[R] test for proportion or concordance

Adrian Johnson oriolebaltimore at gmail.com
Thu Aug 3 16:19:18 CEST 2017


Hello group,

my question is deciding what test would be appropriate for following question.

An experiment 'A' yielded 3200 observations of which 431 are
significant. Similarly, using same method, another experiment 'B' on a
different population yielded 2541 observations of which 260 are
significant.

There are 180 observations that are common between significant
observations of A and B.
(180 are common between 431 and 260).

80 observations are specific to A
251 observations are specific to B.

The question are the 180 observations  that are common between A and B
- are these 180 common observations occurring by  chance?

What test would be appropriate for this scenario.  (if my total
observations are fixed between two experiments A and B, I could use
Cohens kappa for concordance or Chi-square etc.
Since the total observations differ between experiments A and B, I
dont know what test would be appropriate.   I appreciate your help.

thanks



More information about the R-help mailing list