Practice
Exercises
Given the model
logit PðXÞ¼aþbEþg 1 ðSMKÞþg 2 ðHPTÞþd 1 ðESMKÞ
þd 2 ðEþHPTÞ;where SMK (smoking status) and HPT (hypertension sta-
tus) are dichotomous variables.Answer the following true or false questions (circle
T or F):
T F 1. IfEis coded as (0¼unexposed, 1¼exposed),
then the odds ratio for theE, Drelationship that
controls for SMK and HPT is given by
exp[bþd 1 (ESMK)þd 2 (EHPT)].
T F 2. IfEis coded as (1, 1), then the odds ratio for
theE, Drelationship that controls for SMK and
HPT is given by
exp[2bþ 2 d 1 (SMK)þ 2 d 2 (HPT)].
T F 3. If there is no interaction in the above model
andEis coded as (1, 1), then the odds ratio
for theE, Drelationship that controls for SMK
and HPT is given by exp(b).
T F 4. If the correct odds ratio formula for a given cod-
ing scheme forEis used, then the estimated odds
ratio will be the same regardless of the coding
scheme used.Given the model
logit PðXÞ¼aþbðCHLÞþgðAGEÞþdðAGECHLÞ;where CHL and AGE are continuous variables,Answer the following true or false questions (circle
T or F):
T F 5. The odds ratio that compares a person with
CHL¼200 to a person with CHL¼ 140
controlling for AGE is given by exp(60b).
T F 6. If we assume no interaction in the above model,
the expression exp(b) gives the odds ratio for
describing the effect of one unit change in CHL
value, controlling for AGE.Suppose a study is undertaken to compare the lung cancer
risks for samples from three regions (urban, suburban, and
rural) in a certain state, controlling for the potential con-
founding and effect-modifying effects of AGE, smoking
status (SMK), RACE, and SEX.96 3. Computing the Odds Ratio in Logistic Regression