Samplingandsamplesizecalculation样本量计算方法全.ppt

* * 8 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Given the above assumptions: (a) calculate the minimum number of severe anaemia cases needed in each treatment group. (b) if the prevalence of severe anaemia is approximately 20% (based on previous work in this population), obtain the number of women that need to be screened at the antenatal clinics. (c) verify that the same results in (a) would be obtained for a cohort study if the relative risk of the cohort study corresponds to the ratio of the proportions here (i.e. p2/p1). (Note: you can use Epi Info, and indicate in your answer all the parameter values, e.g. power, p1, p2, etc, that you have specified in the process). Practical on Determining Sample Size (2) * * * * The universe is the blue cloud * * * * * * * * * * * * * * * * * * * * * * * 8 * * * * * Multi-stage sampling Principle repeated sampling: can employ different sampling methods at various stages of sampling example: final sampling unit = household 1st stage : sampling districts 2nd stage : sampling wards 3rd stage : sampling villages or hamlets 4th stage : sampling households * Multi-Stage Cluster Sampling A “multistage cluster sample” is performed in two or more steps. For example, to carry out an immunization survey of school children in a given district, the following steps might be followed. Select m wards from the M mutually exclusive and exhaustive wards composing the district. Select a sample of villages or other minor civil divisions (e.g. hamlet) within each of the wards selected at the first step. For each village that was selected, select a sample of schools within the village. Select a sample of classrooms within each of the schools selected at the third stage. Take every child within the classrooms selected at the fourth stage. * Bias in Sampling: Possible Sources Non-response Studying volunteers only Sampling registered units (patients) only Missing cases of short duration Seasonal bias “Tarmac” bias * END OF