Homework No.

4
Experimental design and Two-Way ANOVA
Engineering Graduate School
Universidad de Cartagena
Prof. José Colina Márquez, PhD

Universidad de Cartagena
José Colina
Exercise Márquez, PhD
No.1
A mechanical engineer studies the pushing force of a drill. He suspects that the drill speed and the material feed rate are the most relevant factors. He selected four (4) feed rate and two (2) levels for the drill speed in order to represent the most
extreme operating conditions. He obtain the following results:

Analyze the results with α = 0.05.

Material feed rate

Drill speed 0.015 0.03 0.045 0.06
2.7 2.45 2.6 2.75
125
A: drill speed 2 speed 2.78 2.49 2.72 2.86
B: material feed rate 4 rate 2.83 2.85 2.86 2.94
200
Y: pushing force (n) 2 tests 2.86 2.8 2.87 2.88
N : 16

Material feed rate

Drill speed 0.015 0.03 0.045 0.06

125 2.74 2.47 2.66 2.805 2.66875 0.0741125 0.00303

200 2.845 2.825 2.865 2.91 2.86125 0.0741125 0.05522
2.793 2.648 2.763 2.858 2.765 0.148225
0.00002
0.03423
0.09250

SSTotal SSError
0.00422 0.09922 0.02722 0.00022 0.00160 0.00040 0.00360 0.00302
0.00023 0.07562 0.00202 0.00903 0.00160 0.00040 0.00360 0.00303
0.00423 0.00723 0.00903 0.03063 0.00022 0.00062 0.00003 0.00090
0.00903 0.00123 0.01103 0.01323 0.00023 0.00063 0.00002 0.00090
SST 0.30340 SSE 0.02080

SSA 0.14823 SSE 0.02080

Resumen SSB 0.09250 SSAB 0.04187

SST 0.30340

Ho: τ1=τ2=τ3=τ4=τ5=τ6=τ7=τ8= 0
H1: al menos una τ ≠ 0
Analysis of variance table for the model with a single factor and fixed effects

Variation Sum of Degrees of Mean

Fo Fc P-value
source squares freedom square
Given that each factor and its interactions have their Fo> Fc, the null hypothesis that
Treatments A: establishes that the treatments are equal is rejected. In this way it can be stated that both the
0.148225 1 0.14823 57.00962 5.31766 0.00007
drill speed drill speed, the material feed rate and their interactions have a significant effect on the
pushing force of a drill.
Treatments B: Furthermore, since 3 P-values ​are less than 0.05, these factors have a statistically significant
0.092500 3 0.03083 11.85897 4.06618 0.00258
material feed rate effect on the pushing force of a drill, with a 95.0% confidence level.
Interaction 0.041875 3 0.01396 5.36859 4.06618 0.02557
It should be noted that it is evident that the drill speed (125 and 200) is the one that affects
the most or has the greatest impact (p-value = 0.00001) on the pushing force of a drill.
Error 0.020800 8 0.00260
Total 0.303400 15
Homework No.4
Experimental design and Two-Way ANOVA
Exercise No.2
An experiment is carried out to study the influence of the operating temperature and three (3) types of plates of crystal c
obtained:

Estimate the incidence of the two factors and the interaction by using an ANOVA with a confidence level of α = 0.05.

Temperature
Crystal type
100 125 150

580 1090 1392

1
568 1087 1380
570 1085 1386
550 1070 1328
2 530 1035 1312
579 1000 1299
546 1045 867

3 575 1053 904

599 1066 889

A: Crystal type 3
B: Temperature 3
Y: luminous outlet (n) 3
N : 27

SSTotal
129733.368 22444.479 204136.627
138521.812 21554.590 193437.071
137037.071 20971.331 198750.849
152244.479 16851.886 150400.331
168251.886 8989.849 138246.257
130454.738 3577.812 128748.071
155381.960 10986.145 5356.071
133360.219 12727.182 1309.368
116407.331 15829.368 2619.923
SST 2418330.074

SSA 150864.519
 Resumen SSB 1970334.519

SST 2418330.074

Ho: τ1=τ2=τ3=τ4=τ5=τ6=τ7=τ8= 0
H1: al menos una τ ≠ 0
Analysis of variance table for the model with a single factor and fixed effects

Variation Sum of Degrees of

Mean square
source squares freedom
Treatments A:
150864.519 2 75432.259
Crystal type
Treatments B:
1970334.519 2 985167.259
Temperature
Interaction 290551.704 4 72637.926
Error 6579.333 18 365.519
Total 2418330.074 26

With Statgraphics
ree (3) types of plates of crystal coating, on the luminous outlet of an oscilloscope tube. The following results were

confidence level of α = 0.05.

Temperature

Crystal type 100 125 150

1 572.667 1087.333 1386.000 1015.333
2 553.000 1035.000 1313.000 967.000
3 573.333 1054.667 886.667 838.222

566.333 1059.000 1195.222 940.185

50825.198 1257886.864
6471.309 127052.642
93568.012 585395.012
150864.519 1970334.519

SSError
53.778 7.111 36.000
21.778 0.111 36.000
7.111 5.444 0.000
9.000 1225.000 225.000
529.000 0.000 1.000
676.000 1225.000 196.000
747.111 93.444 386.778
2.778 2.778 300.444
658.778 128.444 5.444
SSE 6579.333

SSE 6579.333
 SSAB 290551.704

d fixed effects

Fo Fc P-value Given that each factor and its interactions

establishes that the treatments are equal is re
206.371 3.55456 0.000000000 ttypes of plates of crystal coating and the ope
the luminous outlet
Since 3 P-values ​are less than 0.05, these fa
2695.259 3.55456 0.000000000 luminous outlet of an oscilloscope
198.726 2.92774 0.000000000 It is important to note that the operating tem
the greatest impact (they have an Fo statistic
an oscillo
or and its interactions have their Fo> Fc, the null hypothesis that
ments are equal is rejected. In this way, it can be stated that both the
l coating and the operating temperature and their interactions affect
the luminous outlet of an oscilloscope tube
ess than 0.05, these factors have a statistically significant effect on
tlet of an oscilloscope tube with a 95.0% confidence level.
hat the operating temperature (100, 125 and 150) is the one that has
y have an Fo statistic well above the others) on the luminous outlet of
an oscilloscope tube