binary regression, RR and RD

binary regression, RR and RD - LOGISTIC REGRESSION EXAMPLES...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
LOGISTIC REGRESSION EXAMPLES Case control design over-samples cases from the population Cases are higher proportion of sample than they are in the population For logistic model: binary outcome Y indicates case status p= P[Y=1|x’s] is probability a subject in the sample is a case In model: logit (p) = β 0 + β 1 x 1 , β 0 is the value of logit (p) in the unexposed in the sample – higher than it would be for the population β 0 does not have a population interpretation When there is no selection bias (cases and control are sampled without regard to exposure, there is same π in numerator and denominator) Example: In sample p= ½ In population, p=1/10,000 How many times the p is bigger? Sample Odds = π * Population odds Here, π = 10,0000 We have seen that logistic regression provides estimates of ORs : logit (p) = β 0 + β 1 X - Exponentiated coefficients e β are ORs and adjusted ORs - Useful for case-control studies Can also fit regression models to binary data that yield adjusted RDs and RRs - (p) = β 0 + β 1 X for RDs - Log (p) = β 0 + β 1 X for RRs Generalized in two sense
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 02/09/2012 for the course STAT 513 taught by Professor Barbaramc.knight during the Spring '11 term at University of Washington.

Page1 / 5

binary regression, RR and RD - LOGISTIC REGRESSION EXAMPLES...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online