RobertNiles.com
About Robert
Reporting Help
Finding Data on the Internet
Questions for Candidates
Stats Lessons
Mean
Median
Percent
Per capita
Standard Deviation
Margin of Error
Data Analysis
Sample Sizes
Stats Tests
Help Board
Bookstore


A Mathematician Reads the Newspaper

Read about reporters who blew it when trying to write about stats. A mathematician leads you through the tricks, citing contemporary examples.
More information
-->

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

linear correlation
Message posted by sanja on June 16, 2000 at 12:00 AM (ET)

I need several examples of variables in which there is a significant linear correlation either positive,negative, or both. Discuss the relationship between the variables and what makes them correlate to such a degree ?


READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: linear correlation
Message posted by Stuart Jones on June 23, 2000 at 12:00 AM (ET)

I need several examples of variables in which there is a significant linear correlation either positive,negative, or both. Discuss the relationship between the variables and what makes them correlate to such a degree ?

1st things 1st. You can only have either a positive correlation, a negative correlation, or no correlation at all.

An obvious linear correlation is that of height and weight. It stands to reason that the taller you are, then, on average, the more you will weigh. Therefore the correlation is positive. If you do a scatter graph of your friend's weights against heights, this should be confirmed. Basically what a positive correlation says is "As one set of data eg height increases, then the other set of data increases too"

An example of negative correlation is this. Suppose we had a pest control problem. It stands to reason that the more poison we put down, the more pests we kill. So if we plotted on a graph Amount of Poison (y-axis) aganst Pest population, we would expect to see that the more poison we use the smaller the pest population will become.
Basically what a negative correlation says is "As one set of data eg Amount of poison increases, then the other set of data decreases too"

In statistics we use what is termed a "correlation coefficient" to measure the relationship between the data. This number is always between the numbers of -1 and 1. 1 is a perfect positive correlation and -1 is a perfect negative correlation. If you get a correlation coefficient of 0 this means that there is no relationship between the 2 sets of data. Statisticians often use hypothesis testing to test whether there are relationships between the 2 different types of data.


Re: linear correlation
Message posted by k banna on June 26, 2000 at 12:00 AM (ET)

i stole this from an old stat professor, but i think it is a good illustration of the old "correlation does not imply causation" rule... two things that are positively correlated are number of cavities and reading level... students will try to come up with a hundred reasons about how tooth decay can lead to increased reading skills.... however the likely reason that the two are correlated is due to a third factor: age. the older you are, the more cavities you are likely to have, and the more reading skills you (hopefully) have...



Your $5 contribution helps cover part the $500 annual cost of keeping this site online.

Niles Online Sites:RobertNiles.comTheme Park InsiderViolinist.com

RobertNiles.com™, the site, content and services 咀opyright 1996-2002, Robert Niles.
All rights reserved. Questions? Comments? Read my Privacy Policy, or E-mail me!