Documente Academic
Documente Profesional
Documente Cultură
Bolt Chemical 1 2 3 4 1 73 73 75 73 2 68 67 68 71 3 74 75 78 75 4 71 72 73 75 5 67 70 68 69
4-3. Plot the mean tensile strengths observed for each chemical type in Problem 4-1 and compare them to an appropriatety scaled t distribution. What conclusions would you draw from this display? 4-9. Assuming that chemical types and bolts are xed, estimate the model parameters i and i in Problem 4-1. 4-11. Suppose that the obersvation for chemical type 2 and bolt 3 is missing in Problem 4-1. Analyze the problem by estimating the missing value. Perform the exact analysis and compare the results.
Hypotheses: Treatment: H0 : 1 = 2 = 3 = 4 = 0 H1 : i = 0 at least one i Block: H0 : 1 = 2 = = 5 = 0 H1 : j = 0 at least one j Level of signicance: = 0.05 Test Statistic: FT = Rejection Region: FT > 3.4903 FB > 3.2592 MSTreatments MSE FB = MSBlocks MSE
Computations:
SST
i =1 i =1
y2 ij
i =1 j =1
yij
ba 14352 20 (1)
SSTreatments
= =
i =1 j =1
yij
ba
(2)
SSBlocks
= =
1 a
j =1
y2 .j
i =1 j =1
yij
ba
SSE
ANOVA Table Source of Variation Treatments Blocks Error Total SS 12.95 157 21.8 191.75 DF 3 4 12 19 MS 4.317 39.25 1.817 F FT =2.376 FB =21.602
Decision: Since the FT is less than the tabulated value F0.05,3,12 = 3.4903. Then, the null hypothesis of treatment is not rejected at = 0.05. Moreover, since 21.602 of FB is greater than than F0.05,4,12 = 3.2592. Then, the null hypothesis of block is rejected. Conclusion: Hence, the four chemical agents tested by the chemist on the strength of a particular type of cloth is not signicant, which means it has no effect. Furthermore, the bolt of cloths has a signicant effect on the strength of a particular type of cloth.
There will be no multiple comparison test to happen in block (bolt), since the chemist wishes only to test the effect of four chemical agents on the strength of a particular type of cloth. 4-3. Plot the mean tensile strengths observed for each chemical type in Problem 4-1 and compare them to an appropriatety scaled t distribution. What conclusions would you draw from this display? Solution: Refer to Figure 1. Conclusion: there is no obvious difference between the means. This is the same conclusion given by the analysis of variance. 4-9. Assuming that chemical types and bolts are xed, estimate the model parameters i and i in Problem 4-1. Solution: If both treatments and blocks are xed, we may estimate the parameters in the RCBD model by least squares. Recall that the
Scaled t Distribution
0.4
qqqq q q q q q q q q q q q q q q q q q q q q q q q q q q q qq qq qqq qqqqqqqq qqqqqq
Figure 1: Tensile strength averages from the Chemical experiment in relation to a t distribution with a scale factor MSE = a 1.82 = 0.675 4
0.3
q q q q
P(x)
0.2
q q q
q q
0.1
q q q q q q q q q qq qq qqq qqqqqqqqqqqqq
0.0
linear statistical model is i = 1, 2, . . . , a yij = + i + j + ij j = 1, 2, . . . , b Applying the rules in section 3-9.2 (Montgomery, D. C., Design and Analysis of Experiments. Fifth Edition) for nding the normal equations for an experimental model, we obtain : 20 + 51 + 52 + 53 + 54 + 4 1 + 4 2 + 4 3 + 4 4 + 4 5 = 1435 1 2 3 4 1 2 3 4 5 : : : : : : : : : 5 + 51 + 1 + 2 + 3 + 4 + 5 = 353 5 + 52 + 1 + 2 + 3 + 4 + 5 = 357 5 + 53 + 1 + 2 + 3 + 4 + 5 = 362 5 + 54 + 1 + 2 + 3 + 4 + 5 = 363 4 + 1 + 2 + 3 + 4 + 4 1 = 294 4 + 1 + 2 + 3 + 4 + 4 2 = 274 4 + 1 + 2 + 3 + 4 + 4 3 = 302 4 + 1 + 2 + 3 + 4 + 4 4 = 291 4 + 1 + 2 + 3 + 4 + 4 5 = 274
The long equation array of s, s, and s can actually be simplied, using the following solutions below, i j
= = =
i =1
i = 0 j = 0
j =1
(7)
= = =
Using the equations 4, 5, and 6, and applying the constraints in equation 7, we obtain 2 4 2 3 4 5
= = = = = = =
35 20
4-11. Suppose that the obersvation for chemical type 2 and bolt 3 is missing in Problem 4-1. Analyze the problem by estimating the missing value. Perform the exact analysis and compare the results. Solution: Bolt Chemical 1 2 3 4 1 73 73 75 73 2 68 67 68 71 3 74 x 78 75 4 71 72 73 75 5 67 70 68 69
a(y i. ) + b(y .j ) y ..
