REGRESYON ANALĐZĐ
1. Bu veri ile ilgili olarak bazı dönüşümler daha iyi sonuçlar verebilir. Logx, sqrtx, x-sq ve 1/x gibi.
Sales(Y) Expandt.(X)
11 39
17 49
26 76
24 68
23 59
29 91
4 34
32 116
33 141
31 149
32 105
30 99
37 171
33 124
Expandt.(X)
Sales(Y)
175 150
125 100
75 50
40
30
20
10
0
S 4,60830
R-Sq 77,7%
R-Sq(adj) 75,9%
Fitted Line Plot
Sales(Y) = 7,535 + 0,1942 Expandt.(X)
The regression equation is
Sales(Y) = 7,535 + 0,1942 Expandt.(X)
S = 4,60830 R-Sq = 77,7% R-Sq(adj) = 75,9%
Analysis of Variance
Source DF SS MS F P Regression 1 888,88 888,877 41,86 0,000 Error 12 254,84 21,236
Total 13 1143,71
logX
Sales(Y)
2,3 2,2
2,1 2,0
1,9 1,8
1,7 1,6
1,5 40
30
20
10
0
Scatterplot of Sales(Y) vs logX
sqrtX
Sales(Y)
14 13
12 11
10 9
8 7
6 5
40
30
20
10
0
Scatterplot of Sales(Y) vs sqrtX
x*x
Sales(Y)
30000 25000
20000 15000
10000 5000
0 40
30
20
10
0
Scatterplot of Sales(Y) vs x*x
1/x
Sales(Y)
0,030 0,025
0,020 0,015
0,010 0,005
40
30
20
10
0
Scatterplot of Sales(Y) vs 1/x
The regression equation is Sales(Y) = 42,9 - 1271 1/x
Predictor Coef SE Coef T P Constant 42,8587 0,7695 55,69 0,000 1/x -1271,32 50,92 -24,97 0,000
S = 1,34159 R-Sq = 98,1% R-Sq(adj) = 98,0%
Analysis of Variance
Source DF SS MS F P Regression 1 1122,1 1122,1 623,44 0,000 Residual Error 12 21,6 1,8
Total 13 1143,7
Unusual Observations
Obs 1/x Sales(Y) Fit SE Fit Residual St Resid 7 0,0294 4,000 5,467 0,892 -1,467 -1,46 X 10 0,0067 31,000 34,326 0,494 -3,326 -2,67R
Residual
Percent
2 0
-2 -4
99 90
50
10 1
Fitted Value
Residual
30 20
10 1
0 -1 -2 -3
Residual
Frequency
2 1 0 -1 -2 -3 8 6 4 2
0
Observ ation Order
Residual
14 13 12 11 10 9 8 7 6 5 4 3 2 1 1 0 -1 -2 -3
Normal Probability Plot of the Residuals Residuals Versus the Fitted Values
Histogram of the Residuals Residuals Versus the Order of the Data
Residual Plots for Sales(Y)
2.
Satış(Y) FiyatX1) Reklam(X2)
10 13 9
6 20 7
5 17 5
12 15 14
10 16 15
15 12 12
5 16 6
12 14 10
17 10 15
20 11 21
The regression equation is
Satış(Y) = 16,4 - 0,825 Fiyat(X1) + 0,585 Reklam(X2)
Predictor Coef SE Coef T P Constant 16,406 4,343 3,78 0,007 Fiyat(X1) -0,8248 0,2196 -3,76 0,007 Reklam(X2) 0,5851 0,1337 4,38 0,003
S = 1,50720 R-Sq = 93,2% R-Sq(adj) = 91,2%
Analysis of Variance
Source DF SS MS F P Regression 2 217,70 108,85 47,92 0,000 Residual Error 7 15,90 2,27
Total 9 233,60
Residual
Percent
4 2
0 -2
-4 99 90
50
10 1
Fitted Value
Residual
20 15
10 5
2 1 0 -1 -2
Residual
Frequency
2 1
0 -1
-2 2,0 1,5 1,0 0,5
0,0
Observ ation Order
Residual
10 9 8 7 6 5 4 3 2 1 2 1 0 -1 -2
Normal Probability Plot of the Residuals Residuals Versus the Fitted Values
Histogram of the Residuals Residuals Versus the Order of the Data
Residual Plots for Satýþ(Y)
3.
