Running head: CHILDREN'S ATTRIBUTIONS OF BELIEFS



Children's Attributions 10

example, graph 4 can be seen as a plausible representation of people living in a society where the
concept of God as omniscient is not very widespread.

Finally, moving on to the last graph in Figure 1, a non-similarity perspective would
predict that children being tested on the false-belief task would start differentiating between
humans and God very early in development. This is the position that Justin Barrett and
collaborators have been advocating (Barrett et al., 2001; Richert & Barrett, 2003; see also Atran,
2002). Their main idea is that young children do not need to conceptualize human agency first
and then use it as a basis to understand supernatural agency; rather young children have already
the potential to think independently about different types of agents and reason accordingly. In
graph 5, the God line remains close to floor level, which signifies that children from an early age
attribute mostly true beliefs to God - i.e., that God knows that there are rocks in the box. The
human line, on the other hand, starts at the same level as the God line but then by the age of 4
steeply climbs - children increasingly say that humans believe that the box contains crackers, as
their capacity to attribute false beliefs improves.

Experimental data from the United States supports the prediction of this non-similarity
position (Barrett et al., 2001). In figure 2, the results of a surprising contents experiment run with
a sample of American children recruited from Reformed and Lutheran Protestant churches are
presented. Children in the US sample can be seen to treat humans and God in the same way up to
age 4. By age 5, they already sharply differentiate between the two agents. The divergence
between God and the mother took place as children started to attribute false beliefs to the latter.
A Wilcoxon Signed-Ranks Test for matched pairs comparing “crackers” responses between
mother and God at each specific age detected significant differences only for 5- and 6-year-olds
(z = 2.37, p = .018, N = 17 and z = 2.93, p = .003, N = 9, respectively).



More intriguing information

1. The name is absent
2. ISSUES AND PROBLEMS OF IMMEDIATE CONCERN
3. On the Relation between Robust and Bayesian Decision Making
4. Education as a Moral Concept
5. Growth and Technological Leadership in US Industries: A Spatial Econometric Analysis at the State Level, 1963-1997
6. National urban policy responses in the European Union: Towards a European urban policy?
7. Palvelujen vienti ja kansainvälistyminen
8. ISSUES IN NONMARKET VALUATION AND POLICY APPLICATION: A RETROSPECTIVE GLANCE
9. The name is absent
10. The name is absent