Magic Statistics

A comprehensive new study on religious liberty around the world shows that countries with Christian backgrounds generally have the highest levels of religious freedom. The study, directed by esteemed expert on religious persecution Paul Marshall, also found that officially atheist or Islamic countries tend to have the least religious freedom.

The four countries given the highest religious freedom rating of one are Hungary, Ireland, Estonia, and the United States.

On the other hand, countries run by atheist government such as communist China, Vietnam, and North Korea were ranked in the bottom two tiers (ratings of six and seven).

Officially atheist countries were joined at the bottom of the religious freedom pole by countries with Islam background such as Pakistan, Palestinian areas, Iran, Iraq, Saudi Arabia, Sudan and Turkmenistan.

Canada was ranked in the second-highest tier.  A few anomalies turned up, for example, the Muslim nations of Mali and Senegal both ranked in the second tier, above France, Germany, and Greece—all ranked in the third.

Nevertheless, as Dr Marshall points out, obvious patterns emerged.

"If we want to classify the 'worst states' or the most egregious persecutors," he says, "they tend to be either communist (North Korea, China), nationalist (Burma, Eritrea), or radical Islamist, such as Iran and Saudi Arabia."

Marshall adds the greater Middle East is the most religiously repressive part of the world. The region is increasingly threatened by a trend of growing fundamentalism. "Radical Islam is the fastest-growing threat to religious freedom in the world."

To add a note of surrealism, a commenter at Christian Post recommends Gregory S Paul’s thoroughly debunked study purporting to show that religious belief is detrimental to societal health.  The same commenter even more absurdly recommends the credulous article in Skeptic touting Paul’s misguided study.

Previous related posts:

• 200 million Christians suffering increasing persecution
• Islamic extremism poses greatest threat to Western culture
• Dogmatic secularism versus reality
• Paul Marshall on the Mohammed Cartoons

Gregory S Paul’s study on social dysfunction and religiosity has taken a beating from several bloggers, myself included.  Now it emerges that the Journal of Religion and Society, which published Mr Paul’s paper, has also printed a thorough refutation.

Gerson Moreno-Riaño, Mark Caleb Smith, and Thomas Mach of Cedarville University take Mr Paul to task on just about every aspect of his paper.  His concepts and definitions were ambiguous, inadequately explained, and, in some cases, excessively flexible.  He focused his analysis exclusively on the individual level and failed utterly to recognise, never mind take into account, the complications arising from the fact that national polities do not necessarily reflect the prevalence of religious belief among citizens.  Thus, for example, most Western European nations favour one particular religion or denomination, yet have great majorities of non-religious individuals.  The United States, by contrast, is an officially secular nation with a highly religious populace.

Moreno-Riaño et al. also discuss data quality problems in Mr Paul’s fundamental data source, the International Social Survey Program, especially with regard to the religion questions.  The concerns are widely known and well-documented in the academic literature but Mr Paul gives no indication that he is aware of the controversy.

And then there’s the fact that Paul’s article contains no actual statistical analysis.

Paul does not provide the reader with a meaningful opportunity to evaluate his findings, for he provides no correlation coefficients. He also fails to determine or report the significance of these correlations, so the reader is left to trust Paul’s judgment that a negative correlation between theism and indicators of societal health has been established. Statistics exist so that we are not required to trust the subjective judgment alone, particularly in regard to matters as weighty as theism, democracy, and the social good.

Mr Paul apparently expected us simply to trust his judgment.

h/t: Telic Thoughts and Thinking Christian.

Previous related posts:

• I can’t believe that atrocious article is still being touted
• Dogma bites man
• From our bulging “How not to do statistics” file

Philanthropy expert Arthur C. Brooks of Syracuse University has determined that religious conservatives give far more than secular liberals to charities of all kinds.  That finding holds true across all income levels. The research is set forth in a new book, Who Really Cares: The Surprising Truth About Compassionate Conservatism, to be released next week.

In the book, he cites extensive data analysis to demonstrate that values advocated by conservatives — from church attendance and two-parent families to the Protestant work ethic and a distaste for government-funded social services — make conservatives more generous than liberals.
. . .
When it comes to helping the needy, Brooks writes: "For too long, liberals have been claiming they are the most virtuous members of American society. Although they usually give less to charity, they have nevertheless lambasted conservatives for their callousness in the face of social injustice."

