CHAPTER 5 SUMMARY STATISTICS The summary statistics module of the SPPC System enables you to produce a variety of elementary descriptive statistics for each variable in a raw data file. It allows you to enter the name of your raw data file, use all variables or select specific variables from your data file, and then produce the following statistics for each of the variables you've selected or retained. Summary Statistics: Variable Name N Mean Std Dev Std Err Mean Biased Std Dev Skewness Kurtosis Minimum Maximum Range Coeff of Variation Geometric mean Harmonic mean Sum X Sum X^2 Sum x^2 Note: x = (X - Mean X) Sum X^3 Sum x^3 Sum X^4 Sum x^4 Missing values code USING THE PROCEDURE The summary statistics module of the SPPC is very easy to use. The first thing to do is select the "Summary Statistics" option shown in the Master Menu of the SPPC program. When you do that, the summary statistics procedure will then be called into your computer memory. Once the module is loaded into memory you will see the opening menu which asks if you would like to compute geometric and harmonic means. This choice is provided only because it is relatively rare that one needs such means. Moreover, their computation is time consuming on systems that do not have the high speed math co-processor. The module will then ask you to enter the name of the data file that you wish to use. When entering the file name, be sure to include the disk drive and any necessary directory paths. This feature of the program means that you may fetch data files from any available diskette or any defined area on any hard disk. Just in case you are using data files stored on diskettes, the module will instruct you to insert your data diskette in the drive which you indicated when you provided the "filename". If your data file resides on a hard disk, just press the space bar. The module will then search for your raw data file. If it cannot find it in the location you specified, you will be so notified and then given another opportunity to enter the name of your file. If your data file is found, it will be described for you on screen. You must then indicate whether you wish to use the file that is described. Once you have indicated that you will use the located data file the module will ask if you wish to delete any variables. If you indicate that you do not wish to do that, the module will provide summary statistics for all of your variables. If you indicate that you wish to delete variables, the module will then present the name of each variable in your data file and ask whether you wish to RETAIN it. When you have finished the selection of variables the module will ask you to enter the value that will be used to check for "missing values" for each variable within your data file. The system allows only one missing values code and it asks that you confirm the one you entered. None of the summary statistics will be sent directly to your printer. However, all of your results are saved for you as you work along. Then, when you exit the SPPC you will be given the option of reviewing your work on screen, printing it, or storing it in a disk file of your choice. Once you have indicated the foregoing options and choices, the module will begin to read and process your data file. When the entire data file has been read and processed, the system will ask whether you wish to read more data from another file. At that point you may indicate your choice and provide any file description information that the system requires. It will guide you completely. Your final results will now be presented to you on screen. When you are finished with the display of your output, the original summary statistics screen will reappear. You may then use the procedure again with the same or a different file or you may exit and return to the Master Menu of the SPPC program. DATA REQUIREMENTS The summary statistics procedure requires the use of a raw data file that has been created for use by the SPPC program. If you are not familiar with the structure of such files, please read Chapter 2 of the User's Manual. NUMBER OF CASES There is no practical limit to the number of cases that you may have in a raw data file. The theoretical limit is 30,000 cases. You may fill an entire diskette with a single data file and you may continue a data file onto two or more diskettes. You may use as many diskettes as you like. ESTIMATED CASES Under normal circumstances, the second line of your data file will contain the number of cases, the number of variables, and the letter R. If you show the number of cases to be zero (0), the procedure will count the number of cases in your file and then report that number to you. CAUTION: You may NOT estimate the number of variables. NUMBER OF VARIABLES The maximum number of variables that you may have in any raw data file is 200. However, if you have more than 50 variables in your file, you must then select any subset of 50 or fewer variables for actual processing. The program will automatically place you in the "variable selection" mode if you have more than 50 variables in your file. SAMPLE OUTPUTS The following is sample output that was obtained from the first three variables in the Longley data. Those data are stored on this diskette in the file named LONGLEYR.DAT which you may use as an exercise to test your use of the summary statistics procedure. SUMMARY STATISTICS (1) Variable Name = Employed Mean = 65.31700 Std Dev = 3.51197 N = 16 SE Mean = 0.87799 Biased S.D. = 3.40045 Skew = -0.09431 Kurtosis = -1.35148 Minimum = 60.17100 Maximum = 70.55100 Range = 10.38000 Coeff of var = 65.49403 Geometric mean = 65.22806 Harmonic mean = 65.13879 Sum X = 1045.07200 Sum X^2 = 68445.97665 Sum x^2 = 185.00883 Sum X^3 = 4494794.96843 Sum x^3 = -59.33140 Sum X^4 = 295946338.81767 Sum x^4 = 3526.62700 Missing value = -1.00000 (2) Variable Name = GNP_Deflator Mean = 101.68125 Std Dev = 10.79155 N = 16 SE Mean = 2.69789 Biased S.D. = 10.44888 Skew = -0.14640 Kurtosis = -1.17419 Minimum = 83.00000 Maximum = 116.90000 Range = 33.90000 Coeff of var = 102.75499 Geometric mean = 101.13503 Harmonic mean = 100.58128 Sum X = 1626.90000 Sum X^2 = 167172.09000 Sum x^2 = 1746.86437 Sum X^3 = 17350841.58300 Sum x^3 = -2672.19977 Sum X^4 = 1817971236.89010 Sum x^4 = 348220.26132 Missing value = -1.00000 (3) Variable Name = GNP Mean = 387.69844 Std Dev = 99.39494 N = 16 SE Mean = 24.84873 Biased S.D. = 96.23873 Skew = 0.02528 Kurtosis = -1.11807 Minimum = 234.28900 Maximum = 554.89400 Range = 320.60500 Coeff of var = 411.58787 Geometric mean = 375.26336 Harmonic mean = 362.58960 Sum X = 6203.17500 Sum X^2 = 2553151.55993 Sum x^2 = 148190.30489 Sum X^3 = 1105119772.78482 Sum x^3 = 360602.98270 Sum X^4 = 498279105976.54724 Sum x^4 = 2582992122.52588 Missing value = -1.00000