Play and Learn 2/19941.ZIP/19941/EDUCMATH/STATS/RICSTS.DOC

Text File | 1994-02-04 | 26.4 KB | 758 lines

INTRODUCTION ..........................................................1
QUESTIONNAIRE ANALYSIS PROGRAMS
        FILEIT - Program to create a questionnaire data file...........2
        COLTAB - Single Column Tabulation Program .....................3
        CRSTAB - Cross-tabulation Program .............................4
        CHISQ1 - Chi Square Test of Independence ......................5
SAMPLE DESCRIPTION PROGRAMS
        DSCRPT - Frequency Distribution, Percentiles and T-Scores......6
        MEANSD - Means, Standard Deviations and Standard Errors........7
        INTRVL - Mean and Standard Deviation Using Grouped Data........8
CORRELATION
        CORREL - Product Moment Correlation............................9
        SPEARM - Spearman Rank Correlation............................10
STATISTICAL TESTS
        ANOVA1 - One-Way Analysis of Varience.........................11
        ANOVAR - One Factor Repeated Measures Analysis of Varience....12
        ANOVA2 - Two Factor Analysis of Variance......................13
        CONFIN - Confidence Interval Program..........................14
        TTESTR - Student's t-test for Correlated Means................15
        TTESTI - Student's t-test for Independent Means ..............16
.PAGE
                                INTRODUCTION

     The SHAREWARE STATISTICAL programs contained on this disk have been
designed in a way such that the user need only identify the statistic
desired and supply the data.  All formulas required for computation
are supplied by the program.  In addition, procedures have been included
that allow the user to store data in a disk file and retrieve it at a
later date for further processing.  This document file contains brief
descriptions of each program and an example of its use.

     All of the programs on this disk are copyrighted by Dr. Robert D.
Cloward, Department of Psychology, Rhode Island College, Providence, R.I.
02908.  The disk may be copied as shareware with the understanding that
the user will send $20.00 to the author, a modest amount for programs that
would go for a hundred dollars or more from a commercial vendor.
.PAGE
        FILEIT:  A PROGRAM TO CREATE A QUESTIONNAIRE DATA FILE

     This program enables the user to create a data file consisting of
alphabetic and numeric codes to identify respondents and their responses
on a questionnaire.  The file will consist of one record for each respondent
and assumes that the same kind of information will be coded in respective
positions (columns) of each record.  The maximum record size is 200 positions.
The program will accommodate up to 1000 records.  The user must specify the
filename and record size at the beginning of the program.  Column markers
are provided to assist the user in entering data in the correct columns.
When the record size is greater than 60, input for each record will occur
over several lines.  Options are provided for displaying, changing, and
deleting records, as well as extending the file with new records.  The
file created is conpatable with the other Questionnaire Analysis Programs
on this disk.

     WARNING: the new file is not created (or the old file updated with
corrections) until the end of the program.  If the program is aborted
prematurely, the data entered will be lost.  Consequently, large amounts of
data should be entered in stages, exiting the program normally at the
end of each stage, and saving that much data.
**************
     The following is an example of its use to create a file:

     *****  FILEIT PROGRAM  *****

     ENTER YOUR FILE LABEL - ANY 6 LETTERS AND .DAT
     ? QUES82.DAT

     HOW MANY CHARACTERS IN YOUR RECORD - RECORD LENGTH?
     ? 50

     FILE QUES82.DAT CONTAINS 0 RECORDS

     SELECT A PROCEDURE:
              1  ENTER DATA
              2  DISPLAY DATA
              3  DELETE RECORDS
              4  CORRECT RECORDS
              5  WRITE DATA FILE AND EXIT
              6  ABORT (EXIT WITHOUT WRITING FILE)
     ? 1

*********

     ENTER YOUR DATA     ENTER **END** IF NO MORE CASES.

