db@cstvax.UUCP (Dave Berry) (09/20/85)
My thesis will involve behavioural research experiments, aimed at
finding if using a certain debugger helps people learn the corresponding
language. I've read various texts on this sort of work in Computing
(Shneiderman, Weinberg, etc. etc.) and articles (e.g. by Brooks, Shiel),
which have been very helpful so far, but I need a good introductory text to
the statistics required, and how the choice of statistical tests affects
the experiment design.
The problem is not so much finding a book - there are hundreds of them -
but I would like a recommendation for a good introductory book, rather
than plucking ones at random of the library shelves. I need to know
about the T-test, analysis of variance, use of covariants, PVA, measuring
transfer effects (etc.) and how to make practical use of them. I have asked
some people locally but they've obviously been the wrong people to ask,
as they weren't able to help!
Thanks in advance for any info.
As usual, mail & I will summarise if requested.
--
Dave Berry. CS postgrad, Univ. of Edinburgh
...mcvax!ukc!{hwcs,kcl-cs}!cstvax!dblmiller@ucla-cs.UUCP (09/22/85)
>...I would like a recommendation for a good introductory [statistics] book. >I need to know about the T-test, analysis of variance, use of covariants, >PVA, measuring transfer effects (etc.) and how to make practical use of them. Winer's book is what you're looking for. Thorough, complete with lots of worked out examples. For a more unified point of view, also see Kerlinger & Pedhazur, "Multiple Regression in Behavioral Research." L. Miller