Looking for a Universal Measure of Religiosity: A Latent Variable Approach
A report by Kirill Zhirkov at the regular LCSR seminar
On December 5, 2013 Kirill Zhirkov (LCSR, research fellow) delivered a report on “Looking for a Universal Measure of Religiosity: A Latent Variable Approach” at the regular LCSR seminar.
The aim of current project of Kirill Zhirkov is to analyze relation between religiosity and various social characteristics on the one hand and religiosity and wide range of value attitudes on the other. His analysis is based on revised modernization theory (in particular, “Sacred and Secular” by R. Inglehart and P. Norris) and on previous empirical research on this topic. In former literature negative relationship between existential security and religiosity was established as well as persistence of cultural zones based on religious heritage and negative influence of religiosity on gender equality and liberal democracy.
Basing on these prerequisites, Kirill created a theoretical model which connects cultural characteristics of different countries, level of democratization and personal religiosity and gender attitudes of an individual. According to Kirill's hypothesis human development index and belonging to certain cultural zone (Inglehart and Norris typology) affect levels of religiosity level and level of democratization. There is evidence that support of gender equality on individual level is determined by age, religious denomination, level of religiosity and educational level, whereas level of religiosity is affected by age, educational level and religious denomination.
Current stage of Kirill's project is related to methodological specificity of measuring religiosity. This analysis is based on empirical evidence from 80 countries, and it is reasonable to doubt that all the indicators are perceived in the same way in different countries, or, speaking formally, whether the measurement invariance can be approved. Previous research clearly showed that if we use data from more than 25-30 countries for a test on such an issue, multi-group structural equation modeling will not be fruitful. Firstly, it will be difficult to find a proper interpretation for the results. That's why Kirill used religious zone as grouping variable. There are 5 groups: Catholic, Protestant, Orthodox, Muslim and Eastern religions. Such reduction can be criticized for oversimplification, but Kirill additionally checked measurement invariance for those groups which seem the most controversial (e.g. Muslim males and females, countries from different domains classified as Eastern).
Several variables were used as the indicators of individual religiosity, such as importance of God (consists of 10 categories; highly skewed), importance of religion (4 categories), religious attendance (7 categories; skewed) and self-assessed religiosity (3 categories). These variables are traditionally used in quantitative analysis as conditionally qualitative, however, in order to test measurement invariance Kirill had chosen a more conservative way and took these variables as categorical. An aggregated dataset of 4 and 5 waves of Word Value Survey were used as an empirical basis.
The results of correspondent tests revealed that there is no significant difference in perception of religiosity by representatives of different cultural zones. There are several problems with some indicators, however. Consequently the question about the possibility of using religiosity measure as a universal one is not answered yet. Kirill had also emphasized that cultural zone classification described in “Sacred and Secular” needs to be refined. The main concern is related to unification of Eastern countries disregarding their cultural differences.
by Olesya Volchenko