The adjusted coefficient of determination (R
2
adj), the root mean
square error (RMSE), and % of sum error of prediction (%SEP) were
calculated to determine the goodness-of-fit. The existing growth
data for S. aureus in different meat products and milk collected from
literature (Table 1) were compared with the models developed in
the current study to evaluate the performance of developed
models. While substrates, methods, and presentation of results
varied among the published data, for overall consistency, only the
data from the primary model (Baranyi model) were fitted to hyperbola
model for
l and square root model for mmax. The hyperbola
model was exclusively used because the development was significant
for all published data. Only secondary models developed
herein with low inoculum size (3.0 log CFU/g) were used for
comparison purpose. In order to compare the observed and predicted
growth parameter during storage, the acceptable prediction
zone (APZ) approach was used by calculating relative error (RE) of
each independent growth curves as suggested by Oscar (2002).
Prediction was considered acceptable when at least 70% of the RE
values were inside the corresponding acceptable zone. Different
widths of APZ are used to assess the performances of individual
growth parameters because of differences between experimental
errors associated with different kinetic parameters. Therefore, the
RE should be considered more wide for
l (from 60 to 30%) and
narrow for
mmax (20 to 20%) (Min & Yoon, 2010; Oscar, 2002).