gender, and hours and revenue. Table 10.14 shows the significance of each
added variable to the model containg hours using type 1 analysis. None of
these three variables meets the 0.10 level for entry into the model.
EXERCISE
10.1 Inflammation of the middle ear,otitis media(OM), is one of the most
common childhood illnesses and accounts for one-third of the practice of
pediatrics during the first five years of life. Understanding the natural
history of otitis media is of considerable importance, due to the morbid-
ity for children as well as concern about long-term e¤ects on behavior,
speech, and language development. In an attempt to understand that
natural history, a large group of pregnant women were enrolled and
their newborns were followed from birth. The response variable is the
number of episodes of otitis media in the first six months (NBER), and
potential factors under investigation are upper respiratory infection
(URI), sibling history of otitis media (SIBHX; 1 for yes), day care,
number of cigarettes consumed a day by parents (CIGS), cotinin level
(CNIN) measured from the urine of the baby (a marker for exposure to
cigarette smoke), and whether the baby was born in the fall season
(FALL). Table E10.1 provides about half of our data set.
(a)Taken collectively, do the covariates contribute significantly to the
prediction of the number of otitis media cases in the first six months?
(b)Fit the multiple regression model to obtain estimates of individual
regression coe‰cients and their standard errors. Draw your con-
clusions concerning the conditional contribution of each factor.
(c)Is there any indication of overdispersion? If so, fit an appropriate
overdispersed model and compare the results to those in part (b).
(d)Refit the model in part (b) to implement this sequential adjustment:
URI!SIBHX!DAYCARE!CIGS!CNIN!FALL
(e)Within the context of the multiple regression model in part (b), does
day care alter the e¤ect of sibling history?TABLE 10.14
Variable LRw^2 pValue
Residency 0.817 0.3662
Gender 1.273 0.2593
Revenue 0.155 0.6938372 METHODS FOR COUNT DATA