On May 5th, modern-day debates hosted a debate between myself and SJ Thomason. During that debate, SJ presented an article she wrote, as well as numerous studies, in defense of her position. I challenged a fair number of the studies, and the article itself. I was asked by multiple people to support my claims, both during the debate and in subsequent followups. I agreed to offer my notes as support, and present them here.
There are a great many notes, and they can be divided into three groups:
- Failure to account for covariants or the intent of the authors: These are studies where, while they found a correlation between religion and health, found that it was not the religion, per se, that accounted for it, and counseled against those conclusions, or found the correlation in a group far narrower than SJ portrays
- Misrepresentation: These are studies which were misrepresented, such as studies of extrinsic motivation either being suggested or outright explained as intrinsic ones.
- Contradiction: These are studies where the conclusions were counter to the intent or the article, and in the worst cases, explicitly contradicting the claims she was using them to support.
A breakdown of these notes by category will be made available. In the meantime they are presented here by the order in which they are used in the original article.
Given the number of notes, I’ll direct you to too of the larger issues, Yeager et al., (2006) and Greenfield & Marks (2007).
My raw notes as used in the debate can be found here.
My notes to the articles I referenced during the debate, their conclusions, and how those account for these older findings, will also be posted soon.
The complaints here
“Atheism has it’s drawbacks”
This is a claim suggesting causality, that being an atheist will have set repercussions. However it is not demonstrated it is atheism that has these drawbacks, as opposed to third causal factors (such as stigmatization of being in a minority group).
“Dervic and colleagues (2004) compared depressed inpatients, finding that the religiously unaffiliated had significantly more lifetime suicide attempts and more first degree relatives who had committed suicide than subjects who endorsed a religious affiliation.”
While true, Dervic and colleagues (2004) was later directly addressed and clarified by Alee Robins and Amy Fisk(2015) who found that “Suicidal ideation was significantly related to public religiousness but not to private religiousness, total religiousness or religious beliefs. Suicidal ideation was significantly related to both perceived social support and general social support”. So while technically her claim is true, the suggestion, as made here, that it’s the religiosity that is the factor, seems unsupportable.
Christians, Jews, and Muslims in studies of youths and adults who scored higher on measures of religiosity were more likely to see their dentists, take their vitamins, and wear their seatbelts and less likely to drink and smoke than their less religious counterparts (Wallace & Forman, 1998; Hill, Burdette, Ellison & Musick, 2006; Islam & Johnson, 2003; Shmueli & Tamir, 2007).
First, this is a gross misrepresentation of Wallice and Forman (1998), and significant cherypicking. Once again this study is being used to tout the benefits of religiosity, but they claim in their conclusions that it wasn’t the level of religiosity or intrinsic religion, but rather “regular religious attendance”, in other words social interaction (much more on that later), that helped. Further, they expressed serious concerns over intrinsic religious motivations: “nor is it clear that these other (intrinsic) aspects of religion may not negatively affect adolescent health… [Conflicting religious orientation] may serve as a risk factor for a variety of negative health outcomes…[Further] teachings of some denominations may expect, if not require, gardens not to pursue medical treatment, maintain prescribed health regimens, or engage in practices that can, at least potentially, protect health (eg Condom use)”. SJ ignores these concerns, which explicitly and dramatically undermine the promise she is positing both in this paragraph and in the paper.
She is also ignoring other Shmueli (2007) findings such as “religiosity had a consistent adverse effect on reported health among Israeli urban Jews aged 45 – 75 years”
In a meta-analysis of 147 independent studies, findings indicated religiosity is associated positively with psychological well-being (Smith, McCullough & Poll, 2003). Findings further indicated positive religious coping, positive God concepts, and intrinsic religious motivation were negatively associated with depressive symptoms. In other words, people who fall back on their religion to cope with events and who sincerely believe that religion is an active, directive force in their lives showed fewer symptoms of depression.
First, it’s worth noting that, as with the vast majority of these studies, the findings were correlated more to the social benefits which was being provided by church attendance, and not to the religion itself. Indeed, In a 2009 followup, McCullough, Michael, Friedma, Howard, Enders K.C. & Martain (2009) found that mortality correlations with religion were “largely attributable to cross-sectional and prospective between class differences in personality traits, social ties, health behaviors, and mental and physical health” and not religiosity.
But as to her direct claims in this study, this is a gross misrepresentation. First, they found was only “mildly associated”. Second, and this is critical, they found that it worked only with extrinsic religious orientation, and that intrinsic orientation was actually “associated with higher levels of depressive symptoms”.
This is critical, because not only does this re-enforce the pattern of studies that suggest it’s the social interaction and not the religion that’s the factor, but her claim here is explicitly to intrinsic religious. She claims that they found that “God concepts, and intrinsic religious motivation were negatively associated with depressive symptoms”. But they explicitly found the opposite, saying that intrinsic orientation “associated with higher levels of depressive symptoms”.
