HW08 Independence
Homework #8
Due before 7:30 AM, Nov 3
Convert to pdf format and upload into eCampus1
The purpose of this homework is to test the statistical properties of the Composite Random Number generator developed for Homework #6. Open the solution to HW #6 found on the G Drive. Let our data set of interest be the first 2000 random numbers generated using the seed 7 and 17 as shown on the spreadsheet. (In other words, the data set will be the values in the range O4:O2003.) To refresh your memory for the Chi-Square and K-S tests, you should use both of those tests to determine if it is reasonable to assume that the data set comes from a uniform distribution between 0 and 1. Then test to determine if the lag-1 correlation coefficient can be assumed to be zero and report those results below. (Note: before you would actually accept these as random numbers, you would also want to check the lag-j correlation coefficients for several more values of j
Fill in the blanks below:
Estimate for lag-1 correlation coefficient:
0.016457823
Variance of estimator:
0.006502
Test statistic (i.e., ratio of lag-1 estimate to its standard deviation):
0.204107
p-value of hypothesis that lag-1 correlation coefficient is zero:
0.83827
1
Note that with eCampus, you must click the Submit button to complete the submission of an assignment. There is no “auto-‐submit” feature for assignments. If you must ask permission for a second submission, there is a one point penalty. Also, to avoid early morning network problems, you should submit your homework the night before it is due.
Due before 7:30 AM, Nov 3
Convert to pdf format and upload into eCampus1
The purpose of this homework is to test the statistical properties of the Composite Random Number generator developed for Homework #6. Open the solution to HW #6 found on the G Drive. Let our data set of interest be the first 2000 random numbers generated using the seed 7 and 17 as shown on the spreadsheet. (In other words, the data set will be the values in the range O4:O2003.) To refresh your memory for the Chi-Square and K-S tests, you should use both of those tests to determine if it is reasonable to assume that the data set comes from a uniform distribution between 0 and 1. Then test to determine if the lag-1 correlation coefficient can be assumed to be zero and report those results below. (Note: before you would actually accept these as random numbers, you would also want to check the lag-j correlation coefficients for several more values of j
Fill in the blanks below:
Estimate for lag-1 correlation coefficient:
0.016457823
Variance of estimator:
0.006502
Test statistic (i.e., ratio of lag-1 estimate to its standard deviation):
0.204107
p-value of hypothesis that lag-1 correlation coefficient is zero:
0.83827
1
Note that with eCampus, you must click the Submit button to complete the submission of an assignment. There is no “auto-‐submit” feature for assignments. If you must ask permission for a second submission, there is a one point penalty. Also, to avoid early morning network problems, you should submit your homework the night before it is due.