Correlation studies attempt to show a relationship between two or more variables. For example, there are many questions that one can ask in the social AND health sciences that touch upon the possibility of a relationship existing between two variables and the nature of that relationship if it exists. Some examples of the kind of questions that seek this type of information are:
Is there a relationship between smoking and any number of ailments such as heart disease, lung cancer, high blood pressure and ulcers?
Are persons who are exposed to domestic violence as children more likely than those who aren't to grow up to become abusive to family members are?
Are residents of certain regions of the United States more at risk for skin cancer?
Not only can we identify whether or not the relationship exists between the variable, by understanding how to interpret a correlation coefficient we can also learn something about the direction and magnitude of the relationship. Both extremes represent perfect (perfect direct and perfect inverse) relationships between the variables. For example, the finding that r = 0 represents the absence of a linear relationship. In other words, we don't worry about our r being LOW (close to 1.00) we worry about it being close to 0 on either the positive or negative side.
A high correlation does not in and of itself establish a causal link (the old example of people's weight and their vocabulary - the more a person weighs, in general, the bigger his or her vocabulary is. If you plot this on a scatter-plot, and do a correlational statistical procedure, you will get a strong positive "r". But it's not the weight that makes the vocabulary go up, is it? The more someone weighs, generally the older they are. It's not the weight that drives the vocabulary up, it's AGE.)
Can't decide on a topic? Need ideas or inspiration for research in your field of study? Paper Masters provides topic suggestions to make your decision easier.