PCjs Machines

Home of the original IBM PC emulator for browsers.


PC-SIG Diskette Library (Disk #863)

[PCjs Machine "ibm5150"]

Waiting for machine "ibm5150" to load....

Information about “STATMATE/PLUS 3 OF 3 (ALSO 861, 862)”

A statistical analysis package for handling numerical data, operated by
entering one-line commands and subcommands.  Command ``batch'' files can
be created for automatic execution, along with explanatory screen

STATMATE operates on information contained in a database, generated by
the program.  A user ID is required before entering a database, and for
every new user ID, an empty database is created.  This feature permits
multiple users to work with STATMATE while keeping the data files

Extract data from an ASCII text file and load it into the database for
operation.  Data is stored in columns and rows, and you can extract
portions of the data according to your specifications.  As you
manipulate the data, the results can be displayed on the screen,
printed, or saved on a disk file.

The main analytic features are elementary statistics, scatter plots,
cross tabulations, histograms, data comparison using the T-Test,
correlation, arithmetic operations, distribution functions, curvilinear
regression, multiple regression, nonlinear regression, data recoding,
and data transformation and manipulation.  An on-line help facility is
included to give you a detailed description of all the STATMATE


Disk No  863
Program Title:  STATMATE/PLUS version 1.3 (Disk 3 of 3)
PC-SIG version 1.1
This is the third disk of the STATMATE package, disks #861-63, and
contains the five-part documentation for the program.  Please refer to
disk #861 for full information.
Usage: Statistics Analysis
System Requirements: 128K memory and two disk drives.
Suggested Registration: $50.00
File Descriptions:
SMPART1  DOC  Documentation, part 1.
SMPART2  DOC  Documentation, part 2.
SMPART3  DOC  Documentation, part 3.
SMPART4  DOC  Documentation, part 4.
SMPART5  DOC  Documentation, part 5.
README        How to get started.
1030D E Duane Avenue
Sunnyvale Ca. 94086
(408) 730-9291
(c) Copyright 1987,88 PC-SIG Inc.


║            <<<<  Disk #863 STATMATE/PLUS (Disk 3 of 3)  >>>>            ║
║ To copy the documentation to your printer, Type:                        ║
║                        PRINTDOC (press enter)                           ║



                            (A Statistical Package)

                                  Version 1.3

                            Shareware User's Guide

                                       August 1, 1988

                                       The Software Hill
                                       1857 Apple Tree Lane
                                       Mountain View, Ca.  94040
                                       Copyright (C), 1987


         The  STATMATE/PLUS   statistical   application   package   is
         copyrighted (C) 1987,  by  The  Software  Hill.   All  rights
         reserved.  Non-registered users are granted a limited license
         to use this product on a trial basis, and to copy the program
         for trial use by others subject to the following limitations:

                 1.  STATMATE/PLUS is distributed in unmodified
                     form, complete with documentation.

                 2.  No fee, charge or other consideration is
                     requested or accepted.

                 3.  STATMATE/PLUS is not distributed in
                     conjunction with any other product.

         If you intend to use STATMATE/PLUS on a regular basis, please
         show  your  support  by registering the program for a nominal
         fee.  Registration information is  give  below.   Commercial,
         business  or  governmental  use  by  non-registered  uses  is

         If you are interested in multiple copies  for  use  at  work,
         site  and corporate licenses are available.  Please write for


         STATMATE/PLUS is a trademark of The Software Hill.

                               TABLE OF CONTENTS

         USER-SUPPORTED SOFTWARE.........................2
         PRODUCT SUPPORT.................................3
         INTRODUCTION TO STATMATE/PLUS...................4
         STATMATE Example................................7
         Command Summary.................................9
         INTERACTING WITH STATMATE.......................11
         Commands and Subcommands........................11
         STATMATE DATABASE CONCEPTS......................14
         STATMATE Database and Directory.................14
         Organization and Manipulation of Data...........14
         Variable Names..................................15
         Ways of Referencing Data--Variables and Cases...16
         ENTERING DATA INTO THE SYSTEM...................17
         External Data Entry--Files......................17
         Creating ASCII Files............................17
         COMMAND DESCRIPTIONS............................21
         APPENDIX A: Computation Methods.................69
         APPENDIX B: Sample Data.........................70
         APPENDIX C: Installation and Miscellanea........73
         APPENDIX D: STATMATE Size Limitations...........77
         APPENDIX E: HELP................................79
         APPENDIX F: Suggested Diskette Organization.....81
         APPENDIX G: Invoice and Order Form..............82
         References .....................................85

         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         Feedback on STATMATE/PLUS is an important part of  developing
         a  useful and successful software package.  Please share your
         impressions, suggestions and comments by writing to us.

         STATMATE/PLUS is distributed as User-Supported Software.  You
         are encouraged to try the program  and  share  it  with  your
         friends and colleagues as long as:

                 1.  STATMATE/PLUS is distributed in unmodified
                     form, complete with documentation.

                 2.  No fee, charge or other consideration is
                     requested except by The Software Hill.

                 3.  STATMATE/PLUS is not distributed in con-
                     junction with any other product.

         If you use STATMATE/PLUS on a regular basis, please show your
         support  by  registering  the  program.   You may register by
         sending a check or money order for $45 to:

                            The Software Hill
                            1857 Apple Tree Lane
                            Mountain View, Ca. 94040

         Registered  users  will  receive    (1) notification of major
         releases of STATMATE/PLUS, newsletters and other  information
         supporting the package and (2) two sort  utilities and a high
         resolution scatter  plot  utility  (for use with  CGA, EGA or
         Hercules graphics cards).  Program disks are not included  in
         the registration fee. Note that when you register you receive
         a $10 coupon applicable to additional purchases.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                            USER-SUPPORTED SOFTWARE

         User-supported  software  is  a  means  for  users to receive
         quality software while directly supporting software  authors.
         It is based on the ideas that:

                 a.  Immediate assessment of the package through
                     hands-on use to detemine whether the package
                     satisifies the user's personal application
                     needs and operational tastes.

                 b.  Creation and support of independent personal
                     computer software is important and desirable
                     by interested application users.

                 c.  Copying of programs should be encouraged,
                     rather than restricted to promote the widest
                     possible development, interest and support
                     by the application's community.

         Under  the  concept  of  user-supported  software, anyone may
         request a copy of  STATMATE/PLUS  by  sending  a  blank,  DOS
         formatted,  5-1/4  inch  diskette  to The Software Hill along
         with a self-addressed, postage-paid return mailer.  You  will
         receive  STATMATE  and  program  documentation on the disk by
         return mail.

         The program carries a  notice  suggesting  registration,  but
         registration is strictly voluntary.

         You  are  encouraged  to  copy   and   distribute   STATMATE,
         regardless of whether or not you register,  for  private  and
         non-commercial use of others.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                                PRODUCT SUPPORT

         As of this date, we are able to provide support of registered
         owners  by  mail  or  phone.   A bulletin board system is not
         presently   available.   In  order  to  answer  questions  or
         comments,  please  provide  as complete a description of your
         problem   as   possible.    Include  a  description  of  your
         configuration  (hardware  and  operating system), steps taken
         before   a   problem  occurrence  and  any  printed  material
         identifying the problem.

         The  latest  version  of STATMATE/PLUS may always be found on
         the PC SIG bulletin board system.  As the popularity  of  the
         program  grows,  it  will  be  found  on  your local bulletin
         boards, shareware distribution disks and a  number  of  other
         computer environments.

                             MACHINE REQUIREMENTS

         STATMATE/PLUS  requrires  128K  of  memory (RAM).  It is best
         operated from a hard disk but may  be  operated  from  two  5
         1/4-inch  floppies.   See  the  appendix  for  suggestions on
         tailoring STATMATE to your system.   It  operates  under  DOS
         version  2.0 or higher.  STATMATE may be used on a PC, XT, AT
         or compatible.   There  is  no  dependence  no  the  type  of
         terminal  used,  whether monochrome, composite or color.  Any
         terminal type is satisfactory.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         This guide briefly describes  some  of  the  capabilities  of
         STATMATE/PLUS.   Enough descriptive material is given in this
         guide   so   that  you  should  be  able  to  understand  the
         capabilities  of  the package and put STATMATE to use in your
         applications.  For registered owners,  a  complete  guide  to
         STATMATE/PLUS is available for $35.

         Probably  the  best  place  to  start  is by reading over the
         material in this guide.   When  you  are  ready  to  try  the
         program,  read  the  section  on operating STATMATE.  Try the
         example discussed there and read  appendix  F  regarding  the
         suggested  organization  of  STATMATE  program files on disk.
         Once you have completed the example, try the package with the
         EXECUTE command on the DEMO file.  After  entering  STATMATE,
         give  the three character ID required, just enter in response
         to the command prompt:

                         EXECUTE DEMO

         This will  cause  STATMATE  to  run  through  a  sequence  of
         commands.  A description of what is happening is given as the
         program  proceeds  through  the  commands.  The demonstration
         will pause and give you the chance to read what is  happening
         at your own pace.

         The command summary given later indicates commands which will
         operate in the demonstration program.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         A  brief  summary  of  some  of the many STATMATE operational
         features are given below.

                             -Operational Features-

                  * Help facility
                  * Data extraction from external files
                  * Placement of output on results files
                  * User named variables
                  * Default names for variables
                  * Data transformation and manipulation
                  * Display of selected data
                  * Missing or not applicable data values
                  * Multiple user operation
                  * Database maintainence operations
                  * Data selection by specified conditions
                  * ASCII output files to other applications

         The analytic features available with the STATMATE package are
         given below.

                             -Statistical Features-

                  * Elementary statistics        * Scatter plots
                  * Cross tabulations            * Histograms
                  * T-Test                       * Correlation
                  * Curvilinear regression       * Random Number Generation
                  * Arithmetic operations        * Distribution functions
                  * Multiple regression          * Group statistics
                  * One-way ANOVA                * Two-way ANOVA
                  * Control chart calculations   * Nonparametric methods
                  * Nonlinear regression         * Polynomial regression
                  * Data recoding


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         Operation of STATMATE is begun by entering:


         You will be prompted with the message:

              ENTER ID-

         The program expects a three alphabetic*, upper or lower  case
         letters,  character  identification  as  a   response.    The
         characters are  used  as  a  database  identification.   Your
         initials are usually the simplest ID to use.  For example,

              ENTER ID-XYZ

         In  this  case XYZ is used as an ID and XYZ is the identified
         database.  Use of another ID would identify another database.
         This mechanism allows you to create databases  for  different
         purposes,  for  example,  one  might belong to your data, and
         another to a colleague.  (IMPORTANT:  Each time you supply  a
         different  ID,  STATMATE  creates  an  empty  database   file
         corresponding to the ID you supplied.   These  files  may  be
         large.   It  is  best  to  use  the same ID each time you use
         STATMATE or you will quickly exceed your disk capacity.   See
         the  use of databases in section STATMATE DATABASE CONCEPTS.)

         After you provide the ID, the program will issue the  prompt:


         At this point, a STATMATE command must be entered in order to
         continue  the  operation of STATMATE.  A carriage return must
         be entered at the completion of each line  of  input.   After
         each  command is completed, another 'Command:' prompt will be
         issued.  Entering EXIT will terminate the program.

         * By supplying a fourth character, q or Q, with the  ID,  the
         shareware  banner output after entering the ID is suppressed.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         The following sample illustrates the operation  of  STATMATE:

                     1860 ,  6.217 , 25.227
                     1870 ,  9.902 , 28.656
                     1880 , 14.130 , 36.026
                     1890 , 22.106 , 40.841
                     1900 , 30.160 , 45.835
                     1910 , 41.999 , 49.973
                     1920 , 54.158 , 51.553
                     1930 , 68.955 , 53.830
                     1940 , 74.924 , 57.246
                     1950 , 88.927 , 61.770

         The data used in the example problem is the U. S.  Population
         data  for  rural and urban areas from 1860 through 1950.  The
         data is on a file  called  USPOPDEM.DAT  (a  portion  of  the
         USPOP.DAT  file supplied with STATMATE) and consists of three
         fields:    year,   urban  population  and  rural  population;
         population data is in millions.

         The  example of STATMATE operation is shown on the next page,
         and the description of the operation  is  described  in  this
         paragraph.   From  the  next  page,  the  ENTER ID- prompt is
         answered with ABC.  This establishes the user's ABC  database
         as  the  database  which is to be used in the example.  Next,
         the ERASE command clears all data from the database.  Data is
         then extracted from the data file USPOPDEM.DAT by  the  INPUT
         command.   In the INPUT command, the clause OMIT 2 causes the
         first two fields of the data file to be ignored.  The  clause
         KEEP  1  causes  the  third  field,  rural  population  to be
         extracted.  Hence, only one variable, that is, one column  or
         field  of data, containing rural population, is placed in the
         database.  STATMATE only  operates  on  data  placed  in  its
         database.   There  are  10  data  points  or  cases  for  the
         extracted variable as is reported by STATMATE as 10 CASES  at
         completion of the INPUT command.  Initially this variable has
         the name #1 assigned to it.  The GIVE NAME command is used to
         give  #1 an alternate name, RURALPOP.  The STATISTICS command
         is applied  to  RURALPOP  to  derive  the  simple  statistics
         produced  by  the  command.  The program is terminated by the
         EXIT command.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         ENTER ID-ABC

         Command:  ERASE #1 THRU END

           10 VARIABLES ERASED

         Command:  INPUT USPOPDEM.DAT OMIT 2,KEEP 1

             1 VARIABLE INPUT AT #1
             10 CASES

         Command:  GIVE NAME #1,RURALPOP


         Command:  STATISTICS RURALPOP


            10 CASES
             0 MISSING

         CENTRAL TENDENCY       SPREAD                      DISTRIBUTION
         -------------------    ------------------------    --------------------
         MEAN          45.10    STD. DEV.          12.17    MINIMUM        25.23
                                VARIANCE          148.15    MAXIMUM        61.77
                                RANGE              36.54
                                COEFF. VAR.         0.27

         SUMMATIONS                 HIGHER MOMENTS
         -----------------------    --------------------
         TOTAL            450.96    SKEWNESS       -0.37
         SUM SQ         21669.59    KURTOSIS        1.93
         SUM SQ(DEV)     1333.37

         Command:  EXIT



         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command Summary

         Commands are the basic form of communicating  with  STATMATE.
         A  summary  of  the commands available for both STATMATE/PLUS
         and STATMATE are shown below.

