5. One-way ANOVA: Tensile Strength (MPa) versus Slag % (in Flux)
Source DF SS MS F P
Slag % (in Flux) 5 135794 27159 87.82 0.000
Error 54 16700 309
Total 59 152494
S = 17.59 R-Sq = 89.05% R-Sq(adj) = 88.03%
Individual 95% CIs For Mean Based on
Pooled StDev
Level N Mean StDev ---+---------+---------+---------+------
0 10 548.00 6.25 (--*-)
20 10 495.20 21.68 (-*-)
40 10 446.30 9.65 (-*-)
60 10 460.60 25.09 (-*-)
80 10 433.20 20.18 (--*-)
100 10 398.60 14.71 (--*-)
---+---------+---------+---------+------
400 450 500 550
Pooled StDev = 17.59
1. Inferences for Tensile Strength
6. 100806040200
550
500
450
400
350
Slag % (in Flux)
TensileStrength(MPa)
398.6
433.2
460.6
446.3
495.2
548
Individual Value Plot of Tensile Strength (MPa) vs Slag % (in Flux)
100806040200
550
500
450
400
350
Slag % (in Flux)
TensileStrength(MPa)
400.5
439.5
462
447.5
495
547
398.6
433.2
460.6
446.3
495.2
548
Boxplot of Tensile Strength (MPa)
7. at most a 60% chance of detecting a difference of 12.074.
least a 90% chance of detecting a difference of 47.438, and
Based on your samples and alpha level (0.05), you have at
100%
47.438
90%
12.074
60%< 40%
12.074 9.7 - 60.0
33.856 60.0 - 100.0
37.489 70.0 - 100.0
41.713 80.0 - 100.0
47.438 90.0 - 100.0
Difference Power
with your sample sizes?
What difference can you detect
0 10 548 6.2539 (543.53, 552.47)
20 10 495.2 21.684 (479.69, 510.71)
40 10 446.3 9.6500 (439.40, 453.20)
60 10 460.6 25.092 (442.65, 478.55)
80 10 433.2 20.176 (418.77, 447.63)
100 10 398.6 14.714 (388.07, 409.13)
Slag % (in Flux) Size
Sample
Mean Deviation
Standard
95% CI for Mean
Individual
Statistics
41.713, consider increasing the sample sizes.
Power is a function of the sample sizes and the standard deviations. To detect differences smaller than
One-Way ANOVA for Tensile Stre by Slag % (in F
Power Report
Power
What is the chance of detecting a difference?
8. One-way ANOVA: Impact Strength (J) versus Slag % (in Flux)
Source DF SS MS F P
Slag % (in Flux) 5 4737 947 2.01 0.091
Error 54 25397 470
Total 59 30134
S = 21.69 R-Sq = 15.72% R-Sq(adj) = 7.91%
Individual 95% CIs For Mean Based on
Pooled StDev
Level N Mean StDev -------+---------+---------+---------+-
-
0 10 107.40 30.86 (---------*--------)
20 10 98.20 19.41 (--------*---------)
40 10 90.40 12.42 (--------*--------)
60 10 96.50 8.51 (--------*--------)
80 10 91.10 24.99 (--------*--------)
100 10 78.20 25.33 (--------*--------)
-------+---------+---------+---------+-
-
75 90 105 120
Pooled St Dev = 21.69
2. Inferences for Impact Strength
9. 100806040200
150
125
100
75
50
Slag % (in Flux)
ImpactStrength(J)
78.2
91.1
96.5
90.4
98.2
107.4
Individual Value Plot of Impact Strength (J) vs Slag % (in Flux)
100806040200
150
125
100
75
50
Slag % (in Flux)
ImpactStrength(J)
73
86.5
95.5
88.5
100
106.5
78.2
91.1
96.5
90.498.2
107.4
Boxplot of Impact Strength (J)
10. at most a 60% chance of detecting a difference of 15.786.
least a 90% chance of detecting a difference of 57.112, and
Based on your samples and alpha level (0.05), you have at
100%
57.112
90%
15.786
60%< 40%
15.786 10.9 - 60.0
40.768 60.0 - 100.0
45.146 70.0 - 100.0
50.233 80.0 - 100.0
57.112 90.0 - 100.0
Difference Power
with your sample sizes?
