... confidence). For example, we can consider 90%, 95%, 97%, 99% and so on. If confidence level (degree of confidence) is the probability that the confidence interval actually does contain the population parameter, why not use 99% degree of confidence all the time in which gives us more confidence about our results?