Home » 7 most commonly asked questions on Correlation

7 most commonly asked questions on Correlation

• tvmanikandan says:

Tavish,
I tried the following in R. I get pearson correlation as 0.88 and not 0.25 as you explained in the post. Please clarify.

> x=c(1:5)
> y=exp(x)
> y
[1] 2.718282 7.389056 20.085537 54.598150 148.413159
> cor(x,y,method=”pearson”)
[1] 0.8862751
> cor(x,y,method=”spearman”)
[1] 1

• Tavish Srivastava says:

Try with
x <- c(1:100) you should get 0.25.

• Abhi says:

Very informative article! Thank you for writing this. If possible, could you also publish an article giving insights into various kinds of hypothesis testing techniques?

• Gaurav Kant Goel says:

Read this one and “difference between correlation and causation” back to back. My concepts on corelation have been solidified. Thanks for writing such informative articles.

• kalyanischakravarthi says:

Falling in love with “AV” . You guys are rocking.

• Nimesh Jha says:

Excellent Article . Thanks for writing in. it helped a lot.

• dunk says:

What’s about the sample size ?
Is there any rule of thumb regarding the sample size to measure correlation? I believe that different sample size can result in a different direction of correlation eg. N=100 correlation >0 but when N = 200 correlation <0

• M.Silambu says:

How to apply Correlation in retail business domain.could you please list some of the x and y values for retail business .Eg.total sales vs day of the week like that.