|
Abstract:
|
In 2004, Statistics Canada conducted the Canadian Nutrition Survey. The main goal of the survey was to estimate usual intake distributions of key nutrients for 15 age-sex groups at the provincial level. A sample of 35,000 respondents was selected to do a first dietary recall and 10,000 of those respondents were asked to do a second recall. It was decided to make a weighting adjustment to reduce the bias due to the day of the week the recall was given. However, this adjustment does not solve every problem related to the day of the week the recall was given. The distribution of respondents in terms of socio-demographic characteristics, from one day to another, was not controlled. This paper presents the results of an analysis measuring the effect of several factors on the estimation of usual intake distributions for several key nutrients.
|
