Statistical significance of size measurements.

P values are from a Student's t test with the null hypothesis that the mutant (or temperature) is the same as the N2 [20°C] control, except for daf-2(e979) where N2 [15°C] is the control, and they are Bonferroni adjusted. Significant (at p<0.01) values are in bold. Values that were adjusted to greater than 1 were set to 1. Sample skewness and excess sample kurtosis are reported for volume. In general, these skewness and kurtosis values justify the normality assumption. daf-2(e979) at 48 hours is exceptional, presumably due to the fact that this mutant is undergoing dauer formation at this time, which includes radical change in shape (Figure 7). Nevertheless, the difference in volume for daf-2(e979) at 48 hours is qualitatively compelling and significant (p<2.2e-16) using Kruskall-Wallis (a non-parametric test).