                   Summary of STATMATE/PLUS Commands

         Command Name    Description             Modifiers/Descriptors
         ------------    --------------------    ----------------------
         BREAKDOWN       Statistics by groups    none
         CHART           X- and R-Charts         TYPE, CONTROL, KCENTER,
                                                 HMARK, VRANGE, TITLE,
                                                 HLABEL, KSIGMA, DISPLAY,
                                                 HFILLER, VPOSITION,
         COMPUTE         Fit and forecast        TYPE
         CORRELATE       Pair-wise correlation   none
         CROSSTABS       Two-way cross tabs      none
         CURVE           Ten curve fits          TABLE, BEST, EQUATION
         CUSUM           Cusum chart             TARGET, DISPLAY, RESET,
                                                 HMARK, HFILLER, VRANGE,
                                                 VPOSITION, TITLE,

         EDIT            Database editing        none
         ELSE            Reverse WHEN condition  none
         END             Remove WHEN condition   none
         EXIT            End STATMATE operation  none
         ERASE           Remove database data    none
         EXECUTE         Multiple command entry  none
         GIVE            Give data attributes    none
         HELP            Provide command help    none
         HISTOGRAM       Histogram               TITLE, RANGE, BARS,
         INPUT           Data input              KEEP, OMIT
         KOLMOGOROV      Kolmogorov tests        DISTRIB, SPARAM, UPARAM
         LET             Arithmetic operations   none
         NONLINEAR       Nonlinear regression    MODEL, MAXITER, REPORT,
                                                 TYPE, CONVERGE
         ONEWAY          One-way ANOVA           METHOD,ALPHA
         ONPARAM         Nonparametric methods   none
         PLOT            Scatter plot            TITLE, HRANGE, VRANGE,
                                                 HPOS, VPOS, HLAB, VLAB
         POLYNOMIAL      Polynomial regression   TABLE,
         PRINT           Data display            none
         QUERY           Database status         none
         RCORRELATION    Rank correlation        TEST
         RECODE          Recode crosstab data    none
         REGRESSION      Multiple regression     TABLE, INTERCEPT, DURBIN
         REMARK          Allows documentation    none


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                   Summary of STATMATE/PLUS Commands (Continued)

         Command Name    Description             Modifiers/Descriptors
         ------------    --------------------    ----------------------
         SET             Set output type         COPY
         SHOW            Show internal status    none
         STATISTICS      Summary statistics      TABLE
         STEPWISE        Stepwise regression     TABLE, MAXSTEP, FORCE,
                                                 FENTER, FREMOVE, METHOD
         TNPARAM         2-way nonparam ANOVA    TEST
         TTEST           Student T-test          none
         TWOWAY          Two-way ANOVA           DESIGN
         WHEN            Select database view    none
         WRITE           Put variables to file   none


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         Commands and Subcommands

         The primary means by which you communicate with  STATMATE  is
         through  commands.   For  example,  the  PRINT  command tells
         STATMATE to list data which is specified in the command.   In
         some instances, a command requires the use of a subcommand to
         enter additional information about the operation requested by
         the  command.   In  STATMATE,  subcommands are available, for
         example, with the CHART, CUSUM, STEPWISE and  PLOT  commands.

         Information  related  to specific commands and subcommands is
         found in the command description portion of this manual.   If
         you  need  help  while  actually  entering   a   command   or
         subcommand, an on-line HELP command is available.


         A command contains a command name  and  a  reference  to  the
         variables that it is to operate on.  For example,


         calculates statistics for the variable URBANPOP, representing
         urban  population data.  A command name may be abbreviated by
         using the first three  characters  of  its  name.   That  is,
         STA URBANPOP is an acceptable command.

         Some commands operate on several variables.  For example, the
         scatter  plot  command, PLOT, requires variables for plotting
         on the vertical and horizontial plot axis.  In order to  help
         distiguish  between the use of variables, some commands use a
         keyword which essentially divides the list of variables  into
         easily identifiable pieces.  For example, in


         ON is a keyword that separates the list of vertically plotted
         variables,  SALES  and  WAGES, from the horizontially plotted
         variable, YEAR.

         Another type of item used in a  command  is  a  modifier.   A
         modifier  supplies  some additional information about how the
         command should operate.  For example,


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         TITLE is a modifier indicating that the  title  shown  should
         appear  on  the  scatter  plot.  Modifiers are followed by an
         equal sign (=) and are separated from the list  of  variables
         and from one another by commas.  For example,


         Some  modifiers contain option names following the equal sign
         to designate  which  options  are  to  be  selected  for  the
         modifier.   For  example,  TABLE=FIT,PARAMETERS indicates the
         FIT  and  PARAMETERS  options  are  selected  for  the  TABLE
         modifer.   As  with  command  names,  only  the  first  three
         characters of a modifier or option name need be used.

         Although some commands contain modifiers,  the  modifiers  do
         not  need  to  be  entered.   If a modifier is not entered it
         assumes a default value.  For example, if the TITLE  modifier
         is  not given for PLOT, the title is assumed to be blank.  In
         most instances, you cannot specify your own  default  values;
         however, with a few commands, STEPWISE and PLOT, for example,
         you  are  allowed  to change the default settings.  This is a
         very useful feature.  For example, in the event that you  use
         the  same  scatter plot title for much of your work, the plot
         title can be set and not changed until necessary.


         Subcommands are used to specify additional operations  for  a
         command.   Subcommands  are  available only with a few of the
         commands.  Usually a subcommand permits you an alternate  way
         of  entering  information about modifiers.  Other subcommands
         permit   the  values  of  current  modifier  settings  to  be
         displayed or saved.

         Entering and Using Commands and Subcommands

         STATMATE will prompt you with a message to enter a command or
         subcommand.  In the case of a command, the prompt is:


         To  cause  STATMATE to execute a command, you only need enter
         the command as in:

         Command:  PLOT URBANPOP ON YEAR

         In some instances, the text of a command may be so long  that
         it  will not fit on a single line.  Commands may be continued


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         by breaking them after a comma.  That  is,  anywhere  that  a
         comma is allowed in a command is a point at which the command
         may  be  broken.   When  a  line  is  terminated  by a comma,
         STATMATE will ask you for an additional line of text with the
         following prompt


         For example,

              Command:  PLOT URBANPOP ON YEAR,
              Continue:  TITLE='URBAN POPULATION FROM 1790 TO 1950'

         As many as 250 characters  of  text  may  be  included  in  a

         Command  names,  keywords and other elements of a command may
         be entered in either upper or lower case letters.

         In the case of the PLOT and STEPWISE commands,  for  example,
         it  is  not  necessary to enter the information following the
         command name in response to the command prompt.  An alternate
         method is available  that  some  users  may  find  easier  is
         available.   For  example,  if  only the command name PLOT is
         entered, STATMATE will then prompt you for the names  of  the
         variables  to be plotted.  After the names have been entered,
         you  will  be  placed  in  the  subcommand  mode.   With  the
         subcommands,  you  may enter any of the PLOT modifiers.  When
         you  have  entered  the  modifiers  you  need,  entering  the
         CONTINUE  subcommand  causes  the PLOT command to be executed
         and a plot to be produced.  The following is  a  sequence  of
         commands  and  subcommands  used  to  enter  the PLOT command
         discussed above.

         Command:  PLOT

         Enter Y-axis variables:  URBANPOP

         Enter X-axis variable:  YEAR

         Subcommand: TITLE='URBAN POPULATION FROM 1790 TO 1950'

         Subcommand: CONTINUE

         Note that  TITLE  and  CONTINUE  are  subcommands.   CONTINUE
         causes  STATMATE to leave the subcommand mode and execute the
         PLOT command.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         STATMATE Database and Directory

         An important concept in the  operation  of  STATMATE  is  the
         STATMATE  database.   This  is  a file containing the data on
         which the program operates, and  is  generated  by  STATMATE.
         Data  is  placed  on the file by the INPUT command and may be
         displayed by the PRINT command.

         A user identifies which of his databases he wants to  operate
         on when responding to the ID prompted by the program.  The ID
         is  associated  with  a  database.   Usually  one database is
         sufficient for most applications.  Since these database files
         can be occupy a lot of disk space, some care should be  taken
         in utilizing a number of different IDs.  An empty database is
         created  each  time  a  new  ID  is specified.  A database is
         reused when the ID given corresponds to an existing database.
         When a database is created, it is large enough to accommodate
         10 variables with as many as 250  cases  per  variable.   The
         STATMATE  install program, SMINSTLL may be used to change the
         size of the databases created by STATMATE.  With  a  database
         of  10  variables, the user may store, modify, manipulate and
         repeatedly use up to 10 variables for analysis.  This ability
         reduces the re-entry of data for each analysis.

         In situations where several users work  with  STATMATE,  they
         may  want  to create their own databases by using a different
         ID.  This feature provides additional security when  multiple
         users operate the package.

         Associated with the database is a program generated directory
         which  contains  the  names  of  the  variables   and   other
         attributes of  the  data  contained  in  the  database.   The
         directory may be examined and manipulated by such commands as
         GIVE, QUERY and ERASE.

         Organization and Manipulation of Data within the Database

         Data  is  organized  by  variables   within   the   database.
         Variables represent  a  collection  of  data  on  which  some
         analytic  or  manipulative  operation is to be performed, for
         example, a variable could be the number of houses built  each
         year  for  15  years.   Each database has a maximum number of
         variables that can be placed in it, and a maximum  number  of
         data values that can be placed in any variable.  Initially, a
         database  is empty, but contains space for the maximum number
         of variables and data values.

         Variables in a database are  assigned  to  specific  database


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         locations in a sequential fashion, starting the with variable
         1  and  proceeding  to  the  highest  numbered variable.  For
         example, the first variable is assigned to database  location
         1,  the second to location 2, and so on.  Variables are given
         predefined names according to their ordered location  in  the
         datatase.   Each  predefined name is prefixed with a # and is
         followed by a number.  For example, #7 is  the  name  of  the
         seventh  variable.   The  user  may  assign his own alternate
         names to the predefined variable names, as well.

         Data is usually brought into a database with either the INPUT
         or EDIT commands.  The INPUT command allows you to enter data
         contained on external files.  EDIT allows you to  enter  data
         from  the  keyboard.  When bringing variables into a database
         with the INPUT command, variables are assigned to  successive
         variables  in a database.  For example, assume that there are
         variables in the  first  five  database  variables.   If  two
         variables are input, then the new variables will reside in #6
         and  #7.   New  variables are generally placed after existing
         variables.  One form of the INPUT command allows you to place
         variables at specific locations in a database.  See the INPUT
         command for further details on  the  entry  of  data  into  a
         database.   Use  of the LET command, permits data to be moved
         from one variable to another.  Variables may be removed  from
         a database by erasing them, using the ERASE command.

         Variable Names

         A  particularly  useful feature of STATMATE is that it allows
         you to assign names to variables, Thus,  it  is  possible  to
         assign  more  meaningful  names  to  variables.  For example,
         are  valid  names.   These  names  consist  of  from  1 to 10
         characters.  The first character must be alphabetic  but  the
         remaining  characters  may  be  either numeric or alphabetic.
         Alphabetic characters must be in  upper  case  letters.   All
         variables  have  alternate  predefined  names  of the form #n
         where n is the location number of the variable.  For example,
         #4 and #9 are predefined variable names.   The  GIVE  command
         allows  alternate  names  to  be  given  to  a variable.  The
         variable #4 might, for example, be given the  alternate  name
         AGE.   Either  #4  or AGE could be used to reference the same

         A special variable, #0, is available, which provides the data
         values 1, 2, 3, ....  This variable does not occupy any space
         in the database.  It may not be given an alternate name.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Ways of Referencing Data--Variables and Cases

         Data is generally regarded as arranged in rows and columns by
         STATMATE.   Although  this  generalization  is  not  entirely
         adequate,  it  is at least a starting point for understanding
         how STATMATE deals with sets of data.  A  simple  example  of
         such  an  arrangement  is the U.S.  population data discussed
         earlier  and  contained  in  data  file  USPOP.DAT  (data  in

                         Year     Urban Pop    Rural Pop
                         1790       0.202        3.728
                         1800       0.322        4.986
               ^         1810       0.525        6.714
               |         1820       0.693        8.945
               |         1830       1.127       11.739
             Rows        1840       1.845       15.224
               |         1850       3.544       19.648
               |         1860       6.217       25.227
               V         1870       9.902       28.656
                         1880      14.130       36.026
                         ...         ...          ...

         The data is arranged so that each column represents some item
         (variable) that is to be examined in detail.  For example, it
         might  be  of  interest to determine the average value of the
         item.  A row represents some common element that each of  the
         columnar  items  have  in   common,   in   this   case,   the
         corresponding population for a given year.

         In STATMATE the data in a column to be examined or studied is
         called a variable.  The data in a row is  referred  to  as  a
         case or observation.  In the above example we have:

                       <------- Variables------>
                      Year       Urban Pop    Rural Pop
             ^        1790
             |        1800

         Year,  urban  population  and rural population are variables.
         The data 0.202 and 3.728 represent the case  data  for  1790.
         Note  that  there is no reason why the dates themselves could
         not be considered as a variable and the individual  years  as
         belonging to a case.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         External Data Entry--Files

         Files  are  the  basic  way  of  entering data into STATMATE.
         (Another method for entering  data,  with  the  STATMATE/PLUS
         EDIT  command,  is  available).  Data files have names within
         the Operating System, CP/M, MS DOS or PC DOS.  Any  of  these
         Operating  Systems  allows a variety of characters for names;
         however, STATMATE only recognizes file names  and  file  name
         extensions  which  are  composed  of alphabetic characters or
         numbers.  File names and extent  names  must  begin  with  an
         alphabetic  character.   ABCDE,  uspop.DAT,  HIST1980.DT  and
         MYDATA are examples of valid file references within STATMATE.
         STOCK/82.DAT (/ is a special character and is  not  allowed),
         SAL$DATA  ($ is a special character) and YEARS.DATA (the file
         extension   DATA   is  too  long)  are  examples  of  invalid
         references.  Your Operating System may allow these names, but
         STATMATE  will  not  accept them if they are used in commands
         where a file name is needed.  Use of your  Operating  Systems
         renaming  capabilities  will solve any difficulties with file
         names.  A file  name  may  be  prefixed  with  a  disk  drive
         identifier  as in A:XYZDATA.DAT.  Data files are entered into
         the STATMATE database with the INPUT command.

