SPC Manual
SPC Manual
SPC Manual
Guide
SPC KISS
Keep It
Simple Statistically
Digital
Computations
inc.
Table of Contents
Chapter 1: INSTALLATION ............................................................ 3
2
Chapter 1: Installation
3
You can also run SPC KISS from the Start menu in Windows 95 and Windows NT. The default
folder is SPC KISS.
Either of these methods will launch Microsoft Excel and SPC KISS. See the Program Overview
(following) for more information on how SPC KISS works.
Note: You will not see the SPC KISS menu option at the top of Excel unless you have a worksheet
open in Excel.
Adding In SPC KISS
If you would like to have access to SPC KISS at all times from Excel, you can manually add-in SPC
KISS to Excel's list of add-ins. To accomplish this, follow these directions:
1. Start Excel using the Excel icon.
Very Important: Do not for any reason start Excel using the SPC KISS icon for this
procedure! Be sure you start Excel using the normal Excel icon or from the Start menu at
Microsoft Excel. Do not launch any program or icon titled SPC KISS at this time!
2. From the Excel menu bar, select Tools then Add-Ins.
3. Use the Browse button to locate the directory in which SPC KISS is installed.
4. Select the file Spc97.xla, then OK.
5. SPC KISS 97 is now listed in the Add-Ins available window. Select OK to add-in SPC KISS and
return to Excel.
SPC KISS will now be available every time Excel is run.
Note: You should not start Excel from the SPC KISS icon after you add in SPC KISS. We
recommend that you delete the SPC KISS icon to avoid future confusion.
Installing to a Network version of Excel
If you are using a network version of Excel and you want all users to have access to SPC KISS,
simply install SPC KISS on the server. Each workstation will then have to create a shortcut to the file
Spc97.xla or add-in SPC KISS using the directions given above.
If you are using a network version of Excel and you do not want all users to have access to SPC
KISS, ensure the you install SPC KISS to each local hard drive.
Overview
This section takes you step by step through the Diagram Expert, a four step process to select your
diagram type, enter your data range, select your diagram options, and title your chart. You may begin
with selecting your data range by clicking on the Excel cell in the upper left corner of your data and,
while holding the left mouse button down, drag the mouse down or down and to the right until you
have highlighted all of your data (release the mouse button). Then go to the SPC KISS pull-down
menu item and select Diagrams.
5
Step 2 of 4 - Selecting Your Data
This window allows you to select the range of data for your diagram. If you have already chosen the
correct data range (see the instructions in the Overview of this chapter), select Next to proceed.
If you need to change the range, simply select the new range of data using the mouse, then Next to
continue. (Drag the Step 2 dialog box out of the way to see all of your data if necessary.)
You may also type in the correct range in the edit box of the Step 2 window; be sure you input the
range in the proper format.
Example: $A$1:$B$13 denotes cells A1 through and including B13 on the worksheet. You can type
A1:B13 to specify that range.
Remember: Each diagram accepts data in a different format. For instance, some will accept rows or
columns, and some only one row or column. See the on-line help system or the appropriate sections
for information on a specific diagrams data format.
6
Step 3 of 4 - Diagram Options
You have several options for viewing your data in the completed diagram. While each diagram has
its own set of options, the buttons at the bottom of the window are the same for all diagrams. The
following window is an example of the options for the Pareto diagram. Each button is explained
below.
The Save Options button allows you to save the settings you have chosen in this
window so that each time you create the same type of diagram, you will automatically get the same
options. For instance, if you select a descending sort order and 2-D bars with a cumulative line (like
the example above), and then select Save Options, each time you create a Pareto diagram these will
be the default options. These options will be saved for the Pareto diagram only; if you create a
Scatter Plot next, you will get the Scatter Plots default options.
The Help button links you directly to the on-line help for the options of the diagram you
are creating.
The Cancel button stops the SPC KISS process and puts you back at your Excel sheet.
The Next button accepts the option settings and sends you forward one step.
The Finish button accepts the current option settings and ends the Diagram Expert
process without going on to any remaining steps. There will be titles on your diagram; however, you
wont get the Step 4 Diagram Expert window allowing you to change the titles before the diagram is
finished.
