Re: Statistics in Psychology?
- From: "Reef Fish" <Large_Nassau_Groupen@xxxxxxxxx>
- Date: 19 Jun 2006 01:42:48 -0700
seeker wrote:
Can anyone give me a run-down on the status of biostatistics in psychology?
Yes. But it's not necessary. They run themselves down to the pits.
:-)
I've done a critique of
a psychology paper in class before, and a lot of psychology papers that are
currently being published do seem to suffer from poor statistical
techniques.
You are right on the mark there. You'll find the same in other areas
social sciences, economics, engineering, and biological and medical
sciences, in their ill-practice to malpractice of statistics.
I get the sense that at present, most biostat departments treat
psychology as more of a fringe specialty.
Statisticians in statistics departments treat MOST other departments
that makes use of statistics as in the fringe of statistics proper.
This will NOT change until the other departments take the subject
of Statistics MUCH MORE SERIOUSLY, and not as something that
can be learned in a course or two with a few SPSS or SAS runs.
Statistics is Statistics!
There ain't such an animal as business statistics, biostatistics,
statistics for the social sciences or engineering statistics.
What I mean is that there is a CORE of statistical methods
that are COMMON to ALL disciplines that make use of statistics.
Without trying to be comprehensive or even anywhere close to
complete, here are a few of those topics:
General Descriptive and Graphical methods
Interval Estimation and Hypothesis Testing: Theory and Methods
Inference and Testing of the equality of means, variances
Design and Analysis of Experiments as in various ANOVA
designs and other experimental designs.
Then there are various standard Multivariate methods
such as Multivariate Distributions and their related
marginal and conditional distributions; Principal Components
Analysis, Eigenvalue-Eigenvector and Factor Analyses, and
other more specialized methods.
In short, those merely form the BASIS or MINIMAL prerequisite
for the proper understanding and application of statistical
methods for the serious practitioners.
The role of a DISCIPLINE such as biostatistics, or business,
or psychology, or social sciences and engineering sciences
merely have certain SPECIAL additional requirements in
topics in those disciplines that are NOT part of the standard
toolbox for statisticians.
That is why the Harvard joint medical-statistical Program,
intended for medical researchers using statistical methods,
is an 8-year Program, requiring a 4-year Harvard M.D.
AND a 4-yr. Harvard Ph.D. in the Statistics Department!
You simply CANNOT take the brightest M.D.s from the
Harvard medical school and hope to make a top-notch
research statistician out of them by anything less than a
full Ph.D. program in Statistics.
That single program should give everyone some sobering
thoughts about their own training (or lack thereof) in the
Statistics areas.
The current state of statistical practice is that FAR, FAR
too often (as I've seen in these sci.stat.* groups) the
practitioners are ill-equipped to deficient in their CORE
TRAINING in Statistics -- to the extent that the very
simple and straightforward subject of multiple regression
analysis and Linear Models have brought THOUSANDS
of posts (since I joined these groups in 2005) to discuss
(and for the most part teach) the materials that are
standard in a junior-senior level to beginning graduate
level courses in those topics.
But could this change, as there
is growth in psychopharmaceuticals, and more demand
for better statistics?
It can change ONLY when people in those disciplines take
Statistics much more seriously in both their INSTRUCTION
and PRACTICE -- as Harvard being the extreme example.
Just imagine an unlikely combination of a psychophama-
ceutical degree with competence in brain surgery. Would
the current statistical training courses be adequate if they
were subsituted for brain surgery courses instead?
IMO, most of the posters and discussants in these sci.stat.*
forums LACK the training equivalent to that of a good
UNDERGRADUATE training in statistics.
If the psychology, social sciences, biostatistics, and other
disciplines take that as their FIRST STEP in training those
who are likely to have to use statistics, and use it properly,
that would go a LONG WAY toward steering the present
course back to something that can be considered reasonably
REASONABLE.
----------
My normal posting ID for was sabbotaged by posters in a travel
newsgroup who filed FALSE abuse reports. To retrieve ANY of
my Reef Fish posts, simply go to the "advanced groups search"
of http://groups.google.com and search for "Reef Fish" as author,
using any other keywords.
-- Reef Fish Bob.
-- For example, with "Reef Fish" as author, and between dates of
-- June 16 and June 18, 2006, you'll find ALL of my postings in 5
-- different newsgroups between those dates (with different addresses).
.
- Follow-Ups:
- Re: Statistics in Psychology?
- From: Old Mac User
- Re: Statistics in Psychology?
- From: Brett Magill
- Re: Statistics in Psychology?
- References:
- Statistics in Psychology?
- From: seeker
- Statistics in Psychology?
- Prev by Date: Re: Quizzes for math "language" skills?
- Next by Date: Re: Statistics in Psychology?
- Previous by thread: Re: Statistics in Psychology?
- Next by thread: Re: Statistics in Psychology?
- Index(es):
Relevant Pages
|
