Procedure for Experiments 1-6
First we calculate the sum of your lab profits in all rounds of all sessions. Then for each type (A,B,C,D,E,F) we look at the distribution of winnings of people in all sections who were of that type in that section. For each type, we define the number TOP to be the score that is at the 80th percentile and BOT to be the score that is at the 20th percentile for that type. For example, for type A, TOP is the score such that 80% of the type A's in the entire class will have classes lower than TOP and BOT is the score such that 20% of the type A's have scores lower than BOT.
Everybody who participates in the experiment and has total profits of
less than or equal to BOT for their type will get a normalized score of
2. Everybody who has total profits greater than or equal to TOP will
get a normalized score of 10.
Scores between BOT and TOP will be between 2 and 10 and will be scaled
proportionately to how much you earn.
The average normalize score for each type on each experiment
will be about 6.
The formula for scaling is:
Normalized Score=2+ 8*((Profits-BOT)/(TOP-BOT))
Reminder:
Your normalized score depends on how well you do compared to all other
students of your type in the whole class, not just in your section.
The more profits that you make in class, the better your normalized
score. If you want to score well, it is important to get the
best price you can.
Procedure for Experiments 7 and 8
Because of the nature of these experiments, market conditions varied a great deal between sections. Therefore in experiments 7 and 8, we computed normalized scores by comparing your scores with those of other participants or your type in your own section, rather than with all other participants of your type.
BOT is the score at the 20th percentile and TOP is the score at
the 80th percentile. Top will get you 10 points,
BOT gets you two points. Scores between TOP and BOT get you scores
in proportionately in between.