Healthy eating and its determinants

Main outcome measures: The difficulty of each recommendation was measured using a Rasch model and the determinants of individual ability to undertake the dietary recommendations as a whole using regression analysis. [...] Strata quotas reflected the demographic profile for these dimensions across the population of the City of Guelph based on the 2006 population census of Canada. [...] The fitness of the Rasch model is indicated by the infit and outfit mean-square statistics which have expected values of one. [...] In order to support the unidimensionality assumption, the variance explained by the Rasch dimension in the data should be greater than 60 percent, while the proportion of the remaining unexplained variance that is explained by the first contrast (second dimension) should be less than five percent (Linacre, 2006). [...] Note that the higher the value of the converted scale, the higher the person ability in undertaking the dietary behaviors considered here.