dr.carroll@optonline.net

Steps in hypothesis testing for dissertation samples



 

There are steps in hypothesis testing for your dissertation sample that you need to follow:

1. State the statistical hypothesis in the null form. The null can test for either statistical differences or statistical relationships for your dissertation sample. If for statistical differences, the null is either non-directional or directional, but you must be aware of which type you are using.

2. Select your level of statistical significance or level of probability either .05 or .01. .05 establishes a 95% confidence level and is more liberal statistically . .01 establishes a 99% confidence level and is more conservative statistically.

3. Compute your statistical analysis. Determine whether you have a statistically significant result

No statistically significant result: Accept your null as true.

Yes, a statistically significant result: Reject your null as false.

4. Determine the statistical significance of your results. Is the statistical difference meaningful? Or is this a "so what?" finding. Concerning the last step, don't let your ego overtake common sense when it comes to your statistical results. Look at the mean scores for your sample groups. Are the differences in the sample groups' mean scores large enough to warrant meaning? Or are they so infinitesimal that they lack practical significance?

Return from hypotheses testing to dissertation research questions and hypotheses.