         There are two ways of creating files for input.  Only one  of
         these  will be described here, the use of ASCII files.  ASCII
         data files may be created using a text editor  program,  such
         as  WORDSTAR  or  EasyWriter.   An  ASCII  file can be easily
         printed or listed.

         Creating ASCII Data Files

         Often the simplest way of producing an ASCII file is to use a
         text editor.  All that is required for preparing  a  STATMATE
         input file with a text editor is a basic understanding of how
         to  arrange the text representing the data.  This is a simple
         task and it is addressed below.

         The following shows the contents of a file prepared  using  a
         text editor.


         If the file were called ENERGY.DAT, it could be read with the
         STATMATE INPUT command by entering:

                 INPUT ENERGY.DAT KEEP 3


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         See the INPUT command description for additional information.

         The following rules apply to preparing a data file:

              1.   The  word  DATA  (uppercase or lowercase) must
                   appear as the first word of  the  first  line.
                   Blanks  or any other characters may not appear
                   before DATA.  Any comments  may  appear  after
                   DATA on the line.  MY ENERGY ...  is a comment
                   describing the data in the above example.

              2.   Data  items  follow  on  each subsequent line.
                   Each item is separated from another by a comma
                   or blank.

              3.   All lines  must  be  followed  by  a  carriage
                   return and line feed, including the last line.

              4.   Alphanumeric  information  must be enclosed in
                   single quotes (') or begin with an  alphabetic
                   character.   If  alphanumeric data contains an
                   embedded blank, the data must be surrounded by
                   quotes.  The MOTOR.DAT file in Appendix  B  is
                   an  example  of  using  alphanumeric data in a
                   DATA file.

              5.   Data must be arranged on a case by case basis.

         It should be noted that although  many  word  processors  and
         editors  will  produce  files which STATMATE will read, there
         are some  word  processors  and  editors  which  place  extra
         characters  at the beginning of a file.  Users of EasyWriter,
         for example, must use  the  TRANSFER  utility  to  produce  a
         proper  ASCII file.  EasyWriter users must also use the ENTER
         key to generate carriage returns after each line.   If  extra
         characters  are  placed  before  DATA,  STATMATE will issue a
         message that the file is invalid.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                       This page deliberately left blank




         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                       This page deliberately left blank


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         Each  command  is  described  in  detail in this section.  In
         order to determine which commands are executable in  each  of
         the  three STATMATE packages, observe that at the top of each
         page a subtitle lists the  packages.   When  a  package  name
         appears  there,  then  the  package  includes   the   command

         An important part of the command description is the syntax or
         format of the command.  When describing  the  syntax  of  the
         command the following general rules are followed:

              1.   Uppercase characters should be entered as

              2.   For  brevity, only the first 3 characters
                   of commands, modifier names,  etc.   need
                   be entered.  If additional characters are
                   entered,  they  should  always  match the
                   name in every  position  given  from  the
                   first  to  last   character   entered.For
                   example, STATISTICS is matched by STA  or
                   STATI but STATS does not match it.

              3.   Lowercase is used to describe the type of
                   entry  that  you   must   provide.    For
                   example,  in  a  description  var1  might
                   represent a variable name.

              4.   Punctuation is entered as shown.

              5.   An  ellipsis  (...)   means  repeat   the
                   previous item as  needed.   For  example,
                   num,...,  where  num represents a number,
                   indicates either that a single number  or
                   a  list of numbers with separating commas
                   is acceptable.

         In addition to a command's syntax, each  command  is  further
         clarified  with  descriptive  material  and detailed examples
         concerning its use.  Input to STATMATE that is entered by the
         user is shown in boldface in the detailed examples.

         Remember to enter a carriage return after entering a  command


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  CROSSTAB

         Purpose: Used to cross tabulate data on two variables.

              A. CROSSTAB classavar ON classbvar
                   classavar = classification variable A
                   classbvar = classification variable B

         Syntax Examples:


         CROSSTAB peforms a two-way cross tabulation on two variables.
         Each  variable  contains  data  which  divides  the data into
         classes.  For example, the data below collected on the season
         and observed color of a botanical specimen.  SEASON might  be
         thought  of  as defining classes for each of the four seasons
         and COLOR as defining five classes:  BLUE, GREN (green), RED,
         BLCK (black) and YELL (yellow).

                         SEASON          COLOR
                         ------          -----
                         FALL            BLUE
                         FALL            GREN
                         WINT            GREN
                         SUMR            RED
                         SPNG            BLCK
                         FALL            GREN
                         SUMR            BLUE
                         SPNG            YELL

         In a two-way cross tabulation, the data might appear as:


                         BLUE    BLCK    GREN    RED     YELL
                         ----    ----    ----    ----    ----
                 SPNG             1                       1
          SEASON SUMR                             1
                 FALL     1               2
                 WINT                     1

         Classes may be defined for  either  numeric  or  alphanumeric
         data.   CROSSTAB  tabulates  data for the class A variable by


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         rows and the class B variable by  columns.   As  many  as  20
         different  classes  may  be  contained  in  a variable.  If a
         variable contains more than  20  classes,  the  STATMATE/PLUS
         RECODE  command  may  be  used  to reduce the total number of

         CROSSTAB produces a table of statistics.  Each  table  entry,
         or  cell,  contains  a  frequency,  row  percentage,   column
         percentage,   and  total  percentage  for  the  corresponding
         classes.   Percentages  and  totals  for rows and columns are
         reported.  Fisher's exact probabilities for the special  case
         of  2x2  tables  are  calculated  when  there are 32 or fewer
         entries in the table, These probabilities are used to produce
         Tocher's correction.  If a 2x2 table is  produced  with  more
         than  32  entries,  Yate's  correction  to  the Chi-square is
         output.  The  Chi-square  statistic  is  produced  for  other
         tables.   A Phi value, a measure independent of the number of
         cross tabulation entries, is produced for all tables.

         If the number of columns and rows in the output table  exceed
         a  reasonable  page  width and length, the table is output in
         conveniently divided sections.  Basically, the  entire  table
         is  output  in  sections  from  left to right and from top to
         bottom.  Classes are sorted in ascending  order  before  they
         are output, so classes placed in the output table are easy to

         Detailed Example

         The  following  example  uses  CROSSTAB  to perform a two-way
         cross tabulation on the class data contained in  FACTORA  and
         FACTORB.   Note  that  some  of  the class data in FACTORA is
         defined by a -2 value, and note that any values may  be  used
         to  identify  a class.  Class identifiers are, numeric values
         in this instance, are printed before each row and above  each
         column.   FACTORB has four classes:  -2, 4, 6 and 8.  FACTORA
         has three classes:  2, 3 and 4.


         FACTORA     FACTORB
         ----------- -----------
                3.00        4.00
                2.00       -2.00
                3.00        6.00
                2.00        8.00
                2.00        6.00
                4.00        8.00
                2.00        4.00
                2.00        8.00
                2.00        4.00
                2.00        8.00
                4.00        6.00


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                3.00       -2.00



            12 CASES
             0 MISSING

             0 NOT TABULATED

                  COUNT  |
                 ROW PCT |        |        |        |        |
                 COL PCT |        |        |        |        |  ROW
                 TOT PCT |   -2.00|    4.00|    6.00|    8.00|  TOTAL
         FACTORA         |        |        |        |        |
                    2.00 |     1  |     2  |     1  |     3  |     7
                         |  14.3% |  28.6% |  14.3% |  42.9% |  58.3%
                         |  50.0% |  66.7% |  33.3% |  75.0% |
                         |   8.3% |  16.7% |   8.3% |  25.0% |
                    3.00 |     1  |     1  |     1  |     0  |     3
                         |  33.3% |  33.3% |  33.3% |   0.0% |  25.0%
                         |  50.0% |  33.3% |  33.3% |   0.0% |
                         |   8.3% |   8.3% |   8.3% |   0.0% |
                    4.00 |     0  |     0  |     1  |     1  |     2
                         |   0.0% |   0.0% |  50.0% |  50.0% |  16.7%
                         |   0.0% |   0.0% |  33.3% |  25.0% |
                         |   0.0% |   0.0% |   8.3% |   8.3% |
                 COLUMN        2        3        3        4       12
                  TOTAL     16.7%    25.0%    25.0%    33.3%   100.0%

         CHI SQ:      3.7381 DF:     6
         PHI =    0.558


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  ERASE

         Purpose: Erases data variables from the EZPACK database.

              A. ERASE vname THRU END
              B. ERASE vname
                   vname = name of first variable where data is
                           to be erased

         Syntax Examples:
              ERASE #6 THRU END
              ERASE WAGES THRU END
              ERASE SALES


         In order to erase data from the database, the  ERASE  command
         must  be  used.   For  the  variables  specified, the command
         resets the number of cases  to  zero,  removes  any  assigned
         name,  and  resets  the  missing  value to 1.0E30.  Variables
         where data is erased become numeric variables.  Data  may  be
         erased variable by variable, or from a given variable through
         the  end  of  the database.  Erasing data does not affect the
         amount of data stored in the database (since erased  data  is
         simply ignored), but does affect the use of the INPUT command
         (See the INPUT command for details).

         Detailed Example:

         Below,  the  ERASE  command  is  used  to erase data from the
         database.  Erasing begins with the variable TOTALPOP, #4, and
         proceeds to the last variable, #10.

         Command:  ERASE TOTALPOP THRU END



         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  EXECUTE

         Purpose: Allows execution of commands from an EXECUTE file.

              EXECUTE fname
              where fname is the name of a file

         Syntax Examples:
              EXECUTE MYCOMFIL


         The EXECUTE command allows commands placed on an EXECUTE file
         to be executed by  STATMATE.   This  feature  simplifies  the
         entry  of  often used sequences of commands.  A discussion of
         this feature may be found in the introductory information  on
         the use of EXECUTE files.

         An EXECUTE command may not appear in an EXECUTE file.

         When  an  EXECUTE command is entered, the subsequent Command:
         prompts that usually are issued by STATMATE are  replaced  by
         !Command:   until  the  commands  on  the file are completely
         processed.  The Command:  prompt is then issued  to  indicate
         that  STATMATE wants you to enter a command when all commands
         on the file have been processed.

         Detailed Example:

         In the following example, EXECUTE refers to the EXECUTE  file
         REGSTUDY.EXC, found on the C disk (directory).

         Command:  EXECUTE C:REGSTUDY.EXC


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  EXIT

         Purpose: Returns the user to the Operating System


         Syntax Examples:


         The EXIT command returns the user to the Operating System.

         Detailed Example:

         EXIT is entered in response to an STATMATE command request to
         return to the Operating System.

         Command:  EXIT



         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  GIVE

         Purpose: Assigns user defined attributes (names, missing values,
                  and data types) to variables.

              A. GIVE NAME aname,vname1,...
                   aname = user or predefined variable name at which
                        assignment begins
                   vname1 = user name for aname
                   vname2 = user name for next variable
                        after aname

              B. GIVE MISSING aname,missv1,...
                   aname = user or predefined variable name at which
                         assignment begins
                   missv1 = missing value to be assigned to
                         variable aname
                   missv2 = missing value to be assigned to next
                         variable after aname
              C. GIVE TYPE aname,atype1,...
                   aname = user or predefined variable name at which
                         assignment begins
                   atype1 = A (alphanumeric) or N (numeric) to be
                           assigned to variable aname
                   atype2 = type to be assigned to next variable
                            after aname

         Syntax Examples:
              GIVE NAME #4,AGE,SCORE
              GIVE NAME AGE,NEWAGE
              GIVE MISSING #1,200.0,200.0,-1000.0,'NA'
              GIVE MISSING QUARTER,'Q?'
              GIVE MISSING AGE,0.0,0.0
              GIVE TYPE,#15,A,N,A,N,N


         The  GIVE  command  allows  a specific name, missing value or
         data type to be assigned  to  a  variable.   Initially,  each
         variable  has  a  default  name,  missing value and data type
         until you change these attributes  with  GIVE.   Recall  that
         each  database  contains  a  fixed  number of variables, even
         though you have not used INPUT to bring  any  data  into  the
         database.   Hence,  GIVE  can  be  used  to  modify  variable
         attributes before data is placed  in  a  variable.   GIVE  is
         often just for this purpose just before using INPUT.

         A  variable name (user assigned name) is used as an alternate
         name to the predefined  name  given  to  every  variable.   A


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         numeric  value  of  1.0E+30  is  assigned  to  every  numeric
         variable, if it is not changed by GIVE  MISSING.   A  missing
         value  of  blanks is assigned to every alphanumeric variable,
         if it is not changed by GIVE MISSING.   Alphanumeric  missing
         values  must  be  enclosed in single quotes (').  The MISSING
         attribute may not be changed  once  data  is  placed  in  the
         variable;  the  ERASE  command  must  be  used  first  before
         assigning a new MISSING attribute to a  variable.   The  data
         type  may  be  numeric  (N) or alphanumeric (A).  In order to
         place alphanumeric data in the database with INPUT, you  must
         make the appropriate variable an alphanumeric type.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Detailed Example:

         GIVE  NAME is used to assign the names YEAR and AGE to #3 and
         #4, respectively, then GIVE MISSING is used to assign missing
         values of 0 and -1 to YEAR and AGE.

         Command:  GIVE NAME #3,YEAR,AGE


         Command:  GIVE MISSING YEAR,0,-1



         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  HELP

         Purpose: Displays descriptive information regarding the use of
                  all commands.

              HELP cname
         Syntax Examples:
              HELP CURVE
              HELP REGRESSION


         The HELP command displays descriptive  information  regarding
         the  use  of  commands  currently  implemented.  This command
         provides a reminder of available commands  and  formats  when
         working  interactively  at the terminal.  If HELP is followed
         by the name of a command, specific help for  the  command  is
         displayed.   If  HELP is given without a command name, a list
         of commands is given  along  with  some  general  information
         about the package.

         Help  information  is  contained  on  a  file  with  the name
         IFHELP.TXT, which can be modified by the  user  with  a  text

         Detailed Example:

         The  following  illustrates  the  output of the HELP command.
         Only a part of the output is shown.

