We presented each subject with a series of four tests: neophobia,
innovation, social learning, and priority of access
to food. We chose to keep the order of tests constant (as
in Whittle 1996; Seferta et al. 2001; Webster and Lefebvre
2001) because a gradual habituation to novel stimuli
and captivity as testing proceeds could potentially mask any
association between social learning and innovation if the order
of these tests were counterbalanced or randomised. For
example, if gradual habituation occurred, birds tested first
on innovation would do poorly on that test, and better on
the social learning test; birds given the tests in the opposite
order would show the reverse pattern, creating a spurious
negative correlation between social learning and innovation
(Beauchamp personal communication).