Correlation analysis is used to identify how closely related two variables are to each other. A numeric value ranging from -1 to +1 indicates if the correlation between the two variables is positive or negative and the strength of the relationship. The closer the correlation is to negative or positive 1 the stronger the relationship. A correlation of zero would indicate that no relationship exists.
One of the more common uses of correlation in marketing research is customer satisfaction studies.
Example: A multi-level marketing company selling home décor products and accessories conducted a survey of its in-home demonstrators regarding their level of satisfaction with several key factors related to their experience with the company.
Q: Using the following 1 to 7 scale, where 1 means “totally disagree” and 7 means “totally agree,” please indicate your level of agreement with the following statements:
The call center answers my questions
The demonstrator support website is easy to use
The training I have received is effective
I am supported by my upline
I am satisfied with the variety of products available for me to offer
Products are shipped in a timely manner
I feel like [company name] cares about me personally
[Company name] is responsive to problems
A question related to overall satisfaction is also included using the same scale.
Q: Using the following 0 to 7 scale, where 0 means “totally dissatisfied” and 7 means “totally satisfied,” please rate your overall satisfaction with [company name].
Running a correlation analysis between overall satisfaction and each of the other variables will allow us to measure the strength of the relation each variable has to overall satisfaction, and thus identify which ones appear to have the greatest impact/influence on overall satisfaction.
If we were to plot the results of the correlation between two variables it may look something like this:
As you can see there is a degree of linearity between the two. In this case the correlation is .63, indicating a moderately strong relationship.
A simple correlation analysis can be run using an Excel spreadsheet. However, a more advanced statistical analysis program like SPSS will provide more complete data. It will also produce a “Sig.” value to indicate statistical significance.
It is important to realize that correlation does not establish cause and effect. Correlations are often used erroneously to state a relationship. A correlation may suggest, but not imply a causal linkage between the variables. Correlations that are real, but obviously not related are known as spurious correlations.
Partial Correlation: A partial correlation coefficient measures the association between two variables after controlling for the effects of one or more additional variables. For example, we may want to measure the correlation between sales of a product and the amount spent on advertising that product while controlling for price.