Remember: Each diagram has its own set of options. (Summary Stats has no options.) See the on-line
help system or the appropriate sections for information on a specific diagrams options.
7
Chart Title
You may change the title of your chart in the first edit box. The default chart title is the name of the
diagram you have chosen, unless you have created this diagram before and saved your titles using the
Save Options button.
Include Legend
You may decide whether or not to include a descriptive legend on your diagram. Even if in Step 3
you chose to use the first row as legend text, you could turn the legend off in Step 4.
Axis Titles
This allows you to change the title of the Category (X) axis and the Value (Y) axis.
Buttons
The buttons at the bottom of the window are the same for each diagram type.
The Save Options button allows you to save the titles you have chosen in this
window. By selecting the Save Options button, you create new default titles for a particular diagram
type. For instance, each time you create a Pareto diagram these will be the default titles. These titles
will be saved for the Pareto diagram only; if you create a Scatter Plot next, you will get the Scatter
Plots default titles.
The Help button links you directly to the on-line help for chart titles.
The Cancel button stops the SPC KISS process and puts you back at your Excel sheet.
The Finish button accepts the title settings and ends the Diagram Expert process.
Remember: Each diagram has its own default titles to begin with, and you can change those defaults
by saving your chart titles with the Save Options button.
Pareto Chart
A Pareto chart is a bar chart for non-numerical categories in which the vertical axis is usually
frequency or monetary loss.
If you can identify all of the factors that cause a defect in a process, a Pareto chart will give you a
graphical representation of these factors and allow you to see which ones need attention first.
Example: You are examining the frequency of defects in the process of manufacturing IC boards. You
find the following information about the defect types and their frequency:
8
From this data, your Pareto chart would look like the example below. It is clear that the majority of
the defects occur in the soldering process.
3DUHWR'LDJUDPRI,&%RDUG'HIHFWV
6ROGHULQJ (WFKLQJ 0ROGLQJ &UDFNLQJ 2WKHU
In this example, the text in the first column would automatically be used as a legend on your
diagram. This text is optional.
When you have a combination of text and numeric data, the text needs to appear in the first column
or row to appear as labels for the X Axis. If any non-numeric data appears in the Y column or row,
SPC KISS will ask you to choose to abort the operation, re-enter your data range, or ignore the non-
numeric data when creating the Pareto diagram.
Data in two rows:
9
Data in two columns:
Pareto Charts will also accept non-contiguous rows or columns of data, as long as there are not more
than two rows or columns selected. For more information on making non-contiguous data selections,
turn to Chapter 10: Tips.
Pareto Options
Below is an example of the window that allows you to change the options for a Pareto diagram. Each
Pareto-specific option is explained here.
10
Pareto Chart Titles
Except for the name of the chart and the axis titles, this step is the same for all the diagrams. Please
refer to Step 4 of 4 - Chart Titles for more information.
The Histogram from this data (shown below) shows you pictorially the number of cars that fall within
each gas mileage range.
+LVWRJUDPRI*DV0LOHDJH'DWD
*DV0LOHDJHPSJ
11
You may use non-contiguous data in creating a Histogram. For instructions on selecting non-
contiguous data, see Chapter 10: Tips.
Histogram Options
Bar Type
You may change the type of bars used in the Histogram by selecting one of these four options. The
change will be seen immediately on your chart in Excel.
Anchor Point
The anchor point sets the left end point of the leftmost bar.
Class Width
This option defines the width of each bar.
Number of Classes
This option changes the number of classes shown on your Histogram. Use the up and down arrows to
increase or decrease the number of classes. The default number of classes is equal to the square root
of N, where N equals the number of data points.
Gap Between Bars
Change the amount of space between the bars of your Histogram using the up and down arrows to
widen or lessen the space. There are actually two gap values, one for 2 Dimensional plots and one for
3 Dimensional plots.
Regenerate Histogram
This option must be used when you change the anchor point or class width to update your chart in
Excel.
Survey Data
This option is typically used for survey data where the width of each bar is one unit (e.g., a Likert
scale of 1, 2, 3, 4, 5). In the example below, the gas mileage data shown previously was used to
create the Histogram. However, instead of a class width of six, the Survey Data option was selected.
