Smertin Pavel Romanovich

Introduction. The goal of the article is to compare different algorithms for calculating the frequencies of dental nonmetric traits.

Materials and methods. The experiment was conducted on the materials of a modern group of Tuvinians (Eastern Siberia), the sample size 135 individuals. The frequencies of 11 traits have been calculated using several methods common for paleophenetic studies. The obtained values were tested using a 2-tailed exact Fisher criterion.

Results. It has been established that the frequencies of traits obtained by different methods are in good agreement with each other. The exception is the increased values obtained with one of the calculation options «per individual»: namely, when the calculation includes cases of the presence of only one tooth, the anti-mere, on which the feature is only positive. In addition, two sample size 25 and 10 observations were randomly generated from individuals who possess the trait on both sides. The frequencies in the sample size 25 individuals turned out to be stable, and the data are in good agreement with the full series, regardless of the method of calculation. A sample of 10 individuals does not agree with the characteristics of the full series.

Discussion. It is most preferable to calculate the frequencies «per individual» when observation is possible from both sides, but this is problematic with a small group of fragmentary materials. For small series, an individual calculation is applicable, when the feature can be observed on both or on one of the sides (all cases are taken into account, regardless of the presence of the trait). The results obtained are valid for a specific series and traits. This work begins a series of articles on feature calculating methods in dental anthropology. © 2025. This work is licensed under a CC BY 4.0 license