Compare Chart

QW5 Help Files

Overview


The Compare Chart is an important tool in identifying and understanding changes in a process or product.  The Chart allows you to compare two separate time frames for an individual variable to help identify if a change has occurred in a product or variable.  The Chart can be used to understand the effects of a raw material change, new equipment, speed or personnel changes in a process.


The Compare chart can also compare two variables in the same time frame by going to Options / Compare / Two Variables.


Using the screen


Changes to a process can be viewed or detected through the use of the Compare Chart, Histogram, Statistics Box and also through the information box in lower left corner.  This box will highlight a difference in averages or variation if detected in the two time frames selected with a 95% confidence level.


It is important to remember that this chart is designed to detect and display differences if they can be proven with a high degree of confidence.  If the Average or Variation item is not highlighted in the box it does not mean that a difference is not present, it means only that the difference can not be proven with a high degree of confidence.


The Compare Chart is made up of six (6) distinct information areas which are described below.


Understanding the screen


The Compare Charts - top of screen


The left hand control chart displays the data from the Before time frame.  The right hand control chart displays the data from the After time frame.  These time frames can be discontinuous (missing time between before and after), continuous (following time frames), overlapping (containing the same time frames and historical (containing old data not necessarily current data).   The scaling on the charts are the same to best picture any changes.  Control charts graphically answer questions such as "How has the pattern of data changed" and "How has the control of data changed"


It should be understood that the color bands displayed on the Compare Chart are based on either the current Fixed or Calculated Limits for this variable.  If the limits are calculated then the limits are based on the number of points currently contained in memory (maximum 100).


The Capability Bars - to the left of the Histogram


To the left of the histogram are a set of bar charts.  The light blue bar represents before.  The dark blue bar represent after.  The spread represents capability in terms of average +/- 3S.  A black box outlines both bars to infer that these two bars are meant to be compared and are scaled the same.  Bar charts graphically answer questions such as "How has the capability changed"?


The Histogram - to the right of the Capability Bars


Again light blue and dark blue indicate before and after time frames.  The black lines outline bars that are comparable in terms that they share the same interval on the histogram.  The histogram displays the actual data in terms of how often the data repeats within a specific interval of the data.    Histograms answer questions such as "Is the distribution still fairly normal in shape".


The Statistics Box Area - to the right of the Histogram


Before and After Statistics are the statistics that would be generated if one was viewing the control chart screen of a given time frame.  Capability statistics such as Average +/- 3S have been added in this box for ease of reading.

The triangle column is the delta or difference column.  This is simply "After minus before" results.  A negative number indicates the after data is less than the before data.  A .5 in the Cr row indicates that the Cr has increased by .5 with the change.

The triangle % column is the delta % or the % difference column.  This is ((After-Before)/Before) times 100.  A -20 in the C OOL PPM row indicates your calculated PPM defect level is 20% lower with the change.


The Test Box Area - bottom left of the screen


The first test is a test to determine if the averages are statistically different at 95% confidence using the  t-test.  The t calculated is compared to the t criteria.  When the t-calculated is higher than the criteria, a highlighted message below indicates that the averages are different.

The second test is a test to determine if the variance is different at 95% using the F test.   The F calculated is compared to the F criteria.  When the F calculated is higher than the F criteria, a highlighted message below indicates that the Variation is different.