For both models, a backward selection process with all potential confounders was used to reach the final multivariate model, and the i th day of stay was introduced as a discrete or a continuous variable with possibly mathematical transformation according to the minimum deviance criteria.
Minimum deviance (MPE) and absolute deviance (MAPE) of predicting the value of the response variable (health-care costs) considered as adequacy of methods.
Once φ is fixed, the linear model has a unique minimum Deviance.
Computation involves a search for local minima, selection of the minimum Deviance at φ, and then a comparison amongst these minima for different φ values.
Mean scores were chosen because they designate minimum deviance (i.e., variance) of the scoring patterns subsumed.
In any case the minimum Deviance at any particular φ can be no higher than the values found here, so any missing minima could only strengthen the evidence of non-linearity.
For example, by visual inspection of the dose response curve, we may identify that the potential MMT might be within 25 28°C, and we then examined the potential MMT from 25 to 28°C (by 0.1°C) to identify the model with a minimum residual deviance.
