Please help me about the sigma level accuracy calculation of a data sampling. I have a spread sheet gives me the calculation, but i would like to know how the calculation is done (shift and things like that). Please let me know any books or references on the same.

Dear Madhan,

It is not clear what you are looking for.

Can you explain what type of data you have and what exactly is it that you are trying to find out from the data?

The formula for calculating the sigma level is

Sigma Level = (upper spec.Limit - Lower Spec.Limit)/(2 * std.dev)

Hi,

Can anyone help with details about how to decide about the project whether it is six sigma project or lean project or a Kaizen.

Is there some tool or calculation for the same ?

Sorry to bother you guys in between of your discussion.

 

Dear Ashwini

The basic difference between Six Sigma and Lean is that while Lean focuses on continuous improvement, Six Sigma is a problem-solving methodology used to identify problems of unknown reasons. For example:

"Finding the reasons for low sales in 3rd quarter in Mumbai area and increasing sales revenue by 20% in 18 weeks." is an example of a Six Sigma project.

While "Improving the workspace and movement area around the conveyor belt to reduce transportation and motion time" is a lean process.

Warm regards

AbhishekM

Hi,

I find a difference when calculating the Sigma Level using the formula and when I input the values in the sigma calculator. Can someone explain this difference? The details are below.

Process mean - 4.54, UCL- 5.11, LCL- 3.97

USL-4.63, LSL- 4.36, Std. dev- 0.19

Using formula, Z= USL- mean/ Std.dev, the sigmal level is 0.48

But, if I use the Sigma calculator, the sigma level comes out to 1.56....Why this difference?

Dear Nageshwar,

 

When you use these values : USL 4.63, LSL 4.36, Std Dev 0.19, Process Mean 4.54, and assuming that this is long term data, Sigma Level Calculator takes into account the defects that are happening both when you get the process output more than USL and also when the process output is less than USL. whereas if you use only Z=USL-Mean/Std Dev,you are accounting for the defects happening because of the USL being exceeded. The latter ignores the defects occurring on the LSL side, hence is only half of the picture.