COLLEGE OF AGRICULTURE

DEPARTMENT OF PLANT SCIENCE

M.Sc in horticulture

BY:

Getahun Tefera (SGSR 0013/13)

Assignment

Submitted to Department Plant S ciences college of Agriculture,Wollo University, in Partial Fulfillment of

the Course Advance biometry

Submitted to:Yohannes G. (PhD.)

Submission date: 27/01/2020

JANUARY, 2020

DESSIE,ETHIOPIA
Assume the following table is grain yield of rice from varietal trials in 5*5 five varieties of rice.

C1 C2 C3 C4 C5 R TOTAL
R1 E C D B A 168
26 42 39 37 24
R2 A D E C B 148
24 33 21 32 38
R3 D B A E C 183
47 45 31 29 31
R4 B A C D E 173
38 24 36 41 34
R5 C E B A D 165
41 24 44 26 30
CTOTAL 176 168 171 165 157
837
Randomized Complete design (CRD)

Treatment replication Treatment

Total
1 2 3 4 5
A 24 24 31 24 26 129
B 37 38 45 38 44 202
C 42 32 31 36 41 182
D 39 33 47 41 30 190
E 26 21 29 34 24 134
R total 168 148 183 173 165 837 (grand

total)
Step .1 Construct ANOVA tables

Source of Degree of Sum square Mean square F calculated F tabulated at

variance freedom
1% 5%
Treatment
Error
Total
+

Step 2 calculate degree freedom(d.f) for treatment( t),error and total(T)

Treatment d.f= t-1=5-1=4

Total d.f = n-1= 25-1=24

Error d.f = T d.f- t d.f =24-4=20

()

() ( )

Error sum square (Ess)=Tss-tss= 1376.24-902.24=474

Step 4 calculate mean square for each source of variation by dividing each ss by its degree of
freedom

Treatment MS=
Error MS=

Step 5 calculate F value for treatment BY dividing mean square of treatment by mean square of
error

F calculated =

Step 6 Obtain F tabular value with error d.f = 20,and treatment df=4 at 1% and 5%

The F tabulated value from Appendix E with treatment d.f=4 and error d.f =20 are 2.87 at
5%significant level and 4.43 at 1%significant level.

Step 7 enter the all calculated value from step 2 to 6

Source of Degree of Sum square Mean square F calculated F tabulated at

variance freedom
5% 1%
Treatment 4 902.24 225.56 9.5** 2.87 4.43
Error 20 474 23.7
Total 24 1376.24

Step 8; compare the F calculated value and F tabulated value for each sour ce of variance

As a result indicates, the difference among the types of rice variety has highly significant
,because the F calculated for the type of rice greater than the F tabulated value; at 1% and
5%significant level .

sTEP 10. Compute the grand mean and the coefficient of variation cv as f ollows:

grand mean =

v
v

The experiment is reliable by reduced the experimental error IN THE experiment

Mean comparison by using LEAST significance difference (LSD)

Step 1 calculates the mean difference between dji and di of the treatment, then calculated as
follow;

Treatment mean in ascending order

Treatment A E C D B
Mean 25.8 26.8 36.4 38 40.5

25.8 26.8 36.4 38 40.5

40.5 14.7 13.7 4.1 2.5
38 12.2 11.2 1.6
36.4 10.6 9.6
26.8 1
25.8

Step 2. Calculate the LSD (least Significant Difference) value at a level of significance then
calculated as follow;

for equal replication

v v 8

at 5% and 1% is2.086 and 2.845 respectively

LSD (0.05 )= * =3.07*2.086 =6.42

LSD (0.01)= *at1%* =3.078*2.845=8.75

Where

=is the standard error of the mean differences

; the tabular t value at a and error d.f.

Step 3. compare the mean differences calculated in step 1 to the LSD

value calculated in step 2
and declare between the ith and jth treatments to be significantly
different at25.8
Treatment
the26.8
a level of
36.4 38 40.5
significance,
mea

40.5 14.7** 13.7** 4.1ns 2.5ns

38 12.2** 11.2** 1.6ns
36.4 10.6** 9.6**
26.8 1
25.8
Where **=significant at 1% significant level ,*=at 5% ,ns=non significant

Interpretation

The experiment is reliable. As a result indicate, among the variety A and B, D, C has
significance difference, but within A and E there is no significant difference at 1% significant
level. Among variety E and B, D, C have significance 1% significant level difference at 1%
significant level. Then within variety B, D and C of rice have no significance difference at 1 %
significant level and they are highly yielder than variety A, and E.
RCBD

Sep 1: the data grouped by treatment and the treatment total (T), block total (B) and grand total
(G) calculated as follow;
Treatment Replication
Treatment
Total
1 2 3 4
A 24 24 31 24 26 129 25.8
B 37 38 45 38 44 202 40.4
C 42 32 31 36 41 182 36.4
D 39 33 47 41 30 190 38
E 26 21 29 34 24 134 26.8
R total 168 148 183 173 165 837(grand