         Command:  HELP


           COMPUTE    CORRELATE  ELSE       END        ERASE
           EXECUTE    GIVE       HELP       INPUT      LET
           PLOT       PRINT      QUERY      REMARK     SET
           STATISTICS WHEN       WRITE

                      ... (more follows but is not shown)


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                             Let's look at the record.
                                  Al Smith (1938)

         Command:  INPUT

         Purpose: Allows entry of data into the STATMATE database
                  from external files.

              A. INPUT fname descriptor1, descriptor2, ...
              B. INPUT fname descriptor1, descriptor2, ... AT vname
                   fname = a file name
                   descriptor = one of the following field extraction
                                  KEEP n
                                  OMIT n
                   vname = variable name

         Syntax Examples:
              INPUT MYDATA KEEP 1
              INPUT B:XYZ KEEP 5 AT #20


         INPUT provides one way of entering  data  into  the  STATMATE
         database.   Data  is  read in (kept) or not read in (omitted)
         from the specified file according to field descriptors  given
         in  the  command.  The descriptors permit any field of a file
         record to be selectively  extracted  and  inserted  into  the
         database  as  a  variable.  Data read from the file is placed
         either beginning at  the  variable  specified  after  the  AT
         keyword  in  the  command,  or  at the start of the rightmost
         block of  erased  or  empty  variables  at  the  end  of  the

         STATMATE  assumes  that  a file is record oriented.  A record
         consists of one or more successive fields .  Each record must
         contain exactly the same number of fields.  A  file  contains
         one  or  more  records.   See  the  introductory  section  on
         specific details of the two file types (DATA and PROG)  files
         that may be read by the INPUT command.

         When  INPUT is used to read such files, it must be told field
         by field, for every field of a record, whether the field that
         is to be transferred to a variable in the database.  The OMIT
         and KEEP descriptors are used for this purpose.   If  several


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         successive  fields  are  to  be kept or omitted, a number may
         follow OMIT and KEEP to indicate the number of  such  fields.
         If  a  number does not follow a descriptor, a value of one is
         assumed for the number of fields the descriptor applies.  The
         total number of fields kept and omitted must be  exactly  the
         number of fields in a record.

         In terms of fields and records, a case corresponds roughly to
         a  record  entry  and  a  variable  to  all  the entries in a
         specific field.

         Fields extracted from a data file are  placed  in  successive
         variables  in  the database.  The location of the first field
         in which data is to be placed may be specified by  using  the
         form  containing  the  AT keyword.  If this form is not used,
         INPUT will placed the first variable extracted from the  file
         at  the  start  of  the  rightmost  block  of erased or empty
         variables at the end of the database.  For  example,  if  the
         last  variable in the database is #25, and all variables from
         #12 through #25 are  erased,  then  #12  will  be  the  first
         variable  to  receive  data of the INPUT command, when the AT
         form is not used.

         Detailed Example:

         INPUT is used to extract data from the States and  population
         density fields of the MOTOR.DAT file given in the Appendix B.
         The  file  contains 8 fields, and the first and fourth fields
         contain the data of interest.  All variables from #4  through
         the  end  of the database are assumed erased or empty.  Since
         the States field contains alphanumeric data, #4 is given  the
         alphanumeric  attribute  with  the GIVE TYPE command prior to
         using INPUT.  When INPUT is executed, the two data fields are
         placed in variables #4 and #5.  Each field contains 50 values
         or cases.  WHEN is used to select the first five  cases,  and
         PRINT  is  used  to  list  the  first  five  cases of the two
         variables read into the database.

         Command:  GIVE TYPE #4,A


         Command:  INPUT MOTOR.DAT KEEP 1,OMIT 2, KEEP 1,OMIT 4

             2 VARIABLES INPUT AT #4
             50 CASES


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  WHEN #0=1 THRU 5

             5 OF    250 CASES

         Command:  PRINT #4,#5

         #4          #5
         ----------- -----------
             AL            64.00
             AK             0.40
             AZ            12.00
             AR            34.00
             CA           100.00


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  LET

         Purpose: Allows simple arithmetic operations and other
                  manipulations on data.

              A. LET vname = simpexp
              B. LET vname = fcn(arg)
              C. LET aname1 = aname2
                   vname = numeric variable name
                   simpexp = simple arithmetic expression
                   fcn = STATMATE function name
                   arg = list of arguments for the function
                   aname1 = alphanumeric variable name
                   aname2 = alphanumeric variable name

         Syntax Examples:
              LET NEWAGE = AGE-21
              LET SALES = COST*AMTSOLD
              LET WAGES = SALARY-TAXES
              LET LAGSALES  = LAG(SALES,1)
              LET #4 = #0
              LET LENGTH = SQRT(AREA)
              LET NEWVAR = #2/3.0
              LET RESPONSE = LOG10(EXPOSURE)


         LET is  a  very  useful  command  for  performing  arithmetic
         manipulations on variables, and moving data from one variable
         to another.  In addition to the simple arithmetic operations,
         several  simple  functions are available to deal with lagging
         and other useful operations.  Numeric and  alphanumeric  data
         may  be  moved from one variable to another by using a simple
         assignment of one variable to another.

         The   entry   of   LET NEWYEAR = YEAR + 1900   produces   the

                 NEWYEAR      YEAR
                 -------      ----
                   1979        79
                   1980        80
                   1981        81
                   1982        82


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Other simple arithmetic operations that may be performed are:

              A+B        A-B     A*B     A/B

              A+cnst     A-cnst  A*cnst  A/cnst  A**cnst

              -A         +A      A       -cnst   +cnst     cnst

         where  A  and B represent any variable names, and cnst refers
         to any numeric constant.  The symbols +, -, * and / represent
         their normal  meanings,  *  indicating  multiplication.   The
         symbol ** indicates the raising of a number to a power.  Note
         that A**-1 is the same as the reciprocal of A (1/A).

         The  LET  also  accomodates a number of arithmetic functions.
         The following functions are avaiable:

                 Function  Description
                 --------  -------------------------------------------
                 LAG       Lag or shift time period data a specified
                           number of periods
                 LOG       Logarithm (base e)
                 LOG10     Logarithm (base 10)
                 SQRT      Square root of data
                 SAM       Select a random sample
                 MOV       Moving average
                 EXP       Raise a number to a power of e
                 NUM       Produce a sequence of numbers
                 PRD       Produce a repeated sequence of numbers
                 STP       Produce a sequence of numbers at a given
                 NOR       Normally distributed random numbers
                 UNI       Uniformly distributed random numbers
                 DNOR      Normal distribution points
                 DEXP      Exponential distribution points
                 DWEI      Weibull distribution points
                 DCAU      Cauchy distribution points
                 ABS       Absolute value
                 CUM       Cumulative sum
                 INT       Integer or whole number

         Functions have the form:

                 LET vname = fname(a1,a2,a3,...)
                   vname is a variable
                   a1, a2 and a3 are arguments

         For example,

                 LET GROWTH = LOG(DOSAGE)
                 LET RESPONSE = NOR(HEIGHT,0,2.0)


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         The first argument of any function  is  always  the  variable
         name  for  which  the  function  is to be applied.  The first
         example causes the logarithm, base e, to be taken on  all  of
         the data in the variable DOSAGE, and placed into the variable
         GROWTH.   In  the  second example, note that the NOR function
         has   three   arguments.   The  second  and  third  arguments
         represent  the  mean  and  standard deviation of the normally
         distributed random numbers to be generated.

         Although it is always necessary to provide a variable as  the
         first  argument of a function, the values of the variable may
         not be  used  by  the  function.   For  many  functions,  the
         variable  is  used  to  determine  how  many values are to be
         generated by the function.  For example, in the  use  of  NOR
         cited  above, exactly as many normally distributed values are
         produced as there are cases in DOSAGE.  However, if a case in
         DOSAGE is missing,  a  missing  value  will  be  assigned  to
         corresponding  case in the resulting variable.  In fact, this
         is true in general.  That is, a  missing  value  found  in  a
         variable  used on the right side of a LET expression produces
         a missing value in the assignment variable.

         Descriptions of the individual functions and  their  argument
         lists are given below.

         Function        Description
         -----------     ----------------------------------------------
         LAG(v,p)        Lag v by p periods. See the NOTE below for
                         more on LAG.
                         LAG(YEAR,3)        -- lag 3 periods

         SAM(v,n)        Select exactly n items randomly from v.
                         The values of the cases selected are placed
                         in the assignment variable.  Items not
                         selected are marked with a missing value.
                         SAM(STATES,12)     -- randomly select 12
                         items from STATES

         MOV(v,n)        Compute an n-term moving average.  If v
                         contains the values a,b,c and d in the
                         first four cases, a two-term moving
                         average produces four cases in the
                         resulting variable: mv, (a+b)/2,
                         (b+c)/2, (c+d)/2, where mv is a missing


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Function        Description
         -----------     ----------------------------------------------
         SQRT(v)         Take the square root of v.

         LOG(v)          Compute the log of v using base e.

         LOG10(v)        Compute the log of v using base 10.

         EXP(v)          Raise e to the power of the data in v.

         PRD(v,b,i,p)    Produce a repeated sequence of numbers
                         beginning with b.  Increment b by an
                         amount i exactly p times, and then repeat
                         the sequence beginning with b again.
                         PRD(SALES,1,1,12)   -- 1,2,3,4,5,6,7,

         STP(v,b,j,p)    Produce the number b exactly p times, then
                         increment b by j and produce b+j exactly
                         p times. Continue adding j to the last
                         sequence of p numbers produced.
                         STP(RESP,4,1,2)     -- 4,4,5,5,6,6,...

         NUM(v,b,s)      Produce the sequence of numbers b, b+s,
                         b+2*s, b+3*s, ...
                         NUM(MONTH,1,2)      -- 1,3,5,7,...
                         (1=start, 2=step)

         NOR(v,m,s)      Generate random numbers from a normal
                         distribution with a mean of m and standard
                         deviation of s.
                         NOR(DOSE,5.0,1.4)   -- Normal random nos.
                         (mean=5, sd=1.4)


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Function        Description
         -----------     ----------------------------------------------
         UNI(v,a,b)      Generate random numbers from a uniform
                         distribution on the interval from a to b.
                         UNI(COST,20,44.5)   -- Uniform random nos.
         DNOR(v,m,s)     Compute the probability F(x), given that
                         F is a normal distribution with mean m
                         and standard deviation s.  See the KOLMOGOROV

         DEXP(v,s)       Compute the probability F(x), given that
                         F is an exponential distribution with mean s.
                         See the KOLMOGOROV command.

         DWEI(v,u,s)     Compute the probability F(x), given that
                         F is a Weibull distribution with location
                         parameter u and scale parameter s.  See the
                         KOLMOGOROV command.

         DCAU(v,u,s)     Compute the probability F(x), given that
                         F is a Cauchy distribution with median given
                         by the parameter u and the 1st quartile
                         given by u-s.  See the KOLMOGOROV command.
         ABS(v)          Take the absolute value of v.

         CUM(v)          Produce the cumulative sum of v.  For example,
                         if 10, 20 and 15 are the first three cases of
                         v, the cumulative sum is 10, 30, 45.

         INT(v)          Take the integer portion of v.  For example,
                         if the first two cases of v are 22.4 and
                         -15.8, then the result is 22 and -15.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         NOTE:   The  LAG  function  provides  a  very  useful  way of
         examining relationships between one time period and  another.
         For  example, assume that annual inventory and sales data are
         available and that a comparison of a year's  inventory  to  a
         previous  year's  sales  is  to be made.  A lag period of one
         year is needed.  A lag has the effect of advancing data  from
         earlier  periods to recent periods by the specified number of
         lag periods.  Consider the following data, where INVLAG is  a
         variable created by using

                        LET INVLAG = LAG(INV,1)

                 Case (Year)     SALES   INV     INVLAG
                 ----------      -----   ---     ------
                   1975           40     115       ?
                   1976           55     135      115
                   1977           65     185      130
                   1978           90     140      185
                   1979           70     200      140
                    ...           ...    ...      200

         A  comparison of SALES and INVLAG shows a simple relationship
         for any year; SALES is about one-half  the  inventory  figure
         for  the  previous year.  Ah, to be so lucky!  Since data for
         1974 did not exist, a  missing  value  (shown  as  a  ?)   is
         indicated.  The concept of lagged variables is a particularly
         useful one in curve fitting, forecasting and regression.

         A LET operation sometimes involves missing values.  For those
         variables  on  the  right  side of the equal sign whose cases
         contain a missing value, a missing value is created  for  the
         corresponding cases in the left variable.

         When  a  LET  operation is performed, new cases or values are
         formed for the variable appearing on the left  of  the  equal
         sign.  The length of this variable, or the number of cases it
         contains,  can  be  affected by the operation.  Generally the
         effect is of little concern.  In some instances, some insight
         into how the length is determined may be useful.

         In most instances, the length is  found  by  determining  the
         longest  variable  involved  in the operation.  If a variable
         does not appear on the  right,  as  in  the  case  of  simply
         assigning a constant, the length is the same as the length of
         left  variable.   If,  in  assigning  a  constant,  the  left
         variable has a zero length or was erased or never  used,  its
         length  becomes  the  maximum  length that can be used in the
         database.  If the variable to the left of the equal sign, the
         assignment variable, has fewer cases  than  other  variables,


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         its  length is not increased.  An already existing variable's
         length can be increased only by erasing the variable and then
         assigning data to it with the LET.

         Detailed Example:

         LET is used to adjust the YEAR variable  from  the  USPOP.DAT
         file  given  in  Appendix B.  Instead of beginning with 1790,
         the data  is  transformed  to  begin  with  the  year  0  and
         continues  in  intervals of 10.  The new variable is ADJYEAR.
         The name ADJYEAR was assigned  earlier  with  the  GIVE  NAME
         command to some variable.

         Command:  LET ADJYEAR = YEAR-1790

           ADJYEAR    MODIFIED

             17 CASES
              0 MISSING


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                         This page deliberately blank




         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  PLOT

         Purpose: Produces a plot at the terminal.

