In a recent Gallup poll the following question was asked to those that have access to a computer at work or school.

Do you, personally, use the Internet or other computer on-line service at your place of work or school?

66% of the respondents answered yes.  If this were exactly true, then the probability of choosing a person that uses the Internet at work or school from those that have access to a computer is 0.66.  We will let Minitab simulate some random samples from this sort of population.

1. Simulate drawing 20 people that use computers.  These are called Bernoulli trials.  (To get data for a Bernoulli trial use Calc > Random Data > Bernoulli.)  A window will then appear asking for the number of rows of data, where to store them, and the probability of success.
1. Find the proportion that answered yes.  (You can either count the number that answered yes (they appear as 1s) or you can use Stat > Tables > Tally Individual Variables.)
2. Repeat this with a sample of size 100 and a sample of size 500.
3. Which trial gave a proportions closest to 0.66?  Which trial would you expect to give a proportion closest to 0.66?
2. Simulate drawing 100 samples of 20 people that use computers.  You could repeat what you did earlier 100 times (sound like fun?) or you could generate binomial data.  Repeated Bernoulli trials form a binomial distribution.  )To get data for a binomial distribution trial use Calc > Random Data > Binomial.)  A window will then appear asking for the number of rows of data (100), where to store them, the number of trials (20), and the probability of success (0.66).  After you get your data, convert the counts to proportions.  (Do this using the Calc > Calculator.  Divide your counts by 100 to change them to proportions.)
1. Make a histogram of these 100 proportions and describe the shape, center, and spread of this distribution.
2. Repeat part (a) by drawing 100 rows of data of with a sample size (or number of trials) of 500.  Again, make a histogram of these 100 proportions and describe the shape, center, and spread of this distribution.
3. In what ways are the two distributions (in part (a) and part (b)) alike and in what ways are they different?
3. Repeat what you did in question 2 using 500 samples of 500 people.
1. Theoretically the mean for this should be 0.66 and the standard deviation for this should be approximately 0.02118.  Find the mean and standard deviation for your sample.  How close are they to the theoretical?
2. Your distribution should also be approximately normal.  Make a histogram along with a normal curve.  (To do this use Histogram  > With Fit.)