1.Minitab 18 recommended for this exercise.It may be more difficult with Express.

This exercise involves creating a 90% empirical confidence interval for a population proportion.

-In this exercise, we are studying the sampling and interval estimation process.So, we will use a known population.Suppose in a 2-person race for governor of Indiana, our candidate has 49% of the vote.

Step 1: Use the directions here to take a sample of 1000 voters.Calc - Random Data - Binomial - Number of rows (1000) - Store in column (c1) - Number of trials (1) - Event probability (0.49). Note that we use draws from the binomial distribution to simulate this type of survey. So, your data consists of 1's (Supporter, or Success) and 0's (Non-supporter or Failure). (Note: There is nothing here to copy to your homework paper.)

a.For your sample, have Minitab compute the percentage of "voters" who support (1's) our candidate.The simplest is probably:Stat - Tables - Tally individual variables - Variables (c1) - check Counts and Percents. Your percentage of 1's is called the point estimate; here, it is the point estimate of our candidate's support. TYPE IT HERE. Is it exactly 49%?Briefly explain (perhaps in one or two words) why it is not.

b.Knowing that the actual support is 49%, is your point estimate reasonably close to the actual support?

c.What range of values do you consider to be reasonably close?

d.To answer the question in c we need to construct an interval estimate (confidence interval).Here, we will construct a 90% empirical confidence interval.

-Use Minitab to take 100 samples of 1000 voters and to compute the point estimates for each sample. This can be done very efficiently using the instructions below:

Calc - Random Data - Binomial - Number of rows (100) - Store in column (c2) - Number of trials (1000) - Event probability (0.49)

-Note that, with the instructions above, Minitab has done the sampling and the tallying in one step, and records for each sample the number of successes (supporters).

-Use the Calculator to divide c2 by 1000 to get your point estimates (sample percentages).Calc - Calculator - Store results in variable (c3) - Expression (c2/1000).TYPE HERE OR COPY YOUR FIRST 4 RESULTS.

-Rank the point estimates from highest to lowest. Use: Data - Sort - Columns to sort by - Column (c3) - Columns to sort (set to Specified columns) - Columns (c3) - At the end of the current worksheet. The interval we want has endpoints at the 6^th value and at the 95^th.This is a range of values that can be considered reasonable at the 90% confidence level.

-Of course, everyone's interval will be different.We need a mathematical method to get these intervals, as we've seen.Exercises 2 & 3, use the mathematical method to reach results.