REMEMBER:
Research design = overall plan for a study.
The 2 major categories of research study design are:
- Non-experimental, observation-only studies, &
- Experimental testing of an intervention studies.
Correlation study designs are in that first category. Correlation studies focus on whether changes in at least one variable are statistically related to changes in another. In other words, do two or more variables change at the same time.
Such studies do not test whether one variable causes change in the other. Instead they are analogous to the chicken-and-egg dilemma in which one can confirm that the number of chickens and eggs are related to each other, but no one can say which came first or which caused the other. Correlation study questions may take this form, “Is there a relationship between changes in [variable x] and changes in [variable y]?” while a correlation hypothesis might be a prediction that, “As [variable x] increases, [variable y] decreases.”
An example of a question appropriate to this design is, “Are nurses’ age and educational levels related to their professional quality of life?” Sometimes a yet-unidentified, mediating variable may be creating the changes in one or all correlated variables. For example, rising nurse age and education may make them likely to choose certain work settings with high professional quality of life; this means the mediating variable of work setting—not age or education—might be creating a particular professional quality of life.
| Alert! Correlation is not causation. |
Although “to be simple is difficult,” that’s my goal. L
et me know what’s difficult for you in research, because it probably is for others as well. Let’s work on the difficult together so that you can use the BEST Evidence in your practice.
Discovering Your Inner Scientist 