Sas stat statistical software package

The food neophobia and liking data were first analyzed using an analysis of variance (ANOVA) test that considered the school (school 1-3), the age (6-9 years), the gender, and the stimulus (the F&V provided) as factors.
The relationship among the BMI, food neophobia and liking data was evaluated using principal component analysis (PCA). Autoscaling was performed on the data prior to any modeling. Cross validation was chosen as the validation method. To further interpret the relationship between liking of F&V and food neophobia, the children were divided according to their level of neophobia into 3 groups: 'low' (children with scores in the lower 25 th percentile of food neophobia scores, score ≤ 17, n=141), 'medium' (children with scores between the 25 th and 75 th percentiles, score ≥ 18 and ≤ 24, n=234) and 'high' (children with scores in the upper 25 th percentile, score ≥ 25, n=154). The data were subjected to a GLM ANOVA that considered the Neophobia level ('low', 'medium' and 'high') , Stimulus category (fruits and vegetables) and their interaction as factors and liking as a dependent variable.
All ANOVAs were conducted using SAS/STAT statistical software package version 9.3.1.
(SAS Institute Inc., Cary, USA). PCA modeling was performed using The Unscrambler X software (CAMO Software AS, Oslo Norway).