(8)
Thus, the estimated value of x is 75.25. The usual analysis of variance may now be performed using x = y23 and reducing the error and total degrees of freedom by 1. The new computation of sum of squares are shown below using the estimated value,
Bolt Chemical 1 2 3 4 Blocks Total 1 73 73 75 73 294 2 68 67 68 71 274 3 74 75.25 78 75 302.25 4 71 72 73 75 291 5 67 70 68 69 274
SST
i =1 i =1
y2 ij
i =1 j =1
yij
ba 1435.252 20 (9)
= SSTreatments = 12.95
[y .j ( a 1) x ]
t ( t 1)
SSBlocks
= =
1 a
j =1
y2 .j
i =1 j =1
yij
ba
ANOVA Table Source of Variation Treatments Blocks Error Total SS 12.82 158.89 21.72 193.43 Degrees of Freedom 3 4 11 18 MS 4.27 39.723 1.97 F F1 =2.17 F2 =20.16
The results for both ANOVAs are very close, but with the estimated value of x and an adjustment of sum of squares of treatment, the FComputed now becomes smaller, which means getting far from the rejection region.
a. Analyze the data completely and interpret your results. b. Obtain the treatment means, treatment effects, standard deviation of a treatment mean and treatment mean difference, and the CV of the experiment. c. Obtain the efciency of this design with respect to CRD and with respect to RCB i. if columns were used as blocks; ii. if rows were used as blocks and interpret your results. Solution: i. Hypotheses: H0 : The ve fertilizers have equal effects on the yields of sugar cane. H1 : At least one of the fertilizers has an effect on the yields of sugar cane. ii. Level of Signicance: = 0.05 iii. Test Statistics: F= MSTreatments MSE
iv. Rejection Region: Reject the null hypothesis if, F > F,p1,( p1)( p2) that is F > ( F0.05,4,12 = 3.2592)
10
v. Computation: Rows 1 1 2 3 4 5 y..k y.j. 14(A) 19(B) 23(D) 21(C) 23(E) 100 A=80 2 22(E) 21(D) 15(A) 25(B) 16(C) 99 B=101 Columns 3 20(B) 16(A) 20(C) 24(E) 23(D) 103 C=93 y2 ... N 4 18(C) 23(E) 18(B) 21(D) 17(A) 97 D=113 5 25(D) 18(C) 23(E) 18(A) 19(B) 103 E=115 99 97 99 109 98 502 yi..
Where: A=no fertilizer C=10 tons manure/ha E=30 tons manure/ha B=complete inorganic fertilizer D=20 tons manure/ha
SST
y2 ijk
i j k
(502)2 25
(13)
= = =
1 p
i =1
y2 i..
y2
...
= 19.04
SSColumns
= = =
1 p
k =1
y2 ..k
= 5.44
SSTreatments
= = =
1 p
j =1
y2 .j.
