), they usually fall near the middle.
Yes, but the most important aspect is that humans can’t be random. For example, if you try to win at Rock, Paper, or Scissors with a “plan,” then you will probably lose against a machine. But if you use a tool for randomization, you can win: http://www.nytimes.com/interactive/science/rock-paper-scissors.html?_r=0
One teacher once told me that, for this reason, he hardly ever used C as the correct answer on multiple-choice exams he made.
This problem seems eminently solvable. First show the witness a page of known innocents, and then proceed to the actual experiment.
I’m not quite sure I followed that. How would you use a control group in this instance?
The first page they see controls for absolute (red in this comic) vs. relative (yellow) witnesses, since the pictures are all from the control group. If the witness rejects that lineup, then they’re good to go on a lineup with the actual suspect (as well as the rest of the control group).