This diagram shows how the previous example would be broken down.
12
+LVWRJUDPRI*DV0LOHDJH'DWD
*DV0LOHDJHPSJ
13
... and the scatter plot of that data would look like this:
6FDWWHU'LDJUDPIRU9HKLFOH:HLJKWYV0LOHDJH
9HKLFOH:HLJKW
Bolt
Diameters ...into this information for the scatter plot.
is legend
title and is
optional.
If, in your data, you put a text label at the beginning of the column or row, the text label at the top of
the Y coordinates will be used as legend text; any text for the X coordinate will be ignored.
14
If there is any text in the X coordinate column or row, your data will be plotted as though you
provided the Y coordinate only. SPC KISS will plot dates and times, and use them as labels for the X
coordinate.
Regression Line
You may choose to have a linear regression line, a quadratic regression line, a cubic regression line,
or no regression line at all on your scatter plot. You may have only one type of regression line at a
time. Regression is computed using the least squares method.
Forecast
Forecasting allows you to extend your regression line forward or backward without the input of new
data. Care should be taken if assuming the actual process follows the predicted line.
Display Equation and R squared
By selecting this option, you can have the regression equation and R squared displayed on the
diagram.
15
Defining the type of Process Capability Chart
SPC KISS can create two types of process capability charts. The difference is in how the mean and
standard deviation are obtained. You may use a range of data selected from a spreadsheet, or you
may use direct entry of the mean and standard deviation.
If you choose a range of data selected from a spreadsheet, you will be asked to select the range in
Step 2 of the Diagram Expert (see Step 2 of 4 - Selecting Your Data).
If you choose direct entry of the mean and standard deviation, you will be given the following
window. Simply type in the appropriate numbers for the mean and standard deviation. You may also
change the Lower and Upper Specification Limits here. Pressing the Recompute Process Capability
Chart button allows you to quickly compute the capability measures for any mean/standard
deviation/LSL/USL specified. Selecting OK completes the process.
16
Process Capability Measures (Cpk) Options
17
2JLYH&KDUW
&ODVVHV
Percentile line
Increase or decrease this number to effect the position of the percentile line on your diagram.
Display kth Percentile lines
This option displays the kth percentile line on your diagram.
Number of Classes
This option simply changes the number of classes or points shown on the Ogive. Use the up and
down arrows to increase or decrease the number of classes. The default number of classes is equal
to the square root of N, where N equals the number of data points.
Number of Rounding Digits
This option changes the number of significant digits in the numbers used to determine the start and
stop of each of the classes. SPC KISS makes an attempt to compute the optimal number for your
18
data set; however, you may change it for any reason. A positive number reflects the number of digits
after the decimal used in rounding. A negative number reflects the number of digits before the
decimal that are used to round. Use the up and down arrows to increase or decrease the number.
Example: One of the classes was computed to be from 2387.289417 to 4376.673692. The table below
shows the actual values used to compute the upper and lower limits of that class.
19
Chapter 3: Control Charts
Overview
This section takes you step by step through the Control Chart Expert, a three step process to select
your chart type, enter your data range, and select options for your chart. Begin with selecting your
data range by clicking on the Excel cell in the upper left corner of your data and, while holding the
left mouse button down, drag the mouse down or down and to the right until you have highlighted all
of your data (release the mouse button). Then go to the SPC KISS pull-down menu item and select
Control Charts.
20
Step 2 of 3 - Selecting Your Data
This window allows you to select the range of data for your control chart. If you have already chosen
the correct data range (see the instructions in the Overview of this chapter), select Next to proceed.
If you need to change the range, simply select the new range of data using the mouse, then Next to
continue. (Drag the Step 2 dialog box out of the way so you can see all of your data if necessary.)
You may also type the correct range in the edit box of the Step 2 window; be sure you input the range
in the proper format.
Example: $A$1:$B$10 denotes cells A1 through and including B10 on your Excel worksheet. You
can type A1:B10 to specify that range.
Remember: Each control chart accepts data in a different format. For instance, some will accept rows
or columns, and some only one row or column. See the appropriate sections for information on a
specific diagrams data format.
21
also allows you to define a range other than the plotted range upon which to compute the control
limits (not applicable for user defined control limits).