              A. PLOT yvar1,yvar2,... ON xvar
              B. PLOT yvar1,yvar2,... ON xvar,mod1,mod2,...
              C. PLOT
              D. PLOT ?
                   yvar1=name of variable to be plotted vertically
                   xvar=name of variable to be plotted horizontally
                   mod1=one of the following:
                        HRANGE=h1 THRU h2
                        VRANGE=v1 THRU v2
                        HPOSITIONS=number of spaces
                        VPOSITIONS=number of lines
              HRANGE = DATA (i.e., use minimum and maximum of xvar data)
              VRANGE = DATA (i.e., use minimum and maximum of yvar data)

         Syntax Examples:
              PLOT SCORES ON AGE
                TITLE='SALES HISTORY'


         PLOT produces a plot of one to five variables against another
         variable  in  the  form  of  a  scatter  plot.   A  number of
         modifiers provide ways of titling, labeling, and  sizing  the

         Multiple curves, variables, are plotted vertically.  Vertical
         and  horizontal axes, scale, and legend information are added
         to the plot.

         If the ranges are not specified with HRANGE  or  VRANGE,  the
         range  over which the data for either axis is to be displayed
         is derived from the data.

         The TITLE, VLABEL and HLABEL modifiers provide a way  to  add
         title  and  axis  descriptions  to  the  plot.  Note that the
         descriptive information for these modifiers must be  enclosed
         by single quotes.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Physical  sizing  of  a  plot  is  aided by the HPOSITION and
         VPOSITION modifiers.  They permit the height and width  of  a
         plot  to  be  specified  in  terms of the number of lines and
         characters   per  line.   A  maximum  of  60  lines  and  120
         characters  may  be  specified.  These modifiers apply to the
         portion of the plot exclusive of  the  axes  and  descriptive
         information placed on it.

         Plot  points  are  shown  as  +,  *,  X, @and # for the first
         through the fifth curve plotted.  If points  are  coincident,
         only  the  point  for  the  rightmost  curve  in  the list is

         Before actually producing the plot, a  pause  occurs  at  the
         terminal  to  allow positioning and placement of paper in the
         printer.  Once the paper is adjusted, depressing return  will
         cause  the  plot  to appear.  Don't forget to enter ctrl-P if
         you want the output on the printer!  See the SET command  for
         an alternate way of placing output on the printer.

         If  PLOT  is  used  without  specifying any variable names or
         modifiers, you will  be  prompted  for  the  variable  names.
         After  entering the variable names, STATMATE will then prompt
         you for subcommands.  The subcommands  allow  you  to  change
         modifier values.  The following subcommands are available:

            Subcommands                          Purpose
            --------------------------------   ----------------------------
            CONTINUE                           Execute PLOT or exit command
            SAVE                               Save modifier default values
            SHOW                               Display modifier values
            HPOSITION = x                      Set HPOS modifier
            VPOSITION = x                      Set VPOS modifier
            HRANGE    = x.x THRU x.x or DATA   Set HRAN modifier
            VRANGE    = x.x THRU x.x or DATA   Set VRAN modifier
            TITLE     = 'xxxx'                 Set TITLE modifier
            HLABEL    = 'xxxx'                 Set HLABEL modifier
            VLABEL    = 'xxxx'                 Set VLABEL modifier

            where x.x is a decimal number, x is whole,
                  xxxx is a string of characters

         Some examples of subcommands are:

                 HRANGE=30 THRU 200.5
                 TITLE='POPULATION DENSITY FOR 1980'
                 VLAB='RESPONSE VALUES'

         Entering  CONTINUE in response to a subcommand prompt, causes
         STATMATE to produce the desired plot and  return  to  command


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         prompting.  If you used PLOT ?  to enter the subcommand mode,
         as  explained  below, CONTINUE will simply return STATMATE to
         command prompting without executing the PLOT.  SHOW  displays
         the current setting of the modifiers.

         When  a  modifier  is  set  to  a  particular  value  using a
         subcommand, its value is only temporary.  That is,  when  the
         plot  is  produced,  the  value  for the modifier will be the
         temporary value that you assigned it.  However, the next time
         you use  PLOT  the  modifier  will  revert  to  its  original
         permanent  default  value.  You may change the default values
         permanently   by   using   the  SAVE  subcommand.   The  SHOW
         subcommand displays the current values of the modifers.

         Setting  the  default  values  as  permanent is a very useful
         feature of STATMATE.  It is particularly useful when you have
         titles and labels that you continually use from PLOT to PLOT.

         Entering PLOT ?, allows you to  examine  and  change  default
         settings  of the modifiers without producing a plot.  In this
         case, you will not  be  prompted  for  variable  names.   The
         default  settings will be displayed and the subcommand prompt
         will appear.  Entering CONTINUE causes STATMATE to return  to
         command prompting without producing a PLOT.

         Detailed Example:

         The  use  of  the  PLOT  command  is  illustrated by plotting
         RURALPOP and YEAR data from the USPOP.DAT file found  in  the
         appendix.   Also  plotted  are  the  fitted,  FITRURPOP,  and
         forecast, FORRURPOP, values  obtained  by  fitting  a  linear
         equation to RURALPOP with the STATMATE CURVE command and then
         computing  these  two  variables with the COMPUTE command.  A
         similar set of fitted and forecast  data  could  be  produced
         using  the  STATMATE  STEPWISE regression command.  RURALPOP,
         FITRURPOP and FORRURPOP are plotted against YEAR.

         YEAR has been extended by modifying the USPOP.DAT  file  with
         the  years  1960  and  1970.   The  corresponding  cases  for
         RURALPOP were assigned missing values.   After  the  plot  is
         made,  the  PRINT  command  is  used to display the values of
         these values.

         The forecast values are printed as X on the plot.  The actual
         and fitted points at 1810  coincide  so  only  the  FITRURPOP
         point  (*)  is  printed.  The command is long enough that the
         first   portion   is   terminated  with  a  comma  to  permit
         continuation  of  the  remainder  of  the  command  after the
         CONTINUE:  prompt.

         Continue:  VLABEL='POPULATION(MILLIONS)',
         Continue:  HLABEL='YEARS',


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

             19 CASES
             19 MISSING



         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


                     :                                                  X
                     :                                               X
                61.55+                                            *
                     :                                         *
                     :                                         +
                     :                                      *
         P      52.59+                                      +
         O           :                                    +
         P           :                                 +  *
         U           :
         L           :                              +  *
         A      43.64+
         T           :                              *
         I           :                           +
         O           :                           *
         N           :
         (      34.68+                         +
         M           :                         *
         I           :
         L           :                      *
         L           :                      +
         I      25.73+                   *
         O           :                   +
         N           :                *
         S           :
         )           :             *  +
                     :             +
                     :           *
                     :           +
                     :        *
                 7.82+        +
                     :     *
                     :  +
                     :+ *
                 1790.00   1826.00   1862.00   1898.00   1934.00   1970.00


                                        + RURALPOP
                                        * FITRURPOP
                                        X FORRURPOP


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  PRINT

         Purpose: Prints (lists) data for variables.

              A. PRINT vname1,vname2,...
                   vname1 = name of variable to be printed
         Syntax Examples:
              PRINT STOCK,PRICES
              PRINT #0,WEIGHT


         The PRINT command is useful for printing (listing) data  from
         the  database.   Data is not actually sent the printer, it is
         displayed on the  CRT.   When  variables  with  a  dissimilar
         number  of cases are printed, an asterisk is printed in place
         of cases which do not exist.

         If the output of PRINT is to be placed on a printer, remember
         to enter a ctrl-P before pressing the carriage return at  the
         end of the command.  See the SET command for an alternate way
         of placing output on the printer.

         Detailed Example:

         An example of the PRINT command use is shown below that lists
         the  first  seven cases of the three variables YEAR, URBANPOP
         and RURALPOP.  The WHEN command is first  used  to  select  a
         view of the first seven cases.

         WHEN #0=1 THRU 7

             7 OF   250 CASES


         YEAR        URBANPOP    RURALPOP
         ----------- ----------- -----------
             1790.00        0.20        3.73
             1800.00        0.32        4.99
             1810.00        0.52        6.71
             1820.00        0.69        8.95
             1830.00        1.13       11.74
             1840.00        1.84       15.22
             1850.00        3.54       19.65


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  QUERY

         Purpose: Provides summary information about database variables.

              A. QUERY vname1,vname2,...
              B. QUERY vname1 THRU vname2
                   vnamei = name of variable
         Syntax Examples:
              QUERY LOANS,#5,#3,REGION
              QUERY #1 THRU #36


         The QUERY command displays information about a variable which
         includes  its  name,  predefined  name,  data type, number of
         cases, number of missing cases and missing value code.

         Detailed Example:

         In   this   example,   QUERY  is  used  to  display  database
         information  about  the first four variables in the database.

         Command:  QUERY #1 THRU #4

         ----   ----------  -------  -------  ---------    -------------
         #  1   YEAR        NUMERIC       17          0      1.0000E+030
         #  2   URBANPOP    NUMERIC       17          0      1.0000E+030
         #  3   RURALPOP    NUMERIC       17          0      1.0000E+030
         #  4   STATES      ALPHA         50          0


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  REGRESSION

         Purpose: Performs a multiple regression analysis.

              A. REGRESSION yvar ON xvar1,xvar2,...
              B. REGRESSION yvar ON xvar1,xvar2,... ,mod1,mod2,...
                  yvar=dependent variable name
                   xvari=independent variable name
                    modi=one of the following modifiers
                            where tabname is PARAM,FIT, ANOVA,
                            where opt is YES or NO

         Syntax Examples:


         The   REGRESSION   command  computes  the  linear  regression
         equation  relating  the  dependent  variable, yvar, to one or
         more independent variables, xvar1,  xvar2,  ...,  xvarn.   It
         also provides analysis of variance (ANOVA), standard error of
         the  parameters  (SEPARAM)  and  fit  (FIT)  statistics.  The
         Durbin-Watson   (DURBIN)   statistics   for   testing  serial
         correlation  among residuals may be calculated.  An equation,
         with or without an intercept, may be fit  to  the  data.   In
         combination  with  the  COMPUTE  command,  REGRESSION  allows
         residual computations to be derived from the fitted data.

         The equation derived by the regression method is:

                 Y =  a0 + a1 * X1 + a2 * X2 + a3 * X3 + ...

         Y is the variable that we wish to determine as a  combination
         of  the  variables  X1, X2, etc.  The parameters a0, a1, etc.
         are determined by the  program  using  the  method  of  least

         The  above equation is sometimes referred to as a model.  The
         form shown is the intercept model.  If the  a0  parameter  is


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         removed  from  the  model,  it  is   referred   to   as   the
         no-intercept,   or  origin,  model.   The  use  of  the  word
         intercept  refers  to the fact that the equation represents a
         plane in space which intercepts the coordinate  axes  of  the
         space.   The  no-intercept  form  passes directly through the
         origin of the coordinate system.  REGRESSION computes  either
         form depending on the setting of the INTERCEPT modifier.  The
         above  equation,  with a0, is the one most often used in data
         analysis and forecasting.

         The TABLE modifier allows any  or  all  of  the  four  output
         tables,  PARAM, FIT, ANOVA, and SEPARAM, to be selected.  The
         ALL option selects all four tables.  The  INTERCEPT  modifier
         is used to select whether an intercept form of the regression
         is to be used.

         Detailed Example:

         In  the  following example, the REGRESSION command is used to
         find a regression equation in  the  intercept  form  for  the
         HALD.DAT  data  file  shown in the appendix.  The application
         involves trying to relate the  heat  produced,  HEATPROD,  in
         cement  production to the amount of certain materials present
         in the cement:  ALUMINATE, SILICATE, FERRITE and  DISILICATE.
         Examination  of the PARAMETER table output shows the relation
         to be:

            HEADPROD = 62.41 - 0.144*DISILICATE + 0.102*FERRITE
                             + 0.510*SILICATE + 1.551*ALUMINATE

         The  TABLE=ALL  option  produces  four  output  tables.   The
         command  is  long enough that the first portion is terminated
         with a comma to permit continuation of the remainder  of  the
         command after the CONTINUE:  prompt.

         Continue:  DISILICATE,TABLE=ALL

            13 CASES
             0 MISSING



         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         ---------- ----------------
         ALUMINATE          1.551102
         SILICATE           0.510167
         FERRITE            0.101909
         DISILICATE        -0.144061

         INTERCEPT         62.405403

            FIT TABLE

         R-SQUARE  0.9736   ADJ. R-SQUARE  0.9366   STD. ER. RES    2.4460

            ANOVA TABLE

         ----------  ---- --------------  -------------- ----------
         REGRESSION     4     2667.89900       666.97460   111.4792
         ERROR          8       47.86361         5.98295     0.0264
         TOTAL         12     2715.76200       226.31350

         P > |F| = 0.0000


         PARAMETER    ESTIMATE      T-VALUE     STD. ER.     P > |T-VAL|
         -----------  ------------  ----------  -----------  -----------
         ALUMINATE        1.551102    2.082660      0.74477      0.0758
         SILICATE         0.510167    0.704858      0.72379      0.5037
         FERRITE          0.101909    0.135031      0.75471      0.8964
         DISILICATE      -0.144061   -0.203175      0.70905      0.8448


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  REMARK

         Purpose: Allows information remarks to document a ezstep

              REMARK rmrktxt
                  rmrktxt is any text representing a remark

         Syntax Examples:


         The  REMARK  command  does not cause anything to happen other
         than for ezstep  to  issue  another  Command:   prompt.   Its
         purpose  is to provide a way of documenting a ezstep terminal
         session.  The text following  the  command  may  contain  any

         Detailed Example:

         In  this example REMARK is used to comment on the next use of
         the STATISTICS command.


         Command:  STATISTICS NEWDATA


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  SET

         Purpose: Sets destination of ezstep output.

              A. SET COPY=opt
              B. SET SEED=sno
                   opt = HARDCOPY, or OFF, or FILE fname
                       where fname = file name
                   sno = integer for random number seed

         Syntax Examples:
              SET COPY=HARDCOPY
              SET SEED=1832


         When COPY=HARDCOPY, all output results from executed commands
         are sent to the printer.  The output does not appear  on  the
         terminal.   When  COPY=OFF  is used, output results appear on
         the terminal.  When COPY=FILE fname is used,  the  output  of
         commands is placed in the file name, fname, designated.  This
         file can be edited and displayed by other programs, including
         text  editors  such  as  WORDSTAR.   Unless  changed  by this
         command, output results are  sent  to  the  terminal.   Error
         messages  and  command  prompts  are  always  directed to the
         terminal.  Whenever output is first  directed  to  a  results
         file,  it  replaces  already existing data on the file.  Data
         are not appended to the file.

