cusedataREADME.rtf

{\rtf1\ansi\ansicpg1252\cocoartf1265
{\fonttbl\f0\fswiss\fcharset0 Helvetica;}
{\colortbl;\red255\green255\blue255;}
\margl1440\margr1440\vieww10800\viewh8400\viewkind0
\pard\tx720\tx1440\tx2160\tx2880\tx3600\tx4320\tx5040\tx5760\tx6480\tx7200\tx7920\tx8640\pardirnatural

\f0\fs24 \cf0 Read Me: CUSE expanded \
source: http://data.princeton.edu/wws509/\
\
n(observations) = 16 (originally) \
\
Dependent Variable: \
\
using: a count of the number of observations matching the factor levels in the row and is using birth control. \
 \
not using: a count of the number of observations matching the the factor levels in the row and is not using birth control \
\
Example: in the first row - 6 women were observed to be using birth control, were younger than 25, had low education, and do want more children. 53 women not using birth control met the same conditions. \
\
EXPANDED: \
\
Binomial expansion by factor level and count to produce binomial column such that 0=not using, 1= using, and there is one row for each count in the original 16 observation data set. \
\
UsingMultiClass: a derived column for multi class classification tasks that ranks observations by membership in userate group, so that women observed as part of the 4-30 using\
\
\
Independent Variables: \
\
age: four factor levels for <25, 25-29, 30-39, 40-49. Each of these factor levels has been expanded to a corresponding binomial column. \
\
education: two factor levels - low and high, each with a corresponding binomial expanded column. \
\
Wantsmore: two factor levels - yes and no, each with a corresponding binomial expanded column\
\
\
\
\
}