http://www.unaab.edu.

ng

COURSE CODE: STS 352

COURSE TITLE: EXPERIMENTAL DESIGN 1

NUMBER OF UNIT: 2 UNITS

COURSE DURATION: TWO HOURS PER WEEK.

COURSE COORDINATOR: MR G.A. DAUDU

LECTURER OFFICE LOCATION: AMREC

COURSE CONTENT:
Basic concepts of experimentation, Completely randomized design, Randomised
complete block design, Latin Square Design, Graeco Latin Square Design, Simple
factorial Design

COURSE REQUIREMENTS:

This is a compulsory course for all statistics students. Students are expected to have
a minimum of 75% attendance to be able to write the final examination.

READING LIST:
1.) Statistical Design and Analysis of Experiments by P.W.M. John.
2.) Experimental Designs by Cochran and Cox.
3.) Designs and Analysis of Experiments for Biology and Agric. Students by
Oyejola, B.A.
4.) Statistical Methods by Snedecor and Cochran.
5.) Statistical Procedures for Agricultural Research by Gomez and Gomez.

LECTURE NOTES

Introduction
An experiment involves the planning, execution and collection of measurements
or observations.
Examples of simple experiment
1. Comparison of two teaching methods
2. Comparison of two varieties of maize

The difference among experimental units treated alike is called experimental

error, this error is the primary basis for deciding whether an observed difference is real or
 http://www.unaab.edu.ng

just due to chance. Clearly every experiment must be designed to have a measure of the
experimental error.

Definitions
Experimental Unit/plot
This is the smallest to which a treatment is applied, and on which an observation
is made e.g. an animal bird, an object, a cage, a field plat and so on.
- Definition of a unit depend on the objective of the experiment.
Factors
These are distinct types of condition that are manipulated on the experimental unit
e.g. age, group, gender, variety, fertilizer and so on.
Factor Levels
Different mode of the presence of a factors are called factor levels.
- Factors can be quantitative or qualitative.
Treatments
Each specific combination of the levels of different factors is called the treatment.
Replication
These are the numbers of experimental units to which a given treatment is
applied.

MAIN ASPECT OF DESIGNING EXPERIMENT

a. Choose the factor to be studied in the experiment and the levels of each factor that
are relevant to the investigation.
b. Consider the scope of inference and choose the type of experimental unit on which
treatment are to be applied.
c. From the perspective of cost and desired precision of inference, decide on the
number of units to be used for the experiment.
d. Finally, and most important, determine the manner in which the treatments are to be
applied to the experimental units (i.e. design of the experiment).

PRINCIPLES OF EXPERIMENTAL DESIGN

 http://www.unaab.edu.ng

There are three basic statistical requirements for a good experiment:

™ Randomization
™ Replication
™ Local Control or Blocking

1. RANDOMIZATION: This is the process by which it is ensured that each

treatment has an equal chance of being assigned to any experimental unit e.g.
9 Suppose two maize varieties, Yellow (Y) and White (W) are to be compared
using four experimental units for each

(I) (II)

In layout (II) if the field has fertility gradient so that there is a gradual
productivity from top to bottom. Then the white variety will be at advantage been
in a relatively more fertile area hence, the comparison within the variety would be
biased in favour of variety “W”.
A better layout is obtained by randomization as shown in layout (I).

2. REPLICATION: Each treatment being applied to more than one experimental

unit. Experimental error can be measured only if there are replications. Also the
more the experimental units used for each treatment, the lower would be the
standard error for the estimate for treatment effect and hence, the more precise the
experiment. Precision is the measurement of how close the observed values are to
each other.
3. BLOCKING OR LOCAL CONTROL: This is the process of grouping together
experimental units that are similar and assigning all treatments into each group or
block separately and independently. This allows the measurement of variation
among blocks which can be removed from the experimental error. Blocking is
therefore one of the measure for reducing or minimizing experimental error. The
 http://www.unaab.edu.ng

ability of detecting existing or real differences among treatments increase as the

size of the experimental error decreases.