X1= Score Y=Satış
X2=Age
X3=Anxiety X4=Experience
X5= GPA
Y X1 X2 X3 X4 X5
44 10 22 49 0 24
47 19 23 30 1 26
60 27 23 15 0 28
71 31 24 6 3 27
61 64 23 18 2 20
60 81 22 33 1 25
58 42 24 32 0 25
56 67 22 21 0 23
66 48 22 60 1 28
61 64 23 18 1 34
51 57 21 38 0 30
47 10 23 45 1 27
53 48 22 45 0 28
74 96 25 1 3 38
65 75 23 9 0 37
33 12 21 48 0 21
54 47 22 23 1 18
39 20 21 30 2 15
52 73 21 3 2 19
30 4 20 27 0 22
58 9 23 44 1 28
59 98 21 39 1 29
52 27 23 14 2 32
56 59 22 27 1 27
49 23 23 27 1 24
63 90 22 22 2 26
61 34 24 7 1 34
39 16 21 31 1 23
62 32 24 6 3 40
78 94 25 46 5 36
The regression equation is
Y = - 75,6 + 0,208 X1 + 5,17 X2 + 0,0321 X3 + 0,376 X4 + 0,138 X5
Predictor Coef SE Coef T P Constant -75,56 21,68 -3,48 0,002 X1 0,20779 0,03447 6,03 0,000 X2 5,167 1,081 4,78 0,000 X3 0,03210 0,06157 0,52 0,607 X4 0,3760 0,9453 0,40 0,694 X5 0,1376 0,2010 0,68 0,500
S = 4,80085 R-Sq = 84,7% R-Sq(adj) = 81,5%
Analysis of Variance
Source DF SS MS F P Regression 5 3063,14 612,63 26,58 0,000 Residual Error 24 553,16 23,05
Total 29 3616,30
Residual
Percent
10 5
0 -5 -10 99 90
50
10 1
Fitted Value
Residual
80 60
40 10
5 0 -5
Residual
Frequency
12 8
4 0 -4 10,0
7,5 5,0 2,5 0,0
Observ ation Order
Residual
30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 10
5 0
-5
Normal Probability Plot of the Residuals Residuals Versus the Fitted Values
Histogram of the Residuals Residuals Versus the Order of the Data
Residual Plots for Y
BEST SUBSETS
Response is Y
Mallows X X X X X Vars R-Sq R-Sq(adj) C-p S 1 2 3 4 5 1 53,8 52,1 46,5 7,7254 X 1 45,7 43,8 59,2 8,3733 X 2 84,2 83,0 0,8 4,6013 X X 2 64,1 61,5 32,3 6,9310 X X 3 84,4 82,7 2,4 4,6507 X X X 3 84,3 82,5 2,6 4,6661 X X X 4 84,6 82,1 4,2 4,7193 X X X X 4 84,5 82,1 4,3 4,7304 X X X X
Buradan X1 ve X2 nin dahil edileceği bir modelin en iyilerden biri olacağı sonucuna varırız.
The regression equation is Y = - 81,1 + 0,215 X1 + 5,62 X2
Predictor Coef SE Coef T P Constant -81,12 15,44 -5,25 0,000 X1 0,21483 0,02981 7,21 0,000 X2 5,6248 0,6938 8,11 0,000
S = 4,60135 R-Sq = 84,2% R-Sq(adj) = 83,0%
Analysis of Variance
Source DF SS MS F P Regression 2 3044,6 1522,3 71,90 0,000
Residual Error 27 571,7 21,2 Total 29 3616,3
Source DF Seq SS X1 1 1653,2 X2 1 1391,5
Unusual Observations
Obs X1 Y Fit SE Fit Residual St Resid 4 31,0 71,000 60,536 1,468 10,464 2,40R 9 48,0 66,000 52,939 0,915 13,061 2,90R R denotes an observation with a large standardized residual.
4.
Y X
295 273,4 400 291,3 390 306,9 425 317,1 547 336,1 555 349,4 620 362,9 720 383,9 880 402,8 1050 437,0 1290 472,3 1528 510,4 1586 544,5 1960 588,1 2118 630,4 2116 685,9 2477 742,8 3119 801,3 3702 903,1 3316 983,6 2702 1076,7
X
Y
1100 1000
900 800
700 600
500 400
300 200
4000
3000
2000
1000
0
Scatterplot of Y vs X
The regression equation is Y = - 786 + 4,24 X
Predictor Coef SE Coef T P Constant -786,2 183,4 -4,29 0,000 X 4,2374 0,3100 13,67 0,000
S = 334,294 R-Sq = 90,8% R-Sq(adj) = 90,3%
Analysis of Variance
Source DF SS MS F P Regression 1 20885630 20885630 186,89 0,000 Residual Error 19 2123296 111752
Total 20 23008926
Unusual Observations
Obs X Y Fit SE Fit Residual St Resid 19 903 3702,0 3040,6 133,4 661,4 2,16R 21 1077 2702,0 3776,2 180,8 -1074,2 -3,82RX R denotes an observation with a large standardized residual.