Example: You may want a control chart of 40 samples, but want the control limits based on the first
20 samples only. Or perhaps you want to plot an X bar R chart of data from March, but want to base
the control limits on data from January.
Shewhart Methodology control limits: This option computes the standard Shewhart control limits
based on the data specified.
Standard Deviation control limits: Allows you to set the control limits based on any number of
standard deviations. Simply change the number of standard deviations by clicking on the arrows or
by typing in a new number.
User Defined control limits: There are two different ways to define these control limits. You can
either enter a Manual Center Line or have SPC KISS compute a Mid Center Line. If you select
manual center line, you will be prompted for the upper control limits, center lines and lower control
limits for both the X Bar and Range.
If you select Mid Center Line, SPC KISS will use the mid point of your UCL and LCL as the
center point.
Note: The user defined control limits are not computed from any data provided to the software. If
you use this option you must know what you are doing. Control limits are not the same as
specification limits; therefore, do not confuse the two.
Regardless of what kind of control limits you use, you can change the range of data on which the
control limits are based. For Shewhart control limits and Standard Deviations, you can enter a new
range in the Control Chart Options Window. The default range is the plotted range; however, you can
enter any range here. This can be done one of two ways:
1. Type in the new spreadsheet range. For example A1:B33.
2. Use the mouse and click on the old range. Select the new range on the spreadsheet using the
mouse.
22
Other Options
Display Zones on Chart
SPC KISS will show the lines that delineate the three zones on the control chart.
Zones A, B, and C can be defined by dividing the area between the center and the upper control limit
(or the center and the lower control limit) into three equal parts. Zone C is nearest the center, with
Zone A nearest the UCL (or LCL). The illustration below shows just where Zones A, B, and C fall on
a control chart.
8&/
=RQH $
=RQH %
=RQH &
&HQWHU
=RQH &
=RQH %
=RQH $
/&/
Exception: With the R chart and the S chart, when the lower control limit is zero, the zones below the
center line will be symmetrical to those above the center line.
Example: If the center line of an R chart is 10, the upper control limit is 40, and the lower control
limit is 0, then the zones above the center line will be as follows:
Zone C from 10 to 20 Zone B from 20 to 30 Zone A from 30 to 40.
The zones below the center line should be:
Zone C from 10 to 0 Zone B from 0 to -10 Zone A from -10 to -20.
However, since the lower control limit is set at zero, only Zone C will appear on the control chart.
Perform Out of Control Checking
Note: Out of control checking only has significance if your control limits make sense. If you have
entered manual control limits, SPC KISS will still perform out of control checking; however, if those
control limits are not valid, then the out of control checking will not be valid.
Out of control checking will examine your data to see if any of the following seven symptoms occur:
One or more points are outside the control limits
Seven or more consecutive points are on one side of the centerline
Seven consecutive increasing or decreasing intervals
Two out of three consecutive points fall in a specific Zone A or beyond
Four out of five consecutive points fall in a specific Zone B or beyond
There are 14 consecutive points that alternate up and down
There are 14 consecutive points in either Zone C
23
If you choose to perform out of control checking, and some out of control symptoms are detected,
additional Excel sheets will be added that highlight each symptom found. You can access these sheets
by clicking on the new tabs found at the bottom of your workbook. An example of the tabs is below.
Depending on the out of control symptom, the names of the new sheets will be the name of the chart
followed by:
1. ... Out UCL or LCL 5. ... 4 of 5 Zone B
2. ... 7 Above or Below Mean 6. ... 14 Alternating
3. ... 7 Incr or Decr 7. ... 14 Zone C
4. ... 2 of 3 Zone A
If none of these symptoms are found, SPC KISS will inform you in a dialog window.
Create Cpk Chart
If you check this box, SPC KISS will prompt you for specification limits. SPC KISS will then
compute the process capability of your data based upon the estimates for the mean and standard
deviation derived from the control chart that you have selected. This is especially helpful when
computing the process capability from p charts and c charts which are not based on the normal
distribution.
X Bar R or X Bar S Chart
If you select an X Bar R Chart, SPC KISS will create two separate charts from your data: an X Bar
chart to show the average of each group, and an R chart to show the range of each group. Each chart
has its own tab in your Excel workbook.
