home *** CD-ROM | disk | FTP | other *** search
- STUDENTIZED RANGE TEST
-
- AIM: The program is designed to calculate a number which represents the
- least amount that any two means can be separated for significance at the 5%
- level.
-
- EXAMPLE: The progress of wound healing was compared when 5 different
- postoperative regimens were employed after abdominal surgery. The 30
- patients in the study were randomly assigned. The numbers below represent
- coded data on the duration of healing.
-
- POSTOPERATIVE REGIMEN Test data for Studentized Range
- A B C D E 5 6
- 3 4 2 6 8
- 3 4 2 6 8 5 7 3 3 2
- 5 7 3 3 2 5 6 4 5 4
- 5 6 4 5 4 2 6 3 5 5
- 2 6 3 5 5 4 9 3 2 6
- 4 9 3 2 6 5 7 5 4 6
- 5 7 5 4 6
-
- The table on the right shows the way the data should be entered in a NON-
- DOCUMENT mode in a WordStar file, if you wish to enter the data from a file
- rather than from the keyboard. Line 1 is the title (65 characters of any
- kind); line 2 contains the number of columns, followed by the number of
- rows. Thereafter, the data itself is entered, just as in the table on the
- left.
-
- Data entry from the keyboard consists of a title (65 characters of any
- kind) to identify the data, followed by entry of the raw data (in this
- case, the 30 numbers shown above). The program calculates the totals for
- each group (column), the means, the means squared and other items in the
- Analysis of Variance. The program then selects a number from its built-in
- Studentized Range table (4.17 in this example), based upon the number of
- groups and the degrees of freedom. It uses this extracted number to
- calculate the smallest significant range at the 5% level (2.604 in this
- example). All this is printed out along with the raw data.
-
- The program also rearranges the group means in descending order of
- magnitude for easy inspection of possible significant differences. In the
- ranked example output from the data above, the first mean (Group 2, 6.5)
- does not differ from the second (Group 5, 5.167) by at least 2.6, so these
- two are not different at the 5% level. The same is true for the second
- (Group 5) and third (Group 4) means, and so on. On the other hand, the
- first (Group 2) and the last (Group 3) differ by 3.167, which is more than
- 2.6. Therefore, these two groups are significantly different, and this
- fact is printed out.
-
- The user may select just screen output, or screen and printer output.
- Printer output is preferred because if there are more than 8 groups, there
- is not enough room on the standard 80 column screen to display the columns
- without wrap around. If the printer option is selected, columns are
- automatically printed in condensed typestyle (SG-10 and similar printers),
- which permits all data to fit on the paper without wrap around. Note that
- all groups (Columns) must have the same number of data points.
-
- REFERENCE: Goldstein, A. 1964 Biostatistics: An Introductory Text.
- Macmillan Company, New York. Pages 71-72 and Table 8.
-
- BASIC VERSION: Dr. Stanley Kaplan May 84
- PASCAL VERSION: Dr. Stanley Kaplan Sep 85
- UPDATES: Oct 86 (Version 2.00)
- Feb 87 (Version 2.10)
- Feb 88 (MS-DOS, Version 2.2)