33.48
Grand
total)
mean

Step 2 outlines of the analysis of variance constructed as follow

Source of Degree of Sum square Mean square F calculated F tabulated at

variance freedom
1% 5%
Treatment
block
Error
Total

Step 3 calculate the degree of freedom for each source of variance then

They calculated as follow ;-

Total d.f. = r*t – 1=5*5-1=24

Treatment d.f. = t – 1=5-1=4
Block d.f. = r-1=5-1=4

Error d.f. = (t-1)(r-1)=(5-1)(5-1)=16= or Error d.f =T d.f-t d.f-b d.f=24-4-4=16

 Step 4 the correction factor and the various sums of squares (SS)
are Computed as follows:
Correction factors =

Treatment SS= =902.24

Replication or block SS=

TOTAL SS= ( )
Error SS =TSS-tss-rss =1376.24-902.24-131.44=342.56

Step 5 calculate mean square

Treatment MS =

Replication or block MS=

Error MS = ( )( )

6. Caculate the F value by

F calculated for treatment =

F Calculated for block=

1.53
STEP7.obtain F tabulated from APPENDEX E by using degree of freedom treatment=4, block
=4and error degree of freedom=16.
F tabulated for treatment at5% =3.01

At 1%=4.71
F tabulated for block at5% significance level 3.01
F tabulated for block at 1% significance level 4.71.
Step8. Enter all the calculated value f rom step2 to step7 in the ANOVA table, then entered as
follow;-

Source of Degree of Sum square Mean square F calculated F tabulated at

variance freedom 5% 1%
Treatment 4 902.24 225.56 10.53** 3.01 4.77
block 4 131.44 32.86 1.53ns
Error 16 342.56 21.41
Total 24 1376.24
**=significant at 1% significance level, ns=non significance
9. Compare F tabulated and F calculated
As result showed the treatments are highly significant, because the F
calculated 10.53 greater
than the F tabulated at 1% (4.77) and 5% (3.01) significance level.
step 10. Compute the grand mean and the coefficient of variation cv as follows
grand mean =

v
v

The Experiment is reliable

Mean comparison by using LSD (Least significance difference)

Step 1. Calculate the mean difference (dij) between the ith ( ) and jth ( )
treatment as:
Treatment A E C D B
Mean 25.8 26.8 36.4 38 40.5
25.8 26.8 36.4 38 40.5
40.5 14.7 13.7 4.1 2.5
38 12.2 11.2 1.6
36.4 10.6 9.6
26.8 1
25.8

Step 2. Calculate the LSD (least Significant Difference) value at a level of significance then
calculated as follow;
 then v v

at 5% and 1% is2.12 and 2.92 respectively

LSD (0.05 )= * =2.92*2.12 =6.19

LSD (0.01 )= * =2.92*2.92 =8.52

Step 3. compare the mean differences calculated in step 1 to the LSD value calculated in step 2
and declare between the ith and jth treatments to be significantly different at the a level of
significance,
Treatment 25.8 26.8 36.4 38 40.5
mean

40.5 14.7** 13.7** 4.1ns 2.5ns

38 12.2** 11.2** 1.6ns
36.4 10.6** 9.6**
26.8 1
25.8
**=significant at 1% significance level, ns=non significance
Interpretation
As a result indicate, among the variety A and B, D, C has highly significance diff erence, but
within A and E there is no significant difference at 1% significant level. Among variety E and B,
D, C are significantly different at 1% significant level difference at 1% significant level. Then
varieties B, D and C of rice are significantly different at 1 % significant level and 5% significant
leve. Generally the experiment is reliable by reduced experimental error (cv%is 13.8) and
Variety B,C.D OF RICE are highly yielder than AandE OF RICE VARIETY .

THE relative efficiency of blocking of RCBD design over CRD

As a result indicate in this experiment blocking in RCBD design is not efficient than CRD
because the calculated F value of blocking (1.53)is less than the tabulated VALUE even at 5%
significatleve(3.01).
LS design

Assume the following table is grain yield of rice from varietal trials in 5*5 five varieties of rice.

