-

Statistics in one of the most important instruments of researcher. So understanding the terminologies, statistical logics and methods for those who utilize researches is necessary.
More over there are limitations in application of statistical methods and inferences based on evidences which should be recognized thoroughly.
If we accept that true definition and description of the research hypothesis as given then one of the problematic position of the process of research is how to convert the research hypothesis into statistical hypothesis. Mistakes in this phase is one of the statistical pitfalls which misleads the researcher in choosing the appropriate statistical techniques for testing hypothesis and analyzing data.
In statistical literature frequent endeavors have been taken place for the conversion of research hypothesis into statistical hypothesis, hut the reality if that the current law (rule) is not that much easy and a vast knowledge of statistics is needed. Moreover this method from the stand point of mathematical logic has short-coming perhaps it can he said that one
of the best approach in this matter is a rule that is called “current rule”. In this approach it is held that the refutation of the research hypothesis is described as null hypothesis (H0) and the claim against it to he described as (H1) and then the (H0) to he tested. Althought this approach provides a correct and appropriate method , con t a ins mathematical problems and dues not always yield productively in behavioral researches.
The summary study of some one range and two—range statistical hypothesis crystallized this subtle point that if we consider (H)) (null hypothesis as criterion of test, necessarily the null hypothesis must have a limit. Otherwise testing the null hypothesis (Ho) would he impossible.
For this purpose a rule is recommended which is called “Proposed rule. Based on this rule the behavioral researcher considers the null hypothesis (if the research hypothesis has a limit) as an equivalent of research hypothesis, otherwise the opposited of research hypothesis can he termed as null hypothesis (H0). This algorithm which is very simple causes the behavioral researchers to carry out this very critical phase in a safe manner without having extra statistical information.
In this article different approaches in regard to behavioral researches have been outlined which can he classified as following:
1- descriptive approaches
2- comparative approaches
3- causal approaches.