Our motivated account of class mobility estimates suggests that
overestimates of class mobility are beyond simple errors in statistical
probability calculation. To more directly test this prediction, we examined
mobility estimates in comparison to estimates for our non-mobility
statistics related to soccer players and astronauts. For this analysis,
we conducted a repeated measures analysis of variance (ANOVA)
with non-mobility estimates, general class mobility estimates, and
self-relevant class mobility estimates as three levels of a single factor.
The overall analysis was significant F(2,1456)=275.38, p b .01, and revealed
significant differences between all three groups that aligned with
our predictions (see Fig. 2, three light gray bars). Specifically, though
participants overestimated non-mobility statistics (M = 7.82) they
were more likely to overestimate social class mobility (M = 19.26),
and in particular, when these estimates were described as relevant to
the self (M = 24.30).