Why is ABC based on only two weeks data?
Statisticians advise that providing the ten day sample is spread over two one week periods, and are taken at representative times during the year, there is little to be gained by using longer sampling periods. However, if you consider that increasing the period of sampling would materially enhance your results you are free to do so. You will, of course, be sensitive to the trade off between acquiring more refined results and the extra costs incurred. [Appendix I of the ABC manual illustrates the comparative accuracy of using differing sampling periods]. Picking a period of time that is ‘typical’ of police activities is essential. The major consideration when choosing the period to sample will be local factors, such as holiday seasons for tourist resorts, and a judicious choice is required to maximise the accuracy. For this reason, many forces choose to split their surveys and do two one-week sampling periods or 4 half weeks. Although this is not necessary, rolling studies are often vie