= 168.64
11
SSError
DF 4 4 4 12 24
F F1 =11.303
vi. Decision: Thus, the null hypothesis is rejected since 11.302 is greater than 3.2592. v. Conclusion: Hence, the ve fertilizers are signicantly different, which means that they do have an effect on the yield of sugar cane. vii. Multiple Comparison Test: Solution: Using the Least Signicance Difference, the critical value is, LSD = t ,N p
2
2MSE = 2.086 n
2(3.73) = 2.548 5
Thus, any pair of treatment averages differ by more than 2.548 would imply that the corresponding pair of population means are signicantly different. The differences in averages are, y.E. y.A. = 23 16 y.E. y.C. = 23 18.6 y.E. y.B. = 23 20.2 y.E. y.D. = 23 22.6 y.D. y.A. = 22.6 16 y.D. y.C. = 22.6 18.6 y.D. y.B. = 22.6 20.2
12
y.B. y.A. = 20.2 16 y.B. y.C. = 20.2 18.6 y.C. y.A. = 18.6 16
The starred values indicates pairs of mean that are signicantly different. b. Obtain the treatment means, treatment effects, standard deviation of a treatment mean and treatment mean difference, and the CV of the experiment. Solution: a. Obtain treatment means, y.j. y.j. A=80 y.A. =16 B=101 y.B. =20.2 C=93 y.C. =18.6 D=113 y.D. =22.6 E=115 y.E. =23
b. Treatment Effects 1 2 3 4 5
= =
= y.B. y... = 20.2 20.08 = 0.12 = y.C. y... = 18.6 20.08 = 1.48 = y.D. y... = 22.6 20.08 = 2.52 = y.E. y... = 23 20.08 = 2.92
d. Mean Difference The ascending order of the means y.E. = 23 y.D. = 22.6 y.B. = 20.2 y.C. = 18.6 y.A. = 16
y.E. y.A. = 23 16 y.E. y.C. = 23 18.6 y.E. y.B. = 23 20.2 y.E. y.D. = 23 22.6
13
y.D. y.A. = 22.6 16 y.D. y.C. = 22.6 18.6 y.D. y.B. = 22.6 20.2 y.B. y.A. = 20.2 16 y.B. y.C. = 20.2 18.6 y.C. y.A. = 18.6 16 e. Coefcient of Variation:
CV =
c. Obtain the efciency of this design with respect to CRD and with respect to RCBD i. if columns were used as blocks; ii. if rows were used as blocks and interpret your results. Solution: Completely Randomized Design Data Layout: Fertilizers A B C D E 14 19 21 23 23 15 25 16 21 22 Yield 16 20 20 23 24 17 18 18 21 23 18 19 18 25 23 Total 80 101 93 113 115 502 Means 16 20.2 18.6 22.6 23 20.08
= =
(47)
14
ANOVA Table Source of Variation Treatments Error Total SS 168.64 69.2 237.84 DF 4 20 24 MS 42.16 3.46 F F1 =12.185
i. Decision: Thus, the null hypothesis is rejected since 12.185 is greater than 2.87, for F0.05,4,20 . ii. Conclusion: Hence, the ve fertilizers are signicantly different, implying that they do have an effect on the yield of sugar cane. iii. Multiple Comparison Solution: Using the Tukey Honestly Signicant Difference, the critical value is, T = q( a, f ) MSE = q(5, 20) n 3.46 = 4.23(0.8319) = 3.519 5
Thus, any pair of treatment averages differ by more than 3.519 would imply that the corresponding pair of population means are signicantly different. The differences in averages are, y.E. y.A. = 23 16 y.E. y.C. = 23 18.6 y.E. y.B. = 23 20.2 y.E. y.D. = 23 22.6 y.D. y.A. = 22.6 16 y.D. y.C. = 22.6 18.6 y.D. y.B. = 22.6 20.2 y.B. y.A. = 20.2 16 y.B. y.C. = 20.2 18.6 y.C. y.A. = 18.6 16
(48) (49) (50) (51) (52) (53) (54) (55) (56) (57)
The starred values indicates pairs of mean that are signicantly different. iv. Relative Efciency
15
RE
LS CRD f
= = =
RE
= =
MSR + MSC + ( a 1) MSE (f) ( a + 1) MSC ( f 1 + 1)( f 2 + 3) ( f 2 + 1)( f 1 + 3) (12 + 1)(20 + 3) = 0.949 (20 + 1)(12 + 3) 4.76 + 1.36 + 4(3.73) (0.949) 6(1.36) 2.447
(58)
Randomized Complete Block Design Data Layout if columns were used as blocks:
Treatment Totals
Means
CF SST SSTreatments
= =
= 168.64
b
SSBlocks(Columns)
= =
1 a
j =1
y2 .j
i =1 j =1
yij
ba
16
SSError
DF 4 4 16 24
F F1 =10.58
v. Decision: Thus, the null hypothesis is rejected since 10.58 is greater than 3.0069. vi. Conclusion: Hence, the ve fertilizers are signicantly different, which means that they do have an effect on the yield of sugar cane when columns were used as blocks. vii. Multiple Comparison Using the Least Signicance Difference, the critical value is, LSD = t ,15
2
2MSE = 2.131 n
2(3.985) = 2.69 5
Thus, any pair of treatment averages differ by more than 2.69 would imply that the corresponding pair of population means are signicantly different. The differences in averages are, y.E. y.A. = 23 16 y.E. y.C. = 23 18.6 y.E. y.B. = 23 20.2 y.E. y.D. = 23 22.6 y.D. y.A. = 22.6 16 y.D. y.C. = 22.6 18.6 y.D. y.B. = 22.6 20.2 y.B. y.A. = 20.2 16 y.B. y.C. = 20.2 18.6
17
= 2.6
(73)
The starred values indicates pairs of mean that are signicantly different.
