May 10 2012

Analytic Thought and Religious Belief

A series of psychological studies recently published in Science explores the relationship between analytic thought and religious belief. The studies raise a lot of issues, including how to interpret such studies, but first let me simply convey the results.

In the first experiment researchers Will M. Gervais and Ara Norenzayan assessed subjects with a standard measure of analytical thought – problems in which the initial intuitive answer is incorrect and must be overridden by deeper analysis. Try to solve them yourself, they are:

A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost?
If it takes 5 machines 5 min to make 5 widgets, how long would it take 100 machines to make 100 widgets?
In a lake, there is a patch of lily pads. Every day, the patch doubles in size. If it takes 48 days for the patch to cover the entire lake,
how long would it take for the patch to cover half of the lake?

You can look at the original paper for the answers. The researchers found that giving the analytic (and correct) answer to the questions above negatively correlated with three related measures of religious belief – instrinsic religiosity, intuitive religiosity, an belief in supernatural agents.

The researchers then did four experiments in which they used methods previously demonstrated to encourage analytic thought to see if they affected one of the above measures of religiosity. They included: showing subjects a picture of Rodin’s The Thinker,  exposing them to analytic words like “think” and “reason” (this method was used twice), and changing the font of the questions (for each there was a control group that had similar stimuli not associated with analytic thought). More difficult to read fonts have been shown to trigger analytic thought. In each experiment the group that received the stimuli associated with analytic thought scored lower on one of the measures of religiosity.

All the results were statistically significantly, although mostly modest in magnitude. The most significant was using The Thinker as a priming stimulus (the control group was exposed to the famous sculpture of the discus thrower), which decreased the mean religiosity score from 61 to 41 (out of 100). In a pilot study the same stimulus was associated with scoring higher on analytic tasks.

Taken together these five experiments show a consistent pattern. One simple interpretation is that engaging in analytic style of thinking decreases religious type beliefs. The authors speculate that the effect is likely occurring within the mildly religious members of the study group, with mild believers acting like mild disbelievers. These type of stimuli are unlikely to dislodge strongly held beliefs.

That is a plausible interpretation, but I have two problems with interpreting these types of experiments. The first, which the authors acknowledge, is that there are often numerous ways to interpret such results. Human thought and behavior is multifactorial and very complex, with many factors coming into play and influencing our behavior. It’s very difficult to tease apart all these influences and isolate the one factor you are interested in – in this case the effect of analytic thought on religiosity.  There always seems to be a trail of assumptions leading up to any one interpretation. Are the measures of religiosity really measuring religiosity? Are the measures of analytic thought really measuring analytic thought? Are the priming stimuli affecting analytic thought, or something else that might produce a similar result?

Perhaps, for example, factors such as the willingness to be honest, or go against perceived social convention were influencing answers to questions such as, do you believe in god? People tend to have multiple psychological and situational influences affecting their behavior in the moment, and often have conflicting motivations. People, for example, may simultaneously view themselves as rational and pious, or have a desire to impress their teachers with analytic performance and their parents by following their religious belief. The priming may have given a nudge to one social pressure over another.

The hard to read font (called a disfluency test) seem particularly subject to alternative interpretations. Perhaps the font just makes subjects more attentive and less distracted, forcing them to process the language more consciously. This could explain better performance on analytic tasks. It might also explain lower reported religiosity – perhaps subjects are not really feeling less religious when working harder to read a font, but are simply more engaged with their “student” role or have less mental energy to worry about appearing pious. These are simply speculations, the point is that interpreting the results of these types of studies is anything but straightforward.

My second concern about interpreting such research is that the priming effects are all immediate, and therefore could simply be situational. One thing psychological studies (and performance magic) have clearly shown is that people are easy to manipulate (at least statistically). There are many situational factors that influence our immediate behavior, by engaging or inhibiting one or more of the many factors that are simultaneously influencing our behavior. In other words – these manipulations may not be changing our personality or the balance of social and cognitive influences on our behavior, but are simply emphasizing one or a groups of such influences in the moment of the study.

In the end what we learn is that human behavior is multifactorial – we have lots of buttons that can be pressed. Psychologists are learning what the buttons are and how to press them, having a modest if statistically significant effect on behavior, but that is not the same thing as identifying a dominant or determining influence on a behavior. It is also probable that those subjects in such studies who are affected by the button-pushing have a balance of influences on the thing being measured, so that subtly pushing on one will change their behavior. But most subjects are not affected, because other influences are dominant and not subject to easy manipulation. In these studies, having deeply held religious beliefs or being a confirmed atheist was unlikely to be influenced the priming. Being conflicted in the first place, however, may make someone easy to push in one direction or the other.

Despite all my concerns, I do think the researchers have likely zeroed in on one influence of religious belief (broadly conceptualized). Most people have both the tendency to think intuitively and to think analytically, in some balance, and can engage in either one or both depending on the situation. Some people are obviously more intuitive and others more analytic. Research (such as that in cognitive therapy) also does seem to suggest that people can change the way they think with training and effort. Thinking style is probably a combination of inherent tendency and habit, the latter can be changed over time with effort.

What I really want to know is  – how can we get people to shift their balance of thought style to be more analytic, at least in the appropriate situations. The questions above, for example, have objective correct answers, so this is not just a matter of style or personal choice. Part of the goal of organized skepticism is to give people the tools to think analytically and the motivation to do it when necessary. This research doesn’t tell us anything about how to do this or how easy or difficult it is to do so.  There is other research that sort-of addresses these questions, but it too is subject to interpretation.

This kind of research is difficult, but that doesn’t mean it’s impossible or that the results of such studies are worthless. It means we have to be cautious when interpreting the results, and it takes many many studies to triangulate to and isolate specific variables and the role they play in our thought and behavior. These studies are interesting, but they are one tiny piece in a large and complex puzzle.

Like this post? Share it!

62 responses so far