Statistical Tables for Testing Data Closer to Expectation than Compatible with Random Sampling

Abstract

Background: A p-value of >0.95 literally means that we have >95 per cent chance of finding a result less close to expectations, and, consequently, <5 per cent chance of finding a result this close or closer. Using the traditional 5 per cent decision level, this would mean that we have a strong argument that such data are not completely random.

Objective and methods: To facilitate the assessment of this issue. T-, chi-square-, and f-tables of unusually large p-values are given to calculate a posteriori p-values of study results closely matching their prior expectations. Simulated examples are given.

Conclusions: Clinical trial data producing large p-values may not be completely random. The current paper is a preliminary effort to assess randomness of clinical trial data

Key words: :

• clinical trials
• random sampling
• data cleaning
• p-values
• statistical tables
• evidence-based medicine