RE
LS RCBD f
= = =
RE
= =
MSR + ( a 1) MSE (f) ( a) MSE ( f 1 + 1)( f 2 + 3) ( f 2 + 1)( f 1 + 3) (12 + 1)(20 + 3) = 0.926 (16 + 1)(16 + 3) 4.76 + 4(3.73) (0.926) 5(3.73) 0.977
(74)
Randomized Complete Block Design Data Layout if rows were used as blocks:
Treatment Totals
Means
CF SST SSTreatments
= =
= 168.64
18
SSBlocks( Rows)
= =
1 a
j =1
y2 .j
i =1 j =1
yij
ba
DF 4 4 16 24
F F1 =10.58
ix. Decision: Thus, the null hypothesis is rejected since 10.58 is greater than 3.0069. x. Conclusion: Hence, the ve fertilizers are signicantly different, which means that they do have an effect on the yield of sugar cane. xi. Multiple Comparison: Using Least Signicance Difference, the process is just the same with (ii.). xii. Relative Efciency of RCBD with rows as blocks
RE
LS RCBD f
= = =
RE
= =
MSC + ( a 1) MSE (f) ( a) MSE ( f 1 + 1)( f 2 + 3) ( f 2 + 1)( f 1 + 3) (12 + 1)(20 + 3) = 0.926 (16 + 1)(16 + 3) 1.36 + 4(3.73) (0.926) 5(3.73) 0.808
(80)
19
Treatment (Times) Totals: A = 118, B = 78, C = 94, D = 75, E = 65. Catalyst Totals: = 83, = 85, = 91, = 82, Computation of Sum of Squares: CF = G.. 4302 = = 7396 25 a2 = 89.
Total SS or TSS = (262 + 162 + + 172 + 142 ) - 7396 = 436 (892 + 882 + 922 + 832 + 782 ) - 7396 = 24.4 Acid SS or ASS = 5 2 + 892 + 862 + 832 + 822 ) (90 Batch SS or BSS = - 7396 = 10 5 2 + 782 + 942 + 752 + 652 ) (118 Times SS or TrSS = - 7396 = 342 5
20
Catalyst SS or CSS =
(832 + 852 + 912 + 822 + 892 ) - 7396 = 12 5 Error SS or SSE = TSS - BSS - ASS - CSS - TrSS = 436 10 24.4 342 12 = 47.6
ANOVA table for 55 Graeco-Latin Square (p=5) SV Times Batch Acid Catalyst Error Total DF p-1=4 p-1=4 p-1=4 p-1=4 (p-1)(p-3)=8 24 SS 342 10 24.4 12 47.6 436 MS 85.5 2.5 6.1 3 5.95 F 14.37 0.42 1.025 0.504
Decision: All FComputed of each Source Variation is less than the critical value, F,4,8 = 3.8379, except for the treatments which is 14.37. And thus, the following decision is obtain, a. The ve standing times are signicantly different. b. The ve batches of raw materials have no signicant difference. c. The ve acid concentrations have no signicant difference. d. The ve catalyst concentrations have no signicant difference. xiii. Multiple comparison for ve standing times, Using Tukey Honestly Signicant Difference, the critical value is obtain, T = q( a, f ) MSE = q(5, 8) n 5.95 = 4.89 5 5.95 = 5.33 5
Treatment Means: y A = 23.6, yC = 18.8, y B = 15.6, y D = 15, y E = 13 Thus, any pair of treatment averages differ by more than 5.33 would imply that the corresponding pair of population means are signi-
21
cantly different. The differences in averages are, y A yE = 23.6 13 y A y D = 23.6 15 y A y B = 23.6 15.6 y A yC = 23.6 18.8 yC y E = 18.8 13 yC y D = 18.8 15 yC y B = 18.8 15.6 y B y E = 15.6 13 y B y D = 15.6 15 y D y E = 15 13
The starred values indicates pairs of mean that are signicantly different.