     CASE 1
       POSITION 1         2         3         4         5
       ....:....0....:....0....:....0....:....0....:....0
     ? SMITH      ANDREW   121212321232143151345432323234

     CASE 2
       POSITION 1         2         3         4         5
       ....:....0....:....0....:....0....:....0....:....0
     ? JONES      JOHN     213456619825453475654346565322

     CASE 3
       POSITION 1         2         3         4         5
       ....:....0....:....0....:....0....:....0....:....0
     ? **END**


************

SELECT A PROCEDURE:
              1 ENTER DATA
              2 DISPLAY DATA
              3 DELETE RECORDS
              4 CORRECT RECORDS
              5 WRITE DATA FILE AND EXIT
              6 ABORT (EXIT WITHOUT WRITING FILE)
? 5

FILE - QUES82.DAT - NOW CONTAINS 2 RECORDS.

.PAGE
                  SINGLE COLUMN TABULATION PROGRAM

     This program tabulates the frequency of codes 0 through 9 in specified
positions of each record in a data file.  The output includes the total
number of records, number of records with codes alphabetic, blank, or
non-zero through nine codes, and the median code for that position
[column].  Up to 100 positions can be tabulated in one run, providing
that the positions to be tabulated are sequential.  The following is an
example of its use with a file containing coded responses to a questionaire:

     SINGLE COLUMN TABULATION PROGRAM

     ENTER NAME OF DATA FILE    ? B:CLOW1.DAT

     FOR HOW MANY COLUMNS ARE TABULATIONS TO BE RUN?      ?5

     ENTER STARTING POSITION ON DATA RECORD          ?3

     STARTING TALLY ROUTINE
      50 RECORDS TALLIED
     TALLY COMPLETED
************

TABULATIONS FOR DATA FILE **CLOW1.DAT**  51 RECORDS

POSI  CODE CODE CODE  CODE CODE CODE CODE  CODE  CODE CODE OTHER  TOTAL
TION  ZERO  ONE  TWO THREE FOUR FIVE  SIX SEVEN EIGHT NINE  CODE  0 - 9

  3          40    6          4    1                                 51
            78%  12%         8%   2%                       MEDIAN   1.1

  4          12         25    9    5                                 51
            23%        49%  17%  10%                       MEDIAN   3.0

  5          16    2    16   14    1                            2    49
            32%   4%   33%  28%   2%                        MEDIAN  2.9

  6          10   11     8    3    1                           18    33
            30%  33%   25%   9%   3%                        MEDIAN  2.9

  7           9   11     3    6    3                           19    32
            28%  34%    9%  19%   9%                        MEDIAN  2.1
.PAGE
                    CROSS-TABULATION PROGRAM

     This program cross-tabulates up to 80 single digit fields
with one field being used to define groups.  The fields are
cross-tabulated in serial order, and need to be contiguous for
any given run.  The program assumes that the data are stored
in a file with each record as much as 200 characters long.
Characters other than 1 thru 9 are automaticlly excluded from the
tabulations.  The following is and example of its use:
************

     SINGLE DIGIT CROSS-TABULATION PROGRAM

     ENTER NAME OF DATA FILE  ?  B:LIV.SRC
     HOW MANY RECORDS ARE TO BE TABULATED   ?  100
     ENTER THE POSITION ON THE FILE RECORD
          OF THE ITEM USED TO FORM GROUPS   ?  76
     HOW MANY GROUPS (COLUMNS) ARE THERE    ?  2
     DO YOU WANT TO POOL GROUPS             ?  NO
     DO YOU WANT TO LABEL YOUR COLUMNS      ?  YES
     ENTER  2 LABELS EACH 12 CHARACTERS ON A LINE
     ? BOYS
     ? GIRLS
     ENTER THE NUMBER OF ITEMS TO BE TABULATED  ?  1
     ENTER THE POSITION ON THE FILE RECORD
             OF THE FIRST ITEM TO BE TABULATED  ?  2
     ENTER THE MAXIMUM NUMBER OF ROWS
                IN THE CROSS-TABULATION TABLES  ?  4
     DO YOU WANT CELL FREQUENCIES PRINTED       ?  YES
************

