Solve this question

[Hazux A]

Find the question at http://expertstatistics.blogspot.com/p/statistical-problems.html

Give answers in comment. Also become a member in this blog.

[Allabux]

Chi square :

i just consider that some patient were tested their illness pre and
post medicine. suppose following is the data.

Category Scale
Pre Moderate
Pre Moderate
Pre Mild
Post Moderate
Pre Mild
Pre Severe
Post Severe
Pre Mild
Post Moderate
Post Mild
Post Mild
Post Severe


the minitab result is:

Rows: Category Columns: Scale

Mild Moderate Severe All

Post 2 2 2 6
2.500 2.000 1.500 6.000

Pre 3 2 1 6
2.500 2.000 1.500 6.000

All 5 4 3 12
5.000 4.000 3.000 12.000

Cell Contents: Count
Expected count


Pearson Chi-Square = 0.533, DF = 2, P-Value = 0.766
Likelihood Ratio Chi-Square = 0.541, DF = 2, P-Value = 0.763

the interpretation would be there is no significant change after
getting medicine also.

[Padmanaban G.V]

Hi,

 

 First we have make contigency table and we can use chisquare test. see the attachment.

 

On Fri, May 28, 2010 at 2:21 PM, Hazux A <hazux.st...@gmail.com> wrote:

Find the question at http://expertstatistics.blogspot.com/p/statistical-problems.html

Give answers in comment. Also become a member in this blog.

--
Regards,
Padmanaban Venugopal

[Hazux A]

THE POINT TO BE NOTED IS THAT THE WE HAVE COMPARE PRE AND POST
MEASURES. IT IS A PAIRED CASE INVOLVING ORDINAL VARIABLES.

CHI-SQUARE TEST CANNOT BE USED IN THIS CASE. ANY OTHER SUGGESTIONS?

PLEASE POST THE REPLY HERE ITSELF.

PLEASE BECOME A MEMBER ALSO.

On May 28, 4:04 pm, "Padmanaban G.V" <padmanaban...@gmail.com> wrote:
> Hi,
>
>  1.JPG
> 26KViewDownload

>
>  First we have make contigency table and we can use chisquare test. see the
> attachment.
>

[Hazux A]

THE STUDY VARIABLE CONTAINS ORDINAL DATA CONTAINING 3 CATEGORIES. A
PAIRWISE COMPARISON SHOULD BE MADE ON THIS VARIABLE.

CHI-SQUARE TEST IS USED ONLY FOR DISCRETE DATA. HENCE IT IS NOT
SUITABLE. WE HAVE TO DO A PAIRWISE TEST ALSO.

TO COMPARE PAIRWISE ORDINAL DATA, WILCOXON SIGN TEST IS USED.

On May 28, 4:04 pm, "Padmanaban G.V" <padmanaban...@gmail.com> wrote:
> Hi,
>
>  1.JPG
> 26KViewDownload
>

>  First we have make contigency table and we can use chisquare test. see the
> attachment.
>

[Allabux]

can you make the problem more practical or which i gave an example is
enough practical situation? if yes in this situation I compared pre
and post effect only. chi square is giving enough result for this
situation

[Hazux A]

READ THIS BOOK. IT IS INTERESTING
http://groups.google.com/group/statistical-experts/web/Sample_HBStatistic.pdf

> > Padmanaban Venugopal- Hide quoted text -
>
> - Show quoted text -

[Allabux]

if you use minitab it will make the crosstabulation and give the
chisquare results. no need to worry about tabulation