Friday 23 October 2009

Correlating friends

It probably won’t surprise you hear that I have a spreadsheet that I update on a weekly basis that details my rating of each friend out of 10. It's quite an arbitrary score and can be based on more or less anything.

Not replying to an e-mail or missing your round of drinks in the pub may get you a 6 rather than a 7. Or ironing my work shirts on a Sunday night may get you a 10 rather than a 9…

I’ve been keeping it up to date for nearly four years now so have amassed a fair amount of data so when bored at work yesterday I decided to attach a correlation matrix off the back of it to see if different friends (or my ratings of friends) correlate with each other.

It’s interesting to see that certain clusters of friends do correlate, it must have something to do with me being happy (or not) with that group as a whole at any given time. Interestingly certain groups of friends also negatively correlate against each other. I guess if one group is moving down the scale I make more time and effort with another so they raise in my expectations.

All very interesting, or at least useful as I found a new technique for excel where I use an indirect expression so that I can use named ranges in the correlation function.

If you want to see YOUR average score over the last 4 years let me know…

No comments:

Post a Comment