Professor of Psychology and Education |
![]() |
Here's an SPSS
macro for univariate and multivariate tests of skew and kurtosis. (from
my 1997 Psychological Methods
article).
Note: this version is updated slightly to deal with some syntax changes in more recent versions of SPSS.
Instructions:
Save the macro to your hard drive. Open the data you wish
to analyze, here's some sample data
(Fisher's Iris data). Open
a syntax window
and enter the commands given in the header of the macro, here's
an example of a syntax
file. You simply need to edit the INCLUDE statement to give the
directory where the
macro (normtest.sps) is located and the
VARS statement to give the names of the variables in your data file. Have fun
analyzing!
NOTE: to get the plot of Mahalanobis distances, you need to have permission to write to the hard drive in the directory you run the macro in.
Here's an SPSS
macro for Mardia's multivariate skew (more computationally intense).
Here's syntax for a Latent Gold 4.5 program that fits an item signal detection model (see
my 2011 JMP article).
Note: save the file and you can use wordpad to open it. Here are data for the recognition memory example
discussed in the article, with 21 subjects and 120 trials per subject, with 60 old words and 60 new words.
Here's syntax for a Latent Gold 4.5 program that fits the re-parameterized DINA model (see
the Appendix of my 2011
Applied Psychological
Measurement article). Here are Tatsuoka's fraction subtraction data for the example discussed
in the article. Here's a LEM program for the RDINA model and the fraction subtraction data (recoded for LEM).
Here's a LEM program for the Higher-Order RDINA model and the fraction subtraction LEM data.
Here's syntax for a Latent Gold 4.5 program that fits a mixture signal detection model (see
my 2010 JMP article). Here are data
for the example, a subject from Swets, Tanner, & Birdsall, 1961 (note: trials are not actual trials, just observation numbers).
Here's syntax for a Latent Gold 4.5 program that fits a random coefficient SDT model by introducing a continuous latent variable
(see my 2010 JMP article). Try it with the Swets et al. data given above and compare results to the unequal variance model.
Here's an Mplus program that fits a random coefficient SDT model as a multilevel model for a single subject (!) for instructive
purposes (see
my 2010 JMP article). Here are Swets et al.'s data for Mplus.
Here's SPSS syntax to fit the unequal variance SDT model using the PLUM procedure (see
my 2003 BRMIC article).
Here are Swets et al.'s data for SPSS (in frequency form).
Here's SPSS syntax to fit the unequal variance SDT model for the race and memory example given in my 2003 BRMIC article).
Here are the data as given in Table 4.
Here are LEM and Mplus programs
for a mixture signal detection model (see my 2002 Psychological
Review article).
Here are LEM and Mplus programs for a latent class signal detection model (see my 2002 Multivariate Behavioral Research article).
Here's an aML program for a multivariate signal detection model with sample selection (see my 2003 Journal of Mathematical Psychology article).
Here's an Mplus program for a multivariate signal detection model (see my 2003 Journal of Mathematical Psychology article & 2000 EMPG presentation).
Here are SAS and SPSS programs for signal detection models expressed as generalized linear models (from my 1998 Psychological Methods article).
Pictures:
Venice
#1
#2
#3
#4 Vienna
#1
#2 Graz
Stockholm #1
#2 Spain Las
Vegas New
Orleans
World Trade Center: #1 Before/After
During Now
World Financial Center #1 #2 Central
Park: The
Sheep Meadow Pearl Harbor Indoor
rubber-powered model airplanes
Outdoor rubber-powered models:
Flying Aces Moth Pacific
Ace: #1
Flying 3/4 Jim Cahill's Clodhopper
Flying in Mitchell
field in 1971
Here's a 2012 project - a rubber-powered rudder-only radio controlled version of Dick Korda's Dethermalizer, a classic oldtime rubber job! Korda
Another recent project - a peanut scale (13") Mitsubishi Zero from the Peck-polymer kit. Zero
A 2012 project - a 52" span radio controlled (rudder only) rubber powered Mulvihill - Jim O'Reilly's Tubestake
Printing Internet Files: To print files downloaded from the Internet, you'll probably need:
One type of file has a .pdf
extension and requires Adobe Acrobat reader; instructions are
given at the website. You'll need this to read and print copies
of some of the papers given in my publications, presentations,
and course syllabi sections.
POSTSCRIPT
Some people use Postscript files (.ps
extension), which can be viewed and printed using Ghostscript and Gsview.
Here you can download self-extracting
files that contain Gsview and Ghostscript for Windows.