> On Sat, 23 Dec 2000, G. Marc Turner wrote:
>
> I have been teaching statistics to undergrads for many years, and the
> amount and content of what I teach has been changing. I have come to
> several conclusions. First, most of what we cover the students will not
> remember. Thus, cover only those topics which the students are likely to
>
I haven't taught stats in years, but do teach experimental psych and
labs in social psych and personality that are intended to expose students to
research in these areas while giving them hands-on practice for the
skills/knowledge they presumably, were exposed to in the stats/experimental
design classes. Students really need to understand power and effect size,
and the use of confidence intervals. The emphasis on NHST is waning, but
they still need to understand what a statistical sig. test says and does not
say. It sounds like many of us should ideally be doing two semester stats
classes and two semester research design classes, but I don't think this is
feasible for us at present.
Along with the origins of these postings, I try to get some simple
arithmetical operations involved in class exercises or homework. Like Rip's
use of inter-observer agreement or simple calculations of means, or even,
rhos, these can be useful to illustrate class ideas and is a way to
continually point out that arithmetical, measurement, and analytical ideas
are central to serious study. Ideas about variability, rates of change,
distributions, differences, effect size, etc., can all be useful and
reinforce stat concepts. That is, in addition to, and certainly not
incompatible with, the psychic, the intuitive, the
transpersonal-hermeneutic, mystical, and semi-spiritual dimensions most
people associate with psychology (I am being facetious? ;-) Gary Peterson
Gerald (Gary) L. Peterson, Ph.D.
Professor, Department of Psychology
Saginaw Valley State University
University Center, MI 48710
[EMAIL PROTECTED]
1-517-790-4491
