In: Statistics and Probability
Spending on credit cards decreases after the Christmas spending season (as measured by amount charged on a credit card in December). The accompanying data set contains the monthly credit card charges of a random sample of 99cardholders. Complete parts a) through e) below.
December on left and January on the right
1542.99 |
902.92 |
|
4301.97 |
7208.23 |
|
4229.48 |
4240.16 |
|
202.62 |
79.93 |
|
3298.47 |
4040.63 |
|
874.08 |
89.25 |
|
3806.31 |
3293.15 |
|
1934.11 |
2419.43 |
|
99.25 |
83.86 |
|
503.91 |
6.43 |
|
410.93 |
0.00 |
|
683.66 |
563.93 |
|
2160.65 |
2713.77 |
|
1123.17 |
187.02 |
|
2506.62 |
3266.62 |
|
1838.22 |
1522.32 |
|
9.94 |
1358.47 |
|
2332.42 |
732.94 |
|
78.51 |
75.03 |
|
101.31 |
70.21 |
|
598.23 |
634.71 |
|
648.87 |
1040.59 |
|
235.97 |
553.45 |
|
1266.34 |
1017.34 |
|
2123.35 |
1305.54 |
|
3.66 |
249.23 |
|
306.35 |
48.71 |
|
1902.49 |
872.77 |
|
558.68 |
485.62 |
|
2447.96 |
616.65 |
|
2799.33 |
1573.35 |
|
531.52 |
422.78 |
|
536.58 |
770.33 |
|
766.98 |
56.54 |
|
1958.44 |
1486.71 |
|
1678.16 |
495.43 |
|
2062.18 |
1065.31 |
|
397.08 |
510.51 |
|
5646.52 |
5640.14 |
|
5.51 |
5.51 |
|
2281.11 |
870.86 |
|
3820.38 |
1635.04 |
|
89.14 |
92.22 |
|
1450.51 |
669.75 |
|
527.35 |
829.29 |
|
105.86 |
69.26 |
|
1403.84 |
831.72 |
|
4234.61 |
2300.57 |
|
632.74 |
270.51 |
|
970.88 |
210.38 |
|
348.42 |
1011.16 |
|
0.00 |
1043.93 |
|
49.96 |
298.57 |
|
29.99 |
−29.99 |
|
471.78 |
1636.98 |
|
1115.95 |
1733.31 |
|
70.66 |
0.00 |
|
31.07 |
31.41 |
|
4.95 |
4.95 |
|
2523.13 |
1088.45 |
|
16.94 |
26.89 |
|
40.52 |
120.15 |
|
259.18 |
2006.79 |
|
123.05 |
291.22 |
|
0.00 |
104.07 |
|
109.71 |
52.99 |
|
5053.41 |
2839.36 |
|
3675.48 |
675.63 |
|
139.88 |
221.54 |
|
75.96 |
37.76 |
|
3150.92 |
533.41 |
|
2987.39 |
1932.47 |
|
651.55 |
692.88 |
|
9128.77 |
6810.41 |
|
916.81 |
393.47 |
|
2875.81 |
1308.63 |
|
797.06 |
796.87 |
|
34.56 |
0.00 |
|
44.16 |
1039.53 |
|
478.48 |
564.97 |
|
762.14 |
339.33 |
|
2349.94 |
5275.61 |
|
44.29 |
40.07 |
|
43.25 |
43.37 |
|
1339.34 |
653.49 |
|
1127.91 |
1072.21 |
|
2801.17 |
2336.41 |
|
52.09 |
91.46 |
|
1294.14 |
1434.02 |
|
328.19 |
720.64 |
|
28.31 |
28.58 |
|
598.68 |
980.71 |
|
4283.98 |
1576.08 |
|
568.23 |
0.00 |
|
479.75 |
162.04 |
|
1616.93 |
493.82 |
|
285.44 |
533.55 |
|
1283.98 |
462.53 |
|
3761.71 |
1477.77 |
a) Build a regression model to predict January spending fromDecember's spending.
Jan=____+____DEC (Round to four decimal places as needed.)
Check the conditions for this model. Select all of the true statements related to checking the conditions.
A. All of the conditions are definitely satisfied.
B. The Randomization Condition is not satisfied.
C. The Nearly Normal Condition is not satisfied.
D. The Equal Spread Condition is not satisfied.
E. The Linearity Condition is not satisfied.
b) How much, on average, will cardholders who charged $2000 in December charge in January?
$ ____ (Round to the nearest cent as needed.)
c) Give a 95% confidence interval for the average January charges of cardholders who charged $2000 in December.
($____,$____)(Round to the nearest cent as needed.)
d) From part c), give a 95% confidence interval for the average decrease in the charges of cardholders who charged $2000 in December.
($____,$____)(Round to the nearest cent as needed.)
e) What reservations, if any, would a researcher have about the confidence intervals made in parts c) and d)? Select all that apply.
A. The data are not independent, so the confidence intervals are not valid.
B. The data are not linear, so the confidence intervals are not valid.
C. The residuals show increasing spread, so the confidence intervals may not be valid.
D. The residuals show a curvilinear pattern, so the confidence intervals may not be valid.
E. A researcher would not have any reservations. The confidence intervals are valid.
Ʃx = | 132284.26 |
Ʃy = | 104474.62 |
Ʃxy = | 309198402.4 |
Ʃx² = | 419276460.8 |
Ʃy² = | 303002692.7 |
Sample size, n = | 99 |
x̅ = Ʃx/n = 132284.26/99 = | 1336.204646 |
y̅ = Ʃy/n = 104474.62/99 = | 1055.299192 |
SSxx = Ʃx² - (Ʃx)²/n = 419276460.7956 - (132284.26)²/99 = | 242517617.9 |
SSyy = Ʃy² - (Ʃy)²/n = 303002692.677 - (104474.62)²/99 = | 192750710.6 |
SSxy = Ʃxy - (Ʃx)(Ʃy)/n = 309198402.3583 - (132284.26)(104474.62)/99 = | 169598929.7 |
a) Slope, b = SSxy/SSxx = 169598929.67667/242517617.92946 = 0.699326223
y-intercept, a = y̅ -b* x̅ = 1055.29919 - (0.69933)*1336.20465 = 120.8562438
Regression equation :
JAN = 120.8562 + (0.6993) DEC
--
Conditions:
C. The Nearly Normal Condition is not satisfied.
D. The Equal Spread Condition is not satisfied.
--
b) Predicted value of y at x = 2000
JAN = 120.8562 + (0.6993) * 2000 = $1519.51
--
c) Predicted value of y at x = 2000
ŷ = 120.8562 + (0.6993) * 2000 = 1519.5087
Significance level, α = 0.05
Critical value, t_c = T.INV.2T(0.05, 97) = 1.9847
Sum of Square error, SSE = SSyy -SSxy²/SSxx = 192750710.61494 - (169598929.67667)²/242517617.92946 = 74145731.76
Standard error, se = √(SSE/(n-2)) = √(74145731.75588/(99-2)) = 874.29342
95% Confidence interval :
Lower limit = ŷ - tc*se*√((1/n) + ((x-x̅)²/(SSxx))) = 1330.08
Upper limit = ŷ + tc*se*√( (1/n) + ((x-x̅)²/(SSxx))) = 1708.94
--
d) 95% Prediction interval :
Lower limit = ŷ - tc*se*√(1 + (1/n) + ((x-x̅)²/(SSxx))) = -226.03
Upper limit = ŷ + tc*se*√(1 + (1/n) + ((x-x̅)²/(SSxx))) = 3265.05
e) C. The residuals show increasing spread, so the confidence intervals may not be valid.