Download presentation
Presentation is loading. Please wait.
Published byHarold Hines Modified over 9 years ago
1
Categorical Data Analysis: When life fits in little boxes AnnMaria DeMars, PhD.
2
What are we going to do today? Basic statistics Logistic regression
3
An actual example Are you going to die soon?
4
Our data Kaiser Permanente Study of the Oldest Old, 1971-1979 and 1980-1988: [California] DEPENDENT VARIABLE: Dthflag = 1 if Died during study period 0 if alive at end of study period
5
Our data PREDICTOR VARIABLES: Gender nursehome = 0 if lived at home continuously 1 = admitted to nursing home at any time
6
STACKED BAR Figure 1
7
STACKED BAR Figure 2
8
Compare the effects
9
…How to get one: STACKED BAR CHART TASKS > GRAPHS
10
Stacked Bar Chart with SAS Enterprise
13
The Syntax PROC GCHART DATA=mydata.oldpeople; VBAR gender / SUBGROUP=dthflag TYPE= PCT INSIDE=PCT ;
14
Now a Brief Look at Logistic Regression
15
Our data PREDICTOR VARIABLES: Nursehome = lived in nursing home during study Age_comp = age when study began ERvisits= number visits to ER during study Gender
16
Odds ratio Value of 1.0 means : The odds are the same for the two groups, e.g., women are just as likely to die as men
19
Receiver Operator Charactistics SENSITIVITY - the percentage of true positives, the people we predicted would die who did SPECIFICITY - or true negatives, the number of people we said would NOT die, who did not
20
PERFECTION!
21
TASKS > REGRESSION > LOGISTIC
22
The Process:
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.