Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Real life situation. I have data from different cities saying on average how fast they build a home. I want to correlate this to the amount of construction activity they do & number of homes constructed per construction personnel. I also have data saying total # of construction personnel & total construction cost in respective cities.

I have ran all this in JMP & Excel and am getting a correlation of up to .20, and if I take out outliers I have correlations up to .60.

My question is: how can I get correlation up to .75 or above, if possible?

I am also trying to work with multivariate regression and see if I can get better results.

share|improve this question
Keep taking out the "outliers." Mathematics guarantees that by the time you have removed all but two cases, you will reach a correlation of $100\%$ :-). I hope it's clear I'm being facetious, but the concern about the lack of validity of this procedure is real: please see the recent discussion (about essentially the same issue) at – whuber Feb 21 '14 at 0:28
I understand I can take out outliers, to get me better results. I have used SAS to do a regression analysis to see with city if I remove would give me a better correlation. But I actaully wanted to somehow (Mathematiclly or Statisticlly) get better correlation without taking out outliers. Keep in mind this is a real world analysis and taking out outliers would not be ideal. I do understand that in real world analysis, if we get even 50% correlation it means great things. – user40675 Feb 21 '14 at 16:40

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.