If you select an X Bar S Chart, SPC KISS creates an X Bar chart to show the average of each group,
and an S chart to show the standard deviations of each group.
Example: You are measuring the salt content in soap. You have twelve subgroups of data which look
like the following:
24
The X Bar chart of the data is below.
;EDU&KDUW6KHZKDUW&RQWURO/LPLWVIRU6DOW&RQWHQWLQ
6RDS
6DPSOH1XPEHU
Data Format
X Bar R and X Bar S charts accept data in either rows or columns, and allow non-contiguous data
(see Chapter 10: Tips). If your data is in only one row or column, you will see the following window
asking you to identify the size of the sample you are using.
If your data is in more than one row or one column, you will be asked if the subgroups are in rows or
columns. Make that designation in the following window.
25
Individuals with Moving Range Chart
An Individuals with Moving Range (IMR) Chart will plot your data and compute the center and
upper and lower control limits.
There will be two separate charts: an Individuals chart which will plot each data point, and a Range
Chart, which will plot the (moving) range between every two successive points on the Individuals
chart.
,QGLYLGXDOVFKDUWZLWK6KHZKDUW&RQWURO/LPLWV
5XQ
26
Individuals with Moving Range Chart Data Format
An Individuals with Moving Range Chart will accept data in one or two rows or columns only. If the
data is in two rows (or columns) the first row will be used as the labels for the X-Axis. The data
must be numeric; non-numeric data will be ignored, but a gap will be placed on the control chart
where the non-numeric data is found.
Data in one column: Data in two columns:
Run Chart
A Run chart plots some metric over time. The run chart will not place any control limits, center lines,
or zones on your chart.
5XQ&KDUW
5XQ2UGHU
27
c Chart
A c Chart plots the number of occurrences observed in an interval and computes the center, upper
control limit, and lower control limit. The c Chart shows the interval or sample number on the X axis,
and the number of occurrences for each interval on the Y axis. It assumes an underlying Poisson
distribution.
FFKDUWZLWK6KHZKDUW&RQWURO/LPLWV
5XQ
Data Formats
The data formats for Run charts and c charts are identical to the format for the Individuals with
Moving Range Chart.
p Chart
A p Chart plots the proportion defective from your data and computes the center, upper control limit,
and lower control limit. The p Chart shows you the sample numbers across the X axis and the p value
on the Y axis. The p value is the proportion of defects in each sample displayed as a decimal fraction.
The p Chart assumes an underlying Binomial distribution.
SFKDUWZLWK6KHZKDUW&RQWURO/LPLWV
5XQ1XPEHU
28
p Chart Data Format
A p Chart will accept numeric data in two rows or two columns. The data may be contiguous or non-
contiguous (see Chapter 10: Tips). SPC KISS will determine which row or column contains the
sample size and which contains the number of defects (the larger value will be used as the sample
size). If either the sample size or the number of defects is left blank or is non-numeric, SPC KISS will
ignore that entry and leave a blank space on the control chart.
Multiple Series
Multiple Series will allow you to plot multiple run charts on the same chart.
0XOWLSOH6HULHV&KDUW
5XQ
29
Chapter 4: Cusum Charts
Overview
This section takes you step by step through the Cusum Chart Expert, a three step process to select a
cumulative sum chart, enter your data range, and select options for your chart. Begin with selecting
your data range by clicking on the Excel cell in the upper left corner of your data and, while holding
the left mouse button down, drag the mouse down or down and to the right until you have highlighted
all of your data (release the mouse button). Then go to the SPC KISS pull-down menu item and
select Cusum Chart.
30
If you need to change the range, simply select the new range of data using the mouse, then Next to
continue. (Drag the Step 2 dialog box out of the way so you can see all of your data if necessary.)
You may also type the correct range in the edit box of the Step 2 window; be sure you input the range
in the proper format.
Example: $A$1:$B$10 denotes cells A1 through and including B10 on your Excel worksheet. You
can type A1:B10 to specify that range.
Remember: Each control chart accepts data in a different format. For instance, some will accept rows
or columns, and some only one row or column. See the appropriate sections for information on a
specific charts data format.
