MON 3/29/99 1:28:56 PM

SYSTAT VERSION 7.0.1
COPYRIGHT (C) 1997, SPSS INC.

Welcome to SYSTAT!

>USE 'C:\SYSTAT7\S209\CRAFT.SYD'
SYSTAT Rectangular file C:\SYSTAT7\S209\CRAFT.SYD,
created Sun Feb 07, 1999 at 12:06:20, contains variables:
 TYPE$        SIZE         SEASON       CLERKS
This program calculates the H matrix using the MATRIX procedure of SYSTAT
>list
  Case number        TYPE$         SIZE       SEASON       CLERKS
        1     STRSEW             12.300       73.000        4.800
        2     SAND                9.900       43.000        7.600
        3     VENT                8.200       29.000       11.700
        4     BRICK               5.500       47.000        3.300
        5     GENCON              6.900       43.000        5.200
        6     SHEET               4.900       29.000       11.700
        7     PLUMB               5.100       20.000       10.900
        8     ELEC                6.300       13.000       12.500
        9     PAINT               2.500       59.000        3.900

>matrix
Current program limits:
     Number of MATRICES = 15.
     Number of row and column NAMES = 230.
     Number of NUMERIC data elements = 4100 (32800 bytes).
     Number of CHARACTER data elements = 1100 (13200 bytes).

>use craft
SYSTAT Rectangular file craft.SYD,
created Sun Feb 07, 1999 at 12:06:20, contains variables:
 TYPE$        SIZE         SEASON       CLERKS
********** Active Matrix: CRAFT **********

>mat x = craft( ;size season)
********** Active Matrix: X **********

>mat const = m(9,1,1)
********** Active Matrix: CONST **********

>mat x = const||x
********** Active Matrix: X **********

>mat y = craft(; clerks)
********** Active Matrix: Y **********

>sho x
 
Matrix: X
                       SIZE       SEASON
         1.000       12.300       73.000
         1.000        9.900       43.000
         1.000        8.200       29.000
         1.000        5.500       47.000
         1.000        6.900       43.000
         1.000        4.900       29.000
         1.000        5.100       20.000
         1.000        6.300       13.000
         1.000        2.500       59.000
********** Active Matrix: X **********

>sho y
 
Matrix: Y
        CLERKS
         4.800
         7.600
        11.700
         3.300
         5.200
        11.700
        10.900
        12.500
         3.900
********** Active Matrix: Y **********

>mat xpxi = inv(trp(x)*x)
********** Active Matrix: XPXI **********

>sho xpxi
 
Matrix: XPXI
                                   SIZE       SEASON
                     1.024       -0.078       -0.010
 SIZE               -0.078        0.017       -0.001
 SEASON             -0.010       -0.001        0.000
********** Active Matrix: XPXI **********

>mat h = x*xpxi*trp(x)
********** Active Matrix: H **********

>sho h
 
Matrix: H
         0.722        0.325        0.105        0.092        0.143       -0.095
         0.325        0.251        0.189        0.037        0.109        0.033
         0.105        0.189        0.210        0.029        0.094        0.106
         0.092        0.037        0.029        0.180        0.124        0.124
         0.143        0.109        0.094        0.124        0.115        0.101
        -0.095        0.033        0.106        0.124        0.101        0.181
        -0.158        0.054        0.160        0.081        0.089        0.199
        -0.144        0.121        0.236        0.018        0.078        0.188
         0.010       -0.121       -0.130        0.316        0.146        0.163
        -0.158       -0.144        0.010
         0.054        0.121       -0.121
         0.160        0.236       -0.130
         0.081        0.018        0.316
         0.089        0.078        0.146
         0.199        0.188        0.163
         0.252        0.281        0.042
         0.281        0.368       -0.145
         0.042       -0.145        0.720
********** Active Matrix: H **********
Next run the regression model and save the diagnostics to confirm that LEVERAGE
is the same as diagonal elements of H

>regress

>model clerks=constant+size+season

>save craftr/data resid idvar=type$

>estimate
 
Dep Var: CLERKS   N: 9   Multiple R: 0.891   Squared multiple R: 0.794
 
Adjusted squared multiple R: 0.725   Standard error of estimate: 1.972
 
Effect         Coefficient    Std Error     Std Coef Tolerance     t   P(2 Tail)
 
CONSTANT            12.948        1.996        0.0        .       6.488    0.001
SIZE                 0.355        0.253        0.277     0.881    1.402    0.211
SEASON              -0.188        0.039       -0.948     0.881   -4.799    0.003
 
                             Analysis of Variance
 
Source             Sum-of-Squares   df  Mean-Square     F-ratio       P
 
Regression                89.835     2       44.918      11.554       0.009
Residual                  23.327     6        3.888
*** WARNING ***
Case            1 has large leverage   (Leverage =        0.722)
Case            9 has large leverage   (Leverage =        0.720)
 
Durbin-Watson D Statistic     2.343
First Order Autocorrelation  -0.229
 
Residuals have been saved.
-------------------------------------------------------------------------------
 

>USE 'C:\SYSTAT7\S209\CRAFTR.SYD'
SYSTAT Rectangular file C:\SYSTAT7\S209\CRAFTR.SYD,
created Mon Mar 29, 1999 at 13:57:42, contains variables:
 ESTIMATE     RESIDUAL     LEVERAGE     COOK         STUDENT      SEPRED
 TYPE$        SIZE         SEASON       CLERKS

>list type$ leverage student cook
  Case number        TYPE$     LEVERAGE      STUDENT         COOK
        1     STRSEW              0.722        1.173        1.119
        2     SAND                0.251       -0.433        0.024
        3     VENT                0.210        0.700        0.048
        4     BRICK               0.180       -1.843        0.178
        5     GENCON              0.115       -1.186        0.057
        6     SHEET               0.181        1.518        0.140
        7     PLUMB               0.252       -0.057        0.000
        8     ELEC                0.368       -0.143        0.005
        9     PAINT               0.720        1.108        1.012



Last modified 29 March 1999