[Rasch] Rasch analysis of interval data

Stephen Humphry shumphry at cyllene.uwa.edu.au
Thu Jul 24 03:05:32 EST 2008


> Don't get me wrong. Psychology needs explicit theories of cognitive
> processes of the kind you have developed. It desperately needs more
> Piagets than it does Fred Lords. The truly disturbing thing about the
> behavioural sciences is the attitude of whilst it is nice to have
> "substantive theory", it's not necessary for measurement. On the
> contrary, if we are genuinely measuring a natural system, then our
> "substantive" and "measurement" theories are one and the same thing,
> from a realist perspective. Instead, psychometricians are always hoping
> the practitioners will provide the substantive theory, whilst the
> practitioners seem to think that psychometric models actually provide
> this. This kind of thinking pervades and dominates the SEM and HLM
> lands, for example.

Hear hear!


More information about the Rasch mailing list