top of page

Research Philosophy

Intellectual Influences

All scientists stand on the shoulders of the giants who came before them. I have been particularly influenced by three traditions. 

​

Behavioral Economics

My interest in the social sciences was originally piqued by discussions of rationality. Economists historically have assumed that people are rational, but behavioral economists like Nobelists Daniel Kahneman and Richard Thaler showed numerous ways in which people are not. First, there are numerous ways in which rationality is “bounded” by cognitive constraints (Ariely, 2008; Gilovich, Griffin, & Kahneman, 2002). For example, people’s mental capacities are limited, so they use heuristic mental shortcuts to make decisions (Gilovich, Griffin, & Kahneman, 2002; Tversky & Kahneman, 1975). Second, studies show how limited self-control leads people to make irrational decisions about using drugs, overeating, and strongly favoring things that increase utility now, while devaluing things that would benefit one in the future (Tangney, Baumeister, & Boone, 2004; Vuchinich & Simpson, 1998). Third, people often would rather an outcome be fair, even if it reduces their own wealth or utility—this concern about fairness violates expected utility theory because the model predicts that people consider only their own utility (Bolton, Katok, & Zwick, 1998; Nowak, Page, & Sigmund, 2000; Thaler, 1988).

​

Free Will

I have worked with three brilliant scientists whose work relates to whether people have free will. I first worked with Bill Banks and then John Bargh, whose work emphasized the power of unconscious forces on human behavior, which limits free will. I later worked with Roy Baumeister on conscious control over behavior, which enables people to have free will. Free will is an important part of rational decision-making. Free will is also an important part of moral character because people can only act morally if they can control and direct their behavior. 

​

The Cultural Animal

What is human nature? The perspective that ties all my interests together is that humans are cultural animals (Baumeister, 2005). Humanity's biological strategy is to cooperate with others to satisfy their needs via culture and society. Thus, all psychological phenomena must be consistent with our biological and cultural evolution. I collaborate with Jon Maner at Florida State, whose work takes a biological evolutionary perspective, and with Joshua Jackson at UNC on projects from a cultural evolution perspective. The survival by cooperation through culture highlights the importance of morality in decision-making, which is why I did postdoctoral research with Kurt Gray, a moral psychologist. 

Methodological Approach

Because there are strengths and limitations to any way of forming knowledge, I use a variety of methods and approaches to answer research questions. Primarily, I conduct experiments, both online and in the laboratory. My laboratory experiments typically measure behaviors that are difficult to elicit online, like whether people will put their hand in a bucket of worms, whether people smile when they see Hitler smiling, and whether people behave differently after engaging in self-control tasks for long durations. More recently, I have also begun testing hypotheses using large datasets like the World Values Survey. With my colleagues, I am also using agent-based modeling to simulate how individual preferences lead to group-level phenomena. For example, I'm looking at how different kinds of leaders provoke different responses in people, and how that shapes how their group develops. 

Replicability Statement

Social psychology has recently begun to question whether our previous methods are reliable and replicable. This has changed how I do experiments in a few ways. Studies don't replicate for many reasons, but two major factors are low statistical power and underspecified theory.

​

If you flip a coin 1000 times, it should land heads roughly half the time. It would be very surprising if your coin landed heads 800 times. So surprising, it would seem fairly certain you had a weighted coin. It would be "statistically significant," and you would be strongly convinced of the results. But if you flipped that same coin 10 times and it landed heads 8 times, the results would be almost significant (p = .051), but much less convincing than the big study. This is just like what happens when you run a small vs. a large research study. It's too easy to find fluke results in a small study, especially if you are smart and creative in how you look for them. Researchers are smart and creative people, so we have to be careful to avoid finding flukes when we do studies. 

​

I preregister my confirmatory studies at aspredicted.org. By preregistering my exact predictions based on theory, I hope to reduce false positives. I recommend aspredicted to my colleagues because it is very easy to use. 

​

I have also dramatically increased the sample size of my experiments. As a general heuristic, I aim to collect about 100 participants per condition, although I modify this approach depending on the predicted size of the effect. For studies with the average effect size in social psychology (d = .4), this is what is needed to achieve significant results 80% of the time. Some of my studies test much larger effects, but as a general rule it can be difficult to predict ahead of time just how large an effect will be. 

bottom of page