For Example 1.1, we repeatedly tossed a coin 16 times to assess whether 15 (the observed result) out 16 tosses would be typical or atypical if the infants were picking equally between the 2 toys. That is, we used 16 coin tosses to model or simulate what could have happened in the study if Buzz was guessing each time which button to push. We repeated this simulation several times to get an idea of how often there would be 15 successes, under the assumption that each outcome is equally likely to be correct or incorrect.

For Bob and Tim, we want to explain why %count1% students (or more) put Tim on the left out of %n1% students. Recall, we have two possible explanations for these results:

  1. College students do NOT use facial prototyping. They were choosing equally between Bob and Tim. In other words, the results from class happened by random chance alone.
  2. College students do use facial prototyping.

Which is the best explanation? To make this determination, let's see whether the results from class are consistent with explanation #1. That is, could class count students (or more) putting Tim on the left out of class sample size, reasonably occur by random chance alone, if the names are being chosen equally often in the long run? 

To answer this question, we want to estimate a probability. Again, to estimate a probability, we can repeat (simulate) a random process many, many times to generate a distribution of what could have happened if each student chooses equally and at randon. In this case, we want to know about the distribution of the number of heads under many, many sets of class sample size coin tosses.

Back Next