COMPLETELY RANDOMIZE DESIGN (CRD)

Introduction
A CRD is a design in which the treatments are assigned completely at random so that
each experimental unit has equal chance of receiving any one treatment. Any difference
among the experimental units receiving the same treatment is considered to be
experimental error.
Model:
 http://www.unaab.edu.ng

yij = μi + eij
= μ + αi + eij
i = 1, 2, …, t and j = 1, 2, …, r
Where is the observed value for replicate j of treatment i, μ i is the population mean

for treatment i, μ is the population mean, is the effect of treatment i and eij is the
experimental error resulting from replicate j of treatment i.
Assumption: are assumed normally distributed about the mean, μi, and variance, σ2

or N (0, σ2) i.e. independently and identically normally distributed with mean 0

and constant variance σ2. Also ∑αi = 0,

Estimation of the Parameters

dS
= −2∑∑ ( y ij − μ − α i )
dμ i j

⇒ −2∑∑ ( y ij − μˆ − α i ) = 0
i j

∑∑ y − ∑∑ μˆ − ∑∑ α
i j
ij
i j i j
i =0

∑∑ y
i j
ij − rtμˆ − r ∑ α i = 0
i

Impose the constrain ∑α

i
i =0

⇒ rtμ̂ = ∑ yij

∴ μˆ =
∑∑ y ij
= y..
rt
dS
= −2∑ ( y ij − μ − α i )
dα i j
 http://www.unaab.edu.ng

− 2∑ ( y ij − μˆ − αˆ i ) = 0
j

∑ y − ∑ μˆ − ∑ αˆ
j
ij
j j
i =0

∑y j
ij − rμˆ − rαˆ i = 0

∑y ij

αˆ i = − μˆ
j

r
= y i. − y..

Randomization Procedure
1. Determine the total number of experimental units or plots (N) where N = rt with r
being the number of replications and t the number of treatments.
2. Assign a plot number to each experimental unit in any convenient manner
consecutively 1 to N.
3. Assign the treatments to the experimental units by any chosen randomization
scheme e.g. using table of random numbers, random number generator, drawing
of lots and so on.

Data Structure
Treatments
1 2… T
1 y11 y12 L y t1 y.1
 http://www.unaab.edu.ng

2 y12 y 22 L yt 2 y .2
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1

r y1r y2r L ytr y.r

y1. y 2. L yt. y y.

Analysis of Variance
The total variation in CRD is partitioned into two sources of variation i.e.
variation due to treatment and variation due to the error. The relative size of the two
variations is used to indicate whether the observed difference among the treatment means
is significant or due to chance, the treatment difference is said to be significant if the
treatment variation is significantly larger than the experimental error.
Total sum of squares, SST,
2
t ni y
SST = ∑∑ ( y ij − y..) 2 = ∑∑ y −
2
..

i =1 j =1
ij N

Sum of squares due to treatment SSB

2 2
t ni t t y y
SSB = ∑∑ ( y ii. − y..) 2 = ∑ ni ( y i. − y .. ) 2 = ∑ i.
− ..

i =1 j =1 i =1 i =1 ni N

Sum of Square due to Error, SSE

t ni
SSE = ∑∑ ( yij − y i. ) 2
i =1 j =1

2
y
C.F = ..
= correcting factor
N
⇒ SST = SSB + SSE
i.e.SSE = SST − SSB

ASSIGNMENT
 http://www.unaab.edu.ng

• Show that:

∑∑ ( y
i j
ij − y.. ) 2 = ∑∑ ( y i. − y.. )2 + ∑∑ ( yij − y i. ) 2
i j i j

ANOVA TABLE

Source of variation Degree of Sum of Means squares F-ratio

freedom Squares
Between treatment t-1 SSB SSB MSB
MSB = Fc = = Fc
t −1 MSE
Error N-t SSE SSE
MSE =
N −t
Total N-1 SST

If there are no differences in the effect of the treatment Fc follows the F-

distribution. Hence, if Fc > FT where FT is the table value from the F-Table with t – 1 and
N – t degrees of freedom at a given significance level, then the effect are said to be
significantly different

Or

Reject H0 if FC FT

COMPARISON OF MEANS
If a significant result is declared then there is need to identified the mean that are
different and this can be done using multiple comparison of means such as
LSD – Least Significant Difference
DMRT – Duncan’s Multiple Range Test Turkey
Scheffee etc.

