Correlation: dissertation research questions about relationships
In addition to differences, research questions are often focused on relationships. Correlations help you to determine if a relationship exists, and if so, what the characteristics of that relationship are. They test the extent to which two variables occur together and how related they are. Correlations can be descriptive and inferential at the same time. They can describe your data and you can also infer relationships from samples to populations.
For a correlation, you need two sets of variables (or paired observations) on the same individuals. The first variable is called x. The second variable is called y. Your data are paired observations of x and y on one person.
There are basically four conceptual areas that correlational procedures address.
The relationship between two variables, and the nature of that relationship, is measured by a correlation coefficient, symbolized by the letter r. The numerical values can range from minus one (–1.00) through zero (0) to plus one (+1.00).
How do you tell the Strength of the correlation?
The actual numeric value of the correlation coefficient tells us the strength of the relationship. The nearer the number is to either +1.00 or –1.00, the stronger the relationship is between the two variables Although there is no hard and firm interpretation of what constitutes strength, here are some correlation coefficients with interpretation of strength suggested.
How do you interpret the Direction of the correlation?
Relationships between two variables can either be positive or negative. That is what is meant by “direction”. Therefore, correlation coefficients can be either positive or negative. A plus or a minus sign before the numeric value indicates direction.
If a correlation is positive, it means that:
* if one variable (x) increases, the other (y) increases, or
* if one variable decreases (x), the other (y) decreases.
If a correlation is negative, it means that:
* if one variable (x) increases, the other (y) decreases, or
* if one variable (x) decreases the other (y) increases.
Return from correlation to statistical tests.