Methodological aspects of the selection of the unit of statistical analysis in kinanthropology research

To decide which unit should be chosen for a statistical analysis has been a difficulty in the methodology of kinanthropology research. The newly created construct "a unit of statistical analysis" is understood as a determinant (the layout of samples and sub-samples) of the research sample on which level the data are statistically processed. A scheme created by Silverman and Solmon (1998) is presented as an aid tool for clearer understanding of the methodological choice. An example is to illustrate how the scheme of the choice of a unit for statistical analysis can be applied in research praxis. A comparison of two data analyses which used different units of statistical analysis - each student or a class of students- has shown that an appropriate choice of the unit is essential for correct data evaluation and interpretation. Aspects that should be considered along with the level of statistical significance are effect size and practical significance. In such a context, qualitative methodological approaches are to support the quantitative approaches and should be enhanced. We perceive the continuity of quantitative and qualitative approaches to be the great research prospects in kinanthropology research.

Keywords: research sample, methodology approach, logical significance, intervention