Female anxiety vs. the parental influence as a predictor of religiosity, Final 1

By Kristin McCartney and Kate Hetrick

April 30^th, 2001

Natural Systems II, Western College Program

It seems that in every place and time, in every human culture religion has played a pivotal role in human life. When speculating on the reasons for the universality of religion, sociobiologists often note the fact that humans are pattern-seeking creatures, who tell ourselves stories to explain our world (Gazzniga 1998). The fact that we are story-telling animals with conceptions of time and consequences means that we may exercise foresight, but also that we may feel deep anxiety and fear (Sapolsky 1998). Religion is one way for us to assuage that fear and come to terms with our own vulnerability. Indeed, several studies have shown that those who are religiously active (attend church regularly, pray often) have lower mortality rates and lower levels of anxiety and depression (Maltby 1998; Hummer, Rogers, and Nam 1999). More specifically, in long term studies it is shown that religiousity is highly protective in depressive disorder to women (Miller, Warren, and Wickramarante 1999; Mirola 1999). Scholars have noted that women are both religious and affected by depressive disorders at rates significantly higher than men (Mirola 1999). As mentioned above they have even noticed the therapeutic nature of religion for women, but few have asked why.

Both men and women experience anxiety because of their vulnerability, but women, physically and often socially, are more vulnerable than men. Biologically, this has to do simply with women’s different reproductive role. Until quite recently, a reproductively fertile woman was bearing or burying a child every year. A woman is not completely incapacitated when pregnant, but there is no doubt that her mobility and strength are compromised. And with every birth, there (until recently) always existed the very real chance of death. Even when not pregnant, a woman has the burden of menstruation. The woman in the ancestral environment was simply more vulnerable than the man and perhaps even more in need of the social and psychic comfort of religion. This could be extended into the lifecycle; it may be no coincidence that when we are at our most vulnerable (near the beginning and end of life) we are the most religious.

For these reasons we believe that there may be a biological basis for women’s greater religiosity; even if women and men come from very similar cultural backgrounds, we believe that women will be significantly more religious. The hypothesis of our research is that sex, not cultural background is the strongest predictor of religiosity. Furthermore, we believe that this might be because biologically women are more vulnerable and have more reason for anxiety than men. So we will also test levels of anxiety. We believe that even if the women and men are from very similar cultural backgrounds, that women will be significantly more anxious than men.

This research is important, because we don’t live in the ancestral environment. Even if men and women have evolved to have different reproductive functions, today we hope to create a more egalitarian society and in reality (for many) it is a more egalitarian society. Childbirth has lost much of its danger and women have gained control over their reproductive lives, but their psyches are still shaped by their ancestral vulnerability. If we understand how women may be inherently different, we can begin to build a society that considers sex differences rather than just viewing men as the norm. Religion and anxiety may be one of those places of essential difference; if so we need to take note. To paraphrase many evolutionary psychologists, first we understand our nature and then we understand how we can make our nature work for our agenda (Wright 1994 and Burnham and Phelan 2000).

Gender is often ignored in studies of the sociology of religion, despite the fact that women are found to be more religious than men by every measure (Walter and Davie 1998). One of the best overviews of the work done pertaining to sex and religion is Walter and Davie sociological literature review, "The religiosity of women in the modern West" published in The British Journal of Sociology. Walter and Davie trace the studies of religiosity and sex and put forward the premise that "whether or not religion is or has been a response to socially-induced vulnerability, it is and always has been a response to the physical vulnerability of the body that is the human condition" (646). They extend the argument to say that because women’s bodies are more vulnerable women should be more religious. Sadly, they also note the dearth of research in the area. Bloom (1995) and Gazzniga (1998) note religion’s connection to human anxiety, but fail to wonder if different sexes might experience religion differently because of it. Bowker’s Is God a Virus? is especially helpful, because he addresses issues of the coevolution of nature and culture and specifically women’s relationship to religion. Taken together these sources and their references are the foundation of our connections between sex, religion, and anxiety.

