Variation and Control Charts A Leadership Perspective Please

Variation and Control Charts A Leadership Perspective Please give credit to www. spcforexcel. com when using this presentation.

How Do We Look at Data? Sales (000) in Jan-10: $11, 229 A number by itself has little meaning. So, we naturally want to compare it to something. But to what? Budget for Jan-10: $10, 000 Sales in Jan-09: $7, 997 Typical comparisons include: q This month’s result to last year’s result q This month’s result to budget q YTD result to last year’s YTD q YTD result to budget

Financial Reports First Quarter Sales: 1 Q 2010 Budget Variance 2 Q 209 Variance 30, 148 30, 000 148 23, 823 6, 325 2 Q 2010 Budget Variance 2 Q 2009 Variance 27, 460 30, 000 (2, 540) 33, 545 (6, 085) 2010 YTD Budget Variance 2009 YTD Variance 57, 608 60, 000 (2, 392) 57, 368 240 Sales $ Second Quarter Sales: Sales $ YTD Sales: Sales $

Variation in Metrics Sales in Jan-10: $11, 229 Sales in Feb-10 $10, 016 Are these two numbers the same? q The measurements we take reflect the variation in the process. q The problem with measurement systems that compare one time frame to another (like the financial report) is that the actual performance (variation) of the process is hidden – we don’t know if there are special causes or just common causes.

Measurements and the Underlying Process q Let’s return to the financial report from before – things did not look too good – especially in the second quarter. Sales $ 2010 YTD Budget Variance 2009 YTD Variance 57, 608 60, 000 (2, 392) 57, 368 240 q Let’s explore another method of examining the data that reveals the “voice of the process” – the true performance.

Measurements and the Underlying Process q The table below shows the monthly data used to make the earlier tables. Month Sales Jan-09 7997 Jul-09 8133 Jan-10 11229 Feb-09 9132 Aug-09 7215 Feb-10 10016 Mar-09 6694 Sep-09 8278 Mar-10 8903 Apr-09 10005 Oct-09 11117 Apr-10 9617 May-09 12055 Nov-09 6481 May-10 8190 Jun-09 11485 Dec-09 6897 Jun-10 9653 q What does this data tell you? q Often, it is difficult to understand the process by looking at a table of numbers.

q It is often better to examine the data as a time series chart. – No major change in monthly sales - some months better than others – Financial report indicated a problem (2 Q 2004) – Time series tells a different story – There is a problem though – sales aren’t increasing

q Adding the average line helps with interpreting the process behavior – – Nine points above, nine points below Monthly sales fairly predictable Range from about $600, 000 to about $1, 200, 000 Can provide an early warning of problems

Control Limits q The time series charts can be further enhanced by adding control limits q These limits define the boundaries for common causes of variation q With common cause of variation: – The upper control limit (UCL) is the largest value you can expect – The lower control limit (LCL) is the smallest value you can expect q As long as the all the points are within the limits and there are no patterns, only common cause of variation is present q The process is said to be “in control” – To improve it, you must fundamentally change the process

q Sales chart with control limits – Process is “in control” – Can predict what will happen in the future – It is not the limits, but the story that the chart tells – Provides feedback on whether the process is changing or staying the same

Interpreting Time Series Charts q Points beyond the limits Special cause UCL Avg LCL Special cause

Zones Test q The zone tests are valuable tests for enhancing the ability of control charts to detect small shifts quickly. – The first step in using these tests is to divide the control chart into zones. q The chart is divided into six equal zones: three above the average and three below the average. UCL A B C Avg C B A LCL

Test for Zone A X= Out of Control Point UCL A X B C Avg C B A X A special cause of variation exists if there are two out of three consecutive points in zone A or beyond (on one side of the average). LCL

Test for Zone B X= Out of Control Point UCL A B X C Avg C B X A LCL A special cause of variation exists if there are four out of five consecutive points in zone B or beyond (on one side of the average).

Test for Zone C X = Out of Control Point UCL A B X C Avg C B A LCL A special cause of variation exists if there are seven or more consecutive points in zone C or beyond (on one side of the average).

Rule of Seven Tests UCL Avg LCL Seven Points in a Row Above the Average Seven Points in a Row Below the Average UCL Avg LCL Seven Points in a Row Trending Down Seven Points in a Row Trending Up

In Control or Out of Control?

Summary