Search the Elsmar Cove!
**Search ALL of Elsmar.com** with DuckDuckGo including content not in the forum - Search results with No ads.

Determining Statistical Equivalence - Our Laboratory Results vs. our Suppliers

ifserav

Involved In Discussions
#1
I have to show a statistical equivalence between our lab's results and the results of the supplyers.

How can i do this?, i try to use ANOVA, but the results although are on the accepted rank are sometimes so different.


Do you know how to do or any advice?

Thank you so much.

Sergio Avila
 
N

NumberCruncher

#2
Re: history development

Hi ifserav

The usual method for comparing analytical results like this is to use a Bland-Altman plot.
There are several threads on this forum. The following pdf file is also interesting. You might want to look it up on google

medlabstats.com/bland/BLAND-AND-ALTMAN-PLOTS-IN-EXCEL.PDF

Also, if you post your data, there may be someone able to help you with the analysis. You don't have to include names of companies or any commercial information. Just the numbers, with some information about what the numbers represent. For example, concentration of a chemical measured by atomic emission, bacterial counts...

NC
 

Statistical Steven

Statistician
Staff member
Super Moderator
#3
Re: history development

Hi ifserav

The usual method for comparing analytical results like this is to use a Bland-Altman plot.
There are several threads on this forum. The following pdf file is also interesting. You might want to look it up on google

medlabstats.com/bland/BLAND-AND-ALTMAN-PLOTS-IN-EXCEL.PDF

Also, if you post your data, there may be someone able to help you with the analysis. You don't have to include names of companies or any commercial information. Just the numbers, with some information about what the numbers represent. For example, concentration of a chemical measured by atomic emission, bacterial counts...
NC
Bland Altman is one of the many options avaialble to you. Without data it is hard to determine the best analysis method.

btw, use this link:
http://medlabstats.com/bland/BLAND-AND-ALTMAN-PLOTS-IN-EXCEL.pdf
 
N

NumberCruncher

#5
Re: history development

Hi ifserav

There's not enough data in the spreadsheet to do a comparison. Can you post all of your data in the same format?

NC
 
Top Bottom