Dear all,:bigwave:

I wonder if anyone knows how Minitab handles outliers when using two sided ANOVA comparancy? If the data has outliers are they in the calculations or does Minitab exclude the outliers? i.e. do you have to screen the data for outliers before running the two sided ANOVA?

If you need to screen for outliers it causes a problem since you can't just delete an outlier, this would cause unbalanced data which Minitab doesn't accept..

Thank you for your help!:agree1:

Regards
Mike

Dear all,:bigwave:

I wonder if anyone knows how Minitab handles outliers when using two sided ANOVA comparancy? If the data has outliers are they in the calculations or does Minitab exclude the outliers? i.e. do you have to screen the data for outliers before running the two sided ANOVA?

If you need to screen for outliers it causes a problem since you can't just delete an outlier, this would cause unbalanced data which Minitab doesn't accept..

Thank you for your help!:agree1:

Regards
Mike
That is why I use SAS. Minitab does not know if data is an outlier. It has diagnostics tools that allow you to review data for outliers, but it does not automatically delete them.

Dear all,:bigwave:

I wonder if anyone knows how Minitab handles outliers when using two sided ANOVA comparancy? If the data has outliers are they in the calculations or does Minitab exclude the outliers? i.e. do you have to screen the data for outliers before running the two sided ANOVA?

If you need to screen for outliers it causes a problem since you can't just delete an outlier, this would cause unbalanced data which Minitab doesn't accept..

Thank you for your help!:agree1:

Regards
Mike

MINITAB recommends reviewing the ANOVA residuals versus the fitted values to identify outliers. If the outlier is confirmed, it then recommends excluding it from the analysis.

I verified that this procedure will work using the One-way ANOVA in MINITAB. MINITAB will flag the outlier using the * symbol. It also accepted the unbalanced data.

Steven and Miner,
:thanx:
Thank you for your help!

It is very much appreciated!

Regards
Mike