Hi everyone,

Lets assume Ho: There is no relationship between the two factors and Ha: there is a relatioship exsiting. And as we all know that Coefficient of correlation quantifies the relationship between the 2 variables/ factors with -1 suggesting a negative correlation and +1 for strong correlation.

Now my question to the team: Is it possible to get a p value > 0.05 with coefficient of correlation coming negative (between -1 to 0)?

Take care,

Shalini

Hello Shalini, 

Correlation coefficient values between -0.8 and +0.8 is considered as poor correlation. This means, if you get correlation coefficient lesser than -0.8 or greater than +0.8, it shows good correlation.

If you have got a correlation coefficient value of -0.3, it means weak negative correlation which is not worth consideration in terms of impact of one factor on another. 

I hope this answers your question. 

Best Regards, 

VK

Dear VK,

Thanks for bringing in clarity on "r".

As VB is suggesting, can we overlook p value? Rather my openion is that it would and should go hand in hand with "r" value. Like as you said if "r" =-0.3 and p=0.09 both are driving home the same point (as clarified in your post on "r" lying between -0.8 to +0.8).

Request your view.

 

 

Dear Shalini,

This subject is more complex than it looks and good that you picked it up.

Are the factors independent? Correlation assumes that any random factor affects only one other factor, and not others.

Are X and Y measured independently? The correlation is not valid if X and Y are intertwined. You'd violate this assumption if you correlate midterm exam scores with overall course score, as the midterm score is one of the components of the overall score.

Were X values measured (not controlled)? If you controlled X values, you should calculate linear regression rather than correlation.

Is the covariation linear? A correlation analysis would not be helpful if Y increases as X increases up to a point, and then Y decreases as X increases further. You might obtain a low value of r even though the two variables are strongly related. The correlation coefficient quantifies linear covariation only.

It is important here to have a look at the four graphs shown on this page at wikipedia. http://en.wikipedia.org/wiki/Correlation_and_dependence

To be more specific, if X and Y are independent, and both are just being measured, we use correlation to understand the linearity of relation between them. Please note linearity is not perfectly understood from correlation coefficient as one outlier can provide misleading interpretation if we only look at r. It is important to look at the scatter diagram. (as you see in wikipedia link above)

If X values are controlled and Y values are measured, we better use regression and p values for inferences. This provides option of considering non linear equations (like quadratic or cubic etc). In all cases, it pays to have a look at the graph.

Best Regards,

VK

Thanks once again VK,

lot of things to look at, really! its not that easy as it was looking.