     CROSS-TABULATION OF ITEM IN POSITION  2 WITH POSITION 76

     GROUPS          1            2
     LABELS        BOYS         GIRLS

     CODE 1          3            1
       PERCENT       16           25
       CHI SQ.       0            0

     CODE 2          12           1
       PERCENT       66           25
       CHI SQ.       0            0

     CODE 3          2            2
       PERCENT       11           50
       CHI SQ.       0            2

     TOTAL CASES INCLUDED =  22
     TOTAL CASES DELETED  =  16
     TOTAL CHI SQUARE     =  4.03098
.PAGE
                 CHI SQUARE TEST OF INDEPENDENCE PROGRAM

     This program performs a 'chi' square test of the independence of a
series of frequency distributions.  Cell frequencies must be entered in
row major order.  The program will accommodate from two to ten columns
and from one to ten rows.

         CHI SQUARE TEST OF INDEPENDENCE PROGRAM

         HOW MANY ROWS, HOW MANY COLUMNS?
         ? 2,4
         ENTER YOUR CELL FREQUENCIES ONE AT A TIME IN ROW MAJOR ORDER.
         ? 67
         ? 87
         ? 97
         ? 45
         ? 27
         ? 37
         ? 33
         ? 22
************

         OBSERVED FREQUENCIES
          67  87  97  45  296

          27  37  33  22  119

          94 124 130  67  415

         EXPECTED FREQUENCIES
          67.0458  88.4434  92.7229  47.788

          26.9542  35.5566  37.2771  19.212

         CHI SQUARE CONTRIBUTION BY CELL
          .0003    .0236   .1972   .1626

          .0008    .0586   .4907   .4045

         TOTAL CHI SQUARE = 1.3375
.PAGE
            FREQUENCY DISTRIBUTION, PERCENTILES, and T-SCORES

     This program produces a frequency distribution and cumulative
frequency distribution of entered scores, and calculates a percentile
rank and T-score value for each score interval with a frequency greater
than zero.  The output includes the mean, the standard deviation (n-1)
of the distribution and the computed median.  The program will accommodate
up to 500 data points.  Options are included for storage of data on
disk, and for data verification.  The following is an example:
**********
     FREQUENCY DISTRIBUTION, PERCENTILES, AND T-SCORES PROGRAM

     ENTER YOUR SAMPLE SIZE
     ? 12
     DOES YOU DATA ALREADY EXIST IN A FILE     ?  NO
     ENTER YOUR DATA, ONE SCORE ON A LINE.
     ? 45
     ? 44
     ? 43
     ? 45
     ? 46
     ? 43
     ? 48
     ? 46
     ? 39
     ? 44
     ? 43
     ? 45
     DO YOU WISH TO VERIFY YOUR INPUT DATA OR CHANGE ANY SCORES?
     ? NO
***********
                              CUMULATIVE    PERCENTILE
      SCORE     FREQUENCY     FREQUENCY       RANK          T-SCORE

      48.00         1             12            96             67
      46.00         2             11            83             58
      45.00         3              9            63             53
      44.00         2              6            42             49
      43.00         3              4            21             44
      39.00         1              1             4             26


      THE MEAN =                        44.2500
      STANDARD DEVIATION =               2.2208
      COMPUTED MEDIAN =                 44.5000

      DO YOU WISH TO SAVE YOUR DATA      ?  NO
.PAGE
               MEANS, STANDARD DEVIATIONS, AND STANDARD ERRORS

     This program computes means, standard deviations and standard errors
for up to 10 variables (scores per case) at a time.  There is a limit of
250 cases per variable.  Data may be permanently stored by a routine that
creates a file at the user's request at the end of the program.  Options
are included for verification and correction of the data.

