Question

2. a. What two conditions for linear regression are violated, based on the residual by predicted plot at right? For each of the two conditions, briefly explain what aspects of the pattern show that a violation occurred. OOO PE 10 15 20 25 Predicted Value Residual Normal Quantile Plot 500 400 b. What condition for linear regression is violated based on the residual normal quantile plot shown at right? Briefly explain your reasoning Income Normal Quantile

Answer 1

2. a) what to conditions for linear regression are violated, base on the residual by predicated plot at right? For each of the two conditions, briefly explain what aspects of the pattern show that of violation occurred....

=> Solution -

Given data :-

a) The two conditions that are violated in the given plot are -

i) linearity of the response predictor relationship

ii) Homoscedasticity ( constant variation of error terms)

=> i) The liner rati of the response predictor :

The relationship between the response and the predictor variable is non- linear, the residuals exhibit a clear U- shape, which provides a strong indication of non-lineraty in the data.

ii) Homoscedasticity ( constant variance of the error terms) :

Violation of the above is term as heteroscedasticity. Non constant very veriances in the errors,or heteroscedasticity, can be interpreted from the presence of a funnel shape in the residual plot.

b) From the plot in b) path we can say that the the residual are not normally distributed.