         With the SEED modifier, you may set the  value  of  the  seed
         used  for  random  number  generation.    The   seed   number
         initializes the random number generators.  Normally, the seed
         is the same number each time you use ezstep.   Hence,  it  is
         possible  to  generate  the same random numbers each time you
         use ezstep.  If this is undesirable, you may want  to  choose
         different  seeds with the SET command.  Use any number from 1
         to 30000.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Detailed Example:

         In this example, SET is used to direct results to the printer
         instead of the terminal.

         Command:  SET COPY=HARDCOPY





         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  SHOW

         Purpose: Displays internal STATMATE settings.

              A. SHOW


         Syntax Examples:


         SHOW  displays  internal information that is set by the user.
         For example, items displayed are  the  value  of  the  random
         number  seed  used  in  LET,  the  current setting of the SET
         command, limits of the database being used and default limits
         of any new database created.

         The default information displayed is taken  from  information
         supplied  through  the  STATMATE install program, EZINST (see
         appendix C).  This information includes  default  limits  for
         any database created by STATMATE and the location of internal
         files.   Since  it  is  possible  to  have  used the STATMATE
         install program with STATMATE, default  information  relevant
         to program groups used in STATMATE is displayed also.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Detailed Example:

         The  following  example  illustrates  the  use  of  the  SHOW

         Command:  SHOW

           MAXIMUM VARIABLES:    50
           MAXIMUM CASES:       250
           HIGHEST USED VAR:     49

           RANDOM NO. SEED:    8632

                INSTALLED SETTINGS

           MAXIMUM VARIABLES:    10
           MAXIMUM CASES:      1000

           GROUP           DISK DRIVE
           --------------  ----------
           NUCLEUS         CURRENT
           STATISTICS      CURRENT
           REGRESSION      CURRENT


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  STATISTICS

         Purpose: Produces elementary statistics.

              A. STATISTICS vname
              B. STATISTICS vname
                   vname=name of variable

         Syntax Examples:
              STATISTICS SALES
              STAT WEIGHT


         A number of useful statistical  quantities  such  as  totals,
         averages, and variances may be computed for a single variable
         from the STATISTICS command.

         The  statistics  may  be  broken  into  the  following   five

                 Summation                       Spread
                   Totals                          Standard Deviation
                   Sum of Squares                  Variance
                   Sum of Squares about Mean       Range
                                                   Coeff. of Variation

                 Central Tendency                Distribution
                   Average or Mean                 Minimum and Maximum

                                 Higher Moments
                                   Kurtosis (thickness of tail)


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Detailed Example:

         The STATISTICS command is used to compute a complete  set  of
         statistics  on  1964 motor death data from the MOTOR.DAT file
         in the appendix.  The variable MTRDEATHS represents the data.



            50 CASES
             0 MISSING

      CENTRAL TENDENCY       SPREAD                      DISTRIBUTION
      -------------------    ------------------------    --------------------
      MEAN         926.76    STD. DEV.         889.32    MINIMUM        43.00
                             VARIANCE       790887.37    MAXIMUM      4743.00
                             RANGE            4700.00
                             COEFF. VAR.         0.96
                             MIDSPREAD         789.00

         SUMMATIONS                 HIGHER MOMENTS
         -----------------------    --------------------
         TOTAL          46338.00    SKEWNESS        2.07
         SUM SQ      81697687.00    KURTOSIS        8.48
         SUM SQ(DEV) 38753481.12


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  TTEST

         Purpose: Perform a comparison of two data sets using the T-Test.

              TTEST vname ON classvar
                 vname=variable name
                 classvar=classification variable name

         Syntax Examples:
              TTEST DOSAGE ON SEX


         A comparison of the means of two data sets (samples)  may  be
         performed  using  TTEST.   The means of two sets are computed
         and compared by computed using the  Student  T-test.   Values
         are  calculated  using  assumptions of both equal and unequal
         variances.  In the case of unequal variances, the degrees  of
         freedom  is calucated using Satterthwaite's approximation.  A
         table of means and standard deviations for each  of  the  two
         sets is produced.

         The  analysis  is  performed  on the data contained in vname.
         The variable classvar determines which of  the  two  sets  or
         classes  the  corresponding value in vname belongs.  That is,
         classvar contains codes that indicate which category or class
         the data in vname belong.  Although not  strictly  necessary,
         class  codes  should be coded as integer (1, 2, ...)  values.

         Detailed Example:

         In the example, the PRINT command is used  to  first  display
         the data to be used in TTEST.  CHEMRESULT contains the weight
         of  a  substance  after a chemical reaction as produced by 20
         individual experiments.  Two laboratories were involved and 4
         experiments were performed at one lab and the remaining 6  at
         the  other  lab.  The laboratory performing the experiment is
         coded   in   LABCODE.   TTEST  is  used  to  determine  if  a
         significant  difference  exists in the procedures used by the
         laboratories in performing the experiment.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         ----------- -----------
               40.50        1.00
               42.30        1.00
               50.80        2.00
               46.00        1.00
               47.80        1.00
               42.20        1.00
               49.00        2.00
               36.90        1.00
               44.40        2.00
               39.00        2.00



         10 CASES
          0 MISSING


    ---------- ----- -----------  ------------  ----------  ----------
           1       6     42.6167        3.8923     36.9000     47.8000
           2       4     45.8000        5.2738     39.0000     50.8000


    VARIANCE     T-VALUE      DF     PROB > |T-VALUE|
    --------  -----------  ------    ----------------
    EQUAL         -1.1055       8        0.3011
    UNEQUAL       -1.0340       6        0.3410


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  WHEN--ELSE--END

         Purpose: Permits data to be selected for analysis from the
                  database according to logical and relational conditions
              A. WHEN relexp
              B. WHEN relexp1 AND relexp2
              C. WHEN relexp1 OR relexp2
              where relexp if one of the following
                 vname relation c1
                 vname = c1, c2, ...
                 vname = c1 THRU c2
                 vname = variable name
                 relation = one of the relations: =, >, <, >=, or <=
                 c1 and c2 = numeric or alphanumeric constants

         Syntax Examples:
              WHEN AGE=13 THRU 30
              WHEN COLOR='RED','BLUE','PINK'
              WHEN WIDTH>25.55
              WHEN STOCK<=4500.00
              WHEN #0=40 THRU 80
              WHEN AGE=21 THRU 29 AND WEIGHT=150 THRU 175


         The WHEN command is a very useful command for analyzing  data
         according  to  specific  selection  criteria.   For  example,
         consider  a  variable  containing  ages  of  individuals  and
         another  containing  their weights.  It might be important to
         find the average weight of this group for all members between
         the ages of 50 and 65.  The  WHEN  command  can  be  used  to
         select  data  according  to  this  criterion,  and  then  the
         STATISTICS command can be used to find the required  average.
         The  ELSE  command  reverses,   or   negates,   a   condition
         established by WHEN.  The END command removes any  conditions
         established by WHEN or ELSE.  Only one WHEN or ELSE may be in

         Simple  forms  of  WHEN  relational  expressions  include the
         following relational operators:

                         Operator        Meaning
                         --------        -------------
                            =            Equality
                            >=           Greater than or equal
                            <=           Less than or equal
                            >            Greater than
                            <            Less than

         For example, YEAR>1950 means select all cases for  which  the


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         year is greater than 1950.

         Another  useful  conditional  expression  is  illustrated  by
         AGE=15,20,25,26,27.  In this example, all cases which have an
         age corresponding to 15, 20, 25, 26 and 27 are selected.

         A frequently useful conditional expression is illustrated  by
         VOLUME=405.44  THRU  650.25.   This  form  of  a  conditional
         expression permits a range of values to be selected.  In this
         example, all cases which have a volume from 405.44 to  650.25
         are selected for use.

         There  are  intances  when  it  is  desirable   to   compound
         conditions with a logical "and" or "or".  STATMATE allows two
         simple expressions to be logically combined in such a  manner
         by  using  an  AND or OR to separate simple expressions.  For
         example, AGE=15 THRU 40 AND WEIGHT>190.55 selects  all  cases
         for  which  the  age  is  from  15  to 40 and whose weight is
         greater than 190.55.

         The WHEN--ELSE commands operate on  a  case  by  case  basis.
         When  data  is  selected with a WHEN, STATMATE effectively is
         restricted to a window or view of your total data.  The  view
         extends  across  all variables on a case by case basis.  That
         is, any variable in the database may be used, but only  those
         cases selected by the WHEN condition are used.

         To  better understand how the data view established by a WHEN
         operates, consider the following  variables  after  the  WHEN
         command  WHEN  AGE=30  THRU  70  has  been  used.   The cases
         designated by the small x are  the  part  of  the  data  view
         established by the command.

                         AGE     WEIGHT
                         ---     ------
                          22      140.5
                 x        33      177.2
                          15      105.4
                          10       88.0
                 x        54      188.3
                 x        38      224.5

         Although  the  selection  was  made on AGE, only the cases in
         WEIGHT corresponding to those selected in AGE are  accessible
         when  WEIGHT  is  used  by  an  analytic  command   such   as
         STATISTICS.  That is, STATISTICS  WEIGHT  would  compute  the
         average  of  177.2,  188.3 and 224.5.  Incidentally, applying
         ELSE now would put the cases corresponding to  ages  22,  15,
         and 10 in the view.

         Some  important  aspects  of  the  WHEN  command  should   be
         understood when an attempt is made to write into the database
         with a COMPUTE, LET  or  RECODE  command  while  WHEN  is  in
         effect.   Perhaps  the  best  way to explain the effect is to


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         consider an example.  Suppose we consider the three variables
         AGE,  WEIGHT,  DOSAGE  and  NEWDOSAGE  shown  below.   Assume
         further  that  WHEN  AGE=30  THRU  70 has been applied to the
         database, as described  earlier.   The  x  symbols  show  the
         selected cases.

                         AGE     WEIGHT  DOSAGE  NEWDOSAGE
                         ---     ------  ------  ---------
                          22      140.5   15.2     empty
                 x        33      177.2   11.4
                          15      105.4   17.8
                          10       88.0   10.4
                 x        54      188.3   11.3
                 x        38      224.5   21.5

         Note  that WEIGHT and DOSAGE contain data, but NEWWEIGHT does
         not.  Assume the following two  LET  commands  are  executed:
         LET  NEWWEIGHT=WEIGHT+100  and  LET  DOSAGE=DOSAGE+2.0.   The
         result is:

                         AGE     WEIGHT  DOSAGE  NEWWEIGHT
                         ---     ------  ------  ---------
                          22      140.5   15.2    missing
                 x        33      177.2   13.4     277.2
                          15      105.4   17.8    missing
                          10       88.0   10.4    missing
                 x        54      188.3   13.3     288.3
                 x        38      224.5   23.5     324.5

         Note that new values have been calculated for  cases  in  the
         view,  but  not for cases outside of the view.  Further, note
         that because NEWWEIGHT did not have data, missing values  are
         placed at cases not within the view.

         While  the  WHEN is in effect, some care must be exercised in
         writing data into variables that have been used to select the
         view.  For example, WHEN  AGE=10  THRU  25  followed  by  LET
         AGE=AGE+5 changes AGE but does not affect the view.  That is,
         the  WHEN  selects  cases whose age is from 10 to 25, but the
         subsequent LET recalculates the ages  so  that  the  selected
         cases  have  ages  between  15  and  30.   In this situation,
         STATMATE does not reselect cases to conform to  the  previous

         Detailed Example:

         In  the example shown below, WHEN is used to select all cases
         of YEAR which are greater than 1890.  The STATISTICS  command
         is  then  executed  on  URBANPOP  to find statistics of urban
         population.  The statistics produced are those for the  urban
         population  from  1900 to 1950.  ELSE is then used to reverse
         the condition, and  statistics  are  computed  on  the  urban
         population  before  1900.   PRINT is used to display YEAR and


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         URBANPOP before and after the use of the WHEN and ELSE.   END
         is used to restore the data view to the full view.

         Command:  PRINT YEAR,URBANPOP

         YEAR        URBANPOP
         ----------- -----------
             1790.00        0.20
             1800.00        0.32
             1810.00        0.52
             1820.00        0.69
             1830.00        1.13
             1840.00        1.84
             1850.00        3.54
             1860.00        6.22
             1870.00        9.90
             1880.00       14.13
             1890.00       22.11
             1900.00       30.16
             1910.00       42.00
             1920.00       54.16
             1930.00       68.95
             1940.00       74.92
             1950.00       88.93

         Command:  WHEN YEAR>1890

              6 OF   17 CASES

         Command:  PRINT YEAR,URBANPOP

         YEAR        URBANPOP
         ----------- -----------
             1900.00       30.16
             1910.00       42.00
             1920.00       54.16
             1930.00       68.95
             1940.00       74.92
             1950.00       88.93

         Command:  STATISTICS URBANPOP


             6 CASES
             0 MISSING

         CENTRAL TENDENCY       SPREAD                      DISTRIBUTION
         ----------------  ------------------------  --------------------
         MEAN       59.85  STD. DEV.          21.85  MINIMUM        30.16
                           VARIANCE          477.63  MAXIMUM        88.93
                           RANGE              58.77
                           COEFF. VAR.         0.37


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         SUMMATIONS                 HIGHER MOMENTS
         -----------------------    --------------------
         TOTAL            359.12    SKEWNESS       -0.07
         SUM SQ         23883.03    KURTOSIS        1.74
         SUM SQ(DEV)     2388.15

         Command:  ELSE

             244 OF   250 CASES

         Command:  PRINT YEAR,URBANPOP

         YEAR        URBANPOP
         ----------- -----------
             1790.00        0.20
             1800.00        0.32
             1810.00        0.52
             1820.00        0.69
             1830.00        1.13
             1840.00        1.84
             1850.00        3.54
             1860.00        6.22
             1870.00        9.90
             1880.00       14.13
             1890.00       22.11

         Command:  STATISTICS URBANPOP


            11 CASES
             0 MISSING

         CENTRAL TENDENCY       SPREAD                      DISTRIBUTION
         ----------------  ------------------------  --------------------
         MEAN        5.51  STD. DEV.           7.14  MINIMUM         0.20
                           VARIANCE           50.92  MAXIMUM        22.11
                           RANGE              21.90
                           COEFF. VAR.         1.29

         SUMMATIONS                 HIGHER MOMENTS
         -----------------------    --------------------
         TOTAL             60.61    SKEWNESS        1.34
         SUM SQ           843.17    KURTOSIS        3.61
         SUM SQ(DEV)      509.17

         Command:  END



         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         Command:  WRITE

         Purpose: Places data from the database onto an external file
              A. WRITE vname1,... ON filenm
                 vname1 = variable name
                 filenm = a file name

         Syntax Examples:


         The  WRITE  command  places  data  from  the  database onto a
         designated file.  The list  of  variables  specified  in  the
         command are written case by case onto the file in ASCII form.
         Files  in  the ASCII format may be printed on your system, or
         modified by word processors, such as  WORDSTAR.   With  minor
         modifications,  the  file  may  be  used  as  input  to other
         application programs, such as dBASE II  (III),  which  accept
         ASCII data.