Fortunately, there are several studies focused on measuring religiosity including "The Religious Orientation Scale" (Trimble 1997) and "National Context, Parental Socialization, and Religious Belief: Results from Fifteen Nations" (Kelly and De Graaf 1997). Kelly and De Graaf’s piece is most useful, because it provides us with a scale for measuring parent socialization and religiosity based on concrete measures such as church attendance: " church attendance, a clear cut behavior, is reliably reported and is generally the key family influence, with strong direct and indirect effects on respondent’s [religiosity]" (644). Bowker’s Is God a Virus? also outlines and critiques several important tools for measuring religiosity.

We have also looked at studies that address the question of sex, religion , and anxiety or depressive disorders. Maltby’s "Church Attendance and Anxiety Change" (1998) suggests that church attendance lowers anxiety levels. Hummer, Rogers, and Nam’s "Religious Involvement and U.S. Adult Mortality" (1999) found that people who attend church weekly to live longer than those who do not attend church at all. Miller, Warner, and Wickramarante’s "Religiosity as a Protective Factor in Depressive Disorder" (1999) noted that, in women, religiosity had a protective affect against depressive illness. Finally, Mirola’s "A Refuge for Some: Gender Differences in the Relationship between Religious Involvement and Depression" suggested that "religious involvement measures tend to have a negative impact on depression for women" (419). The fact that church attendance and involvement is vital to all of these studies of religious benefit and Kelly and De Graff’s article lead us to make church attendance and involvement as our indicators of religiosity.

Finally, there is one more road of investigation and that is into Miami’s demographics. After contacting the college’s religious organizations we found that the ratio of men to women in their membership rolls is very unbalanced; women’s numbers outstrip the men’s in every coed organization (although there are no exact records). This information helps us frame our assumptions, discussion, and conclusions for further research.

Our question (concerning gender, religiosity, and anxiety) is particularly important, because of the patriarchal nature of so many religions. If women have a greater need for the solace of religion, they should not only have patriarchal religions to turn to. As a society we should try to find ways to fulfill women’s unique psychological needs in a society that values women.

For our data, we had 250 students fill out a brief survey that asked them about

their sex, urban/suburban/rural background, religiosity of self and parents (frequency of church attendance), and their level of anxiety. Again and again, studies have found that women are more religious than men, we would go further and say that sex is the strongest predictor of religiosity. What are the possible predictors of religiosity? If we use a sample of people (Miami students) of approximately the same age, same ethnicity, same educational background, and same nationality we can begin to judge the importance of three other variables in determining religiosity: sex, rural or urban background, parental socialization, and anxiety. (Indeed, all the students we surveyed were white, between 18-22 years-old, and of the same educational background (some undergraduate education).) Economic background may also be a predictor, but measuring that is beyond the scope of our research. National background is also a predictor, but again a moot point in this survey of American students. Of course, a woman is not only physically more vulnerable than men, but often socially. By choosing a sample group in which the women and men enjoy close to equal social status (though even this is arguable), we hoped to zoom in on anxiety as a product of biological vulnerability, not social vulnerability.

We have included a copy of our survey in the Appendix, though we recoded our results. Gender and background are self-explanatory. Our method for measuring personal religiousity and parental religiousity was based upon the technique used in "National Context, Parental Socialization, and Religious Belief." In addition, we asked about the number of religious organizations one was in to have an addition measure of religiousity. While some people may attend religious services out of habit, fewer will be actively involved in religious organizations with the same motivation. The two questions measuring anxiety are based on two of the experiences that would have had the biggest effect on women’s vulnerability in our ancestral environment (and that continue to be significant today). These questions measured a person’s anxiety about sexual assault and unwanted pregnancy.

