Question

1. What does “t-distribution have thicker ends” mean?

2. Why do we divide by n-1, but not n in empirical variance?

3. What is the level of significance?

Answer 1

1. t-distribution has thicker ends meaning that probability that t distribution has extreme values or outliers is relatively higher as compared to normal distribution.

2. Dividing by n-1 reduces the bias. Since empirical variance is a sample statistic, for the sample since there is dependency between the units within a sample, knowing only n-1 units help us get the last value. This is possible due to the condition that the sum of differences between sample units and sample mean is zero.

3. Level of significance is probability of rejecting null hypothesis under the condition of true null.