# check if X variable influences Y or Y influences X

##### New Member
Hello all helpers,

I am not a statistics pro but, I wanted to know if it is OK to use correlation to check the connection between X and Y and then to check the correlation between Y and X and to see which of the statistics is closer to 1.

#### hlsmith

##### Not a robit
If you are trying to figure out which variable predicts the other, this is an impossible task to achieve using statistics.

##### New Member
but what about known two variables: 1st is the number of comments and 2nd is the number of people entering my site.
I want to figure out if: the number of comments influence the number of people coming to my site OR if the number of people on site influencing the number of comments on my post on LinkedIn.

##### New Member
I saw in this discussion:
https://stats.stackexchange.com/que...ween-correlation-and-simple-linear-regression

* Correlation is almost always used when you measure both variables. It rarely is appropriate when one variable is something you experimentally manipulate. With linear regression, the X variable is usually something you experimentally manipulate (time, concentration...) and the Y variable is something you measure.

#### Dason

If you're just looking at the two variables and not lagging them or something then the correlation between X and Y is the same as the correlation between Y and X so... Trying to compare those correlations won't be very useful.

##### New Member
If you're just looking at the two variables and not lagging them or something then the correlation between X and Y is the same as the correlation between Y and X so... Trying to compare those correlations won't be very useful.

#### Dason

Without doing something else I don't think linear regression is going to get you anything more than correlation would to answer your question. You might be interested in Granger causality testing though.

##### New Member
Without doing something else I don't think linear regression is going to get you anything more than correlation would to answer your question. You might be interested in Granger causality testing though.
thanks! sorry for the ignorance ) (didn't know this testing existed)