For the record, Brooks, 42, has been registered in the past as a Democrat, then a Republican, but now lists himself as independent, explaining, "I have no comfortable political home."

Religious conservatives who object to government-mandated redistribution of income give the most to charity, while liberals who support such programs give the least.  Even when it comes to non-monetary giving, such as blood donations and volunteering, liberals contribute less.

Such an attitude, he writes, not only shortchanges the nonprofits but also diminishes the positive fallout of giving, including personal health, wealth and happiness for the donor and overall economic growth.

All of this, he said, he backs up with statistical analysis.

Harvey Mansfield of Harvard University, who has not met Dr Brooks but has read his book, had this to say:

"His main finding is quite startling, that the people who talk the most about caring actually fork over the least.”

Dr Brooks admits he is not happy with his own results, in part because he was raised in a liberal household.  He does not consider himself a conservative or a supporter of the “religious right”, yet he knows his book will be deeply unpopular with liberals.

This new research is yet another nail in the coffin of the claim, made by Gregory S Paul and Skeptic magazine among others, that secularism is good (or, at least, not bad) for society.

Dr Brooks is Professor of Public Administration and Director of the Nonprofit Studies Program at Syracuse University's Maxwell School of Citizenship and Public Affairs. 

h/t: Clayton Cramer

Previous related post: Small group of Canadians provides lion’s share of charitable giving

That was my initial reaction when my friend Tom Gilson told me that the latest issue of Skeptic magazine features an article extolling Gregory S. Paul’s paper on social problems and religious belief.  Mr Paul’s study, published last year in the Journal of Religion and Society, is entitled "Cross-National Correlations of Quantifiable Societal Health with Popular Religiosity and Secularism in the Prosperous Democracies”; the html version is posted here and the pdf version here.

Mr Paul’s paper presents itself as a statistical investigation of social problems—including homicide, prevalence of sexually transmitted diseases, abortion, pregnancy among young teenagers—and religious belief.  He collected data from eighteen democracies and claimed to have found positive correlations between social dysfunction and religious belief, thus purportedly showing that religiosity is not necessarily beneficial and secularism is not inherently harmful to social mores.

One year ago next week, I wrote one of my most popular blog posts critically assessing Mr Paul’s paper and showing it to be statistically invalid.  I concluded that, in my professional judgment as a statistician, the study should not be taken seriously.

Now comes the Skeptic article, by one Matthew Provonsha, who merely reiterates—in the most uncritical fashion imaginable—the errors and shortcomings in Mr Paul’s original work.  For Provonsha, evidently, Paul can say no wrong.  (Why was this published in a magazine called Skeptic?)

Thus, Mr Provonsha does not wonder why Mr Paul fails precisely to document his data sources or provide sound reasons for his selection of countries to be included.

The study focuses on the prosperous democracies, because “levels of religious and nonreligious belief and practice, and indicators of societal health and dysfunction, have been most extensively and reliably surveyed” in them. Also, “The cultural and economic similarity of the developed democracies minimizes the variability of factors outside those being examined.” With a database of 800 million people, this study is far more reliable than results based on smaller sample sizes used in other such studies.  The data are also current and extensive, collected in the middle and latter half of the 1990s and early 2000s from the International Social Survey Programme, the UN Development Programme, the World Health Organization, Gallup, and other well-documented sources.

From a statistical perspective, the criteria Paul used to include his eighteen chosen countries and omit all others are arbitrary and statistically unsupported, but Provonsha falls for it.  Also, Paul does not document his data sources, in that we don’t know from which particular source he obtained each of the various measures of social health.

The claim that Paul utilised “a database of 800 million people” is laughable.  The sample size of Paul’s study was precisely eighteen: one data point from each country for each data series.  To claim this represents the combined population of all the countries is like a market research firm conducting a survey of 1200 Canadians and then claiming they accessed a database of over 32.5 million people.

