SCHART Statement: SHEWHART Procedure

Example 17.29 Specifying a Known Standard Deviation

See SHWSEX1 in the SAS/QC Sample LibraryIn some applications, a standard value $\sigma _{0}$ may be available for the process standard deviation $\sigma $. This example shows how you can specify $\sigma _0$ to compute the control limits.

Suppose that the amount of power needed to heat water in the heating process described in Creating Standard Deviation Charts from Raw Data has a known standard deviation of 200. The following statements specify this known value and create an s chart, shown in Output 17.29.1, for the power output measurements in the data set Turbine:

ods graphics off;
title 's Chart for Power Output';
title2 'Using Known Process Standard Deviation of 200';
proc shewhart data=Turbine;
   schart KWatts*Day / sigma0  = 200
                       ssymbol = s0;
run;

The SIGMA0= option specifies $\sigma _0$, and the SSYMBOL= option specifies a label for the central line indicating that the central line is computed from $\sigma _0$. Since all the points lie within the limits, you can conclude that the variability of the process is stable.

Output 17.29.1: Reading in Standard Value for Process Standard Deviation

Reading in Standard Value for Process Standard Deviation


You can also specify $\sigma _0$ as the value of the variable _STDDEV_ in a LIMITS= data set, as illustrated by the following statements:[79]

data Plimits;
   length _var_ _subgrp_ _type_ $8;
   _var_    = 'KWatts';
   _subgrp_ = 'Day';
   _type_   = 'STDSIGMA';
   _limitn_ = 20;
   _stddev_ = 200;
run;

title 'Chart Using Known Process Standard Deviation';
proc shewhart data=Turbine limits=Plimits;
   schart KWatts*Day / ssymbol=s0;
run;

The resulting s chart (not shown here) is identical to the one shown in Output 17.29.1. For more information, see LIMITS= Data Set.



[79] In SAS 6.09 and in earlier releases, it is necessary to specify the READLIMITS option.