Introduction to SAS: Project 3

Project 3

In this project futher analysis is done with the data from Project 2.

Since two sleep latency measurements, given in the file sleep.dat, are taken on each patient, one correct method for testing for effectiveness of the drug is to do a paired comparison t-test. To do this test, the difference of the sleep latency times for each patient must be computed. As is described in Defining New Variables this can be done easily by defining a new variable in the data step.

ASSIGNMENT:

1. Use proc means and the modified data set to test for effectiveness of the drug. Turn in a neat printout of the RESULTS of using proc means. There's no need to print out the actual data. Turn in a copy of your program too. Give a SHORT answer to the question: is the drug effective? Don't forget to use the title statement to put your name and project number on each page of the printout.

2. The file iris.dat contains the famous `Fisher Iris Data' (from Andrews and Herzberg) consisting of sepal length and width and petal length and width measurements for 3 types of Iris: iris setosa, iris versicolor, and iris virginica. There are 50 measurements for each type of iris. Each row in the file contains 12 numbers: the four sepal and petal measurements for one plant of each species in the order specified above. Use the infile statement (not the cards statement) to create a SAS data set which contains only the sepal length measurements for each species. Turn in a nice printout of this data set, with appropriate labels. Use the pagesize option to print about 55 lines on a page. Consult infile statement for the method of reading files. The information in Defining New Variables may also prove useful, as might that regarding drop and keep.

3. Briefly describe how you would conduct a statistical test of whether the sepal lengths for the three species are the same or different.

Contents Index <<Browse<< >>Browse>>