=3.30, CI=(2.93, 3.67)
Problem 1 Can we use Pearson's correlation coefficient to measure the strength of the relationship between Income and Distance? Why or why not? No matter what you answer, find the correlation coefficient anyway.
Problem 2 Can we use the least squares regression method to describe the relationship between Job Level and Satisfaction? Why or why not? No matter what you answer, find the least squares equation (with y=Satisfaction) anyway.
Problem 3 Find the least squares regression equation for (y=)Income by Years for the female employees only. If you got it right you will find b1=1432. What is the meaning of this number?
Problem 4 Find the mean, standard deviation and the five-number summary for the Satisfaction of the employees.
Problem 5 In problem 4 we found the various summary numbers for the Satisfaction of the employees. Because WRInc (at that moment in time) had exactly 527 employees this dataset comes from a census, it is for the whole population, so these are all parameters.
Let's say that instead they would have found a simple random sample, say of size n=50. Then they could have calculated the same numbers, but now they would be statistics. In that case they would also have calculated confidence intervals. Let's do a little simulation to see what might have happened, specifically for the mean.
1) Get a sample of 50 Satisfactions using the Calc> Random Data> Sample from Columns command
2) Find the mean of your random sample and its 90% confidence interval with the Stat> Basic Statistics>1-sample t test
Here is what I found when I ran this simulation once: =3.30, CI=(2.93, 3.67)
a) Now you do this and repeat it 10 times.
b) In your 10 runs, how often did the confidence interval not include the true parameter?