We recoded the scores as following: For religiousity, people who went to religious services several times a week up to more than once a week were "High," people who attended about once a month were "Medium," and those who attended once a year or never were "Low." A person was given either a "Yes" or "No" for belonging to religious organizations. Those who circled one for anxiety about sexual assault were "Low," those who circled two were "Medium," those who circled three to five were "High." Finally, those who circled one for anxiety about unwanted pregnancy were "Low," those who circled two or three were "Medium," those who circled four or five were "High." We also had the category for "Not applicable" for those who chose not to answer the question or were not sexually active.

We gathered our samples by giving out the surveys in public spaces (like university food courts) and by taking the surveys to the meetings of local student religious organizations. In addition, we also used our NS class as a focus group after presenting on the topic of religion and gender and made adjustments to our survey based on student recommendations. From this session, we also got the sense that students agreed with our hypothesis, but like us they thought that that women’s higher religiousity and anxiety could have a cultural basis.

To analyze our surveys, we entered our statistics into Statview and ran several analyses using Chi square. We found that we could not actually test our hypothesis given our methods, but we could measure the correlations between gender, background, religiousity, and anxiety. We first compared gender and religousity, and our result was a bit too high to be statistically significant (p-value .0771). Contrary to our original hypothesis, gender did not seem to be a consistent indicator of church attendance. Because our sample consisted solely of Miami University students, the low correlation between gender and religousity might be partially attributable to the college age bracket. College students overall seem less likely to attend church on a weekly basis, perhaps due to the desire to sleep in on Sunday mornings. Instead of a weekly church activity, they may choose to be involved in some sort of extraneous social organizations.

It is fitting, then, that our results indicated that females were more likely to be involved with a social religious organization (defined on our survey as a religious fraternities-sororities, service organizations, study/support groups, fellowship groups) with a p-value of .0007. When we surveyed specific campus religious groups, both Intervarsity and Catholic Campus Ministry had more females in attendance. In accordance with our introduction and theories of evolutionary psychology, females may feel more vulnerable and anxious and desire the comforting companionship and fellowship a social religious organization provides.

Not surprisingly, females reported a much higher fear of assault (p-value .0001). Theories of evolutionary psychology substantiate this result—females have a higher degree of physical vulnerability than males. Even though it can be argued that females simply have different areas of physical strength, the pregnant woman has less of a chance of successfully escaping or defending an attack than a burly male.

Your background (urban, rural, and suburban) did not have a strong correlation to measures of religousity (p-value .2468). An overwhelming majority of those we surveyed came from a suburban background. It is not especially shocking that background did not contribute highly to religousity, for churches and various religious organizations can be found all over the world, regardless of urban, suburban, or rural locations. We originally expected that those from a rural location might report a higher level of religousity, but our final statistics indicated virtually no relationship to background and religousity.

In contrast, parental religousity and personal religousity were statistically significant at a p-value of .0001. Though we cannot say that parental religousity is a stronger indicator than any of our other measurements (gender, background) because they all influence one another; it yielded the highest correlation. Due to routine, ritual, comfort level, and mere familiarity, those raised in a religious environment are probably more likely to continue attending a religious institution than those raised with little to no religious exposure.

Analysis indicated that anxiety or fear of assault was not a strong indicator of religiousity, though the p-value of .0088 does indicate at least a slight correlation. Women overwhelmingly reported a higher fear of assault, and thus it stands to reason that they may turn to the support of religion and religious social organizations to assuage those anxieties and feelings of vulnerability.

Again, not surprisingly, our results indicated that females had a much higher fear of unwanted pregnancy than males (p-value .0001). Though several males also reported fearing unwanted pregnancy, ultimately, the woman has to carry the baby for nine months or make the decision to have an abortion. The fear of unwanted pregnancy more directly affects the female than it does the male. Evolutionarily, the female body goes through a monthly cycle that reminds the female of the possibility of pregnancy. A woman cannot escape the reality of her own capabilities, while the male connection is not quite as intimate or traumatizing.

