Complete Question:
Attached below as picture.
Answer:
From first graph there is no linear pattern so here linearity assumption violated.
From second graph there is observation is in some pattern like funnel or v shape so there is no constant variance occur that is there is no constant variance for error.
Constant variance for error occur when in residual plot all observation are in scatter everywhere.
From third graph we can say there is positive distribution but for regression analysis we need symmetric that is normal distribution.
Step-by-step explanation:
See graphs attached below.
