# Applications of deduction to science

Combining theories produced by induction, as the the two examples presented at the section of Applications of induction to science, or previous deductive thinking made, and using deductive methodology, one can produce a new model predicting the counts of the pathogen in the final product of the food matrix. Continuing the above two examples and, given that the starting conditions for the relationship defined in the second theory are derived from the distribution from the first theory, the regression model can be fed with the inputs of a distribution regarding counts or time, and build a stochastic model that could incorporate in this way also stochasticity. In this case the results will be also distributions, that can help in the estimation of the risk of the final product and the effect of the intervention strategies.

This route is common in predictive microbiology, epidemiology, modelling and risk assessment. Theories are made after experimental work and conclusions were derived from field experiment data. Based on these theories and conclusions, modellers build their models to simulate realistic conditions and test them with observations, that could be either experimental or monitoring, in order to validate the model. For example making a model that predicts the prevalence of Salmonella at pigs at the slaughterhouse, could be validated either with experimental data, or with the results from national Salmonella monitoring programs.

More info on the deductive approaches and applications to modelling will be soon available.

--Ilias.soumpasis 14:06, 21 November 2008 (UTC)