         When  the output file is written, the very first line of data
         contains information which enables the file  to  be  read  by
         STATMATE,  if  it  is  desirable  to  re-enter  the data into
         STATMATE again.  The format of the first line is the same  as
         discussed  in  the section ENTERING DATA INTO THE SYSTEM (See
         the discussion of ASCII  files).   Also,  the  names  of  the
         variables  placed  on  the file are listed on the first line.
         Each subsequent line of output contains  one  case  for  each
         variable specified in the command.

         Detailed Example:

         In  the  example  below, WRITE is used to place the data from
         variables YEAR and URBANPOP on the file POPDATA.DAT.


              17 CASES AND   2 VARIABLE(S) WRITTEN


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                         This page deliberately blank




         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                APPENDIX A: COMPUTATIONAL METHODS (Abbreviated)


         The  computations  of  CORRELATE  may  be   found   in   many
         statistical texts; see in particular Ostle in the references.


         Most of the statistics computed by STATISTICS may be found in
         any standard text on statistics.  Sums and sums of powers are
         calculated using provisional methods.


         The basis for the uniform random number generator used in the
         LET   functions   is   the  Wichmann  article  cited  in  the
         references.   Normally  distributed  numbers are generated by
         summing twelve numbers generated from a uniform  distribution
         and  applying appropriate transformations to scale the result
         to the desired mean and standard deviation.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                            APPENDIX B: SAMPLE DATA

         This appendix describes three data files which  are  included
         with the STATMATE package:

              1.  U. S. Population Data
              2.  Motor Vehicle Death Data
              3.  Hald Cement Data

         1.  U. S. Population Data

         The  data listed below represents U. S.  Population data from
         1790 through 1950.

                 Column  Description
                 ------  ----------------------------
                   1     Year
                   2     Urban population in millions
                   3     Rural population in millions

                 1790 ,  0.202 ,  3.728
                 1800 ,  0.322 ,  4.986
                 1810 ,  0.525 ,  6.714
                 1820 ,  0.693 ,  8.945
                 1830 ,  1.127 , 11.739
                 1840 ,  1.845 , 15.224
                 1850 ,  3.544 , 19.648
                 1860 ,  6.217 , 25.227
                 1870 ,  9.902 , 28.656
                 1880 , 14.130 , 36.026
                 1890 , 22.106 , 40.841
                 1900 , 30.160 , 45.835
                 1910 , 41.999 , 49.973
                 1920 , 54.158 , 51.553
                 1930 , 68.955 , 53.830
                 1940 , 74.924 , 57.246
                 1950 , 88.927 , 61.770


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         2.  Motor Vehicle Death Data

         The data below may be found in Draper and  Smith,  page  191,
         given  in  the  references.   Note  the presence of a missing
         value in the rural road mileage for Washington, D. C..

                 Column  Description
                 ------  ---------------------------------------------
                   1     States
                   2     (Y) Motor vehicle deaths in 1964
                   3     Number of drivers (in units of 10,000)
                   4     Persons per sq. mi. in 1960
                   5     Rural road mileage (1,000)
                   6     Percentage of males greater than females
                         (0=no, 1=yes)
                   7     January maximum temperature
                   8     Highway fuel consumption in 1964 (10 million)

                 DATA MOTOR VEHICLE DATA
                 AL,   968, 158,    64.0,  66.0, 0, 62, 119.0
                 AK,    43,  11,     0.4,   5.9, 1, 30,   6.2
                 AZ,   588,  91,    12.0,  33.0, 1, 64,  65.0
                 AR,   640,  92,    34.0,  73.0, 0, 51,  74.0
                 CA,  4743, 952,   100.0, 118.0, 0, 65, 105.0
                 CO,   566, 109,    17.0,  73.0, 0, 42,  78.0
                 CT,   325, 167,   518.0,   5.1, 0, 37,  95.0
                 DE,   118,  30,   226.0,   3.4, 0, 41,  20.0
                 DC,   115,  35, 12524.0, 1E+30, 0, 44,  23.0
                 FL,  1545, 298,    91.0,  57.0, 0, 67, 216.0
                 GA,  1302, 203,    68.0,  83.0, 0, 54, 162.0
                 ID,   262,  41,     8.1,  40.0, 1, 36,  29.0
                 IL,  2207, 544,   180.0, 102.0, 0, 33, 350.0
                 IN,  1410, 254,   129.0,  89.0, 0, 37, 196.0
                 IA,   833, 150,    49.0, 100.0, 0, 30, 109.0
                 KS,   669, 136,    27.0, 124.0, 0, 42,  94.0
                 KY,   911, 147,    76.0,  65.0, 0, 44, 104.0
                 LA,  1037, 146,    72.0,  40.0, 0, 65, 109.0
                 ME,   196,  46,    31.0,  19.0, 0, 30,  37.0
                 MD,   616, 157,   314.0,  29.0, 0, 44, 113.0
                 MA,   766, 255,   655.0,  17.0, 0, 37, 166.0
                 MI,  2120, 403,   137.0,  95.0, 0, 33, 306.0
                 MN,   841, 189,    43.0, 110.0, 0, 22, 132.0
                 MS,   648,  85,    46.0,  59.0, 0, 57,  77.0
                 MO,  1289, 234,    63.0, 100.0, 0, 40, 180.0
                 MT,   259,  38,     4.6,  72.0, 1, 29,  31.0
                 NB,   450,  89,    18.4,  97.0, 0, 32,  61.0
                 NE,   215,  23,     2.6,  44.0, 1, 40,  24.0
                 NH,   158,  37,    67.0,  13.0, 0, 32,  23.0
                 NJ,  1071, 329,   807.0,  21.0, 0, 43, 231.0
                 NM,   387,  54,     7.8,  62.0, 1, 46,  48.0
                 NY,  2745, 744,   350.0,  84.0, 0, 31, 439.0
                 NC,  1580, 226,    93.0,  71.0, 0, 51, 177.0
                 ND,   185,  38,     9.1, 102.0, 1, 20,  24.0


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                 OH,  2096, 530,   237.0,  84.0, 0, 41, 358.0
                 OK,   785, 137,    34.0,  94.0, 0, 46, 107.0
                 OR,   575, 108,    18.0,  73.0, 0, 45,  81.0
                 PA,  1889, 570,   252.0,  89.0, 0, 39, 353.0
                 RI,   100,  46,   812.0,   1.3, 0, 38,  27.0
                 SC,   870, 122,    79.0,  52.0, 0, 61,  86.0
                 SD,   270,  40,     9.0,  87.0, 1, 23,  28.0
                 TN,  1059, 177,    85.0,  67.0, 0, 49, 135.0
                 TX,  3006, 515,    37.0, 196.0, 0, 50, 448.0
                 UT,   295,  57,    10.8,  32.0, 0, 37,  38.0
                 VT,   131,  20,    42.0,  13.0, 0, 25,  15.0
                 VA,  1050, 208,   100.0,  50.0, 0, 50, 150.0
                 WA,   730, 160,    43.0,  59.0, 1, 46, 109.0
                 WV,   467,  88,    77.0,  32.0, 0, 43,  54.0
                 WI,  1059, 207,    72.0,  87.0, 0, 26, 141.0
                 WY,   148,  22,     3.4,  67.0, 1, 37,  20.0

         3.  Hald Cement Data

         The data shown below may be found in Draper and  Smith,  page
         630, given in the references.

                 Column  Description
                 ------  --------------------------------------------
                   1     Amount of tricalcium aluminate (% clinker wgt)
                   2     Amount of tricalcium silicate
                   3     Amount of tetracalcium ferrite
                   4     Amount of dicalcium silicate
                   5     (Y) Heat produced in hardening cement

                  7,26, 6,60, 78.5
                  1,29,15,52, 74.3
                 11,56, 8,20,104.3
                 11,31, 8,47, 87.6
                  7,52, 6,33, 95.9
                 11,55, 9,22,109.2
                  3,71,17, 6,102.7
                  1,31,22,44, 72.5
                  2,54,18,22, 93.1
                 21,47, 4,26,115.9
                  1,40,23,34, 83.8
                 11,66, 9,12,113.3
                 10,68, 8,12,109.4


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         STATMATE Components

         The  STATMATE  package  is   supplied   on   several   disks.
         Generally, the disks contain a set of compiled programs  with
         file  extensions  of  .OVL and .COM (.OVL and .EXE for the PC
         DOS-MS DOS version).  A set of sample  data  sets  with  file
         extensions  of .DAT are included.  As well, the disks contain
         an SMSA$ file, an SMHELP.TXT file and an  SMINSTLL.COM  (.EXE
         for  DOS)  program  file.   It is advisable to make a copy of
         these disks.  Use the copies as your working disks.

         Once you have made backup copies, you  are  ready  to  begin.
         Make  sure  that  the  disk  containing  SMATE.COM is in your
         A-drive when  you  execute  STATMATE.   See  the  section  on
         STATMATE  Operation  for  instructions  on how to operate the
         package.  If you have space problems fitting the package onto
         your   system,  see  the  section  in  this  appendix  titled
         Tailoring STATMATE to Your System.

         STATMATE Internal Files

         In  order  to operate, STATMATE creates several files for its
         internal  use.   These  include  SMSA$,  xxxSM$DI,  xxxSM$DB,
         xxxSM$S1,  xxxSM$PL,   xxxSM$WH,   xxxSM$EQ   and   xxxSM$PL,
         xxxSM$CH, SM$CU, where xxx represents the ID provided by  the
         user.   Some  versions  of  STATMATE  may  produce additional
         files, but they are always prefaced by  xxSM$.   SMSA$  is  a
         control  file  containing  installation  information (problem
         size, etc.).

         Tailoring STATMATE to Your System

         STATMATE contains an SMINSTLL program  which  allows  you  to
         tailor  STATMATE  to  meet  various  disk  space needs and to
         modify problem and data size parameters.  Let us examine  how
         SMINSTLL can be used to help solve disk space needs.

         In  order  to  accomodate  different  disk  space  needs, the
         STATMATE package provides a way of distributing  the  various
         programs  over  several  disks  by  dividing STATMATE program
         files into five groups.   These  groups  designate  the  disk
         location  of  the  programs for particular commands and files
         for internal STATMATE use.  The  following  table  shows  the
         five  groups  and  the  commands  and files controlled by the


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                 Group           Controls Disk for
                 --------------  ----------------------------------
                 Internal Files  STATMATE files
                 Nucleus         ERASE, EXECUTE, GIVE, HELP, WRITE
                                 INPUT, PRINT, QUERY, REMARK
                                 SET, EXIT, WHEN, ELSE, END
                 Statistics      STATISTICS, BREAKDOWN
                                 ONEWAY, TTEST, TWOWAY,
                                 CROSSTAB, KOLMOGOROV,
                                 TNPARAM, ONPARAM, RCORR
                 Miscellaneous   PLOT, HISTOGRAM, EDIT, LET, RECODE
                                 CHART, CUSUM
                 Regression      REGRESSION, CURVE, POLYNOMIAL,
                                 NONLINEAR, COMPUTE, CORRELATE,

         For   example,  the  Regression  group  controls  which  disk
         STATMATE  expects  to  find  the  programs  for the CURVE and
         REGRESSION   commands.   Through  the  use  of  the  SMINSTLL
         program,  discussed in the next paragraph, the user may alter
         the  disk  locations  where  STATMATE  expects  to  find  the
         programs for its commands.

         In  order  to  modify  the  delivered  configuration,  it  is
         necessary to use the SMINSTLL program provided with STATMATE.
         SMINSTLL is an interactive program that asks for the  desired
         disk  drive  of  the groups shown above.  Once the drives for
         these groups are specified, SMINSTLL will  give  instructions
         on  how  STATMATE programs should be distributed on your disk
         drives.  Information about the new configuration is placed on
         the SMSA$ file, which is used by STATMATE to  determine  what
         configuration  is  to  be  used  during  execution.   It   is
         necessary   to   operate   STATMATE  in  the  specified  disk
         configuration.   STATMATE   only   knows   of   the   current
         configuration as specified by SMINSTLL.

         A second need solved by SMINSTLL, is the  ability  to  change
         database  and  problem  size  parameters.   As delivered, the
         maximum number of variables allowed in the database is 10 and
         the maximum number  of  cases  that  can  be  placed  in  the
         database  is  250.   Use  SMINSTLL  to  change  these values.
         SMINSTLL will  query  you  for  the  values.   For  the  best
         results,  change  the  maximum  number  of  variables to some
         multiple of 5, for example, 15.

         Once you  use  SMINSTLL  to  change  these  values,  all  new
         databases will be of the specified sized.  STATMATE maintains
         these  size parameters with each database.  Hence, previously
         created databases can be used without  any  difficulty.   The
         size of an existing database cannot be changed.

         Some  care should be exercised in specifiying the size of the
         database.  A database with a maximum of 10 variables and  250
         cases  occupies  about 5K bytes of file space.  This space is


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         assigned  immediately,  and  not  as  you  add  data  to  the
         database.   If you specify a database of 50 variables and 500
         cases, then you will use about 50K bytes of file  space.   If
         you  only  work  with at most 200 cases and 5 variables, only
         part of the  database  will  be  used.   The  remainder  will
         nevertheless occupy file space on your disk.

         Note  that  there  is  no  way  to  change individual command
         limitations on problem sizes.  For example, there is  no  way
         to  increase the size of problems that can be accommodated by

         User's with a hard disk should use SMINSTLL, and specify that
         all groups belong on the  current  directory.   All  STATMATE
         files should then be placed in a single directory.

