When should i reject a null hypothesis?

Answer 1

When we state that the data analysis suggests that we "Reject the null hypothesis" we are stating that the sample statistic is sufficiently different from our assumed value of the population that it is unlikely to be explained by chance. If we use for example, that under the null hypothesis that engineers make on the average $120,000 per year. If we consider that the test statistic (size n) is normally distributed, we can use a two-tail test with an level of significance "alpha" to identify the lower and upper rejection zones on the normal distributon. If the test statistic falls in the non-rejection zone, we state that the "null hypothesis is not rejected." There are many good websites on hypothesis testing. Wikipedia provides a good summary of controversy on hypothesis testing. I note that some of the controversy stems from the idea that hypothesis testing will prove or validate population parameters, which is really beyond the scope of hypothesis testing theory. http://en.wikipedia.org/wiki/Statistical_hypothesis_testing A second way to determine whether the null hypotheis is to calculate p-values. For this, please see: http://en.wikipedia.org/wiki/P-value