
Modeling dynamic exposure–response relationships 179
Num Den
Effect DF DF F Value Pr > F
loglag 1 543 413.52 <.0001
DRUG*hour1 3 543 86.51 <.0001
DRUG*hour2 3 543 3.92 0.0087
Estimates
Standard
Label Estimate Error DF t Value Pr > |t|
a vs p 0.1899 0.03162 543 6.01 <.0001
c vs p 0.2367 0.03190 543 7.42 <.0001
a vs p 2h 0.02590 0.03169 543 0.82 0.4140
c vs p 2h 0.02159 0.03168 543 0.68 0.4958
In general, it seems the first model we fit is reasonable.
6.7 Testing hysteresis
A proper assessment for hysteresis is advisable before fitting a complex dy-
namic model to deal with hysteresis. Even when the hysteresis is statisti-
cally s ignificant, sometimes it may not ...