co efficient of correlation

if line is fit in positive slope then we have correlation coefficient +1 and if the line fit in negative slope we have correlation coefficient -1 and zero would be the coefficient of correlation if two lines are perpendicular to each other.i would like to have figure of different way of regression line giving different correlation co-efficient

bhuvaneshnick said:

if line is fit in positive slope then we have correlation coefficient +1 and if the line fit in negative slope we have correlation coefficient -1 and zero would be the coefficient of correlation if two lines are perpendicular to each other.i would like to have figure of different way of regression line giving different correlation co-efficient

Click to expand...

Rather than say 'if line is fit in positive slope then we have correlation coefficient +1' is is more appropriate to say 'if line is fit in positive slope then we have correlation coefficient is positive'. It is possible to have the slope of a linear regression line be positive without the correlation coefficient being plus one.

The sign of the slope does not indicate how good the fit is, it only says how one variable changes with another. For example, suppose you were to buy some apples to fill up your bag. The more apples you bought the fuller your bag would become (a positive correlation) but the more apples you bought, the less money you would have (a negative correlation).

You might enjoy looking through
http://www.stat.purdue.edu/~xuanyaoh/stat350/xyApr6Lec26.pdf