Those who were from a rural background were more likely to be involved in a religious organization than those from a suburban or urban background (p-value .0003). Perhaps this result stems from fewer activities for involvement in rural areas, and thus rural and suburbanites are more likely to turn to a religious activity.

When we analyzed a possible relation to background and fear of assault, we did not find a statistically significant relationship (. p-value .9131). Fear of assault seemed to largely depend on gender, again substantiating Wright’s theory of evolutionary psychology.

Our data indicated that those with a lower fear of pregnancy were more likely to report a high level of religiousity (.0001). This results seems initially rather shocking, but closer examination reveals that those who reported a "high" level of religousity were much less likely to participate in sexual activity. Participation in sexual activity greatly increases the fear of unwanted pregnancy—those not having sex (likely for a religious reason) have no reason to fear an unwanted pregnancy.

Finally, a comparison of background and personal religousity did not indicate statistically significant results (p-value .9311). When background is compared to our results regarding parental religousity and gender, it does not yield a strong correlation. Both gender and parental religousity seem to be a much more consistent determiner of personal religousity. The full analyses including observed and expected frequency charts are available in the Appendix.

As two young college students doing a very limited survey, we cannot conclusively draw any ultimate conclusions. That said, our data indicated a few broad trends in measures of religiosity, gender, and anxiety. While women were more likely than men to be highly religious, gender was not a consistently accurate predictor of religiosity. In comparison to our results on gender, parental religiosity proved to be a better predictor of personal religious activity. Females were overwhelmingly more likely to be involved a social religious organization. We were not surprised by the results that females reported a much higher fear of assault and unwanted pregnancy. Finally, background did not seem to be a strong indicator of religiosity, though those from a rural background were more likely to express involvement in a social organization.

After some consideration of evolutionary psychology throughout the WCP 222 class, we believe there may be an evolutionary link to female religiosity. Females in our 250 surveys did show a higher degree of anxiety and vulnerability, and may turn to religion in an attempt to sooth those fears. Perhaps women also just enjoy the camaraderie of other religious women, or feel that in a patriarchal society, religion is an area where they can take an active leadership role. Ultimately, though, we feel that our survey indicates that religious activity is largely cultural. Parental religiosity directly related to personal religiosity. We cannot conclusively say that this eliminates all evolutionary speculations. However, it does indicate that the family environment plays a very strong role in religious activity. We also felt that the fact that women were more likely to be involved in a religious activity might be a cultural factor. Stereotypically, women are viewed as more emotional and may turn to religious fellowship as a social support network. Finally, the association between rural background and religious activity seems to be explicable by culture. In rural areas, there is simply not as much of an opportunity for involvement in a wide variety of activities. In addition, as we both come from rural backgrounds, we noted that there is a high societal pressure to be a church-going family. Those who did not attend church in our home communities were definitely "suspect," and the smaller community and the "everybody and their monkey knows everyone" phenomenon may cause those from a rural area to attend church due to cultural pressure.

We agreed that the biggest strength of our study was the relatively homogeneous sample population. All 250 survey-takers were white, between the ages of 18 and 22, and students at Miami University. Demographic information for Miami students indicates that most come from a fairly similar economic and educational background. We eliminated some possibility for severely skewed statistics by sampling from a homogenous group. Though we would have liked to have surveyed more individuals, we also felt that 250 would be a stronger number to display trends, than, say 100. We modeled our own survey on previous models, thus using methods that had been tried and tested before. We attempted to also get a random sample by surveying a wide variety of campus groups and "random" individuals. Our survey was brief and easy to fill out, and people seemed willing to participate in our study.

In retrospect, we would have eliminated the "sexual" aspect of the question on assault and changed it to, "What is the likelihood that you will fall victim to assault in the next five years?" We felt that by using the phrase "sexual assault," women were automatically more likely to report a higher level of anxiety. Simply using "assault" might have yielded a higher population of males with reported fears. Another problematic question was the question on unwanted pregnancy. We worried that people who reported attending church regularly may not feel comfortable admitting to sexual activity. When handing out our surveys, we made attempts to assure anonymity by asking recipients to stick their completed survey in the middle of the stack and by giving the survey-taker some "space" to fill out the survey. Despite our precautions, we believe the data on this question may be somewhat skewed.

