According to the depicted table of multiple regression models, deviation from the actual output and the predicted output for all the data points are reflected clearly. Hence, the deviation of the estimated result from the predicted result can also be assessed from the multiple regression models. For example, the first point of sales is cost of sales and advertisement, the residual output is £310315. It implies that the forecasted and actual sales may have the variation of £310315 in according to the calculated data of multiple regression models.
Ethical issues associated with the regression modelling
The key ethical issue of the regression analysis is the misinterpretation of the regression assumptions. In that case, Agresti and Kateri (2011) stated that the awareness of the regression is the main problem that in turn underlines the least square regression. The analysts have the little knowledge regarding the evaluation of the assumptions as most of the cases the regression statistics are used mechanically without considering the assumptions of the regression. Moreover, Dimaggio (2013) added to the previous statement that people do not have sufficient knowledge regarding the alternative option while one of the key assumptions is violated by one of the factors, considered as variable. Therefore, only the calculations and manipulation of the outcomes are performed by the analytics diverting the decision towards a wrong direction due to deviation from actual result.