I have at least one main dependent (outcome) variable and at least one independent (explanatory) variable I think is affecting it

For example: For the research question 'Do women drink more coffee than men?', we think gender (the independent) is affecting the amount of coffee drunk (the dependent). In the case of assessing whether a diet is effective, we may measure weight before and after the diet. Here weight is the dependent variable and time (before/after) is the independent.

I want to explore or describe associations or patterns between variables or subjects

Choose this option if you have no clear outcome variable. For example, if you wanted to see whether two variables are generally related, but neither is thought to predict the other. Other situations include reducing a large number of variables to a few and grouping variables or cases.

I wish to test reliability or equality of observers, methods or scales

Choose this option if you are assessing agreement between observers or instruments. For example: Assessing whether two observers using the same instrument give the same scores or whether a questionnaire scale is reliably constructed.

I only have one variable of interest

Choose this option if you're seeing whether your data fits a certain distribution or if you are comparing one statistic with a population value.