Competing Failure modes- Different results- same example

[Temanag]

Hello!
In the Training Guide there is an example for Competing Failure modes.
The column " Number in State" was used for the entering of the data.
According to the calculation b= 0,6711, T=449,4272. I agree

I found the same example in the "Life Data Analysis Reference" page 490.
On page 493 the data are entered without using the column "Number in State".
I entered the data in Weibull++ in the same way. (exactly how they are shown on page 493) I became different results. b= 0,722, T= 354,6271
I am sure that I entered the data correctly, because I really checked the data many times. I become different results not only at this example, also when i solve other similar examples.

Can you tell/ explain me why is there a difference in the results?
Thanks

[Pantelis]

Depending on your user and analysis settings W++ can let you treat grouped and ungrouped data differently when using regression (not MLE).

Specifically:

Under regression, which I am assuming is what you are doing given the differences noted, you can use all points on the plot or just the group upper point (see discussion at http://www.weibull.com/LifeDataWeb/grouped_data_analysis.htm (http://www.weibull.com/LifeDataWeb/grouped_data_analysis.htm) for more details on this).

This behavior can be controlled from the [User Setup]|[Analysis]|[Default Grouped Data Settings].

If the ‘Ungroup on Regression’ is unchecked you will notice the behavior as discussed in the referenced link. If checked your results will be identical.

Hope this helps.

P.S. Note that this behavior is the same regardless of the analysis you are doing, i.e. it is not because you are using competing failure modes.

[Temanag]

thanks

the results are fully identical now :)