What is a Correlation? How is the Strength of a Correlation Determined?
Correlation is a term used to describe the relationship of two variables, where one has an effect on the other. A negative correlation describes the relationship between the two variables, where one variable will increase, and the other deceases. Similarly, a positive correlation is where one variable affects the other, but both will move in a positive direction.
A linear correlation describes how variables plotted on a scatter diagram follow a pattern that is a straight line, and a curvilinear correlation shows a pattern too, although it is not linear. A curvilinear correlation is a pattern of curvature, as the word suggests.
It is important to identify correlations, positive or negative, so we can see how much, or possibly why one variable influences the other. Sometimes there will be no correlation whatsoever between two variables, so other tests or experiments to discover answers may be necessary.
For example, there may be no correlation at all between two variables such as Autistic children learning to read new words, and his or her mood on the day those words were learned. However, it could be that mood has everything to do with learning and memory for the child, so a positive correlation would show a pattern suggesting this is indeed, the case. Correlation is about cause and effect, or the lack of it.
How is the strength of a correlation determined?
Correlations can be strong whether they are negative correlations, or positive correlations. When determining the strength of the correlation, the + or – sign is not to be considered. Zero (0) shows that there is no correlation at all. It is important to determine the strength of correlation so that if the relationship between two variables is found to be low, other questions can be asked about why. Sometimes there are outside influences (a third variable) that are affecting the strength of correlation. This cannot always be pinpointed, but a strong correlation may suggest that there is no interference from a third variable, and that one variable is having a strong influence on the other regardless of other possible contributing factors.
Studies can be done to develop new treatment plans in the field of psychology if we use correlation studies. For example, the incidence of panic attacks is high for people who consume caffeine. This is a strong correlation, and cutting out coffee is one step in a treatment plan to help people cope with panic attacks.