X denotes an observation whose X value gives it large influence.
X
Y
1100 1000 900 800 700 600 500 400 300 200 4000
3000
2000
1000
0
S 334,294
R-Sq 90,8%
R-Sq(adj) 90,3%
Fitted Line Plot
Y = - 786,2 + 4,237 XResidual
Percent
1000 500
0 -500 -1000
99 90
50
10 1
Fitted Value
Residual
4000 3000
2000 1000
0 500
0 -500
-1000
Residual
Frequency
400 0
-400 -800
8 6 4 2 0
Observ ation Order
Residual
20 18 16 14 12 10 8 6 4 2 500
0
-500
-1000
Normal Probability Plot of the Residuals Residuals Versus the Fitted Values
Histogram of the Residuals Residuals Versus the Order of the Data
Residual Plots for Y
The regression equation is
Y = 1389 - 10,47 X + 0,02915 X**2 - 0,000017 X**3
S = 170,653 R-Sq = 97,8% R-Sq(adj) = 97,5%
Analysis of Variance
Source DF SS MS F P Regression 3 22513845 7504615 257,69 0,000 Error 17 495080 29122
Total 20 23008926
Sequential Analysis of Variance
Source DF SS F P Linear 1 20885630 186,89 0,000 Quadratic 1 963768 14,96 0,001 Cubic 1 664448 22,82 0,000
X
Y
1100 1000 900 800 700 600 500 400 300 200 4000
3000
2000
1000
0
S 170,653
R-Sq 97,8%
R-Sq(adj) 97,5%
Fitted Line Plot
Y = 1389 - 10,47 X + 0,02915 X**2 - 0,000017 X**35.
Sales Inc. Rate Y(t-1)
8,0 336,1 5,5
8,2 349,4 5,5 8,0
8,5 362,9 6,7 8,2
9,2 383,9 5,5 8,5
10,2 402,8 5,7 9,2
11,4 437 5,2 10,2
12,8 472,2 4,5 11,4 13,6 510,4 3,8 12,8 14,6 544,5 3,8 13,6 16,4 588,1 3,6 14,6 17,8 630,4 3,5 16,4 18,6 685,9 4,9 17,8 20,0 742,8 5,9 18,6 21,9 801,3 5,6 20,0 24,9 903,1 4,9 21,9 27,3 983,6 5,6 24,9 29,1 1076,7 8,5 27,3
Index
Sales
16 14
12 10
8 6
4 2
30
25
20
15
10
Time Series Plot of Sales
The regression equation is
Sales = - 0,014 + 0,0297 Inc. - 0,350 Rate
Predictor Coef SE Coef T P Constant -0,0140 0,2498 -0,06 0,956 Inc. 0,0297492 0,0002480 119,96 0,000 Rate -0,34987 0,04656 -7,51 0,000
S = 0,219930 R-Sq = 99,9% R-Sq(adj) = 99,9%
Analysis of Variance
Source DF SS MS F P Regression 2 738,88 369,44 7637,91 0,000 Residual Error 14 0,68 0,05
Total 16 739,56
Residual
Percent
0,50 0,25
0,00 -0,25
-0,50 99 90
50
10 1
Fitted Value
Residual
30 25
20 15
10 0,4
0,2
0,0
-0,2
Residual
Frequency
0,3 0,2 0,1 0,0 -0,1 -0,2 -0,3 3
2
1
0
Observ ation Order
Residual
16 14 12 10 8 6 4 2 0,4
0,2
0,0
-0,2
Normal Probability Plot of the Residuals Residuals Versus the Fitted Values
Histogram of the Residuals Residuals Versus the Order of the Data
Residual Plots for Sales
Bir başka Model denesek:
The regression equation is
Sales = 0,049 + 0,0264 Inc. - 0,326 Rate + 0,123 Y(t-1)
16 cases used, 1 cases contain missing values
Predictor Coef SE Coef T P Constant 0,0489 0,2720 0,18 0,860 Inc. 0,026412 0,004110 6,43 0,000 Rate -0,32615 0,05615 -5,81 0,000 Y(t-1) 0,1234 0,1528 0,81 0,435
S = 0,230604 R-Sq = 99,9% R-Sq(adj) = 99,9%
Analysis of Variance
Source DF SS MS F P Regression 3 670,42 223,47 4202,32 0,000 Residual Error 12 0,64 0,05
Total 15 671,05