Open AccessLOGLINEAR SMOOTHING: AN ALTERNATIVE NUMERICAL APPROACH USING SAS
Author(s) -
Moses Tim,
Davier Alina A.,
Casabianca Jodi
Publication year - 2004
Publication title -
ets research report series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.235
H-Index - 5
ISSN - 2330-8516
DOI - 10.1002/j.2333-8504.2004.tb01954.x
Subject(s) - smoothing , bivariate analysis , univariate , statistics , mathematics , log linear model , flexibility (engineering) , econometrics , computer science , multivariate statistics , linear model
The purpose of this report is to demonstrate loglinear smoothing using SAS PROC GENMOD. The results from four published examples, which include the smoothing of a) univariate distributions, b) bivariate distributions, c) distributions with teeth, and d) bivariate distributions with structural zeros, are reproduced to show the flexibility of the SAS procedure. Comparisons of graphical displays and likelihood ratio statistics show that the SAS procedure yields results identical to the published results. SAS PROC GENMOD provides an alternative approach to smoothing that is readily available for researchers and graduate students in educational measurement, as well as researchers from other scientific fields.