         MEAN, STANDARD DEVIATION AND STANDARD ERROR PROGRAM

         ENTER NUMBER OF VARIABLES TO BE RUN:
         ? 2
         HOW MANY CASES OR SCORES FOR EACH VARIABLE?
         ? 10
         DOES YOUR DATA ALREADY EXIST IN A FILE   ? NO
***********

         ENTER 2 SCORES FOR EACH CASE ON A LINE.
         ? 23,1.45
         ? 28,2.31
         ? 26,5.54
         ? 25,-4.98
         ? 31,-8.01
         ? 25,-7.77
         ? 24,10.33
         ? 19,-3.25
         ? 26,.25
         ? 30,-1.21

         DO YOU WANT TO VERIFY YOUR DATA   ? NO

         DO YOU WANT TO CHANGE ANY SCORES   ? NO

         VAR           MEAN           STD DEV             STD ERROR

          1           25.7000         3.4657                1.0960
          2           -0.5340         5.8062                1.8361
.PAGE
              MEAN AND STANDARD DEVIATION USING GROUPED DATA

     This program computes the means and standard deviation from a
frequency distribution with no limit on the number of intervals.
The data from the frequency distribution may be entered in any order
providing that the midpoint of an interval is followed immediately by
the score frequency for that interval (the two numbers being separated
by a comma, of course.)  After the last interval and frequency has been
entered, the user must enter an additional interval of zero and a
frequency of zero.  The following is an example of its use:

     MEAN AND STANDARD DEVIATION FROM GROUPED DATA PROGRAM

     ENTER YOUR DATA.   MIDPOINT OF INTERVAL AND FREQUENCY ON A LINE.
     AFTER YOUR LAST SCORE, ENTER A ZERO FOR INTERVAL AND
                                A ZERO FOR FREQUENCY.
     ? 1.5,3
     ? 3.5,5
     ? 5.5,11
     ? 7.5,8
     ? 9.5,4
     ? 0,0
***********

            SAMPLE SIZE=                        31
            SUM OF SCORES=                180.5000
            SUM X SQ=                    1211.7500
            SUM x SQ=                     160.7740
            THE VARIANCE (N-1) =            5.1863

            THE MEAN=                       5.8226
            STANDARD DEVIATION (N-1) =      2.3150
            THE STANDARD ERROR=             0.4158
.PAGE
                  PRODUCT MOMENT CORRELATION PROGRAM

     This program computes the means, standard deviations, and the
Pearson product moment correlations for from two to ten variables.
The program is limited to a total of 1500 data points.  Options are
included for user correction of  input data and for permanent storage
of data in a user file.  The following is an example of its use:

CORRELATION PROGRAM

HOW MANY VARIABLES?
? 4
HOW MANY CASES?
?9
DOES YOUR DATA ALREADY EXIST ON A FILE   ? NO
***********

ENTER 4 SCORES FOR EACH CASE ON A LINE.
? 21,23,6,3
? 23,21,7,5
? 10,22,2,4
? 15,27,0,5
? 19,21,7,6
? 25,19,0,7
? 27,25,3,8
? 23,54,8,5
? 25,31,4,7

DO YOU WANT TO VERIFY YOUR DATA  ? NO
DO YOU WANT TO CHANGE ANY SCORES ? NO
DO YOU WANT REGRESSION COEFFICIENTS AS PART OF THE OUTPUT ? NO
***********

VARIABLE            MEAN          STD DEV           STD ERROR

    1              20.889          5.4416             1.8139
    2              27.000         10.7587             3.5862
    3               4.111          3.0596             1.0199
    4               5.556          1.5899             0.5300

 VR   VR  COEFFICIENT

  1   2    0.1729
  1   3    0.2411
  1   4    0.6293
  2   3    0.4101
  2   4   -0.0512
  3   4   -0.2712
.PAGE
                      SPEARMAN RANK CORRELATION