LSD = tSED
 http://www.unaab.edu.ng

=r

If the observe difference between any two means is greater than the LSD value then those
two means are said to be significantly different.

COEFFICIENT OF VARIATION
This is a measure of precision of the estimates obtained from the data. It is also
used to assess the quality of the management of an experiment. A low coefficient of
variation indicates high precision of estimate or efficient management of the experiment.

Example: In an effort to improve the quality of recording tapes, the effect of four kinds of
coating A, B, C, D on the reproducing quality of sound are to be compared. The
measurements of sound distortion are given below.

A 10 15 8 12 15
B 14 8 31 15
C 17 16 14 15 17 15 18
D 12 15 17 15 16 15
Recommend the best coating for the sound production.

ADVANTAGE OF CRD
1. The design is very flexible
2. The statistical analysis is simple
 http://www.unaab.edu.ng

3. It has high degrees of freedom relative to other designs

4. It is best for small experiments

DISADVANTAGE OF CRD
Design is very inefficient if units are not homogenous.

ASSIGNMENT
1. Analyze the following data from a field experiment with four treatments using 1%
significance level. Carryout mean comparison if necessary. How good is the
management of the experiment.
A 14.3 11.6 11.8 14.2
B 20.7 21.0
C 32.6 32.1 33.0
D 24.3 25.2 24.8

2. Three fertilizer sources A, B, C, were each applied to seven plots chosen at

random in a field of carrot. Analyze the data using 5% significance level.
Carryout mean comparison if necessary. How efficient was the management of
the experiment
A 24 18 18 29 22 17 15
B 46 39 37 50 44 45 30
C 32 20 26 41 36 28 27

RANDOMISED COMPLETE BLOCK DESIGN (RCBD)

 http://www.unaab.edu.ng

Introduction

The design is used when the experimental unit can be grouped such that the
number of units in a group is equal to number of treatments. The groups are called blocks
or replicates and the purpose of grouping is to have units in a group as homogeneous as
possible so that observed differences in a group are mainly due to treatments. Variability
within group is expected to be lower than variability between groups. Since the number
of units per block equal the number of treatments, the blocks are of equal size hence, the
design is a complete block design. The primary purpose of blocking is to reduce the
experimental error by eliminating the known sources of variability.
Model:

yij = μ + αi + βj + eij
i = 1, 2, …, t and j = 1, 2, …, r
where is the observed value for block j of treatment i, μ is the population mean, is
the effect of treatment i, βj is the effect of block j and eij is the experimental error
resulting from block j of treatment i.
Assumption:
- block and treatment effect are additive,

- N (0, σ2)

- ∑αi = 0, ∑βj = 0,

Estimation of Parameters
A procedure similar to that used in CRD can be utilized here to obtain the desired
estimates.
Randomization and Layout
The randomization process for randomised complete block design is applied separately
and independently to each of the block.
- Divide the experimental area into r-blocks.
- Sub-divide the block into t-experimental units. Where t is the number of
treatments.
 http://www.unaab.edu.ng

- Number the plot consectively from I to t and assign the t-treatment at random to
the t-unit within each block following any randomization scheme.
DATA STRUCTURE
Blocks
Treatment 1 2 3…r Total
1
2
3 y 31 y 32 y 33 ... y 3r
1 1 1 1
1 1 1 1
1 1 1 1

t
Total y.3 ... y.r y..