         When  using SMINSTLL, it is probably a good idea to print the
         output   so   that  you  will  have  a  record  of  how  your
         configuration  should  be  installed.   The  names of program
         files which are to be placed on specific disk drives may vary
         with   the   version  of  STATMATE.   The  example  below  is
         representative  of  the  interaction  with  SMINSTLL.  In any
         case,   the  output  instructions  from  SMINSTLL  should  be
         followed when you actually perform the installation.

         SMINSTLL Example

         In  the  following  example,  SMINSTLL  is  used  to create a
         configuration that establishes databases with a maximum of 40
         variables and 500 cases.  The location of internal  files  is
         specified  as  the  B-disk.   We  begin  by entering the word
         SMINSTLL at the terminal.


                     INSTALL PROGRAM

         Note: [ ] denotes default value

            MAXIMUM VARIABLES [10]:  40
            MAXIMUM CASES [250]:  500
            NOTE: Your STATMATE database file will occupy
                  about   80K bytes of disk storage.



                         --- File Distribution Settings ---

         STATMATE Internal File Group Files Will Be Generated on  B Drive

             CP/M        PC DOS/MS DOS
         --------------  -------------


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

           xxxSM$DB       xxxSM$DB
           xxxSM$DI       xxxSM$DI
           xxxSM$S1       xxxSM$S1
           xxxSM$WH       xxxSM$WH
           xxxSM$EQ       xxxSM$EQ
           xxxSM$PL       xxxSM$PL
           xxxSM$ST       xxxSM$ST
           xxxSM$CU       xxxSM$CU
           xxxSM$CH       xxxSM$CH

         where xxx is the user ID

            Press carriage return to continue

         Make sure all of the following STATMATE files are on the same diskette:

                   CP/M        PC DOS/MS DOS
              ---------------  -------------
            1. STATMATE.COM     STATMATE.EXE
            2. All .OVL files   .OVR files
            3. SMSA$ file       SMSA$ file
            4. SMHELP.TXT       SMHELP.TXT

         End of installation


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         Command problem size  limitations  are  shown  in  the  table


         Command  problem  size  limitations  are  shown  in the table

                         Command Limitations

         Command         Cases   Var  Classes    Other Limitations
         ------------    -----  ----  -------    -------------------
         BREAKDOWN       32000   1    100
         CHART           32000   1    100        100 subgroups & 20 items
                                                 per subgroup
         COMPUTE         32000*  1    na**
         CORRELATE       32000   20   na
         CROSSTABS       32000   2    20
         CURVE           32000   2    na
         CUSUM           32000   1    100        100 subgroups & 20 items
                                                 per subgroup
         EDIT            32000   20   na
         ERASE           na      na   na
         EXECUTE         na      na   na
         EXIT            na      na   na
         GIVE            na      na   na         20 attribute
         HELP            na      na   na
         HISTOGRAM       32000   1    20         20 bars
         INPUT           32000   50   na
         KOLMOGOROV        500   2    na
         LET             32000   2    na
         NONLINEAR         250   5    na         5 parameters
         ONPARAM           500   2    xx
         ONEWAY          32000   2    20
         PLOT            32000   6    na         5 on y-axis and 1
                                                 on x-axis
         POLYNOMIAL        250   2    na         Degree less than 11
         PRINT           32000   6    na
         QUERY           na      na   na         20 variables per query
         RCORRELATION      500   2    na
         RECODE          32000   1    na         10 individual values
                                                 when recoding a set
                                                 of specific values
         REGRESSION      32000   20   na
         REMARK          na      na   na

                 *   32,767 for those who can remember or need it
                **   not applicable


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                         Command Limitations (Continued)

         Command         Cases   Var  Classes    Other Limitations
         ------------    -----  ----  -------    -------------------
         SET             na      na   na
         SHOW            na      na   na
         STATISTICS      32000   1    na         500 cases when
                                                 computing quartiles
         STEPWISE        32000   20   na
         TNPARAM            na   2    na         20 sets by 20 groups
         TTEST           32000   2    na
         TWOWAY          32000   3    20
         WHEN            32000   2    na
         WRITE           32000   20   na

         A STATMATE database may contain as many as 32,000 cases,  and
         several  hundred  variables.   When a database size is large,
         the limiting factor becomes the disk capacity of your system.

         Generally, analyses can process as many as 32,000  cases  and
         20  variables.   Specific  limitations are shown in the above


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

         APPENDIX E:  HELP

              1.   Question:  How do I enter alpanumeric data into the
                   Answer:  Read  the  INPUT  and  GIVE  TYPE  command

              2.   Question:  I keep running out of disk space because
                   the  disk  contains  many  STATMATE internal files.
                   What can I do?
                   Answer:  Try to keep your  use  of  identifiers  in
                   response  to  the  ID prompt to a minimum.  Use the
                   install program to create larger databases and  use
                   as few databases as possible.

              3.   Question:  Why won't INPUT read my data.
                   Answer:   Check your input file to see that data is
                   in the correct  fields  and  that  data  items  are

              4.   Question:  What needs to be done to print my output
                   on my printer?
                   Answer:  Use the SET COPY=HARDCOPY command.

              5.   Question:  Why can't the INPUT command find my data
                   Answer:   Use a disk identifier, such as B:, before
                   the file name to specify the disk the file  is  on.

              6.   Question:   Why  aren't  my PLOT modifiers retained
                   for repeated use.
                   Answer:  Use SAVE in the subcommand mode.

              7.   Question:  When I created one of  my  databases,  I
                   forgot  how many variables and cases I specified as
                   the maximum.  How can I find out what  the  maximum
                   is for a database.
                   Answer:  Use the SHOW command.

              8.   Question:   Is it possible to get data from another
                   program, such as dBASE into STATMATE.
                   Answer:  Create an ASCII file with the program  and
                   turn it into a 'DATA' file with your word processor
                   or editor.

              9.   Question:   Almost  every time I use the package, I
                   use the same seven or eight commands for  my  task.
                   Is their any way to reduce the effort.
                   Answer:  Use the EXECUTE command.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

              10.  Question:   PRINT  produces  too many cases to view
                   easily.  Can I reduce the output some way?
                   Answer:  Use the WHEN command to  restrict  output.

              11.  Question:   How  can  I  control  my  output to the
                   screen, it scrolls by so fast that it can't be read
                   Answer:  Use your  operating  system's  ability  to
                   hold  screen output with control keys.  Use the SET
                   COPY=FILE command to output the results to  a  file
                   for examination later.

              12.  Question:   Do  the  STATMATE database and internal
                   files and my data files need  to  be  on  the  same
                   Answer:   No.  Internal file usage is controlled by
                   the  install  program.   Your  data  files  may  be
                   referenced  with the INPUT command by preceding the
                   file name with a disk identifier, if necessary.


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987


         STATMATE/PLUS is distributed on diskettes with  the  contents
         listed  below.   If  you  are  executing  the  program   from
         diskettes, this is the  arrangement  that  you  should  begin
         with.   Disk 1 should be placed in your a-drive and disk 2 in
         your b-drive.  Disk 3 contains the shareware version  of  the
         STATMATE/PLUS  user's  guide  as  listable  files.  It is not
         needed to execute STATMATE.

         The suggested diskette  arrangement  should  leave  you  with
         about  30K of free space on disk 1, enough to use STATMATE as
         configured by the install  program,  SMINSTLL.EXE.   Some  of
         this space will disappear when you use STATMATE for the first
         time.   STATMATE  creates  a database file and other internal
         files as it is used.  If you need more space, consider moving
         SMINSTLL.EXE and the data files (.DAT) to  another  diskette.
         If  you  continue  to have space problems, see appendix C for
         information on installing STATMATE in  other  configurations.

                          Suggested Diskette Contents

                             Disk 1

         SMATE .EXE      --- main program
         SMASKC.OVR  SMATE .EXE  SMBRKD.OVR      --- overlay files
         SMSA$           --- configuration file
         SMHELP.TXT      --- help file
         SMINSTLL.EXE    --- install program
         DEMO            --- example command file
         USPOP.DAT   USPOPDEM.DAT        --- sample data files

                             Disk 2

         SMASUM.OVR  SMCHLX.OVR  SMCHPL.OVR  SMCHRT.OVR  --- overlay files

                              Disk 3

         SMPART1.DOC  SMPART2.DOC  SMPART3.DOC  SMPART4.DOC  --- user's guide


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                      APPENDIX G: INVOICE AND ORDER FORM

         A  sample  invoice and form is enclosed to simplify ordering.
         Use the order form to place an order.  The invoice form is to
         be used within your  organization  to  generate  payment  for
         STATMATE/PLUS registrations.

         STATMATE/PLUS  is  available  on disk for your evaluation and
         convenience for $35.  This fee only  covers  diskette  costs,
         handling  and  postage  (within the U.S.).  It does not cover
         registration.  Please show your support  by  registering  the
         program,  if  you are using it on a regular basis and find it
         of value.  Note that a 190+ page user's  guide  is  available
         for $35 and you can save $10 by purchasing the diskettes  and
         registering  at the  same  time.   Remember  that  registered
         owners receive several utilities when they register. Further,
         when you purchase either  the diskettes or registration,  you
         receive a coupon  worth $10  on your next  purchase.   If you
         purchase both, then you receive $10 off now and a $10 coupon.
         The coupon offer applies only to purchases made directly from
         the Software Hill.

         Business, commercial, governmental or educational institution
         use of non-registered copies  of  STATMATE/PLUS  is  strictly
         forbidden.   Write  for  details concerning site or corporate


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                               ---INVOICE FORM---

                 Remit to:       The Software Hill
                                1857 Apple Tree Lane
                                 Mt. View, CA 94040
                                  (415) 969-4233

         Sold to: ____________________   Ship to:  ______________________

                  ____________________             ______________________

                  ____________________             ______________________

             Date:                                   PO #:
         Items:                                            Qty      Price
          [ ] STATMATE Evaluation disks  $ 35            ______ $______.____

          [ ] STATMATE/PLUS Registration  $ 45           ______ $______.____

          [ ] STATMATE disks & registration $ 70 (save)  ______ $______.____

          [ ] STATMATE/PLUS User's Guide only $ 35       ______ $______.____

          [ ] California Residents add 7% sales tax              $_____.____

                         Total                         (U.S.) $________.____

          [ ] Ship COD via UPS or not U.S. mail; add $15         $______.___

          [ ] Ship outside of North America:

              Add $15 [ ] STATMATE/PLUS disks only           $______.___

              Add $25 [ ] STATMATE/PLUS manual & disks       $______.___

              Add $15 [ ] STATMATE/XG disks and guide        $______.___


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                               ---ORDER FORM---

                                The Software Hill
                               1857 Apple Tree Lane
                                 Mt. View, CA 94040
                                  (415) 969-4233

         Operating System and Computer:
          [ ] PC DOS   [ ] MS DOS  Version _____   Computer ______________

         Items:                                            Qty      Price
          [ ] STATMATE Evaluation disks  $ 35            ______ $_____.___
          [ ] STATMATE/PLUS Registration   $ 45          ______ $_____.___
          [ ] STATMATE disks & registration $ 70 (save)  ______ $______.__
          [ ] STATMATE/PLUS User's Guide only -- $ 35    ______ $_____.___
          [ ] California Residents add 7% sales tax              $____.___

                         Total                         (U.S.) $_______.___

          [ ] Ship COD via UPS or not U.S. mail; add $20     $______.___

          [ ] Ship outside of North America:
              Add $15 [ ] STATMATE/PLUS disks only           $______.___
              Add $25 [ ] STATMATE/PLUS manual & disks       $______.___
              Add $15 [ ] STATMATE/XG disks and guide        $______.___

          [ ] Check enclosed. Amount enclosed    $________._____ (U.S.)

           Make check               The Software Hill
           payable to:             1857 Apple Tree Lane
                                    Mt. View, CA 94040

                 1.  Allow three weeks for delivery
                 2.  Orders outside U.S. send check drawn on U.S. bank
                     or international money order. Amount in U.S. dollars.

             Date:                                   PO #:

                               ---Customer Information---

                 Name ______________________________________________

                 Address ___________________________________________



                 City    __________________ State ______  Country __________

                 Phone   (_____) ______ - _________


         STATMATE/PLUS Shareware User's Guide--Copyright (C), 1987

                             REFERENCES (Partial)

         Cooper,  B. E.,  Statistics  for  Experimentalists,  Pergamon
         Press Ltd., London, Englad, First Edition, 1969

         Draper,  N. R. and  Smith,  H.,  Applied Regression Analysis,
         John Wiley and Sons, New York, New York, Second Edition, 1981

         Duncan, A. J., Quality  Control  and  Industrial  Statistics,
         Richard  D. Irwin  Inc.,  Homewood, Illinois, Fourth Edition,

         Marquardt,   Donald   W.,   An  Algorithm  for  Least-Squares
         Estimation of Nonlinear Parameters, Journal of the SIAM, vol.
         11, no.  2, pp 431-441, June 1963.

         Ostle,  Bernard,   Statistics   in   Research,   Iowa   State
         Univ. Press, Ames, Iowa, Second  (now  in  Seventh  Edition),

         Seigel,  Sidney,  Nonparametric Statistics for the Behavorial
         Sciences, Chapters 6 and 8, New York, McGraw-Hill, 1956

         Tukey, John W., Exploratory  Data  Analysis,  Addison-Wesley,
         Reading, Massachussetts, 1977

         Wichmann,  B. A.   and  Hill,  I. D.,  A Psuedo-random Number
         Generator, NPL Report, DITC, June, 1982

         Winer, B. J., Statistical Principles in Experimental Designs,
         Mc Graw-Hill, New York, 1970



Directory of PC-SIG Library Disk #0863

 Volume in drive A has no label
 Directory of A:\

FILES863 TXT       751   9-07-88   2:18p
GO       BAT        38  11-05-87   3:26p
GO       TXT       463  11-05-87   3:56p
PRINTDOC BAT       789  11-05-87   3:59p
SMPART1  DOC     44886   8-31-88   6:45a
SMPART2  DOC     42240   8-07-87   5:42p
SMPART3  DOC     25216   8-07-87   5:43p
SMPART4  DOC     23808   8-07-87   5:44p
SMPART5  DOC     35845   8-31-88   6:47a
        9 file(s)     174036 bytes
                      143360 bytes free