I do want to thank Mr Provonsha for re-drawing some of Mr Paul’s charts.  They are now larger and much easier to read.  Their statistical value, however, is still negligible.  Paul seemed to hold the belief that plotting points on charts constitutes statistical analysis.  Not in my world.  Paul and Provonsha should both read Darrell Huff’s classic, How To Lie With Statistics, which devotes an entire chapter to the perils of graphs.

Visual representations of data, e.g., plotting data points on a chart, can be valuable as aids to understanding, but they are no substitute for calculating correlation coefficients and other analytical statistics.  Paul did not calculate any analytical statistics—not one—and Provonsha doesn’t call him on it.  Paul’s paper has many descriptive statistics, but no actual statistical analysis, and Provonsha apparently didn’t even notice.  That says it all.

h/t: Tom Gilson, Thinking Christian

Other recommended commentary on Gregory Paul’s study:

• “Dogma Bites Man”, by George H. Gallup, Jr. at Touchstone Magazine
• “Debunking Gregory S. Paul”, by my friends John and Scott at Verum Serum

Previous related posts:

• From our bulging “How not to do statistics” file
• Mystery man
• A primer on crime statistics

In the latest issue of Touchstone Magazine, renowned public opinion pollster George H. Gallup, Jr., takes a hard look at Gregory Paul's study of religious faith and social pathologies, "Cross-National Correlations of Quantifiable Societal Health with Popular Religiosity and Secularism in the Prosperous Democracies". In my opinion, it's a rout across the board. Mr Gallup and research associate D. Michael Lindsay of Princeton University challenge Mr Paul's data, his analytical methodology, and his conclusions.

Paul has made strong claims about the effect of religion upon society without examining all the other factors that might explain the phenomena he wrote about.
. . .
[A] mountain of survey data from the Gallup and other survey organizations shows that when educational background and other variables are held constant, persons who are "highly spiritually committed" are far less likely to engage in antisocial behavior than those less committed. They have lower rates of crime, excessive alcohol use, and drug addiction than other groups.

On the other hand, the "highly spiritually committed" are more hopeful about the future and experience greater joy in life. They contribute more time helping people who are burdened with physical and emotional needs. They are less likely to be racist, and are more giving and forgiving.

At the same time, Mr Gallup acknowledges that not all Americans who profess religious faith are "highly spiritually committed". It is this gap between profession and behaviour that Mr Paul attempts to exploit.

As Paul noted, a huge majority of Americans attest to a belief in God or a higher power, but he did not ask the key question in understanding the effect of religion on American life: How deep is this belief?

Mr Gallup suggests only about 10% of American Christians are deeply commited to the faith; but these Christians are far more likely to exhibit attitudes and behaviours in keeping with the teachings of Christ.

Clearly, Mr Paul is away off the track. Equally clearly, however, Christians could be doing much more to live as salt and light in our societies.

As a final word, I am honoured that Mr Gallup and Touchstone Magazine editor David Mills quoted part of my blog post on this study.

The daily e-mail from Charles Colson's BreakPoint just arrived, and I see that Mark Earley, Prison Fellowship President, devotes today's commentary to a look at Gregory Paul's study that purported to find correlation between religious faith and social pathologies. Mr Earley calls this a "man bites dog" story because it is such an unusual finding. He criticises the study for apparent cherry-picking of data observations so as to ensure the outcome that Mr Paul sought.

His goal doesn’t seem to be an honest appraisal of the role of religion in public life, but rather simply to embarrass American Christians.

That’s why he all-but-commits the oldest mistake in statistics: confusing correlation with causation. Just as a rooster’s crowing when the sun comes up doesn’t mean that the crowing causes the sun to come up, the presence of social problems in religious societies doesn’t mean that religion is the cause.

Yet that’s clearly the inference Paul wants us to draw from his so-called "findings." Why else would he suggest that "Europeans are increasingly repelled by the poor societal performance of the Christian states"?

Amen to that. Mr Earley includes in his list of material for further reading my blog post of 27 September. Thanks for the link, and I hope that new visitors will look around my blog and find some other stuff of interest.

