Chi-Square and Bonferroni

[Tom]

 

Hello

 

Marta GG helped me a lot with the syntax to perform a Bonferroni adjusted comparison (see below) of the five groups.

 

It works fine, but with these datas I got a result, which I can’t explain at all: With a highly significant Chisquare there seems to be no significant differences between the groups!? Part of my problem ist, honestly, that I do not really understand what CTABLES with Bonferroni adjustment does.

 

Thanks for any hint!

Tom

 

 

 

DATA LIST LIST/Gruppe Result Count(3 F8).

BEGIN DATA

1          1          48

2          1          142

3          1          249

4          1          278

5          1          168

1          2          24

2          2          58

3          2          62

4          2          64

5          2          55

END DATA.

WEIGHT BY Count .

VALUE LABELs Result 1'No' 2 'Yes'.

VALUE LABELs Gruppe 1'1-3' 2'4-10' 3'11-20' 4'21-30' 5'31-40'.

* A) Using CTABLES with Bonferroni adjustment *.

CTABLES

  /VLABELS VARIABLES=Gruppe Result DISPLAY=DEFAULT

  /TABLE Result [COUNT F40.0, COLPCT.COUNT PCT40.1] BY Gruppe

  /SLABELS POSITION=ROW

  /CATEGORIES VARIABLES=Gruppe Result ORDER=A KEY=VALUE EMPTY=INCLUDE

  /COMPARETEST TYPE=PROP ALPHA=0.05 ADJUST=BONFERRONI ORIGIN=COLUMN

  INCLUDEMRSETS=YES CATEGORIES=ALLVISIBLE.

[<Joanne Tse>]

Hi,

 

Due to Veteran's Day, our office is closed.  I will reply to your email on Monday, 11/15/10.

 

Thank you.

 

Joanne Han

Marketing Research Manager

UCLA Extension

www.uclaextension.edu

 

[Marta Garcia-Granero]

Sorry for the delay (kinda hectic days...)

Bonferroni adjustment can be too conservative if the number of groups compared is high (in your case, with 5 groups you have 10 pairwise comparisons). Since Marascuilo procedure is even more conservative (Scheffee like) it will not give better results than CTABLES with Bonferroni...

Try ADJUST=NONE instead of ADJUST=BONFERRONI.

I tried with your data and it gives A&B different from C&D. Unfortunately, group E is stuck right in the middle (not significantly different from neither A&B nor C&D).

HTH (somehow),
Marta GG


El 12/11/2010 11:54, Balmer Thomas escribió:

DATA LIST LIST/Gruppe Result Count(3 F8).

BEGIN DATA

1          1          48

2          1          142

3          1          249

4          1          278

5          1          168

1          2          24

2          2          58

3          2          62

4          2          64

5          2          55

END DATA.

WEIGHT BY Count .

VALUE LABELs Result 1'No' 2 'Yes'.

VALUE LABELs Gruppe 1'1-3' 2'4-10' 3'11-20' 4'21-30' 5'31-40'.

* A) Using CTABLES with Bonferroni adjustment *.

CTABLES

  /VLABELS VARIABLES=Gruppe Result DISPLAY=DEFAULT

  /TABLE Result [COUNT F40.0, COLPCT.COUNT PCT40.1] BY Gruppe

  /SLABELS POSITION=ROW

  /CATEGORIES VARIABLES=Gruppe Result ORDER=A KEY=VALUE EMPTY=INCLUDE

  /COMPARETEST TYPE=PROP ALPHA=0.05 ADJUST=BONFERRONI ORIGIN=COLUMN

  INCLUDEMRSETS=YES CATEGORIES=ALLVISIBLE.