What's interesting about this study is that it shows some component of religion does lead to some negative evaluations of people based on race," said Wade Rowatt, associate professor of psychology and neuroscience at Baylor, who led the study. "We just don't know why."

Priming occurs when exposure to one word or concept influences an individual's future response to something else. Studies show that people asked to choose between a string of letters on a computer screen that forms a word -- "banner," for example -- and one that does not -- like "banack" -- respond more quickly the next time they see a similar word. After being primed with "water," for example, subjects will more quickly identify the string "drink" as a word. Psychologists call it "cognitive priming," and it has long been thought to play a role in prejudice.

Priming is why supermarkets stock items in bins marked with signs like "10 for $10" or "limit 12 to customer." Because it's subconscious, shoppers don't think they are influenced to buy larger quantities, but it must work. Otherwise businesses wouldn't continue to do it. It has also been observed that people voting in a school are more likely to support a school funding project than those who vote in another polling place.

In the Baylor study, college students recruited from introductory psychology classes were primed with either religious-word letter strings like "Bible," "faith," "Christ" and "church or neutral words like "shirt," "butter," "switch" and "hammer." Researchers found that religiously primed students demonstrated "a slight but significant" increase in racial prejudice.

Previous studies show a complex relationship between religiosity and racial prejudice. Some dimensions of religion have been shown to increase levels of prejudice, while others reduce it. Those studies all rely on self-reporting, however, and are therefore skewed by "social desirability," meaning that some people report more positive racial attitudes than they actually hold.

The Baylor study is thought to be the first to experiment whether exposure to religious concepts may contribute to racial prejudice.

Rowatt, along with fellow researchers Megan Johnson, a graduate student in social psychology, and Jordan LaBouff, a Ph.D. candidate in social psychology, speculated about possible explanations for what they call the "Christian-racial-prejudice hypothesis."

One is that because America's religious tradition is so influenced by Puritanism, people responding to religious terms may be drawing on ideals like the "Protestant work ethic," which has been shown to activate anti-black attitudes.

Another possibility is that Christian words might evoke feelings of fundamentalism -- with has previously been found correlated with prejudice -- or political conservatism, which tends to justify existing inequalities between blacks and whites.

Another is that religious terms trigger positive feelings toward "in-group" members while denigrating "out-group" members perceived as violating a particular core value.

Groups in the study were ethnically diverse but predominantly white. They were also predominantly Protestant or Catholic.

Researchers said they chose Christianity for the study because it is the most prominent religion in America. African-Americans were selected because they are historically a disadvantaged social group in the United States. They suggested that future studies could explore whether a similar effect occurs in other cultures and religions other than Christianity.

Students primed with religious terms scored higher in both terms of "covert" racism -- where individuals evaluated whether conclusions were supported by certain arguments rather than whether they agreed with those arguments or conclusions themselves -- and "overt" racism -- negative attitudes expressed in responses to questions like how afraid they are of African-Americans as a group or whether or not they like them.

"These experiments are the first to document that activation of Christian concepts by priming affects racial attitudes and provide some insight into the 'paradoxical' relationship between religion and prejudice," the study concludes. "What now remains to be demonstrated is what mechanisms underlie this relationship between religion and racial prejudice."

Rowatt said he doesn't know if the study helps explain why despite progress in race relations since the 1950s, black and white Americans still overwhelmingly worship in separate churches.

Kevin Dougherty, assistant professor of sociology at Baylor, says people tend to be attracted to churches by a "shared culture" that include things like agreement about what defines sound doctrine, worship style and the presence of family and friends. He warns, however, that cultural divisions also promote social divisions, which are part of what the Bible says Christ came to redeem.

Dougherty says factors like race, social class and gender should not divide the nation's 350,000-plus congregations. While not every church can be multi-racial, he says, every church should be a part of "God's redemptive work of reconciling societal divisions."

----Bob Allen is senior writer for Associated Baptist Press.