What difference can you detect
0 10 107.4 30.862 (85.322, 129.48)
20 10 98.2 19.407 (84.317, 112.08)
40 10 90.4 12.420 (81.515, 99.285)
60 10 96.5 8.5147 (90.409, 102.59)
80 10 91.1 24.986 (73.226, 108.97)
100 10 78.2 25.332 (60.078, 96.322)
Slag % (in Flux) Size
Sample
Mean Deviation
Standard
95% CI for Mean
Individual
Statistics
50.233, consider increasing the sample sizes.
Power is a function of the sample sizes and the standard deviations. To detect differences smaller than
One-Way ANOVA for Impact Stren by Slag % (in F
Power Report
Power
What is the chance of detecting a difference?
11. One-way ANOVA: Hardness (HRC) versus Slag % (in Flux)
Source DF SS MS F P
Slag % (in Flux) 5 364.0 72.8 2.95 0.020
Error 54 1332.4 24.7
Total 59 1696.4
S = 4.967 R-Sq = 21.46% R-Sq(adj) = 14.19%
Individual 95% CIs For Mean Based on
Pooled StDev
Level N Mean StDev -------+---------+---------+---------+--
0 10 12.450 4.450 (--------*--------)
20 10 14.930 6.064 (--------*--------)
40 10 16.580 7.017 (--------*--------)
60 10 18.490 1.863 (--------*--------)
80 10 15.110 4.757 (--------*--------)
100 10 11.060 4.016 (--------*--------)
-------+---------+---------+---------+--
10.5 14.0 17.5 21.0
Pooled StDev = 4.967
3. Inferences for Hardness
12. 100806040200
35
30
25
20
15
10
5
0
Slag % (in Flux)
Hardness(HRC) 11.06
15.11
18.49
16.58
14.93
12.45
Individual Value Plot of Hardness (HRC) vs Slag % (in Flux)
100806040200
35
30
25
20
15
10
5
0
Slag % (in Flux)
Hardness(HRC)
12
16.517.75
14.25
13.05
11.75 11.06
15.11
18.49
16.58
14.93
12.45
Boxplot of Hardness (HRC)
13. at most a 60% chance of detecting a difference of 4.6358.
least a 90% chance of detecting a difference of 13.279, and
Based on your samples and alpha level (0.05), you have at
100%
13.279
90%
4.6358
60%< 40%
4.6358 15.7 - 60.0
9.4807 60.0 - 100.0
10.497 70.0 - 100.0
11.678 80.0 - 100.0
13.279 90.0 - 100.0
Difference Power
with your sample sizes?
What difference can you detect
0 10 12.45 4.4500 (9.2666, 15.633)
20 10 14.93 6.0641 (10.592, 19.268)
40 10 16.58 7.0171 (11.560, 21.600)
60 10 18.49 1.8628 (17.157, 19.823)
80 10 15.11 4.7569 (11.707, 18.513)
100 10 11.06 4.0164 (8.1868, 13.933)
Slag % (in Flux) Size
Sample
Mean Deviation
Standard
95% CI for Mean
Individual
Statistics
11.678, consider increasing the sample sizes.
Power is a function of the sample sizes and the standard deviations. To detect differences smaller than
One-Way ANOVA for Hardness (HR by Slag % (in F
Power Report
Power
What is the chance of detecting a difference?
14. One-way ANOVA: Elongation (%) versus Slag % (in Flux)
Source DF SS MS F P
Slag % (in Flux) 5 782 156 1.39 0.241
Error 54 6061 112
Total 59 6843
S = 10.59 R-Sq = 11.43% R-Sq(adj) = 3.22%
Individual 95% CIs For Mean Based on
Pooled StDev
Level N Mean StDev --+---------+---------+---------+-------
0 10 30.50 13.41 (---------*--------)
20 10 28.90 14.75 (--------*---------)
40 10 26.20 4.42 (--------*---------)
60 10 27.20 5.20 (---------*--------)
80 10 25.20 11.55 (---------*---------)
100 10 19.10 9.81 (--------*---------)
--+---------+---------+---------+-------
14.0 21.0 28.0 35.0
Pooled StDev = 10.59
4. Inferences for Elongation
15. 100806040200
60
50
40
30
20
10
0
Slag % (in Flux)
Elongation(%)
19.1
25.2
27.226.2
28.9
30.5
Individual Value Plot of Elongation (%) vs Slag % (in Flux)
100806040200
60
50
40
30
20
10
0
Slag % (in Flux)
Elongation(%)
18
24
29.5
25
28.5
31
19.1
25.227.226.2
28.9
30.5
Boxplot of Elongation (%)
16. at most a 60% chance of detecting a difference of 7.1553.
least a 90% chance of detecting a difference of 28.477, and
Based on your samples and alpha level (0.05), you have at
100%
28.477
90%
7.1553
60%< 40%
7.1553 9.6 - 60.0
20.317 60.0 - 100.0
22.498 70.0 - 100.0
25.035 80.0 - 100.0
28.477 90.0 - 100.0
Difference Power
with your sample sizes?
