Statistical approach to food sampling

Although the U.K. food sampling network is well developed, the number of food poisoning incidents continues on an upward trend. Existing food sampling programmes used by the U.K. local authorities, where they exist, operate in a 'hit or miss' fashion. The use of small sample sizes is also common. This study examined the costs and benefits of current sampling programmes, and proposes an alternative using statistically validated approach to food sampling. Because sampling involves errors due to uncertainties and variations, a statistically validated sampling model was developed in an attempt to determine suitable sample sizes under various sample proportions that would also satisfy good normal approximation in order to reduce margin of error to a minimum.