[R] Reproducibility of experiment

Marc Schwartz marc_schwartz at comcast.net
Wed Dec 12 18:49:36 CET 2007


On Wed, 2007-12-12 at 17:23 +0100, darteta001 at ikasle.ehu.es wrote:
> Dear list,
> 
> I have an experiment that I have run 10 times in order to find out its 
> reproducibility. I wonder if there is any function that I can use for 
> obtaining a significance value of reproducibility or agreement of 
> measurements. I thought of coefficient of variation but, as far as I 
> know, I would have to set a threshold for saying the experiment is not 
> reproducible. Any pointers to something more "objective" would be very 
> helpful.
> 
> Thanks
> 
> David

I suspect that you are going to have to be more specific regarding the
subject matter and the experimental design so that those with the
requisite expertise could comment.

If this is looking at a continuous measure (ie. instrumentation
measurements), you could look at Bland-Altman methods. More information
here:

  http://www-users.york.ac.uk/~mb55/meas/meas.htm

Otherwise, given that Google returns almost a million hits with the
phrase "reproducibility of experiment"...

HTH,

Marc Schwartz



More information about the R-help mailing list