Changing the way you learn

Question	Answer
Heteroscedastic	Fans out around the linear. Increasing variance with Y axis for example
Image: Scat_Strong_Linear (image/png)	Strong, linear, positive relationship (y increases as x increases). Linear because we could imagine a straight line passing through the points
Image: Scat_Weak_Linear (image/png)	Weak, linear, positive relationship (y tends to increases with x )
Image: Scat_Weak_Linear_neg (image/png)	Weak, linear, negative relationship (y tends to decrease with x )
Image: Scat_Strong_Linear_neg (image/png)	Strong, linear, negative relationship (y decreases with x )
Image: Scat_little_No_Rel (image/png)	Little or No relationship
Image: Scat_Strong_Non_Linear_Poss_Quadratic (image/png)	Strong relationship, but non-linear, possibly quadratic.
Image: Scat_Strong_Poss_Outlier (image/png)	Strong linear positive relationship, but with an obvious unusual point known as an outlier. Little variance (Heteroscedastic)
Image: Scat_Strong_Poss_Linear_Outlier_Influ (image/png)	Strong linear positive with Outliner on line of best fit (Influential point)
Image: Scat_Linear_Y_increase_X_Fanspng (image/png)	Some sort of linear relationship as Y tends to increase with X, but the variation in Y increases with x, giving a fan out effect
Image: Scat_rounded_distrete (image/png)	Variance of Y decreases with X (Heteroscadastic). X is rounded or discrete. A few outliers are present. Possibly has a downward trend

Scatter Plot Shapes

	Created by Steve Hiscock almost 12 years ago