31
32
Chapter 5: Analysis Tools
Overview
This section takes you step by step through the Analysis Tools. You may begin with selecting your
data range by clicking on the Excel cell in the upper left corner of your data and, while holding the
mouse button down, drag the mouse down and to the right until you have highlighted all of your data
(release the mouse button). Then go to the SPC KISS pull-down menu item and select Analysis
Tools.
Multiple Regression
This feature computes the multiple regression of a data set with up to 16 independent variables and
one dependent variable. The output is shown on a new Excel sheet titled Multiple Regression.
The output table includes the following information: Factor, Coefficient (Coef), Standard Error (Std
Error), t stat, p Value, R Squared, Adjusted R Squared (Adj R Sq), F Value, Sum of Squares
regression (SS reg), and Sum of Squares residual (SS resid).
33
(16 independent variables and one dependent variable). There is a high limit on the number of rows
of data you can have.
Note: The dependent variable must be in the right-most column in your data range.
If you name the variables on your Excel spreadsheet, those names will be used to identify the factors
in the regression output table.
Example: This table shows a data set with the independent variables named. The dependent variable,
Y, is in the right-most column.
If you do not name the factors, SPC KISS will assign each factor a sequential letter (A, B, C, D, etc.)
to be shown in the output.
If you use non-contiguous data, the last column you select will be used as the dependent variable. If
your last selection contains more than one column, the right-most column in that selection will be
used as the dependent variable.
The table above is an example of non-contiguous data.
The above input table of non-contiguous data would yield the following multiple regression output.
34
Correlation Matrix
The Correlation Matrix returns the correlation coefficient between every possible combination of
columns or rows.
t TEST MATRIX
The p value of a t Test gives the probability that two samples come from two underlying populations
that have the same mean.
SPC KISS always computes the two-tailed t Test, and assumes that the two samples have equal
variance.
35
F Test Data Format
F Test data format is identical to Correlation Matrix data format.
Independence Test Matrix
The Independence Test allows you to test for independence between several random variables. This
feature will create a contingency table and the associated p value.
36
Chapter 6: Six Sigma Reports
Overview
This section takes you step by step through the Six Sigma Reports process. You can create Product
Reports and Process Reports.
The dialog below shows the ranges you would select to create the product summary for this data.
37
DPU is the number of Defects Per Unit.
DPO is the number of Defects Per Opportunity (note that if the # of opportunities is one, DPU
will be identical to DPO).
DPMO is the number of Defects Per Million Opportunities.
Sigma Capability is the Sigma Capability for that characteristic.
38
Statistical Output
The statistics sheet that is created in addition to the charts includes two simple columns: one for long
term and one for short term. The Mean is the mean of the data (not Xbarbar or the target value) for
both Short and Long Term. The Standard deviation for short term is the estimate from the control
chart that was created (i.e., for IMR Rbar/d2); for long term it is the true standard deviation. The Cp
and Cpk use the control chart type standard deviation; they are for short term only. Pp and Ppk are
only for long term. Sigma Capability comes from the respective DPM.
In this example, you are the manufacturer of capacitors. The capacitors should hold a voltage
between 1.7 and 2.7 volts. The data below represents that last 60 measured observations coming from
your capacitors.
To analyze this data you would select SPCKISS - Six Sigma Reports - Process Summary. The
dialog below would appear and you would enter the range and specification limits.
39
From this you would receive output in both table and graphical format. The data from the above
example is shown below.
40
Chapter 7: Distributions
Discrete Distributions
There are six discrete distributions. When you select a discrete distribution, a new worksheet will be
added to the active workbook. This new worksheet will have the appropriate distribution ready for
your use. Below is an example of the binomial distribution.
Binomial
If a trial is binary (i.e., pass/fail) and the probability of passing is (p), then the Binomial distribution
will give you the associated probabilities that the outcome of passing occurs (x) times in (n) trials
(assumes trials are independent and that (p) is constant).
Critical Binomial
Given the number of trials (n), the probability of success on each trial (p), and the criterion value (or
probability, such as 0.95), the Critical Binomial returns the smallest value (the critical value) for
which the cumulative binomial distribution is greater than or equal to a criterion value.
