cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Register for our Discovery Summit 2024 conference, Oct. 21-24, where you’ll learn, connect, and be inspired.
Choose Language Hide Translation Bar
Brice_L
Level I

How F Ratio Value is determined for Brown-Forsythe and Levene's test for Equal Variance

I had evaluate the JMP output for Brown-Forsythe and Levene's test for Equal Variance under Fit X-Y analysis module.  The values from the JMP Fit X-Y are F-Ratio, DOF for factor, DOF for sample, and Prop>F.  I performed Levenes and Brown-Forsythe test in excel using the single factor ANOVA module in the data analysis add-on to complete the Levenes and Brown-Forsythe analysis.  The F-ratio from JMP and the excel F values are different.  If the F-ratio value is different from the F value, can the F-ratio be translated into an F-value?  If F-ratio and F value are the same, why are there discrepancies between excel and JMP values?  Any insight would be greatly appreciated.

7 REPLIES 7
MRB3855
Super User

Re: How F Ratio Value is determined for Brown-Forsythe and Levene's test for Equal Variance

Hi @Brice_L : The excel results you show are for comparing means (via ANOVA), not for comparing variances. There must be some other Excel output that you've missed?  

Brice_L
Level I

Re: How F Ratio Value is determined for Brown-Forsythe and Levene's test for Equal Variance

Thank you for the feedback.  Clarification.  I am manually calculating Levene and Brown-Forsythe using the ANOVA function in excel.  The image of the excel output I provided was the final step of the Levene's and Brown-Forsythe.  I have attached the excel file to further clarify what I am doing.  Youtube video also describes what I am doing   https://www.youtube.com/watch?v=HoYos9IwZNY

This video demonstrates how to conduct and interpret a Levene's Test of Homogeneity of Variances in Microsoft Excel. The result of the Levene's Test in Excel is identical to the result found using SPSS.
MRB3855
Super User

Re: How F Ratio Value is determined for Brown-Forsythe and Levene's test for Equal Variance

Hi @Brice_L :In my data set they match exactly (I tried Levene and Brown-Forsythe) . How exactly did you carry out the tests in JMP? i.e., what did you input into the Fit Y by X dialog?

Brice_L
Level I

Re: How F Ratio Value is determined for Brown-Forsythe and Levene's test for Equal Variance

Dear MRB3855,

 

Thank you for your curiosity and assistance.  I used 'lot' column as 'X' and test value for 'y'.  The lot column data type is character and test value data type is numeric. When you stated they match, are you stating that your JMP output and the excel method matched, or is my JMP output matching yours?  

MRB3855
Super User

Re: How F Ratio Value is determined for Brown-Forsythe and Levene's test for Equal Variance

Hi @Brice_L   In my case, the JMP output and the excel method matched exactly as they should. I don’t know what happened in your situation…and I assume you checked all Excel calculations carefully.

Brice_L
Level I

Re: How F Ratio Value is determined for Brown-Forsythe and Levene's test for Equal Variance

Can you please post a screenshot?

MRB3855
Super User

Re: How F Ratio Value is determined for Brown-Forsythe and Levene's test for Equal Variance

Hi @Brice_L : Sure, I'm not sure how helpful it is though (it is what it is).

And you have a calculation error(s) in your Stacked Levene sheet (cells C33-C46). 

And it appears you have some copy/paste errors when stacking. 

If you clean up these errors you will see that there is no discrepancy.

 

That said, on to the results! (for my data)

1. Levene and B-F. via JMP

MRB3855_0-1724317438293.png

2. B-F via  "Excel" method (I used JMP).

MRB3855_1-1724317510792.png

3: Levene via "Excel" method (I used JMP).

MRB3855_2-1724317567503.png