Dear all
I came across an unpublished paper by Christian Schunn suggesting the use
of what he calls "statistical significance bars" in preference to using
standard deviations, standard errors, or confidence intervals as error
bars. His rational seems sound to me, and he has an Excel file that will
calculate the correct bars and draw the chart available for download along
with his paper at http://www.lrdc.pitt.edu/schunn/SSB/ .
My question is whether anyone would recommend against using these
non-standard error bars in manuscripts for publication (assuming that the
journal editor has not stated a preference for something else) or indeed
in my PhD dissertation? Google and PsychINFO do not produce any critique
of Schunn's method, so I am unsure whether it is an appropriate technique
or not. Any views?
Thanks,
Nicholas Gibson
--
Psychology and Religion Research Programme
Faculty of Divinity, University of Cambridge
West Road, Cambridge, CB3 9BS, UK
tel +44 (0)1223 763010 · fax +44 (0)1223 763003
http://www.divinity.cam.ac.uk/pcp/personnel/nicholas.html
|