1.  Remember: sampling distributions are theoretical.  That means all the probabilities we obtain from the tables or the computer are theoretical. This theory doesn’t hold if  sample doesn’t represent the population. 

2. You can often assess whether a difference is important based on your substantive expertise.  However “eyeballing” a relationship can often result in inaccurate conclusions because whether or not a difference or a relationship exists depends on not just the mean difference but the sample size and the variation within the variables. 

3.  Always calculate descriptive statistics on key demographic variables in your sample before doing any inferential analysis so that you can determine if the sample represents the population, and if the distribution of your variables matches the distributional assumptions of the inferential statistics you plan to compute.