Question

Construct a normal probability plot for the residuals. Comment on the normality assumption.

Use MINITAB package software/other software to compute and plot your data (No manual calculation is required)

​​​​​​​

Value of $1000 Year 2014 1863 2013 1639 Year-X (Independent) Value of $1000 Y-Dependent 2012 1239 2011 1068 2010 1046 2009 909 719 2008 2007 1141 2006 1081 2005 934 2004 890 803 2003 624 2002 801 2001 2000 909

Answer 1

Regression Analysis: Value of $1000 (Y) versus Year (X)

The regression equation is
Value of $1000 (Y) = - 112903 + 56.8 Year (X)

Predictor Coef SE Coef T P
Constant -112903 26676 -4.23 0.001
Year (X) 56.78 13.29 4.27 0.001

S = 222.406 R-Sq = 58.4% R-Sq(adj) = 55.2%

Analysis of Variance

Source DF SS MS F P
Regression 1 902552 902552 18.25 0.001
Residual Error 13 643035 49464
Total 14 1545588

Value of
Obs Year (X) $1000 (Y) Fit SE Fit Residual St Resid
1 2014 1863.0 1441.8 109.3 421.2 2.17R
2 2013 1639.0 1385.0 98.3 254.0 1.27
3 2012 1239.0 1328.3 87.8 -89.3 -0.44
4 2011 1068.0 1271.5 78.3 -203.5 -0.98
5 2010 1046.0 1214.7 69.9 -168.7 -0.80
6 2009 909.0 1157.9 63.3 -248.9 -1.17
7 2008 719.0 1101.2 58.9 -382.2 -1.78
8 2007 1141.0 1044.4 57.4 96.6 0.45
9 2006 1081.0 987.6 58.9 93.4 0.44
10 2005 934.0 930.8 63.3 3.2 0.01
11 2004 890.0 874.1 69.9 15.9 0.08
12 2003 803.0 817.3 78.3 -14.3 -0.07
13 2002 624.0 760.5 87.8 -136.5 -0.67
14 2001 801.0 703.7 98.3 97.3 0.49
15 2000 909.0 647.0 109.3 262.0 1.35

R denotes an observation with a large standardized residual.

(response is Value of $1000 () 95 XD 80 AD C 60 U 50 20 10 500 250 250 500 Residual

From the above plot we observe that the all points are closed to reference line hence the residual follows normal distribution. Hence normality assumption holds.