# Exploratory and Confirmatory Factor Analyses of Religiosity. A Four-Factor Conceptual Model

## TL;DR

This study identifies four core factors of individual religiosity through factor analysis of survey data and confirms their structure, providing insights into their interrelations and causality, with implications for understanding religiosity.

## Contribution

It introduces a validated four-factor model of religiosity and explores causal links between these factors using structural equation modeling.

## Key findings

- Four reliable religiosity factors: Religious formation, Supernatural beliefs, Belief in God, Religious practice.
- Confirmed the four-factor structure with excellent fit measures across data waves.
- Identified causal pathways, such as Religious practice influencing Belief in God.

## Abstract

We describe an exploratory and confirmatory factor analysis of the International Social Survey Programme Religion Cumulation (1991-1998-2008) data set, to identify the factors of individual religiosity and their interrelations in quantitative terms. The exploratory factor analysis was performed using data from the first two waves (1991 and 1998), and led to the identification of four strongly correlated and reliable factors which we labeled Religious formation, Supernatural beliefs, Belief in God, and Religious practice. The confirmatory factor analysis was run using data from 2008, and led to the confirmation of this four-factor structure with very good fit measures. We also ran a set of structural equation models in an attempt to determine the causality links between these four factors. It was found that for the models which provide the best fit Belief in God does not cause Supernatural beliefs, Religious practice can cause Belief in God and that there are multiple paths leading to Belief in God, most of which include Religious formation as a source. The exploratory factor analysis also led to the identification of other factors related to traditional values, confidence in institutions and influence of religious leaders on politics, but these were found to have lower reliability or insufficient number of items to meet the acceptance criteria, and thus were not included in the confirmatory factor analysis and the investigation of causal links. The results obtained in this work have important material implications for the conceptualization of "religiosity," and important methodological implications for the scientific study of religion.