22
Tensile Strength (lb/in2 ) 3129 3200 2800 2600 3000 3300 2900 2700 2865 2975 2985 2600 2890 3150 3050 2765
Step 2
Figure 3: The second step after inputting your data, go to analyze compare means one-way anova.
23
Step 3
Figure 4: Next, enter the variable yield to the dependent list (yield dependent list) and treatment to factor (treatment factor). After that you can click the post hoc.. for choosing the test for multiple comparison.
Table 1: The output of the performed steps. In multiple comparison table, the test performed was Scheff. You can check it in the post hoc.. section of the Step 3
4-1. A chemist wishes to test the effect of four chemical agents on the strength of a particular type of cloth. Because there might be variability from one bolt to another, the chemist decides to use a randomized block design, with the bolts of cloth considered as blocks. She selects ve bolts and applies all four chemicals in random order
24
to each bolt. The resulting tensile strength follow. Analyze the data from this experiment (use = 0.05) and draw appropriate conclusions. Bolt Chemical 1 2 3 4 Solution Step 1
Figure 5: The above table is entered in SPSS in this manner.
1 73 73 75 73
2 68 67 68 71
3 74 75 78 75
4 71 72 73 75
5 67 70 68 69
Step 2
Figure 6: The second step after inputting your data, go to analyze general linear model univariate
25
Step 3
Figure 7: Next, enter the variable yield to the dependent list (yield dependent list) and treatment and block to fixed factor(s) (treatment and block fixed factor(s)). After that you can click the post hoc.. for choosing the test for multiple comparison.
Step 4
Figure 8: Before clicking the ok button, go rst to the model (seen on Step 3). In the univariate: model window, click (custom) then put the treatment and block to the model box, as shown in the gure. Then, change the type to main effects and unchecked the include intercept in model, before the clicking the continue button.
output of the performed test. Refer to Manual Computation and Graphical Illustration Section item 4-1 for the interpretation. 4-11. Suppose that the obersvation for chemical type 2 and bolt 3 is missing in Problem 4-1. Analyze the problem by estimating the
26
missing value. Perform the exact analysis and compare the results. Solution: Bolt Chemical 1 2 3 4 1 73 73 75 73 2 68 67 68 71 3 74 x 78 75 4 71 72 73 75 5 67 70 68 69
Solution: For missing value, just replace x to 75.25 as computed in the Manual Computation and Graphical Illustration Section. After that, perform the above steps in 4-1 of this section.
Table 2: This is the output of the test performed.
Refer to Manual Computation and Graphical Illustration Section item 4-11 for the interpretation. 1. Shown below the yield (ton per 1/4-ha.plots) of sugar cane in a Latin square experiments comparing ve (5) fertilizer levels. Row 1 1 2 3 4 5 14(A) 19(B) 23(D) 21(C) 23(E) 2 22(E) 21(D) 15(A) 25(B) 16(C) Columns 3 20(B) 16(A) 20(C) 24(E) 23(D) 4 18(C) 23(E) 18(B) 21(D) 17(A) 5 25(D) 18(C) 23(E) 18(A) 19(B)
Where: A=no fertilizer C=10 tons manure/ha E=30 tons manure/ha B=complete inorganic fertilizer D=20 tons manure/ha
27
Solution: Step 1
Figure 9: Enter the data to SPSS in this manner.
Step 2
Figure 10: The second step after inputting your data, go to analyze general linear model univariate
28
Step 3
Figure 11: Next, enter the variable yield to the dependent list (yield dependent list) and treatment, row and column to fixed factor(s) (row and column fixed factor(s)). After that you can click the post hoc.. for choosing the test for multiple comparison.
Step 4
Figure 12: Before clicking the ok button, go rst to the model (seen on Step 3). In the univariate: model window, click custom then put the treatment, column, and row to the model box, as shown in the gure. Then, change the type to main effects and unchecked the include intercept in model, before the clicking the continue button.
29
Refer to the Manual Computation and Graphical Illustration Section for the interpretation of these outputs.