I can find no way to avoid the conclusion that, just as in the Yeagar (2006) study, SJ is falsely representing the study, reporting the exact opposite of what the study found, both falsely claiming their extrinsic findings as intrinsic ones, and ignoring the intrinsic findings which contradict her, altogether.
Other benefits accrue for the religious. Numerous studies have found that frequent religious service is related to a 25 – 30% reduction :.in mortality in a variety of countries (Powell, Shahabi & Thoresen, 2003; Musick, House & Williams, 2004; T.D. Hill, Angel, Ellison & Angel, 2005; Teinonen, Vahlberg, Isoaho & Kivela, 2005; Yeager et al., 2006; la Cour, Avlund, & Schultz-Larsen, 2006).
While indeed Powell, Shahabi & Thoresen, (2003) did find “a strong, consistent, perspective, and often graded reduction in risk of morality,” in church attenders, SJ fails to point out that it was reduced significantly after adjusting for confounders. Further, they found that “evidence fails to support a link between depth of religiousness and physical health”, and even evidence that “religion or spirituality impedes the recovery from acute illness”, which seems to ignore, and seems to undermine the argument. Further, once again, the findings were in church attendance, once again suggesting that it’s the social interaction, and not the religion, giving the benefit.
Further, Musick, House & Williams, (2004) like the other studies, found that this only applies to extrinsic “service attendance”. Like nearly all the other studies,they actually this found that intrinsic religious behaviors actually made outcomes worse, and at a minimum “often tend to suppress the association between service attendance and mortality”. Once again, these findings actually seem to undermine her thesis on the benefits of religion.
Not only did Vahlberg, Isoaho & Kivela (2005) limit the benefits they found only to women, but again, said that “health benefits of religious attendance have been explained by the larger social network”, not intrinsic religiosity. Further, since they also made it clear that t “because of different religious cultures, the results of studies cannot be generalized from (Finland) to (America)” or any other cultures, they would appear to be warning against generalizations to larger conclusions about religions or countries, which none the less is implicitly the service in which SJ is applying it.
And the representation of Yeager et al., (2006) appears to be so egregious so as to be utterly jaw dropping. They found that the health benefits of religion “disappears in the presence of controls for health behaviors, social networks, and prior health status”. Once again, as in nearly every other study on this list they found that it was the social networks, and not the religion, that granted the benefits. Whats more, they found that “in all cases, private (intrinsic) religious practices and stronger beliefs are associated with worse health”, (identical to to several studies on this list). What’s more, they found that “even after controlling for prior health, participation in social activities has a more robust effect on health than religious attendance”. This means that not only did they find that it was the social aspect of religious attendance that granted the benefits, but that there were plenty of other social activities which helped even better! As a result they said that “consequently, we question whether the purported health benefits are attributable to religion, or to social activity in general.”
Including Yeager to tout the health benefits of religious seems a gross misrepresentation of their findings.
While SJ uses la Cour, Avlund, & Schultz-Larsen, (2006) to tout the benefits of church service, they, like nearly every other study here, they did note that it was not religion, per say, but the the social engagements found in church service. While they did say that religion had other benefits, they found that, when accounting for covariates, “results decreased and only stayed significant regarding church attendance”. once again, this would appear only to support that it is social interaction, and not religion, that is providing the benefit.
One might question why findings suggest the highly religious have better health, well-being, and social behaviors.
it is extraordinarily difficult to overestimate how shocking this statement is. Nearly every one of these studies offered explanations. Nearly all of the explanations were identical – that the benefits came from social connections, connections you could find in any social activity (Findings which several of these studies point out Emile Durkheim actually predicted in discussing suicide and mental health). It would appear that SJ is ignoring the explanations of these findings offered by the authors, and instead looking for ones that more suit her purpose.
One comprehensive literature review recently published in Psychological Bulletin(McCullough & Willoughby, 2009) identified self-control and self-regulation as two factors. The scholars found that religion promotes self-control and facilitates self-monitoring and self-regulatory strength. Religion further influences how goals are selected, pursued, and organized.
The McCullough & Willoughby, (2009) study has a number of eccentricities which SJ failed to account for. Further, by their own admission, they failed to differentiate between intrinsic and extrinsic religion. They included all together without accounting for the differences. This led them to state that they could in no way conclude if the benefits came from “(a) the fact that religion prescribes sets of rules that are legitimated on the basis of the preferences of an omniscient deity; (b) the conviction that one’s behavior is being monitored by that omniscient deity, who can administer rewards and punishments; (c) the self discipline that is needed to maintain regular involvement in private and public religious rituals; or (d) something else remains an open question.” They also found it worth noting that “some religious phenomena (e.g., ecstatic or mystical experiences, speaking in tongues, and other religiously normative rituals that involve altered states of consciousness) seem to generate losses of self-control.”
