ABSTRACT
The population to which significance tests apply is the population from which the random sample was drawn or the random assignment made. The probability indicated in a significance level is auxiliary probability, the relative frequency with which a given or larger difference or relationship would occur on repeated random generation from the population involved if the null hypothesis in question were true for that population. In contrast is intrinsic probability, the relationship between variables expressed in probabilistic terms. There is a basis for choosing a certain level of significance only when a decision about a hypothesis is required, as in the situation in which action will or will not be taken according to the outcome of the test. In the vast majority of instances in which significance tests are reported in behavioral research no firm decision is required because no specific actions are to be guided by the decisions.
