Salnikova S. The Validity of Religiosity Indices Measurement in the Quantitative Sociological Studies

In the given research the result of methodological principles and methodical peculiarities of religiosity measurement in the quantitative sociological studies suggests the development of a certain strategy of religiosity indices measurement and their validity. We have taken into consideration the main concepts of religiosity measurement and determined that in different concepts the uniting is the position concerning the dimension of the notion under research that has obtained the empirical evidence. Taking into account the diversity of approaches and a number of debatable issues we have established which approach - generalizing or detailing (or specific) - should be used in the religiosity studies. In the focus of the required approach we consider the problem of the concept choice, the formation of the samples and the agreement of different techniques and analyses. We also sum up that the problem of approach choice is explained by the aims and level of the analysis and is solved only in terms of wider issues which we use in our research. In the foreshortened view of the research strategy of the given thesis is the specific approach to the religion studies which envisage the formation of religiosity model in multidimensional with further integration of the indicators into additive indices - the indices of religiosity. In our work we have theoretically and empirically interpreted the notion 'religiosity', and singled out its main components; we have also established that between the structural scheme of the religiosity theoretical model and the structural scheme of the factor analysis there is a certain analogy that allows to use the given analysis as (1) a means of theoretical model verification and (2) as an instrument of its creation. The theoretical model is characterized by a two-level scheme of relation between the notion 'religiosity' and its components, that does not contradict multidimensional the notion under research, represented in various conceptual approaches. The specification of the given approach lies in the triple use of the factor analysis techniques; the one-factor solution is realised twice: first up to the system of indicators that represent the constituent part of the component, for the purpose of detecting those indicators that form one coordinated dimension of a certain component, then - to the constituent parts of components, made up in the form of additive indices for the purpose of corroboration that they also represent one dimension, one notion, that is - religiosity, for the third time the many-factor solution towards indicators is being realised. They have been detected at the stage of the first factor analysis realization for the purpose of establishing relations between them and every constituent part of religiosity. The validity of the measurement model is provided by testing the validity in three directions, that is, content, criterion-oriented and construct validities have been under control. We have checked the reliability of the data twice using the retesting technique. It has been proved that multi-level model of religiosity measurement is reliable and valid; the advantage of its use in the quantitative sociological studies lies in the possibility of making the additive indices that show the level of religiosity proved with the help of factor, regression and reliability analyses. Key words: measurement, concept, indices, validity of measurement, reliability of measurement, religiosity, validity of the measurement model.