# Correlations and Regressions

?
• Created by: eleanor
• Created on: 11-04-15 14:15
• Correlation and Regression
• Correlation
• for positive correlation the points on the scatter diagram increase as you go from left to right.
• for a negative correlation the point on the scatter diagram decrease as you from left to right.
• r is a measure of linear relationship
• r=1 perfect positive linear correlation
• r= -1 perfect negative correlation
• r=0 no linear correlation
• Coding
• you can rewrite the variable x and y using coding
• r is not affect by coding
• Regression
• the equation of the regression line of y on x
• y = a +bx
• where b =Sxy / Sxx
• a =  ?y  -b ?x
• Coding
• to turn a coded regression line in to an actual regression line substitute the codes into the answer
• the purpose of the regression line is to make predictions about values of one variable for some given values of the other
• predictions should only be made in the range of the data
• extrapolaion are predictions made outside the range of data and are unreliable
• outliers do effect the regression line