Column and row total computed as follow

C1 C2 C3 C4 C5 R TOTAL
R1 E C D B A 168
26 42 39 37 24
R2 A D E C B 148
24 33 21 32 38
R3 D B A E C 183
47 45 31 29 31
R4 B A C D E 173
38 24 36 41 34
R5 C E B A D 165
41 24 44 26 30
CTOTAL 176 168 171 165 157 837
Treatment total computed as follow

Treatment replication Treatment

Total
1 2 3 4 5
A 24 24 31 24 26 129
B 37 38 45 38 44 202
C 42 32 31 36 41 182
D 39 33 47 41 30 190
E 26 21 29 34 24 134
R total 168 148 183 173 165 837 (grand

total)
Step .1 Construct ANOVA tables

Source of Degree of Sum square Mean square F calculated F tabulated at

variance freedom
1% 5%
Treatment
Row
Column
Error
Total
2 . Calculate degree of freedom for each source of variance
Row =r-1=5-1=4

Column =c-1=5-1=4

Treatment =t-1=5-1=4

Total n-1=25-1=24

Error =(t-1)(t-2)=(5-1)(5- 2)=12

()

() ( )

RowSS= =

Replication or block SS=

Error sum square (Ess)=Tss-tsss-rss-css=1376.24-131.32-40.24-902.24=300.32

Step 5 calculate mean square

Treatment MS =

RowMS=

Column MS=

Error MS = ( )( )

6. Calculate the F value AS FOLLOW

F calculated for treatment =

F Calculated for ROW= 1.34

F Calculated for column =
0.401

STEP7.obtain F tabulated from APPENDEX E by using degree of freedom treatment=4, row

=4andcolumn =4with err or degree of freedom=16.
F tabulated for treatment at5% =3.26

At 1%=5.41
Step8. Enter all the calculated value from step2 to step7 in the ANOVA table, then entered as
follow;-
Source of
variance

Degree of Sum square Mean square F calculated F tabulated at

freedom
1% 5%
Treatment 4 902.24 225.56 9.001** 5.41 3.26
Row 4 131.44 32.86 1.34ns
Column 4 40.24 10.06 0.401ns
Error 12 300.32 25.026
Total 24 1376.24
**=significant at 1% significance level, ns=non significance
9. Compare F tabulated and F calculated
 As result show a treatments are highly significant, because the
F calculated 9.001 gr eater than
the F tabulated at 1% (5.41) and 5% (3.26) significance level.
Whereas row blocking and column
blocking are not significant al 1% and 5% significance level .
Step 10. Compute the grand mean and the coefficient of variation cv as follows

grand mean =

v
v

The experiment is conducted in the field and their cv is 14.96%.this indicate the reliability of
experiment so it is reliable.
Mean comparison by using LSD (Least significance difference)
Step 1. Calculate the mean difference (dij) between the ith ( ) and jth ( )
treatment as:
Treatment A E C D B
Mean 25.8 26.8 36.4 38 40.5
25.8 26.8 36.4 38 40.5
40.5 14.7 13.7 4.1 2.5
38 12.2 11.2 1.6
36.4 10.6 9.6
26.8 1
25.8

Step 2. Calculate the LSD (least Significant Difference) value at a level of significance then
calculated as follow;

then v v

at 5% and 1% is2.179 and 3.055 respectively

LSD (0.05 )= * =3.16*2.179 =6.89
Relative efficient of ls design over CRD=

()
() ( ) 9.66
LSD (0.01 )= * =3.16*3.055= ()

THE RESULT is 0.9637 this result even less than 1, it indicate the LS design is not efficient in reducing
Step 3. compare
experimental errorthe mean
in the differences
experiment over calculated in step 1 todesignbyusing
complete randomized the LSD valuetwocalculated
way blockingin step
. 2
and declare between the ith and jth treatments to be significantly different at the a level of
significance,
Relative efficient of LS design over RCBD
Treatment 25.8 26.8 36.4 38 40.5
mean
RE Row (RCBD)= ( )
()
40.5 14.7** 13.7** 4.1ns 2.5ns
RE Column (RCBD) = ( )
38 12.2** 11.2** 1.6ns
36.4 10.6** 9.6*
Then if the value error df is less than 20 the v alue should be multiply by precision factor k
26.8 1
25.8
THE K= (( )( ) )(( )( ) ) (( )( ) (((A)(and
) ) B, D, C has highly significance diff erence, but
As a result indicate,
(( )( ) )((( )( ) ) (( )( among
) )(( )( ) ) the variety
within A and E there is no significant difference at 1% significant level. Among variety E and B,
For row =1.068*0.968=1.033;then1.033-1=0.033*100==3.3%
D, are highly significantly different at 1% significant level but with variety c at 5% significant.
Then
For varieties
column B, D of rice are significantly different at 1 % significant level and variety C is
=0.88*0.968=0.85
significant at 5% significant level. Generally the experiment is reliable by reduced experimental
error
As (cv%indicates
aresult is 13.8)that andinVariety B, Dusing
experiment OF RICE are highly
Ls design columnyielder than
blocking notC;efficient
A and over
E OFRCBD
RICEdesign
VARIETY.
beceause the result is less tha 1,but row blocking increase the precision of the experiment over RCBD by
3.3%