Remember that study by one Gregory S. Paul that purported to show a correlation between religious faith and social pathologies? It was quickly trounced by several bloggers, me included. Now Foyle at Verum Serum, "an eclectic Christian blog by two guys from Orange County, CA.", has done a lot of research, both on Mr Paul and his findings.

Foyle provides evidence that Mr Paul had come to his conclusions well before he undertook his little investigation. Quotes from Paul’s earlier books and articles say almost exactly what he wrote in his study. Foyle also utilises more and better data that point to different conclusions than Paul reached. Good work.

Check it out.

The alacrity with which atheists in the blogosphere and beyond latched on to Gregory Paul's study purporting to show a correlation between religious belief and social pathologies has baffled me. As I and others have shown, Mr Paul's "study" was astonishingly simple-minded, statistically incompetent, and utterly lacking any foundation in social scientific theory. Moreover, the fact that it contradicted hundreds of other, statistically credible and scientifically rigorous studies would, one would have thought, give pause to those who agreed with its alleged "findings".

Despite all these rather obvious defects, atheists held it aloft as if it were the Holy Grail. I wondered: "Whatever happened to atheism's much-vaunted critical reason?"

As one who made a contribution to debunking Mr Paul's silly little investigation, I witnessed ad hominem attacks of the most puerile sort: "You're a Christian, so I can’t believe a word you say". It's rather sad–pathetic, really.

A possible explanation appears in this morning's Guardian: Atheists simply aren't as smart as they like to think they are.

Guess who said this: "How much boundlessly stupid naivety is there in the scholar's belief in his superiority, in the good conscience of his tolerance, in the simple, unsuspecting certainty with which his instincts treat the religious man as inferior and a lower type which he has himself evolved above and beyond." Some uppity Christian complaining about warmed-up anti-clericalism in the Guardian? Or the most vociferous atheist of them all, that great genius of anti-Christianity, Friedrich Nietzsche. For although Nietzsche hated Christianity, he also recognised that atheism is prone to a self-satisfied smugness in which religion is written off as a fool's game, practiced by suckers and easily coopted by the wicked.
. . .
The joke is that many who were converted [to atheism] at university via Richard Dawkin's The Selfish Gene think of themselves as agents of some subversive counterculturalism. This is ridiculous to Da Vinci Code proportions. Contemporary atheism is mainstream stuff. As John Updike put it: "Among the repulsions of atheism for me has been its drastic uninterestingness as an intellectual position."

Sounds about right to me.

via titusonenine.

George Gallup, one of the most respected public opinion pollsters in the world, echoes some of the criticisms I made about Gregory Paul's study which purported to find a positive correlation between religious belief and several social pathologies. Mr Gallup has written a letter to the Times of London taking Ruth Gledhill to task for her uncritical news article puffing Mr Paul's little study.

From David Virtue's exclusive report, here are Gallup's two main objections to Paul's work:

First. Paul claims that regressions and multivariate analyses were not used because 'causal factors for rates of societal function are complex', and because he finds enough uniformity across the cases of 18 of the world's most powerful societies to consider them basically consistent and not in need of control variables. Can he identify a single other study published in a major social scientific journal that compared results across countries that did not employ multivariate analysis to control for differences among nations? No, because multivariate analysis is required for cross-national comparisons of this sort.

Secondly. In order for the author's bold claims against religious commitment contributing to society to hold true, he would have to refute the hundreds of volumes that have proven otherwise. From discussions on parenting and fatherhood, to mental and physical health, the weight of empirical evidence is against Paul's assertions: religious commitment has notably positive effects on the individual and collective levels of human society.

That first criticism is the same as one I presented in my post of 27 September. As I said then, regression analysis, or some other kind of multi-variate analysis, is required to assess adequately the factors Mr Paul attempts to correlate on a pair-wise basis.

Mr Gallup's second point has been noted by other bloggers. Mr Paul's work represents a single study with this finding, compared to hundreds of other, far more rigourous studies showing that religious faith is correlated with all kinds of positive individual chacteristics and social outcomes. It is simply not rational to reject all those other findings on the basis of this one simple-minded analysis which, to be honest, seems to have been done on the back of an envelope.

Read the whole thing for an outline of how properly to measure religious belief and practice, and the results that obtain when such measures are correlated with specific behaviours.

