Japan Women’s University Journal: Faculty of Integrated Arts and Social
Sciences, 2015, 26, 99-112. (published in 2016.05)
A Two-Step Analysis with Common Quantification of
Categorical Data
Yasuharu Okamoto
Abstract
A data
from a psychological research may contain categorical variables.
This
study proposes to use common quantification of categorical variables for
various analyses,
so that
results from those analyses can be interpreted according to common
quantification.
The
proposed method consists of two steps.
In step
1, categorical variables are quantified independently of the subsequent
analyses.
After
quantification, in step 2, various analyses are conducted based on common
quantification.
A
hypothetical data set, including two categorical variables and one continuous
variable, was prepared.
Categorical
variables were quantified in step 1; then in step 2, using the quantification
in step 1,
principal
components, regression, and discriminant analyses were successfully conducted.
Full Text (link to JWU's repository),
Full Text (direct link to the pdf file)
A Related Paper (Okamoto, 2015; in Japanese)
An Application of Common Quantification to Real Data (Okamoto, 2017; in Japanese)
Comparison with Quantifications 1 and 2 (in Japanese)
Programs and Explanations (in Japanese)
How to Calculate Quantifications with SPSS (in Japanese)
A Python Program for Common Quantification (in English)