If you have a new account but are having problems posting or verifying your account, please email us on [email protected] for help. Thanks :)
Hello All, This is just a friendly reminder to read the Forum Charter where you wish to post before posting in it. :)
Hi all, The AutoSave Draft feature is now disabled across the site. The decision to disable the feature was made via a poll last year. The delay in putting it in place was due to a bug/update issue. This should serve as a reminder to manually save your drafts if you wish to keep them. Thanks, The Boards Team.
Hello all! This is just a quick reminder to ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere.

Average Correlation Coefficient

  • 22-09-2021 11:02am
    Registered Users Posts: 43,605 ✭✭✭✭ Mitch Connor

    Hey, hoping someone can help me here.... i've tried researching this online but I am lost (and i have basically no background in statistical analysis).

    I have a data set that is made up of admissions per day, per site.

    I have created an overall trend for the sites by adding the admissions for each site for each day.

    I have then found the correlation of each site against that totaled admissions trend.

    What I want to do is then find the average correlation of all the sites, against the total trend.

    What I did was simply add the correlation of each site (for a day) and then devide by the count, but I have been told that because correlation is not additive, this isn't the correct thing to do. But I am not sure it is 'wrong' enough for it to be ignored.

    Is there a relatively easy way of finding the average correlation of the group of sites to the group trend? Does the fact I am finding the correlation between the same variable make a difference (comparing admissions to admissions)

    Ie. I want to find the average correlation of the sites for 2021-09-06 - so I simply added the figures together and divided by 28 (count of sites). I am lead to believe 0.85 is not correct as the average correlation, but I don't know what the correct method would be.