What difference can you detect
0 10 30.5 13.410 (20.907, 40.093)
20 10 28.9 14.746 (18.352, 39.448)
40 10 26.2 4.4171 (23.040, 29.360)
60 10 27.2 5.2026 (23.478, 30.922)
80 10 25.2 11.545 (16.941, 33.459)
100 10 19.1 9.8144 (12.079, 26.121)
Slag % (in Flux) Size
Sample
Mean Deviation
Standard
95% CI for Mean
Individual
Statistics
25.035, consider increasing the sample sizes.
Power is a function of the sample sizes and the standard deviations. To detect differences smaller than
One-Way ANOVA for Elongation ( by Slag % (in F
Power Report
Power
What is the chance of detecting a difference?
18. General Regression Analysis: Tensile Strength (MPa) versus Slag % (in
Flux)
Regression Equation
Tensile Strength (MPa) = 550.09 - 4.45063 Slag % (in Flux) + 0.0721726 Slag %
(in Flux)*Slag % (in Flux) - 0.000428472 Slag % (in
Flux)*Slag % (in Flux)*Slag % (in Flux)
Coefficients
Term Coef SE Coef T
Constant 550.090 5.94031 92.6031
Slag % (in Flux) -4.451 0.57693 -7.7144
Slag % (in Flux)*Slag % (in Flux) 0.072 0.01434 5.0345
Slag % (in Flux)*Slag % (in Flux)*Slag % (in Flux) -0.000 0.00009 -4.5520
Term P
Constant 0.000
Slag % (in Flux) 0.000
Slag % (in Flux)*Slag % (in Flux) 0.000
Slag % (in Flux)*Slag % (in Flux)*Slag % (in Flux) 0.000
Summary of Model
S = 19.1691 R-Sq = 86.51% R-Sq(adj) = 85.78%
PRESS = 23149.0 R-Sq(pred) = 84.82%
19. Statistics
R-squared (adjusted)
P-value, model
P-value, linear term
P-value, quadratic term
P-value, cubic term
Residual standard deviation
85.78%
0.000*
0.000*
0.000*
0.000*
19.169
Cubic
Selected Model
78.71% 80.86%
0.000* 0.000*
0.000* 0.000*
- 0.008*
- -
23.460 22.239
Linear Quadratic
Alternative Models
100806040200
550
500
450
400
350
Slag % (in Flux)
TensileStrength(MPa)
Large residual
Y: Tensile Strength (MPa)
X: Slag % (in Flux)
* Statistically significant (p < 0.05)
Regression for Tensile Strength (MPa) vs Slag % (in Flux)
Model Selection Report
Fitted Line Plot for Cubic Model
Y = 550.1 - 4.451 X + 0.07217 X**2 - 0.000428 X**3
20. Slag % (in Flux) is statistically significant (p < 0.05).
The relationship between Tensile Strength (MPa) and
> 0.50.10.050
NoYes
P = 0.000
be accounted for by the regression model.
85.78% of the variation in Tensile Strength (MPa) can
100%0%
R-sq (adj) = 85.78%
1007550250
550
500
450
400
Slag % (in Flux)
TensileStrength(MPa)
causes Y.
A statistically significant relationship does not imply that X
Tensile Strength (MPa).
correspond to a desired value or range of values for
Flux), or find the settings for Slag % (in Flux) that
to predict Tensile Strength (MPa) for a value of Slag % (in
If the model fits the data well, this equation can be used
Y = 550.1 - 4.451 X + 0.07217 X**2 - 0.000428 X**3
relationship between Y and X is:
The fitted equation for the cubic model that describes the
Y: Tensile Strength (MPa)
X: Slag % (in Flux)
Is there a relationship between Y and X?
