such as the Wilcoxon Rank Sum Test or the
equivalent Mann Whitney U Test.
You could also consider resampling or permutation tests...
WBW
__
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500
Excuse the bad grammar or typo noted below... It's been a long
morning already, and it's still not 9 am...
:)
Bill
On Fri, 15 Feb 2002, William B. Ware wrote:
What are your samples sizes? If there are equal or nearly so, the t-test
*they*
is robust
the high end of its value... Sort of a maximum range... I
prefer not including the additional one unit...
Bill
__
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500
Have you tried simulations? with something like Resampling Stats or
Minitab?
WBW
On 26 Sep 2001, Warren wrote:
Hi,
I've been teaching an introductory stats course for several years.
I always learn something from my students...hope they learn too.
One thing I've learned is that confidence
__
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500 Measurement, and Evaluation
University of North Carolina PHONE (919)-962-7848
Chapel Hill, NC 27599-3500 FAX: (919)-962
Could it be that it was more like Ninety percent of the students in
District X were above the national median?
WBW
__
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500
protection with greater power than a "pure" Bonferroni
approach...
WBW
______
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500 Measurement, and Evaluation
Also of interest are the writings of R.B. Cattell from the '50s and
'60s. There is a nice discussion of his "data box" in the first Handbook
of Multivariate Statistics, published in the mid '60s.
Bill
______
Willi
of variance. ANOVA is robust against such
violations when the cell sizes are equal...
WBW
__
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500 Measurement
__
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500 Measurement, and Evaluation
University of North Carolina PHONE (919)-962-7848
Chapel Hill, NC 27599-3500 FAX: (919
d...
Bill
______
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500 Measurement, and Evaluation
University of North Carolina PHONE (919)-962-7
If the p-level is below your "predetermined" level of significance, then
you would reject the null hypothesis that the data are a sample from a
normally distributed population...
WBW
______
William B. Ware, Professor
Good points by Karl... It does not tell you how it might depart from
normality... you should also look at measures of skewness and kurtosis, as
well as graphic plots... q-q plots and histograms...
WBW
On Tue, 6 Jun 2000, Karl L. Wuensch wrote:
Do not, however, assume that this
Regading the expected mean squares, see books on experimental design by
Kirk, Winer, or Keppel.
WBW
__
William B. Ware, Professor and Chair Educational Psychology,
CB# 3500
14 matches
Mail list logo