To conduct a test of hypothesis with a small sample, we need to be able to make an assumption that: the region of acceptance will be wider than for large samples. the population is normally distributed. a larger computed value of t will be needed to reject the null hypothesis. the confidence interval will be wider than for large samples.