Vindication is sweet.

UPDATE (30 Dec.): Touchstone Magazine has published George Gallup's article criticising Mr Paul's study.

Rosa Brooks, a defender of Gregory Paul's study on religious belief and social problems, was given air time recently on Tucker Carlson's MSNBC show. Based on the show's transcript, however, Ms Brooks made a hash of Mr Paul's study, which was incompetent to begin with.

Mr Paul purported to find correlations between religious belief and social pathologies such as suicide and homicide. Ms Brooks recently wrote an LA Times column promoting Mr Paul's study, saying: "At a minimum, his findings suggest that contrary to popular belief, lack of religiosity does societies no particular harm".

Mr Paul used these data to gauge religious belief: "Bible literalism and frequency of prayer and service attendance . . . absolute belief in a creator . . . [s]elf-reported rates of religious attendance and practice. In her LA Times column, Ms Brooks explicitly endorsed this view of religious belief: absolute belief in God, the frequency of prayer reported by . . . citizens and . . . frequency of attendance at religious services."

Yet, when Mr Carlson proposed that most murders in the 20th century were committed in the name of secular ideologies, Ms Brooks made a complete about-face in her definition of religion:

CARLSON: You say arguably Paul's study invites us to conclude that the most serious threat humanity faces today is religious extremism. And I was just thinking about the last 100 years. All the people who were killed in the 20th Century the vast majority were killed by aggressively secular regimes, fascist and communist regimes. It wasn't religious people.

BROOKS: You know when I think that these have in common though, Tucker, and first of all give religion a chance. Religion's got a lot more millennia behind it of slaughter and mayhem that you can lay at its door than secular ideologies.

That said, I completely agree with you that the major 20th Century crimes many of them had to do with absolutist secular ideologies. I think what these things have in common though is that element of absolutism that once you're out there and you're saying I'm right. You're wrong. I know because I know. Shut up. If you disagree, just get out of the way. I'm going to steamroll right over you. That's when you're getting problems.

CARLSON: So, essentially they're all a kind of religion.

BROOKS: Yes, I do think so. I think that communism in some of its forms during the 20th Century was a kind of religion. I think that Nazism was a kind of religion to its more fanatical adherence [sic], absolutely, sort of a total world view that couldn't be challenged by evidence (or) by logic.

But this definition of religious belief is completely contrary to that analysed by Mr Paul in his study and defended by Ms Brooks in the LA Times.

Now she tells Mr Carlson that Fascism and Communism qualify as religions, despite the fact that these ideologies reject everything Mr Paul used to measure religiosity: Bible literalism, absolute belief in a creator-God, efficacy of prayer, and attendance at religious services.

Ms Brooks, for the sake of your own credibility: if you're going to criticise religion, you should at least keep your story consistent. Just a suggestion.

It’s been a wild week here at Magic Statistics. My daily readership swung away up, then dropped off, and then soared again. Overall, an upward trend is evident. Thanks to everyone who stopped by to take a look and have a good long read. Based on comments and page view counts, I know many of you did just that.

The week started when, as I reported here, Canon Kendall Harmon of titusonenone, posted his recommendation of this blog. I’ve been reading Kendall for quite some time, but I had no idea how many faithful readers he has. He plainly has a very large group of regular readers. That’s as it should be: his is one of the leading blogs with a traditional and orthodox perspective on the world of Anglicanism. Over the next two days, my number of daily readers almost tripled. Then, traffic subsided somewhat for a short while.

Then, on the evening of 27 September, I posted this item—a review of an article on religion and social problems by one Gregory Paul that received an uncritically favourable report in the [U.K.] Times. I reviewed the article from the perspective of technical statistics, and I have to say it wasn’t pretty for those who agree with Mr Paul that religion is bad for social health.