Fitted Line Plot for Cubic Model
Y = 550.1 - 4.451 X + 0.07217 X**2 - 0.000428 X**3
Comments
Regression for Tensile Strength (MPa) vs Slag % (in Flux)
Summary Report
% of variation accounted for by model
21. General Regression Analysis: Hardness (HRC) versus Slag % (in Flux)
Regression Equation
Hardness (HRC) = 11.9504 + 0.229152 Slag % (in Flux) - 0.0023558 Slag % (in
Flux)*Slag % (in Flux)
Coefficients
Term Coef SE Coef T P
Constant 11.9504 1.40101 8.52984 0.000
Slag % (in Flux) 0.2292 0.06589 3.47772 0.001
Slag % (in Flux)*Slag % (in Flux) -0.0024 0.00063 -3.72471 0.000
Summary of Model
S = 4.88827 R-Sq = 19.71% R-Sq(adj) = 16.90%
PRESS = 1496.83 R-Sq(pred) = 11.77%
22. Statistics
R-squared (adjusted)
P-value, model
P-value, linear term
P-value, quadratic term
P-value, cubic term
Residual standard deviation
16.90%
0.002*
0.001*
0.000*
-
4.888
Quadratic
Selected Model
0.00% 16.28%
0.754 0.005*
0.754 0.387
- 0.914
- 0.450
5.404 4.906
Linear Cubic
Alternative Models
100806040200
30
25
20
15
10
5
Slag % (in Flux)
Hardness(HRC)
Large residual
Y: Hardness (HRC)
X: Slag % (in Flux)
* Statistically significant (p < 0.05)
Regression for Hardness (HRC) vs Slag % (in Flux)
Model Selection Report
Fitted Line Plot for Quadratic Model
Y = 11.95 + 0.2292 X - 0.002356 X**2
23. (in Flux) is statistically significant (p < 0.05).
The relationship between Hardness (HRC) and Slag %
> 0.50.10.050
NoYes
P = 0.002
accounted for by the regression model.
16.90% of the variation in Hardness (HRC) can be
100%0%
R-sq (adj) = 16.90%
1007550250
30
20
10
0
Slag % (in Flux)
Hardness(HRC)
causes Y.
A statistically significant relationship does not imply that X
a desired value or range of values for Hardness (HRC).
or find the settings for Slag % (in Flux) that correspond to
to predict Hardness (HRC) for a value of Slag % (in Flux),
If the model fits the data well, this equation can be used
Y = 11.95 + 0.2292 X - 0.002356 X**2
the relationship between Y and X is:
The fitted equation for the quadratic model that describes
Y: Hardness (HRC)
X: Slag % (in Flux)
Is there a relationship between Y and X?
Fitted Line Plot for Quadratic Model
Y = 11.95 + 0.2292 X - 0.002356 X**2
Comments
Regression for Hardness (HRC) vs Slag % (in Flux)
Summary Report
% of variation accounted for by model
33. Total N 30
Subgroup size 1
Mean 423.73
StDev (overall) 1.6522
StDev (within) 1.9436
Process Characterization
Cp 0.86
Cpk 0.64
Z.Bench 1.91
% Out of spec (expected) 2.81
PPM (DPMO) (expected) 28072
Actual (overall)
Pp 1.01
Ppk 0.75
Z.Bench 2.26
% Out of spec (observed) 0.00
% Out of spec (expected) 1.20
PPM (DPMO) (observed) 0
PPM (DPMO) (expected) 12036
Potential (within)
Capability Statistics
429.0427.5426.0424.5423.0421.5420.0
LSL Target USL
Capability Histogram
Are the data inside the limits and close to the target?
Actual (overall) capability is what the customer experiences.
shifts and drifts were eliminated.
Potential (within) capability is what could be achieved if process
Capability Analysis for Tensile Stre
Process Performance Report
34. Total N 30
Subgroup size 1
Mean 14.083
StDev (overall) 0.81201
StDev (within) 0.79176
Process Characterization
Cp 0.84
Cpk 0.81
Z.Bench 2.26
% Out of spec (expected) 1.20
PPM (DPMO) (expected) 11997
Actual (overall)
Pp 0.82
Ppk 0.79
Z.Bench 2.19
% Out of spec (observed) 0.00
% Out of spec (expected) 1.43
PPM (DPMO) (observed) 0
PPM (DPMO) (expected) 14277
Potential (within)
Capability Statistics
1615141312
LSL Target USL
Capability Histogram
Are the data inside the limits and close to the target?
Actual (overall) capability is what the customer experiences.
shifts and drifts were eliminated.
Potential (within) capability is what could be achieved if process
Capability Analysis for Hardness (At
Process Performance Report