Nav view search

Navigation

Search

The alert notifications for products from the European Economic Area and other countries in the Rapid Alert System for Food and Feed (RASFF)

The alert notifications for products from the European Economic Area and other countries in the Rapid Alert System for  Food and Feed (RASFF)

 

Key words:
food safety, RASFF, regression analysis, polynomial, correlation tables.

 

Summary
The purpose of the article was to examine variability of the alert notifications number in the Rapid Alert System for  Food and Feed (RASFF) in the years 1990–2012 for products from all countries, the European Economic Area (EEA)  countries and other countries (using the trend / regression lines and the determination coefficient R 2 ) and to  examine the relationship between the number of alert notifications for products from the EEA countries and other countries and hazards categories within products categories (by means of correlation tables and using χ 2 statistics  and contingency coefficients). The value of the determination coefficient R 2 in the case of polynomial regression  model of sixth-degree was about 0.9 (very good fitting) and about 0.6 (moderate fitting) for the lineare rgression  model (in the three examined groups of countries). Despite a very good fitting, the olynomial regression model of the  sixth-degree was not adequate to make a prediction for 2013. In fact empirical data for this year was higher than the  predicted values y ˆ p . In turn, the values y ˆ p , predicted based on the linear regression model, indicate a  significantly greater increase in the number of alert notifications in RASFF in 2013 and 2014 (than in the case of polynomial model). The values y ˆ p , predicted based on this model were within the prediction standard error S y ˆ p  in case of all and EEA countries. Results obtained from the linear and polynomial models point to, however, the need  of using an another model – the multiple regression model in further studies. It could in fact take into account a  larger number of independent variables, eg.: accession of new countries to the European Union, the introduction of  rejections at the border of the EEA, the population, the number of tourists, the volume of production (not just time). It  would allow to obtain the better model fitting to empirical data (the determination coefficient R 2 ) and a smaller  prediction standard error S y ˆ p . The values of the contingency coefficients allowed observing a high relationship of  the alert notifications for products from the EEA and other countries and hazards categories for the following products  categories: crustaceans and products thereof, fats and oils, cereals and bakery products and soups, broths,  sauces and condiments. Therefore, in the case of these products, traceability within the food origin and  movement is particularly important. 

dr inż. Marcin PIGŁOWSKI
Gdynia Maritime University
Faculty of Business Administration and Commodity
Sciences
Department of Commodity and Quality Science
ul. Morska 81-87, 81-225 Gdynia
phone (+48) 58 6901295
e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.

Additional information