Analysis of Variance
The total variation is partitioned into the variation due to blocks, variation due
treatments, and variation due to error. i.e.
2
y
SSTotal = ∑∑ ( yij − y .. ) = ∑∑ y −
2
2 ..
ij N
2 2
y y
SS Trt = ∑∑ ( y i. − y .. ) 2 = ∑ i.
− ..

r N
2 2
y y
SS Block = ∑∑ ( y ij − y .. ) 2 = ∑
.j
− ..

t N
SSE = SSTotal − SSTrt − SS Block
 http://www.unaab.edu.ng

ANOVA TABLE
Source Df SS MS F
Block r–1 SSB
MSB =
r −1
Treatment t–1 SST
MST =
t −1
Error (r-1)(t-1) SSE

Total rt – 1

Hypothesis

Or

Comparing the calculated F-ratios to the table F-value at a given significance level, we
decide to reject or fail to reject the null hypothesis. i.e. Reject if
Reject if

COMPARISON OF MEANS
Use LSD to compare the treatments if the F-ratio is found to be significant.

where

Coefficient of variation

CAUSES OF MISSING VALUES AND THEIR ESTIMATIONS

A missing data can occur whenever a valid observation is not available for any
one of the experimental units, occurrence of missing data result in two major problems
i.e. loss of information and non applicability of standard analysis of variance.
 http://www.unaab.edu.ng

COMMON CAUSES OF MISSING DATA include:

1. When intended treatment is not applied i.e. improper treatment.
2. When experimental plants are destroyed probably due to poor germination,
physical damage, pest damage etc. This causes total or high percentage of the
plants in a plot to be destroyed such that no meaningful observation can be made
on the plot.
3. Loss of harvested sample: This may result from the fact that some plant characters
cannot be conveniently recorded either in the field or immediately after harvest
due to some other process required. Hence some samples will be lost between the
time of harvesting and actual recording of data.
4. This happens after data have been recorded and transcribed generally referred to a
illogical data. The value may be too extreme as a result of misread observation or
incorrect transcription.

ESTIMATION OF MISSING VALUE FROM RCBD

Let x be the missing value

,
,

⇒ xˆ − μˆ − αˆ i − β j = 0 ,

=
 http://www.unaab.edu.ng

=0

where G0 is the grand total excluding the missing value.

• T0 is the total observed value for the treatment that contained the missing value.
• B0 is the total observed value for the Block that contained the missing value.

Note that: the degree of freedom must be adjusted by the number of missing values i.e.
reduce the number of degrees of freedom by the number of missing values.

ADVANTAGES OF RANDOMIZED COMPLETE BLOCK DESIGNS

1. A reduction of experiment error due to blocking is expected.
2. Estimation of missing value is easy to compute.
3. The ANOVA is also easy to compute.

DISADVANTAGES OF RCBD
1. Not best for large number of treatments.
2. More tasking in the execution of the design than the CRD.
3. Missing value can create problem especially in estimation and non formal
analysis.
4. The precision will be affected due to missing values.

RELATIVE EFFICIENCY
Blocking maximizes the difference among blocks. Hence it is necessary to examine how
much is gained by the introduction of blocking into the design. The magnitude of the
 http://www.unaab.edu.ng

reduction in the experimental error due to blocking over the CRD can be obtained by

computing relative efficiency.

Where is the block mean square and the is the error mean square.

1. Example: In an experiment to examine the respond of maize to nutrient fertilizer

application. Six treatments were used in four blocks. Analyze the data and
recommend the appropriate fertilizer rate.
F1 F2 F3 F4 F5 F6 TOTAL
I 0.42 0.46 0.60 2.63 2.81 3.39 10.31
II 0.22 0.42 0.86 2.39 2.82 2.50 9.22
III 0.31 0.58 0.83 1.68 1.95 2.50 7.85
IV 0.33 0.55 0.89 2.22 2.81 2.10 8.9
TOTAL 1.28 2.01 3.18 8.92 10.40 10.49 36.28
2. The following data are yield of groundnut in a variety trial involving five varieties
of groundnut using four replications in randomized complete block design. The
data has one missing value. Analyze the data and make your recommendation.

