In Brief

While studies have long suggested that American adults become more religious as they age, a new study in the July Journal of Personality and Social Psychology (Vol. 88, No. 7) argues that people's religiousness is variable; whether a person's religiousness increases--or decreases--depends on the rational choices they make throughout their adult life.

However, amid the variability of people's religiousness, the researchers found three distinct trends:

  • People who become increasingly more religious until midlife and progressively less religious in later adulthood.

  • People who are not religious in early adulthood who become even less religious later in life.

  • People who are religious in early adulthood who continue to become more religious with age.

"We expected to find that people become slightly more religious as they go through life," says lead researcher Michael E. McCullough, PhD, a University of Miami associate psychology professor. "But instead what we found were three very different and distinct pathways of development that varied based on the individuals and the decisions they made."

To reach these findings, McCullough and his colleagues used self-report data from the Terman Longitudinal Study, a five-decade longitudinal study of 1,151 mainly white, middle-class children with IQs greater than 135. One rater read the information related to religiousness that participants provided in six waves of data collection (1940, 1950, 1960, 1977, 1986, 1991) and assigned a single value between negative one and four reflecting her perceptions of the participant's religiousness at that point in his or her life.

The results indicated that the personality characteristic of agreeableness, along with the participants' religious upbringing and the decisions they made throughout their lives, such as the size of their family and whether they married, predicted their membership in one of the three pathways.

For example, participants who were unmarried throughout their entire lives were more likely to belong to the increasing-religiousness group. Participants with families larger than two children were less likely to belong to the low-then-decreasing religiousness group than participants with relatively small families.

McCullough acknowledges that the Terman participants were considerably less religious than the general U.S. population, with only 40 percent of Terman participants being church or synagogue members compared with 72 percent of the general population in 1940. However, he believes that although the proportions are different, the same patterns would also appear in analyses of data from the general U.S. population. In future studies, McCullough and his colleagues will investigate how religious development throughout life relates to mental health, longevity and quality of life in old age.