Sample size

SAMPLE SIZE

  • Sample size depends on,
    • Non-statistical consideration
    • Statistical consideration

Non-statistical consideration

  • Availability of manpower and sampling frame

Statistical consideration

  • Precision of the estimate of prevalence
    • The ability of an estimator to determine the true population value of a variable. i.e. the estimator's precision.
    • It is expressed in terms of the bound on the error of estimation that can be tolerated.
    • The error is defined either 'absolute' or 'relative'.
    • Example
      •  Absolute acceptable error: If it is 2% means, a prevalence of 40% represents 38-42%.
      • Relative acceptable error: If it is 2% means, it represents 2% of 40% i.e. range of 39.2-40.8% (2% of 40% = 0.8%).
  • Expected frequency of the disease
    • It is not readily available. However, assumption can be made based upon the previous history, incidence and other associated epidemiological parameters
Last modified: Monday, 23 May 2011, 1:27 PM