Justin Barrett's "Born Believers" is divided into two parts. The first part considers the evidence and the second the implications.
The question is how naturally are children able to identify "agents"? Agents are anything that can intend to do something. Non-agents just sit there and wait for an agent to act on them. These agents can be human, animal or non-material like an imaginary friend. The first evidence is to establish how easy a child can distinguish between an agent and a non-agent.
One of the methods to get information is through "preferential looking" originated by Robert Franz about 50 years ago: https://en.wikipedia.org/wiki/Preferential_looking This evidence measures how long a child pays attention to a set of stimuli.
One of the problems for this research is that it is not clear what is being studied. Do we have innate nature? Can everything be reduced to the environment? Does reduction make sense at all? Elizabeth S. Spelke and Katherine D. Kinzler describe what is being studied as a "core knowledge" which doesn't fit into the previous categories: http://www.psy.cmu.edu/~siegler/710-Spelke2007.pdf It is probably good that one gets past old metaphors. Barrett wants to leave the physical "hardwiring" to the electricians. Also a reductionism to what a culture tells children to believe is questioned because children don't believe everything they are told.
I am currently looking at the concept of imaginary friends and Bradley Wigger is one of the researchers mentioned. That is where I have stopped.
When I read scientific surveys I try to identify whether the author has a bias. Barrett seems to have a theistic bias. So his evidence has to get past this bias. Spelke seems to have a feminist bias. There is nothing wrong with these biases. They motivate research, but one needs to make sure the evidence justifies the bias and not the other way around. Although Barrett is aware that he is writing a popular presentation his audience is mainly professionals. He wants to motivate the way they see their research so further questions can be asked that help answer questions Barrett is interested in. In a footnote to his introduction Barrett writes (p260):
I am the first to concede that alternative glosses of the data to date may be reasonable and new data may change the state of the art as presented here. Nevertheless, for clarity and to put in sharp relief where the critical issues for future scientific research lie, I present a strong version of the thesis that children are naturally born believers.
My view is that all science surveys are literature so I have no problem with that.