1. Consider the defects data from January 2017 through August 2018. Build a scatterplot of the defects over time. Fit (1) a simple linear regression model to the data, (2) a polynomial trendline of order 2, and (3) a polynomial trendline of order 3. Which model is the best fit for the data?

a) simple linear regression

b) polynomial trendline of order 2

c) polynomial trendline of order 3

2. Is the best fit model from question 1 a good predictor (at the time) of future defects? Yes or no?

3) Is linear regression an appropriate model to use for making a prediction of defects between January 2017 and August 2018?

4) Do the residuals in your linear regression model suggest any potential issues with the model?

a) Nonlinear patterns

b) Multicollinearity

c) Changing Variability

d) Correlated observations

e) Excluded variables

f) No problems with the regression residuals

Please show work. thank you

F A B D E 1 Defects After Delivery 2 3 Defects per million items received from suppliers 4 Month 2017 2018 2019 2020 5 January 812 828 824 682 6 February 810 832 836 695 7 March 813 847 818 692 8 April 823 839 825 686 9 May 832 832 804 673 10 June 848 840 812 681 837 849 806 696 12 August 831 857 798 688 13 September 827 839 804 671 14 October 838 842 713 645 15 November 826 828 705 617 16 December 819 816 686 603 17 2021 571 575 547 542 532 496 472 460 441 445 438 436 11 July