As word got around the blogosphere, that post (and its follow-up) generated hundreds of visits, as well as several comments and trackbacks. I’ve tried to visit the blogs of those who mentioned my post and leave thanks in the comment boxes, but now I find that there are more than a dozen that I haven’t visited. So, I want to take this opportunity to thank all those who linked to me by listing their blogs here:

• Telic Thoughts
• Wittingshire
• Thinking Christian
• Prosthesis
• ID in the United Kingdom
• Blogotional
• Uncommon Descent
• Mr Smith In Washington
• Intellectuelle
• Speculative Catholic
• The Revealer
• Confederate Yankee
• Thudfactor
• Glaivester
• 4 The Little Guy
• On The Other Foot
• friends
• Hannah Im
• Dr Albert Mohler's Blog
• Stand To Reason
• Dissecting Leftism

In some cases, the link to my post appeared in a comment left by a reader at the site. I also received a few appreciative e-mails sent to the "Contact me" address on the sidebar [now in the "About" box at the top of the page]. Some readers also posted the link at bulletin board sites. I apologize to anyone I have inadvertently omitted.

Thanks to everyone who posted the link or sent an e-mail. I’m looking forward to visiting your sites on a semi-regular basis at least. And if, in your future net surfing, you come across a statistical analysis that looks questionable or problematic, let me know. I’m always on the look-out for that kind of stuff.

I’ve also added a new feature to my sidebar [now moved to the top of the page]: a section entitled "Popular Posts" with one entry. Here’s hoping there will be more in the not-too-distant future.

By the by, at least one commenter apparently did not get my sense of humour, so I should point out that the title of this post is tongue-in-cheek. (But it could be true anyway.)

UPDATE (3 Oct.): List of those linking to my post augmented with two new entries.
UPDATE 2 (5 Oct.): One more added.

The study by Gregory S. Paul on religious faith and social problems has been spinning around the blogosphere for two days now. So, who is this Gregory Paul anyway? It turns out that he is really a "freelance" paleontologist with no apparent experience in sociology, social science research, or statistical analysis. (Those who have read this post already know that Mr Paul’s knowledge of statistical analysis is, shall we say, sketchy.)

Kathy Shaidle points us to an interview with Mr Paul printed in an Australian newspaper. Even though the reporter lobbed him softball questions, he sounds like he’s not sure what he’s doing analysing social behaviours.

The Journal of Religion and Society (JRS), where Mr Paul’s paper was published, is apparently willing to publish articles contributed by authors with no expertise or qualifications in their area of study. The table of contents of the current issue of the JRS lists affiliations for most contributors, but Mr Paul is listed only as: "Gregory S. Paul, Baltimore, Maryland." One could conclude that he does not hold a relevant position at an academic or other research institution. In view of the poor quality of Mr Paul’s analysis, I’d be interested to know if the JRS is a refereed journal, or do they publish just about anything they receive?

An enterprising blogger named John Williams at Thudfactor went so far as to e-mail the Journal of Religion and Society to ask if they had any further information about Mr Paul’s experience and credentials. I guess I shouldn’t be surprised, but still I was astonished to read that the JRS knows nothing about this fellow—except that the e-mail address he provided to the journal is no longer in service. (Didn’t they even get a mailing address from this guy?) Mr Williams also links to a Wikipedia entry on one Gregory S. Paul that describes him as a "freelance paleontologist, author and illustrator" who "is best known for his work and research on theropod dinosaurs . . . " No mention of any proficiency in social science research or statistical analysis. (Who wrote this entry for Wikipedia anyway? Any bets it was—Gregory Paul?!)

Finally, a Google search for "Gregory Paul" turns up this web page which lists Mr Paul as a speaker recommended by the Council for Secular Humanism for debates with young-earth creationists.

So, what can be pulled together from all this? Gregory Paul has published a study of social problems and religious faith; but he has no apparent expertise or qualifications in social science research so, predictably, said study is statistically invalid. Said study was published by a journal that apparently does not have high standards for articles it publishes, and it does not even know how to contact Mr Paul. Finally, said study makes a ham-handed attempt to portray religious faith as a dangerous and socially destructive force in the U.S., and it transpires that the author is on the Council for Secular Humanism’s list of recommended speakers.

I think that Mr Paul has successfully played a big con game. That, at least, inspires a grudging (not to say perverse) admiration.

UPDATE: John Williams tells me he didn't phone the JRS, he sent an e-mail. The post has been amended accordingly.