In this chapter, first of all, the difference between effect estimation and statistical testing is discussed and it is shown that effect estimation is more important than statistical testing. Furthermore, It is shown that probabilities and probability distributions play a crucial role in the estimation of the confidence intervals around effect estimates and in statistical testing. Furthermore, the t-distribution is introduced as an alternative for the standard normal distribution. Besides that, it is also shown that the binomial probability distribution can be used for dichotomous variables.