     This program calculates the Spearman Rho from paired rankings for
individuals on two variables.  The user must produce the rankings by
sorting cases into rank order and numbering them '1' (highest) to 'N'
lowest on each variable.  The program includes a verification routine
that allows the user to display and correct rankings before calculation
of the correlation coefficient.  The output of the programs includes a
T-value that may be used to test the significance of the correlation
coefficient.  The following is an example of its use:

         SPEARMAN RANK CORRELATION PROGRAM

         HOW MANY RANKS DO YOU HAVE?
         ? 10
**************
         ENTER TWO RANKS FOR EACH CASE ON A LINE.
         ? 7,5
         ? 4,2
         ? 2,1
         ? 9,8
         ? 1,4
         ? 6,3
         ? 3,6
         ? 10,10
         ? 5,7
         ? 8,9

         DO YOU WISH TO VERIFY YOUR DATA  ?  NO
         DO YOU WISH TO CHANGE ANY VALUES ?  NO
***********

                RANK ON         RANK ON
         CASE  VARIABLE A      VARIABLE B     DIFFERENCE    DIFF SQUARED
            1      7.0             5.0             2.0            4.00
            2      4.0             2.0             2.0            4.00
            3      2.0             1.0             1.0            1.00
            4      9.0             8.0             1.0            1.00
            5      1.0             4.0            -3.0            9.00
            6      6.0             3.0             3.0            9.00
            7      3.0             6.0            -3.0            9.00
            8     10.0            10.0             0.0            0.00
            9      5.0             7.0            -2.0            4.00
           10      8.0             9.0            -1.0            1.00

         SPEARMAN RHO =   0.7455               T-VALUE =   3.1632

.PAGE
                     ONE-WAY ANALYSIS OF VARIANCE

     This program performs a one-way analysis of variance on from two to
ten groups.  The program is limited to 1000 data points.  Options are
included for verification of input data and data storage in a file.
The following is an example of its use:

         ONE-WAY ANALYSIS OF VARIANCE PROGRAM

         HOW MANY GROUPS?
         ? 2
         DO YOU HAVE THE SAME SAMPLE SIZE FOR EACH GROUP?   YES
         ENTER YOUR SAMPLE SIZE  ? 5
         DOES YOUR DATA ALREADY EXIST ON A FILE   ? NO
***********
         ENTER SCORES FOR GROUP 1--ONE SCORE ON A LINE
         ? 435
         ? 498
         ? 425
         ? 500
         ? 525
         ENTER SCORES FOR GROUP 2--ONE SCORE ON A LINE
         ? 501
         ? 513
         ? 490
         ? 520
         ? 530
         DO YOU WISH TO VERIFY YOUR DATA   ? NO
         DO YOU WISH TO CHANGE ANY SCORES  ? NO
***********
         GROUP     CASES      MEAN   STD DEV    STD ERROR

            1         5    476.600   43.9920    19.67380
            2         5    510.800   15.7067     7.02424
         TOTAL       10    493.700   35.9816    11.37839

         ANOVA SUMMARY   SUM OF SQUARES   D.F.    MEAN SQ.      F RATIO

         BETWEEN GROUPS     2924.10        1     2924.1000      2.6802
         WITHIN GROUPS      8728.00        8     1091.0000

         TOTAL             11652.10        9

.PAGE
              ONE FACTOR REPEATED MEASURES ANALYSIS OF VARIANCE

     This program performs an analysis of variance over treatments using
a 'within subjects' design, i.e. each subject being measured under every
treatment.  The program is limited to 1000 data points (10 treatments by
100 subjects).  Data for the analysis must be entered in subject-treatment
order, i.e. all of the treatment scores in order for the first subject,
then treatment scores in the same order for the second subject, and so on.
The following is an example of the program:

ONE FACTOR REPEATED MEASURES ANALYSIS OF VARIANCE PROGRAM

HOW MANY TREATMENTS?
? 3
ENTER YOUR SAMPLE SIZE
? 4
DOES YOUR DATA ALREADY EXIST IN A FILE?   NO
***********
ENTER YOUR TREATMENT SCORES FOR EACH SUBJECT ON A LINE.
? 17,20,22
? 18,24,25
? 20,19,26
? 16,18,17
DO YOU WANT TO VERIFY YOUR INPUT DATA  ?  NO
DO YOU WANT TO CHANGE ANY SCORES       ?  NO