Indeed, they did not make a claim that their findings in no way was necessarily a good thing. They highlighted this difference by saying “using religion to sanctify terrorism should be useful for generating additional motivational force to impel individual acts of terrorism as well as to sustain collective will for long campaigns of terrorism… A self-regulation analysis of religion suggests that religion is well suited to motivate any behavior that is predicated on self-control and self-regulation, whether that behavior is studying hard for final exams or donning an explosives belt and then detonating it on a crowded city bus”
However you choose to view it, they make no claim that self-control is inherently a good thing, or that it is the reason there are correlations between extrinsic religion and mental health, nor do they even differentiate between extrinsic and intrinsic religion.
Previous studies have found that religious practices and frequent church attendance have been associated with higher levels of happiness and life satisfaction (Diener & Suh, 2008; Okulicz-Kozaryn, 2010; Lim & Putnam, 2010).
Technically some of this is correct. However, keep in mind that in the context of this paper, this is being pushed into the service of suggesting that religion provides these health benefits, when in fact these studies found that it was social interaction that provided these benefits, of which religion is simply one way (and according to some of the studies, not even a particularly good way) to get it. further, as in most of the studies, there is a tremendous amount of cherry picking going on.
Okulicz-Kozaryn, (2010) found that “The relationship between religiosity and life satisfaction is bimodal. Religious people tend to be either very satisfied or dissatisfied with life. It is also two-dimensional. Forms of religiosity that promote social capital (extrinsic) predict high life satisfaction. People have so called “need to belong” and religion helps to satisfy it. Forms of religiosity that do not promote social capital (intrinsic) do not predict high life satisfaction. Religiosity is also context-dependent. Religious people are happier in religious nations.” so just as in other studies here, the cultural context is critical. If you are in a predominantly religious country, religion will provide more benefit, and principally because you will fit in more. This is identical to the findings of at least four other studies, which explicitly caution against saying that it’s the religion (which of course is the implication SJ is using here). Like so many others, they say explicitly that “it is not religiosity per se that makes people happy, but rather a social setting it offers.”
Further, as with several other cases, this study is being used in support of a a claim that both frequent church attendance and extrinsic “religious practices” are associated with higher levels of happiness and life satisfaction. But as with the prior studies, this study explicitly found that intrinsic religion did not predict high life satisfaction.
Lim & Putnam, (2010) suffers the same problem, having exclusively extrinsic findings, yet being cited as a paragraph about “religious practices and frequent church attendance”. this incorrectly implies that their findings support that full claim, when in fact they do not speak to religious practices other than church attendance. Indeed, while they said that “findings suggest that religious people are more satisfied with their lives because they regularly attend religious services and build social networks in their congregations…” they also pointed out that “We find little evidence that other private or subjective aspects of religiosity affect life satisfaction independent of attendance and congregational friendship”, meaning that this study does not support that “religious practices” outside of church attendance are beneficial.
Further, Lim later said to the American Sociological Association that “our study offers compelling evidence that it is the social aspects of religion, rather than the theology or spirituality, that leads to life satisfaction… We find that friendships.. Are the secret ingredient in religion that makes people happier.” This would appear once again to say that it is not the religion, but the social interaction, that leads to happiness.
Since purportedly this article is on the benefits of religion to happiness, most of these articles thus far not seem to support that conclusion, and indeed in many cases undermine it. They find that social interaction is beneficial, and religion is one way to obtain that interaction (and several suggest it is not the best way).
Other studies have concluded that personal aspects of religiosity relate to intrinsic motivations and beliefs and religious identity corresponds to higher levels of subjective or psychological well-being (Greenfield & Marks, 2007; Pargament, 2002; Green & Elliott, 2010; Laurencelle, Abell, & Schwartz, 2002).
The inclusion of Greenfield & Marks (2007) here is quite staggering. This paragraph says that they concluded that it was the personal aspects of religiosity that was beneficial. However Greenfield and Marks were studying only extrinsic ones. Further, it was not even a study of religion, but of volunteer groups. They found that “participation in voluntary groups (recreational, civic, and religious) buffers individuals against the harmful psychological effects of developing functional limitations”. To, from there, claim that they concluded that it was the personal aspects of religion and intrinsic motivations that led to higher well-being, seems utterly indefensible.
Pargament, (2002) Is nearly as bad. This was a study of an active intervention in the form of regular group meetings over the course of seven weeks, including active discussions, readings, sharing of opinions and life experiences, and general socializing. while indeed a great many of the participants had an intrinsic religious belief, those were not studied or controlled for. The activity was, by definition, and extrinsic activity, similar to church participation. The results that they publish are the result of that social interaction, not of the actual intrinsic beliefs. Indeed, the benefits they found came from “a sense of connection among themselves as well as the facilitators”.
In their conclusion, Green & Elliott, (2010) “In conclusion, this study contributes to the growing body of literature that links religious engagement to physical health and psychological well-being.” since religious engagement is explicitly in extrinsic activity, it would seem grossly inappropriate to use this study to support a claim to benefits of intrinsic beliefs.
It would appear that the overwhelming majority of the studies here can not be pressed into the service she is attempting, and are either misrepresented, or have conclusions which outright contradict her claims.