Speaker's Notes
A third data collection is planned for 2008. Having 3 data collection periods is a necessary condition for the determination of a trend. While it is theoretically possible to find a trend with three data collections periods, the impact of certain limitations related to the observations made of food safety practices and behaviors from one study to the next may require additional data collection points.
At the beginning of the study design it was anticipated that the number of observations for each data item would vary. The total number of observations for each data item is likely to change from one data collection period to another. When analyzing data from two data collections periods, the variation in the number of observations makes it difficult to draw statistical conclusions between overall compliance percentages or risk factor compliance percentages.
Changes in the number of observations of data items may be attributed to such things as sample variation -sampling establishments within the dame facility type that have different food products and/or procedures OR changes in industry practices. For example, establishments shifting from the use of raw shell eggs to using in shell pasteurized eggs will reduce the number of observations related to inadequate cooking from one data collection period to the next.