Example: Use the Critical Binomial distribution to determine the greatest number of defective parts
that are allowed to come off an assembly line run without rejecting the entire lot.
Hypergeometric
Given the population size (N), the sample size (n), the size of subpopulation of interest (D), and the
number of elements from the subpopulation of interest (x), the hypergeometric distribution will return
the probability of getting (x) items from (D).
Hypergeometric is used for problems with a finite population, where each observation is either a
success or a failure, and where each subset of a given size is chosen with equal likelihood.
Example: A box contains 20 items, of which five are defective. A sample of four items is selected.
What is the probability that two of the four are defective?
41
Poisson
The Poisson distribution is used to calculate the probability that a certain number of random events
will occur in a specified interval. Its only parameter, lambda, represents the average number of
occurrences per interval. It can also be used to approximate the binomial distribution.
Negative Binomial
The Negative Binomial distribution returns the probability that there will be a certain number of
failures before a certain number of successes. This is similar to the Binomial distribution, except the
number of successes is fixed and the number of trials is variable. Like the Binomial distribution, the
trials are assumed to be independent.
Example: You need to find 20 people with blue eyes, and you know the probability that a person has
blue eyes is 0.3. The Negative Binomial distribution calculates the probability that you will find a
certain number of people without blue eyes before you find 20 people with blue eyes.
Uniform
Given the upper and lower boundary limits, the Uniform distribution will find the probabilities
associated with the placement of (x) between those limits, assuming that each possible outcome is
equally likely.
Continuous Distributions
There are fifteen continuous distributions.
Chi-squared
This returns the one-tailed probability of the Chi-squared distribution. It is associated with the Chi-
squared test. The Chi-squared test would be used to compare observed and expected values to
determine if the original hypothesis is valid.
Inverse Chi-squared
Instead of providing (x), the Inverse Chi-squared distribution allows you to provide the probability
and will compute (x).
Exponential
The Exponential distribution models the time between events (i.e., how long an automated bank teller
takes to deliver cash). The Exponential distribution can be used to determine the probability that the
process takes at least one minute.
F Distribution
The F probability distribution can be used to determine whether two data sets have different degrees
of diversity.
Example: The F distribution can be used to tell if the variances between two different assembly lines
are different.
Inverse F Distribution
This is the same as the F distribution, but you provide the probability and the Inverse F distribution
computes the (x) value.
42
Gamma
The Gamma distribution is used to study variables that may have a skewed distribution. It is
commonly used in queuing analysis.
Lognormal
The Lognormal distribution is used to analyze data that has been logarithmically transformed.
Normal
A Normal distribution is characterized by a smooth bell-shaped curve. This is similar to a standard
normal distribution, except that you can enter the mean and the standard deviation.
Inverse Normal
This is similar to a normal distribution, except that you enter the probability and the Inverse Normal
distribution computes (x).
Standard Normal
The Standard Normal distribution has a mean of zero and a standard deviation of one. This function
is used in place of a table of standard normal curve areas.
Student's t
The Student's t distribution is used in the hypothesis testing of small sample data sets. Given a test
statistic (t-value) and the appropriate number of degrees of freedom, it returns a probability.
Inverse Student's t
This returns the (t) value for a given probability (p).
Uniform
Returns the probabilities associated with the continuous Uniform distribution.
Weibull
The Weibull distribution is used in reliability analysis, such as calculating a device's mean time to
failure.
43
Chapter 8: Failure Mode & Effect Analysis
Failure Mode and Effect Analysis (FMEA) is a method to identify and prioritize potential failure
modes in a system. The numerical values of interest are severity, occurrence and detection. These are
usually scored using the following table.
6FRUH
:RUVW %HVW
&DWHJRU\
When you select this option, a new worksheet will be added to the active workbook. The user inputs
the Severity, Occurrence and Missed Detection ratings based on the scale defined above, and SPC
KISS calculates the Risk Priority Number (RPN) for each failure mode/cause combination. A Pareto
of the RPNs can then identify high risk areas.
