A t-test comparing differences in means is more likely to be statistically significant when:
a. the variances in the sample are smaller
b. the sample sizes are smaller
c. the differences between observed and expected frequencies are larger
d. the differences in means is smaller