SUBJECT  1       MEAN SCORE=  19.6667
SUBJECT  2       MEAN SCORE=  22.3333
SUBJECT  3       MEAN SCORE=  21.6667
SUBJECT  4       MEAN SCORE=  17.0000

TREATMENT 1       MEAN SCORE=  17.7500
TREATMENT 2       MEAN SCORE=  20.2500
TREATMENT 3       MEAN SCORE=  22.5000

TOTAL           MEAN SCORE=  20.1667
***********

SOURCE       SUM OF SQRS      DF         MEAN SQ          F RATIO

TREATMENTS        45.17        2          22.583          5.0497
SUBJECTS          51.67        3          17.222          3.8510
INTERACTION       26.83        6           4.472

TOTAL            123.67       11          11.242

.PAGE
          FACTORIAL ANALYSIS OF VARIANCE, 2 INDEPENDENT FACTORS

     This program performs an analysis of variance over two independent
factors, with up to ten levels for each factor.  However, the program is
limited to a total of 1000 data points.  The program permits verification
and correction of input data.  An option is included for storage of data
in a virtual array file.  The following is an example of its use:

FACTORIAL ANALYSIS OF VARIANCE, 2 INDEPENDENT FACTORS

HOW MANY LEVELS FOR FACTOR A  ? 2
HOW MANY LEVELS FOR FACTOR B  ? 2
DO YOU HAVE THE SAME NUMBER OF CASES IN EACH CELL  ? YES
ENTER YOUR CELL SIZE  ?  4
DOES YOUR DATA ALREADY EXIST ON A FILE?   NO
***********
ENTER 4 SCORES FOR CELL (1,1) - ONE SCORE ON A LINE
? 2
? 4
? 3
? 3
ENTER 4 SCORES FOR CELL (1,2) - ONE SCORE ON A LINE
? 6
? 5
? 4
? 5
ENTER 4 SCORES FOR CELL (2,1) - ONE SCORE ON A LINE
? 5
? 4
? 5
? 4
***********
ENTER 4 SCORES FOR CELL (2,2) - ONE SCORE ON A LINE
? 7
? 6
? 5
? 6

DO YOU WANT TO VERIFY YOUR INPUT DATA   ?  NO
DO YOU WISH TO CHANGE ANY SCORES    ?   NO

  CELL     CASES      MEAN      S.D.
A1   B1       4       3.000     0.8165
A1   B2       4       5.000     0.8165
A2   B1       4       4.500     0.5774
A2   B2       4       6.000     0.8165

TEST OF HOMOGENIETY OF VARIANCE    F =  1.4142
***********

FACTOR A   CASES      MEAN
LEVEL 1       8       4.000
LEVEL 2       8       5.250

FACTOR B   CASES      MEAN
LEVEL 1       8       3.750
LEVEL 2       8       5.500

TOTAL        16       4.625

SOURCE     SUM OF SQUARES   D.F.   MEAN SQUARE       F RATIO
FACTOR A      6.2500          1       6.2500         10.7143
FACTOR B     12.2500          1      12.2500         21.0000
A x B         0.2500          1       0.2500          0.4286

WITHIN        7.0000         12       0.5833
TOTAL        25.7500         15
.PAGE
                      THE CONFIDENCE INTERVAL TEST

     This program utilizes the mean square for error from the analysis of
variance test to compute a confidence interval for each pair of means
employed in the analysis of variance.  The user must supply the width of
the interval by entering from a table of F the value given at a particular
level of probability for his degrees of freedom.   The program will
accomodate up to ten means.  The sample sizes need not be equal.  The
following is an example of its use with five samples of equal size and
a .05 interval:

         CONFIDENCE INTERVAL PROGRAM

         ENTER DESIRED F-VALUE FROM TABLE ON F.
         ? 2.52
         ENTER THE MEAN-SQUARE-FOR-ERROR FROM YOUR ANALYSIS OF VARIANCE
         ? 1.8312
         HOW MANY SAMPLE MEANS WERE THERE IN YOUR ANALYSIS OF VARIANCE
         ? 5
***********
         ENTER THE SAMPLE SIZE AND MEAN FOR EACH SAMPLE ON A LINE
         ? 10,1.10
         ? 10,1.50
         ? 10,7.20
         ? 10,23.50
         ? 10,23.90

                             CONFIDENCE INTERVAL
         MEAN      MEAN       FROM         TO        SIGNIFICANT

           1         2     -1.5214      2.3214           NO
           1         3      4.1786      8.0214           YES
           1         4     20.4786     24.3214           YES
           1         5     20.8786     24.7214           YES
           2         3      3.7786      7.6214           YES
           2         4     20.0786     23.9214           YES
           2         5     20.4786     24.3214           YES
           3         4     14.3786     18.2214           YES
           3         5     14.7786     18.6214           YES
           4         5     -1.5214      2.3214           NO

.PAGE
                   T-TEST FOR CORRELATED MEANS PROGRAM

     This program computes the student's t-value for a group's difference
in scores at two time intervals.  Each variable or type of score is
entered as a separate problem.  The program accommodates up to 250 pairs
of scores and includes a routine for storage of data on a file as well
as routines for verification and correction of data.  The following is
an example of its use.

         T-TEST FOR CORRELATED MEANS PROGRAM

         ENTER YOUR SAMPLE SIZE
         ? 8
         DOES YOU DATA ALREADY EXIST IN A FILE    ?  NO
***********
         ENTER YOUR SCORES, ONE PAIR ON A LINE.
         ? 34,44
         ? 38,42
         ? 35,35
         ? 33,38
         ? 39,34
         ? 40,46
         ? 39,40
         ? 41,43

         DO YOU WANT TO VERIFY YOUR DATA   ?  NO
         DO YOU WANT TO CHANGE ANY SCORES  ?  NO

                               TIME 1        TIME 2      DIFFERENCE

         MEANS                 37.375        40.250          2.875
         STD. DEV.             2.9731        4.3012         4.4861
         STD. ERRORS          1.05115       1.52069        1.58607
         T-VALUE                                           1.81265
.PAGE
                   T-TEST FOR INDEPENDENT MEANS

     This program computes the student's t-value for two independent
samples.  Each variable or type of score is entered as a sample.
The program accommodates up to 250 scores and includes a routine for
storage of data on a disk and data verification.  The following is
an example of its use.

             T-TEST FOR INDEPENDENT MEANS

         DOES YOUR DATA ALREADY EXIST IN A DISK FILE?  n

         HOW MANY SAMPLES ARE TO BE ENTERED?  2

         ARE YOUR SAMPLES OF EQUAL SIZE?      Y

         ENTER YOUR SAMPLE SIZE -       6

*****************

         ENTER YOUR DATA FOR SAMPLE 1 -- ONE VALUE OR SCORE ON A LINE.
         1         ? 42
         2         ? 38
         3         ? 36
         4         ? 40
         5         ? 40
         6         ? 38

         ENTER YOUR DATA FOR SAMPLE 2 -- ONE VALUE OR SCORE ON A LINE.
         1         ? 36
         2         ? 35
         3         ? 37
         4         ? 38
         5         ? 36
         6         ? 37

         DO YOU WISH TO VERIFY OR CHANGE YOUR DATA?  N

***************


      GROUP     CASES          MEAN       STD DEV      STD ERROR

        1         6          39.000        2.0976        0.85635
        2         6          36.500        1.0488        0.42817



      SAMPLE VS SAMPLE    DIFFERENCE      T-VALUE          DF

        1         2           2.5000       5.4031          10


.PAGE