44
Chapter 9: MSA (gage capability)
Measurement system analysis (MSA) or gage capability is used to measure the variability in a
measurement system. MSA will produce an X bar R chart for each operators data set, as well as the
following statistics: Sigma Repeatability, Sigma Reproducibility, Sigma Measurement, Sigma Total,
Precision to Total Ratio, Precision to Tolerance Capability Ratio (if specification limits are entered).
The purpose of an MSA is to learn how much variability is in the measurement system itself
compared to the total variability in the measured product.
45
Chapter 10: Tips & Troubleshooting
Tips
Select your data first
One of the most important things you can do to make SPC KISS easier to use is to select your data
range before you start the program. Select your data by clicking on the Excel cell in the upper left
corner of your data and, while holding down the left mouse button, drag the mouse down or down
and to the right until you have highlighted all of your data (release the mouse button). Then go to the
SPC KISS pull-down menu item and select the option you want.
Non-Contiguous data
Non-contiguous data is data that is not all located together on a continuous rectangular portion of the
worksheet. That is, you may want to skip over some rows or columns. To select non-contiguous data,
select the first group of data, then hold down the Control (Ctrl) key while you select the next group of
data. You may do this as many times as you wish (except with a Pareto chart, p Chart or a Scatter
Plot, which allow only two columns or rows to be selected). The example below shows three areas of
non-contiguous data selected.
All diagrams, charts and tools accept non-contiguous data selections, except the Individuals with
Moving Range Chart, Run Chart, and c Chart (those that allow only one column or row of data).
46
Not only can you copy and paste within Excel, you can also copy and paste between most
applications. This is useful if you want to incorporate a diagram into your word processing document.
Follow the procedures outlined above, using the Edit-Paste command from the menu of your word
processing (or other) application.
To select and edit or move the text, simply click once on it with the mouse and
begin typing. When you hit the Enter key, your changes will be made on the diagram. When the text
is marked with the box and handles as shown here, you can also move it anywhere on the diagram.
Just grab any side of the box with the mouse pointer and drag it to its new location.
If you want to change more than just the words of the text -- the size or color, for
instance -- you can highlight the text, or you can double-click on it. (If you double-click on the text, a
dialog box will appear in which you can make all changes to the text.) To highlight the text, click on
the text once with the mouse (youll get the box and handles). Position the mouse pointer over the
text. When the arrow pointer becomes the I-beam, you can drag it across the text to highlight it. Then
you can change the font, size, style, or color of the text, as well as the words themselves. Your
changes will be seen as you type.
Changing Graphical Items
To make any changes to the graphical elements of your chart or diagram, simply double-click on the
item you want to change. The following illustration shows the histogram bars selected, and the Excel
window used to change elements such as the weight and color of the border and the color of the area
of the bars. You can also change the X values, names and labels here.
47
Troubleshooting
Excel 97 Users
When opening SPC KISS you may see the following dialog box. To run SPC KISS, you must click
Enable Macros.
Windows Users
If SPC KISS and Excel do not start when you launch SPC KISS (using either method described in the
section Running SPC KISS in Chapter 1), you will need to manually add-in SPC KISS to Excel. To
do this, follow the directions in Chapter 1: Installation under Adding In SPC KISS.
Note: You should not start Excel from the SPC KISS icon after you add in SPC KISS. We
recommend that you delete the SPC KISS icon to avoid future confusion.
Technical Support
Before you call for technical support, check the README file on your distribution disk for up-to-
date information. You may find the answer to your question there.
For installation or compatibility questions, call Digital Computations, Inc. at (719) 338-3793 or send
e-mail to DigitalC@ix.netcom.com. For other technical support, contact Air Academy Associates at
(719) 531-0777 or FAX (719) 531-0778.
Please have the following information available before calling for assistance:
SPC KISS and Microsoft Excel version numbers
Your operating system type and version number, plus any network information
48
More books and software from
Air Academy Associates
Knowledge Based
Management
KBM
3URFHVV&DSDELOLW\0HDVXUHV&KDUW
&ULWLFDO0HDVXUH
Understanding
Industrial Designed
Experiments
Fourth Edition
Distributed by
Air Academy Associates, LLC
1155 Kelly Johnson Blvd.
Suite 105
Colorado Springs, CO 80920
Phone (719) 531-0777
Fax (719) 531-0778
www.airacad.com
e-mail: aapa@airacad.com