



 
You should decide on your procedure, including the rejection region, BEFORE YOU COLLECT THE DATA. Specifically, if when you set up the test, your decision was to reject the null hypothesis when the test statistic is equal to or less than the critical value, then that is what you do. Otherwise, probability has little meaning. Your question, however, points out one weakness of this way of reporting the outcome of an experiment. It is considerably more informative to report the pvalue of the experiment, and to provide an appropriate interpretation. In your case you should probably report that there is evidence for the alternative hypotheses, but more data is required for greater certainty. You might want to check out what Wikipedia has to say about the pvalue: http://en.wikipedia.org/wiki/Pvalue. Chris  


Math Central is supported by the University of Regina and The Pacific Institute for the Mathematical Sciences. 