Variability in Dental Caries Experience and its Implication Upon Sample Size

Abstract

Accurate epidemiological work in the caries field requires not only common-sense evaluation of the accuracy of dental examinations but statistical appraisal of variations which cannot be eliminated. Examining errors are discussed and illustrated. The use of a value for standard deviation is demonstrated in testing the significance for a difference between mean values for dental caries experience (DMF teeth or surfaces). Several examples of standard deviation are given, where even in relatively homogeneous samples the standard deviations are between 4 and 5 DMF teeth per person once the majority of the permanent teeth have been present in the mouth for a few yrs. Where deviations of this sort exist, large samples are needed in order to reduce the probability of chance occurrence of wide differences between means. A formula for minimum sample size is derived and illustrated, for use where certain constant values can be roughly predicted. A method for testing the significance of a difference between proportions is also illustrated. Data in terms of proportion are shown to be less descriptive in certain ways than data in terms of mean values per person..