Introduction
Recently, confidence intervals (CIs) associated with param
eter estimates in the susceptible-
infected-recovered (SIR) epidemiological model have been
developed (Chowell et. al. [4]).
When model assumptions are met and the observation error is r
elatively small, these CIs are
relatively short, as we will illustrate. This work describe
s the behavior of CIs for parameters as
observation and/or equation or model error becomes larger,
and includes a comparison of two
estimation procedures. The first procedure fits a simple line
ar regression relating the per-time-
step response and predictors. This procedure demonstrates
significant bias as observation error
increases. In general, observation error in predictors lea
ds to bias to varying degrees, as has been
illustrated in the “errors-in-variables” literature (Car
roll et al. [3]). For example, bias arising
from using measurements of species abundance rather than tr
ue species abundance has recentl