Though we worked well with our survey questions in persuading individuals to fill out our survey easily and quickly, we feel that a more comprehensive survey would have led us to more conclusive adults. The homogeneous population was both a strength and weakness of our study. It might have been interesting to strive to find more non-religious individuals or those from a wide variety of backgrounds. We also may have been slightly over ambitious from the very beginning. We set out to prove something that virtually no researchers had ever proven. We had very little to base our own work on, and had trouble establishing a strong evolutionary connection. Finally, we want to again reiterate the very limited scope of our study. While we can draw some very basic, broad trends from our data, we have no ultimate proof of determinations of religiosity.

For future study, we would first suggest a much more comprehensive survey. While the brief nature of our survey helped us collect a large sample, it cannot be denied that more questions on factors such as anxiety would have allowed us to report a more definitive analysis. We would also suggest broadening the sample and background and surveying more individuals. Again, this would simply allow for a stronger and more accurate sample for analysis. Future study might also allow for some sure method of anonymous results, thus allowing survey-takers to feel secure that we were not intimately familiar with their sexual activities. Ultimately, we were not at all surprised by our findings, and felt that they were relatively consistent with our original expectations. Nevertheless, because of our inexperience, lack of resources, and past studies to base our own surveys on, we feel that our results can only be very broadly generalized, and our possible reasons for religiosity are merely our best attempts at scientific speculation.

Bowker, J.1995 Is God a Virus?, London : SPCK.

Burnham, T. and Phelan, J. 2000 Mean Genes: From Sex to Money to Food: Taming Our Primal Instincts, New York: Perseus.

Gazzniga, M.S. 1998 The Mind’s Past, Berkeley: U of California P.

Hummer R.A., Rogers, R.G., and Nam, C. B. 1999 "Religious Involvement and U.S. Adult Mortality," Demography 36(2): 273-85.

Kelly, J. and De Graaf, N.D. 1997 "National Context, Parental Socialization, and Religious Belief: Results from Fifteen Nations," American Sociological Review 62: 639-59.

Maltby, J. 1998 "Church Attendance and Anxiety Change," The Journal of Social Psychology 138 (4): 537-538.

Miller, L., Warner,V., and Wickramarante, P. 1999 "Religiosity as a Protective Factor in Depressive Disorder," The American Journal of Psychiatry 156(5): 808-809.

Mirola, W. A. 1999 "A Refuge for Some: Gender Differences in the Relationship between Religious Involvement and Depression," Sociology of Religion 60(4): 419-37.

Sapolsky, R.M. 1998 Why Zebras Don’t Get Ulcers, New York: W.H. Freeman and Co.

Trimble, D.E. 1997 "The Religious Orientation Scale," Educational and Psychological Measurement 57: 970-986.

Walter, T. and Davie G. 1998 "The Religiosity of Women in the Modern West," The British Journal of Sociology 49 (4): 640-660.

Wright, R. 1994 The Moral Animal, New York: Vintage Books.

Male____ Female____

Background: Urban ____ Suburban____ Rural ____

Parents| You

__|__ Never attend church, or have no religion

__|__ Attend about once a year

__|__ Attend several times a year up to once a month

__|__ Attend several times a month up to almost every week

__|__ Attend every week or more often.

____ Number of religious organizations you are a member of (Examples are religious

fraternities/sororities, service organizations, study/support groups, fellowship groups.)

What is the likelihood that you will fall victim to assault in the next five years? (Circle one #)

1_____ 2______3_____4______5

None Likely Certain

(If sexually active) How much do you worry about unwanted pregnancy?

1_____ 2______3______4______5

Never Sometimes All the time