Variable Charts
X-Bar Chart
An X-bar chart is used to control the process
mean. Each point on the chart represents the
average of a subgroup.
Subgroups of 2 to 30 samples may be used when
computing the control limits for the X-bar chart
when based on the range.
Optionally, there is no limit to the subgroup
size when limits are based on sigma. If the
subgroup size varies, the computed limits will
also vary accordingly.
R Chart
The R chart is used to control the process variability.
Each point on the chart represents the range
of a subgroup. Subgroup sizes may be as small
as 2 or as large as 30.
s Chart
The s chart is also used to control the process
variability. Each point on the chart represents
the standard deviation of a subgroup. There
is no limit to the subgroup size.
Median Chart
The median chart is used to control the centering
of the process. Each point on the chart represents
the median of a subgroup.
The limits are based on the median range and
may optionally be changed to use the average
range.
The center line may be computed based on the
average or median of subgroup medians. Subgroups
may be as large as 25 samples.
Individual Chart
An individual chart displays each value. A subgroup
size is used to compute the limits, with value
of 2 being most common, although the subgroup
size may be as large as 30.
Moving Range Chart
The moving range chart is used to control variability
of processes which do not form natural subgroups.
Moving Average Chart
Either uniformly weighted or exponentially weighted
moving average (EWMA) charts may be generated.
Cumulative Sum Chart
This chart may be used to monitor a two-sided,
variable process. The subgroup size may be as
large as 30 samples.
Two tests for out-of-control indication are
available: Run Sum and V-mask. The Run Sum may
either be reset to zero or to a fast initial
response (FIR) value when an out-of-control
value is reached. Optionally, a V-mask may be
specified. A cumulative sum chart for attribute
data is also included.
Process Capability
Analysis
This graph superimposes a normal curve over
a histogram of the data and displays numerous
values such as Cp, Pp, Cr, Cpk, Ppk, CPU, CPL,
both actual and estimated percent beyond specification,
etc. This may be used for data collected in
subgroups or individually.
Attribute Charts
p Chart
The p chart shows either the fraction or percent
of nonconforming units in each sample. The sample
size may vary.
np Chart
The np chart shows the number of nonconforming
units in each sample.
c Chart
The c chart shows the number of nonconformities
per inspection unit.
u Chart
The u chart shows the number of nonconformities
per inspection unit when the inspection unit
changes.
Pareto Chart
The Pareto chart displays a bar chart of discrete
characteristics with their cumulative percentage
of total superimposed. The number of characteristics
is limited only by the resolution of the display
device.
Optionally, if a maximum number of values is
specified, all remaining characteristics will
be automatically included in an OTHER category.
Cumulative Sum Chart
The cumulative sum chart is used for detecting
an increase or decrease in the number of nonconformities
per unit from a known target value.
Additional Features
Most of the variable and attribute charts offer
these features:
Data Table
The Table feature presents a list of items to
display below the chart. The choices include
the underlying data, plotted values, control
limits, etc. Other variables may also be specified.
Data Tips
When the chart is viewed within Statit and a
Data Tip is clicked on, then a pop-up window
is opened, displaying additional information.
Data Tips are defined for the curve markers,
which display, by default, the underlying data,
plotted value and X-axis identification. Other
items may also be displayed, such as the control
limits and other user-specified variables. Other
Data Tips available include "Summary",
which displays the number of points on the chart,
the subgroup size, the number of sigma control
limits, etc.; "Rules Tested" displays
all the rules which were tested; and others.
Custom Data Tips may also be easily added to
the chart.
Data Highlighting
When data highlighting is requested, then a
curve marker may be clicked-on. This results
in the underlying data for the point selected
being highlighted in the Data Window.
Web Settings
Developers may define Web pages/scripts which
will be automatically loaded when a data marker
or out-of-control letter is clicked-on when
the chart is viewed in a Web browser.
Stratification/Phases
This feature is used in conjunction with a categorical
variable. When the value of the categorical
variable changes, then the new value is displayed
above the data space in a block-style legend.
This may be helpful in identifying special causes
of variation or trends. Also, Statit may be
instructed to recalculate the control/warning
limits within each category defined by a categorical
variable.
Automatic Alarms
When rules testing reveals an out-of-control
situation, then several pre-defined alarms may
be generated. These include changing the background
color of the chart and displaying a text message.
As well, a chart can be generated, yet only
displayed if rule violations are detected.
Rules Testing
AT&T/Western Electric rules testing is fully
implemented. Additionally, Statit offers the
ability to add new rules or change the existing
ones.
Points which violate a rule are indicated by
a letter near the point on the chart and the
description of the rule is displayed either
as a Tip or at the bottom of the graph.
Assignable Cause
Points on the chart may be excluded from control/warning
limit calculations and rules testing. This feature
is often used to identify out-of-control points,
for which an assignable cause has been found,
with a text string. The string is displayed
on the chart by the point.
Annotation
Text may be placed by any point on the chart.
This text will automatically follow the point
as it moves across the chart when more data
is collected. This, for example, allows an operator
to identify the point as adjustments are made
to the process.
Control Limits
The control limits may be either calculated
or user-specified. If calculated, the coefficient
of sigma may be easily changed to something
other than the default of 3-sigma. Either of
the limits may be turned off.
Warning Limits
Warning limits may be requested, and are placed
by default at plus and minus 2-sigma. Either
limit may be turned off.
Specification Limits
Upper, center and/or lower specification lines
may be displayed.
Subgroup Identification
Subgroups may be identified either by the constant
subgroup size or by a categorical variable,
such as date/time. The horizontal axis may be
numeric or character. This permits subgroups
to be identified by date, time, lot number,
etc. The size and degree of rotation of the
tick mark labels may also be specified.
Multiple Charts
The QC charts, and all other graphics furnished
with Statit, may be sized and positioned anywhere
on the graph. This permits the design of custom
graphs easily and quickly.
Text Reports
A text report may be requested which lists the
control limits, rules tested and violated, and
outliers. For points which violate a rule, or
are outliers, the raw data is listed along with
the X-axis identification.
Fishbone Diagrams
Fishbone charting can be used to display the
names and relationships of factors which contribute
to a particular problem.
Design Of Experiments
Production processes are affected by many factors
that contribute to the variation in quality.
Statit’s Design of Experiments procedures
may be used to find the optimum settings of
the various factors to yield the best results.
Statit can be used to design and analyze factorial
experiments with up to eleven factors, each
at two different settings. The specified number
of runs may result in full or fractional designs.
Some available features include:
The factor names and levels may be user-specified.
Runs are automatically randomized; blocking
may optionally be requested.
The maximum order of the model may be specified
to remove the high order aliasing.
For central composite or Box-Behnken designs,
Statit can generate the design and then analyze
the data collected. Up to six factors may be
examined. The results consist of parameter estimates,
an ANOVA table, contour plots, and normal probability
plot of residuals.
