3. Refer to the Excel "Viral" data posted on eLearning: (25) *If appropriate, you must share your relevant regression output to demonstrate where your answers are coming from. If you have such output, you may use it directly with limited Excel calculations to answer some questions. a. Estimate a typical, ordinary least squares regression model which explains how many (in thousands) clicks a social media video will receive as a function of two predictors: both the channel (X or Y) that hosts therabbitvideo and the number of paid social media promoters devoted to sharing therabbitvideo. Write out this estimated equation to explain how many clicks are received (use the estimate values and the specific predictors (not 'x1' e.g.)!). b. Provide the residual plots against each of the two explanatory variables you used in "a" above (channel and promoter). Comment substantively on any irregularities that you notice. If you notice any irregularities in part "b", above, attempt other models which utilize quadratic and/or interaction terms withrabbitchannel and promoter as explanatory variables. c. If you find a model with residuals which better satisfy our necessary regression assumptions, then write out this new, estimated model. Comment on the statistical significance levels of the predictors and provide and comment on the new residual plots for your finalized model. If you have a 'better' final model than that estimated in part 'a' then make sure to describe why it is better while considering the 'trade-off' between models that best explain the relationship withrabbitpredictors and clicks and the level of complication of the model.

Video Clicks (in thousands)	Host	Promoters
9	X	2
49	X	4
73	Y	7
64	Y	7
101	X	10
225	Y	13
484	X	13
1089	Y	16
1600	Y	17
1445	X	18
4420	X	19
4489	Y	20
4624	X	23
6084	X	23
7225	Y	24
7569	X	25
9801	Y	28
13400	Y	32
14400	Y	32
16900	X	33
16010	X	33
18720	Y	34
19600	Y	36
21800	Y	38
22300	X	38
25760	X	39
27100	Y	40
28700	X	42
30500	X	44
31400	Y	45