Mohammad,

Pearson's correlation coefficient is usually used in scientific research (See: https://en.wikipedia.org/wiki/Pearson_correlation_coefficient).

I believe you should calculate the correlation only for the exposed group, not your control - as you're trying to determine the relationship between X (the exposed group) and Y.

Dear Armin,

Thank you a lot. I do use Pearson's coefficient with my data, but partial or bivariate??!! Which type should I trust and so write in my paper?

"The bivariate Pearson Correlation is commonly used to measure the following:

- Correlations among pairs of variables

- Correlations within and between sets of variables

The bivariate Pearson correlation indicates the following:

-Whether a statistically significant linear relationship exists between two continuous variables

-The strength of a linear relationship (i.e., how close the relationship is to being a perfectly straight line)

-The direction of a linear relationship (increasing or decreasing)

Note: The bivariate Pearson Correlation cannot address non-linear relationships or relationships among categorical variables. If you wish to understand relationships that involve categorical variables and/or non-linear relationships, you will need to choose another measure of association.

Note: The bivariate Pearson Correlation only reveals associations among continuous variables. The bivariate Pearson Correlation does not provide any inferences about causation, no matter how large the correlation coefficient is."

See: http://libguides.library.kent.edu/SPSS/PearsonCorr

Ultimately your analyses should depend upon your hypotheses.  But, if you collected continuous variables X and Y in an experimental (“exposed”) and control group, here are the analyses you might reasonably do.  First, you could do a split-biserial correlation of control (0) and experimental (1) to predict X (and again for Y).  This would tell you if X or Y is different across the conditions.  You might do t-tests instead, which would be more intuitive, but technically they’ll give you the same p-value.  Second, you can do the (bivariate) correlations between X and Y within each condition separately.  This will tell you if X and Y are related within conditions.  Finally, you could test if the correlations between X and Y are different for your two conditions with a Fisher z-score comparison (link for online calculator below).  Partial correlations are useful for different circumstance, when you might have variables X, Y, and Z, and you would like to know if X and Y remain correlated even after you control for Z.  Best wishes with your thesis Mohammad! ~ Kevin

http://www.danielsoper.com/statcalc/calculator.aspx?id=104

The difference between bivariate correlation and partial correlation is that bivariate correlation is used to obtain correlation coefficients, basically describing the measure of the relationship between two linear variables, while partial correlation is used to obtain correlation coefficients after controlling for one or more variables.

I think you should just do a bivariate correlation between x and y :)

Hello Jamil

My colleagues have already ably explained to you what the two statistical tests are, and what they are meant to be used for. I now want to advise that the choice of whether to use bi-variate or partial correlation measures on a study result you want to publish should really depend on what you want to write about. For instance, if you were interested in reporting on the linear relationship between two variables which you have investigated, a bi-variate correlation test would do. However, if you investigated the correlation between two variables however you have controlled other variables in order to enable you get the relationships between these two given variables, then, you would need to consider using the partial correlation test. Can that do for now?

Hi,