hypothesis testing

Share This

“The statistical process of evaluating data results. A null hypothesis is defined for a neutral state (such as a coefficient equal to zero). An error rate is specified for a Type I error (the probability of rejecting the null hypothesis if it is true)—often set at 5 or 1 percent. The test statistic is defined. When the experiment is conducted, the null hypothesis is rejected if the observations are improbable given the null hypothesis.”

« Back to Glossary Index Download Tooltip Pro
By |February 1st, 2019|Comments Off on hypothesis testing