Historically, the sociology of religion has dealt mostly with macro sociological problems. Perhaps the most important of these is whether religion has been able to continue its existence and influence as a social institution since modernization. These debates have been shaped around secularization theories and studied extensively up to the 1980s. Later on, as sociologists in general began working on micro-level problems, sociologists of religion started working on micro-issues such as how religion affects individuals’ lives. The proliferation of research on micro problems has made ethnographic field research and quantitative methods more visible.

Using quantitative techniques, scientists have tried to answer questions regarding the differences between men and women in their religiosity (Cornwall, 1989; de Vaus & McAllister, 1987), the differences in religiosity between younger and older adults (Young & Dowling, 1987), and the effects of education on religion (Albrecht & Heaton, 1984; Hunsberger, 1978). Similarly, studies examining the effects of religious groups and socialization experiences on individuals’ religiosity can be mentioned within this framework.

Understanding religiosity’s interactions with individuals’ attitudes and behaviors emerged as the next step in quantitative studies on religiosity. In this context, sociologists have studied the effects of religiosity on choosing friends, work, and where to live, as well as political preferences (Malka, Lelkes, Srivastava, Cohen, & Miller, 2012), voting behaviors (Gerber, Gruber, & Hungerman, 2016; Miller & Wattenburg, 1984; Smith & Walker, 2013), consumption habits (Wilkes, Burnett, & Howell, 1986), participation levels in volunteer activities (Park & Smith, 2000; Taniguchi & Thomas, 2011; Wuthnow & Hodkinson, 1990; Yeung, 2004), civic engagement levels (Smidth, 1999; Wuthnow, 1999), life satisfaction (Willits & Crider, 1988), and marital satisfaction (Brown, Orbuch, & Bauermeister, 2008; Dudley & Kosinski, 1990; Gaunt, 2006; Hunt & King, 1974; Shehan, Wilbur, & Lee, 1990).

Because sociology of religion focuses more on theoretical discussions in Turkey, the number of religious studies conducted using quantitative methods is very low. Moreover, the findings from these few studies have been shared only as descriptive statistics, with no competent studies based on advanced analyses being produced.

International studies such as the World Values Survey and the International Social Survey Program (ISSP), which enable comparisons around the world, have also included questions about religiosity. However, given that their primary focus is to make international comparisons, their sample sizes are too small to make national analyses. No study is found in Turkey to have examined the effects of socio-economic and socio-demographic variables on religiosity using multivariate techniques. 

This article attempts to fill this gap in the literature. The effects of demographic variables such as age, gender, and marital status and social class variables such as education level, rural or city origin, family religiosity, and sect on religiosity in Turkey have been examined using multivariate analysis.

Religiosity is a multidimensional concept encompassing several unique but interrelated dimensions such as religious beliefs, religious practices, and saliency of religion. This article uses three different measures of religiosity: self-reported religiousness, frequency of performing the five daily prayers, and attitudes towards Islamic inheritance where male descendants inherit twice as much as female descendants.

Method

This article uses the dataset from the Survey of Religious Life in Turkey (SRLT, 2013), which was prepared by the Directorate of Religious Affairs (DRA) with the support of the Turkish Statistical Institute (TURKSTAT). The purpose of SRLT is to determine the religious tendencies of citizens and their reflections on daily life. The sample consists of 21,490 participants and is representative of the entire Turkish population over 18 years of age. This study is based on the multiple regression analysis, as it aims to reveal the effects of many variables simultaneously and separately. Logistic regression analysis is preferred more than other methods for explaining the relationship between variables in cases where the dependent variable is categorical. Because religiosity is an ordinal dependent variable, this research needs to use ordinal logistic regression analysis. The three models have the same independent variables.

Independent Variables

Self-rated religiosity has been measured by asking the participants the question, “Which of the following statements reflects your sense of religiosity?” The answers given to this question are (1) “I am quite religious,” (2) “I am religious,” (3) “I am neither religious nor non-religious,” (4) “I am not religious,” and (5) “I am not at all religious.” In order to convert this to a positive sequence, the values for this variable have been reverse coded. The second dependent variable, the frequency of performing the five daily prayers, is measured through the question, “How often do you perform the five daily prayers?” The answers are (1) “Always,” (2) “Most of the time,” (3) “Occasionally,” (4) “Rarely,” and (5) “Never.” This variable has also been reverse coded. The third dependent variable is measured by the question “Indicate your views on the following statements: When distributing inheritance, I give men a double share.” The answer categories are (1) “I agree,” (2) “I partially agree,” and (3) “I disagree.” This item has also been reverse coded so that higher values mean higher religiosity.

Independent Variables

This research uses the following variables as the independent variables: (1) gender, (2) education, (3) age, (4) employment status, (5) income level, (6) marital status, (7) settlement (rural-urban), (8) family religiousness, and (9) sect.

Findings

Gender

Gender appears to have a statistically significant effect on religiosity levels for all three models. Even when controlling the effects of other independent variables, women have been observed to be more religious than men in the first two models. Women define themselves as more religious than men. Similarly, women pray more often than men. Unlike the first two models, however, women are less religious than men when taking attitudes towards inheritance distribution into consideration.

Age

Age has a positive effect on religiousness. In the first two models, religiosity gradually increases with age. However, the effect of age on attitudes towards inheritance sharing is slightly more limited.

Settlement

According to the first two models, people living in the city are less religious than those living in rural areas. Indeed, those living in the city feel less religious and pray less frequently than those living in rural areas. However, the third model shows that those who live in rural areas have a more negative attitude toward Islamic inheritance distribution than those living in the city.

Employment Status

In all three models, those who work actively in a job are less religious than those who do not work.

Marital Status

All three models show that married people are more religious than unmarried people are. 

Education

According to the first model, the highest level of religiosity is observed among secondary school graduates, followed by post secondary vocational college and primary school graduates, respectively. Those who have no formal education pray the most frequently, whereas university graduates pray more frequently than high school graduates. Those with no formal education have a more positive attitude towards Islamic inheritance distribution than university graduates.

Family Religiosity

In all three models, family religiosity appears to have a positive effect on religiosity. As family religiosity increases, participants are more likely to feel religious, pray more frequently, and have a more positive attitude towards Islamic inheritance distribution.

Sect

According to the regression results, sect has a significant effect on religiousness levels. In all three indicators of religiousness, Hanafi and Shafi’i individuals are more likely to be religious than those from other sects. Shafi’i individuals are the most religious with respect to all three indicators.

Conclusion

The findings from the three models show level of religiosity in Turkey to not change with respect to income groups. One important finding from this study is that age has significant implications for religious self-assessment and frequency of prayer. Similarly, many studies conducted in Turkey and in the world have shown a positive relationship to exist between age and religiosity (Argue, Johnson, & White, 1999; Taplamacıoğlu, 1962). Consistent with the literature, married people are found to be more religious than unmarried people. Furthermore, family religiosity has a positive effect on religiosity. Another prominent finding from this study is that the factors affecting the attitudes towards heritage shares in Islamic law differ from those that affect the level of self-evaluated religiousness and frequency of prayer. Another result of the research is the significant effect sect has on religiousness. Hanafi and Shafi’i individuals have been found to be more religious than other sects for all three measures.