From the bootstrap samples, the factor ranges are obtained taking into account the correlation between the two parameters of the fitted Gamma pdf.
In individual samples the factor ranged from 2 to 218.
Although the raw scores on the four scales are not comparable between the two samples, the factor structure underlying the short scales is comparable with the factor structure of the original scales (data not shown).
Once the proxy models are constructed, a sampling method is needed to sample the factors and to translate the uncertainty from the input to the response.
While quantitative studies using questionnaires have the advantage of being able to assess factors associated with smoking behaviors in a large sample, the factors that can be explored are limited to what have been included in the questionnaires.
In terms of predicting participation, the evidence suggests that, for this sample, the factors significantly predictive of probability of participation are age at delivery and insulin use during pregnancy.
Experimental design is an efficient method for sampling the factor space and calculating the response with minimum number of runs (Montgomery 2008).
Since the FIOS was originally developed for an adult sample, the factor structure for the adolescent sample was first investigated using a confirmatory factor analysis (CFA).
To the extent that analytical and experimental factors do vary among the samples, these factors should be explicitly included in the model used to develop the classifier.
In several samples, the absorption factor was not in the range 1.5 5.
