[R] Question in Chi-squared test, can I do it with percentage data?

[Dieter Menne]

Jeff wrote:
> 
> I know Chi-squared test can be done with the frequency data by R function
> "chisq.test()", but I am not sure if it can be applied to the percentage
> data ? The example of my data is as follow:
> 
> #############################################
> 
>                           KSL   MHL   MWS   CLGC   LYGC
>  independent (%) 96.22  92.18  68.54   93.80   85.74
> 
> #############################################
> 
> 

No. If this are "measured" number, eg. percent concentration, the test has
no idea if you data are measured with 0.05 or 10% standard error (which
anyway is tricky, because we are so close to 100%).

If these are counts, it makes a difference if these are 96 out of 100 or
960000 out of 1000000. 

Dieter





--
View this message in context: http://r.789695.n4.nabble.com/Question-in-Chi-squared-test-can-I-do-it-with-percentage-data-tp3335312p3335708.html
Sent from the R help mailing list archive at Nabble.com.