> Surely the mean is 14%

In the simplest sense, perhaps, if we simply arithmetically average the two
extremes.   However, I do not know if these extreme cases are part of an
orderly pattern, or isolated freaks.

I also do not know the values of all the samples, or how the samples are
distributed between these end points.  Is there a cluster (or clusters) of
points?... or a random scatter?  This matters much.

> and the scatter is 3-26%, unless you have some other measure in mind.

That is the range, but how the samples are plotted in that range is the
question.

Studying the data tables in studies will sometimes lead one to question the
conclusions reached by the authors, and whether conclusions can really be
made form the data.

Let me hasten to add that I am not suggesting any such thing in this case.

At this point, I simply do not know.

allen

-- Visit www.honeybeeworld.com/bee-l for rules, FAQ and  other info ---