V1 V2 V3 V4 V5
I 2.73 1.73 3.00 2.13 0.27 9.86
II * 2.27 1.53 2.40 0.27 6.47
III 2.33 2.00 2.13 1.87 0.10 8.43
IV 2.53 2.00 2.00 1.47 0.13 8.13
7.59 8.00 8.66 7.87 0.77 32.89

ASSIGNMENT
In an experiment to test the effect of five level of potash (ABCDE) on the yield of
cotton the following strength indices were obtain as given below. One of the data point is
missing. Analyze the data and compare the mean of the five level of potash and make
necessary recommendation. How effectively was the experiment carried out. Was there
any gain in precision in using RCBD over CRD
 http://www.unaab.edu.ng

A (7.62) C (*) E (7.46) D (7.17) B (8.14)

E (7.68) B (8.15) C (7.73) D (7.57) A (800)
C (7.47) E (7.20) A ( (7.93) B (7.87) D (7.80)

LATIN SQUARE DESIGN

Introduction
The major feature of the latin square design is its capacity to simultaneously handle two
known sources of variation among the experimental units. These are commonly referred
to as row blocking and column blocking. It is ensured that every treatment occurs only
once in each row and once in each column. Hence the variation due to row and column
can be estimated and removed from the experimental error.
Note that: the presence of row and column blocking also constitute a restriction. This is
due to the requirement that all treatment appear in each row and in each column. This is
only satisfied if the number of replications equal the number of treatments. Hence, for
large number of treatments, the design is not practicable. Also when the number of
treatments is small the degrees of freedom for the error becomes too small for the error to
be reliably estimated, the design is therefore not generally, widely adopted.
 http://www.unaab.edu.ng

Model:
yijk = μ + αi + βj + δk + eijk
where i, j, k = 1, 2, … , t
where is the observed value from row j and column k receiving treatment i.

- is the overall mean

- is the effect of treatment i
- is the effect of row j

- is the effect of row k

- is the random error component for row j and column k receiving treatment i.

Assumptions:
The model is completely additive i.e. there is no interaction between the rows, columns
and treatments.

N (0, σ2)

and

Estimation of Parameters
A procedure similar to that used in CRD can be utilized here to obtain the desired
estimates.

RANDOMIZATION PROCEDURE
1. Obtain a square field partitioned into t rows and t columns.
2. Arrange the treatment into the unit in a standard form.
3. Randomize between the columns
4. Randomize between the rows

Example: Consider an experiment with four treatments to be compared using latin square
design i.e. 4 x 4 LS
 http://www.unaab.edu.ng

Standard form Randomize btw

col
* A B C D 1 B D C A
2431
* B C D A 2 C A D B
* C D A B 3 D B A C
* D A B C 4 A C B D
1 2 3 4

Randomize between row

1 C A D B
2341
2 D B A C
3 A C B D

4 B D C A
ANALYSIS OF VARIANCE
The total variation is partitioned into components for row, column, treatment and
error. The sum of squares are obtained in the usual form.
ANOVA TABLE
Source Df SS MS F
MSR
Rows t-1 SSR MSR /MSE
MSC
Columns t-1 SSC MSC /MSE
MS
Treatments t-1 SSTrt MSTrt Trt/MSE

Error (t-1) (t-2) SSE MSE

Total t2-1 SST

BLOCKING EFFICIENCY
The efficiency of both row and column blockings in a latin square design indicate the
gain in precision relative to either the CRD or RCBD.
 http://www.unaab.edu.ng

RELATIVE EFFICIENCY OF LSD TO CRD

Relative efficiency of a latin square design as compared to CRD is given by

Where Er, Ec, Ee are the mean squares row, column, and error respectively with t as the
number of treatment. For an R.E = 325% it indicates that the use of LSD is estimated to
increase the experimental precision by 225% while if the R.E is less than 100% means
that there is no gain.
RELATIVE EFFICIENCY LSD TO RCBD
Relative efficiency of latin square design as compared to RCBD can be computed in two
ways i.e. when rows are considered as blocks and when columns are considered as blocks
of the RCBD.

I II III IV I
II
III
IV

1) =

2) =

Example: Suppose we have

Result indicate that, the additional column blocking by use of latin square design is
estimated to have increased the precision over that of the RCBD with row as block by
294%. However, the additional row blocking in the LS Design did not increase the
precision over the RCBD with column as blocks. Hence for this experiment the
Randomized Complete Block Design with column as blocks would have been as efficient
as the LS Design.
MISSING VALUE ESTIMATION
Missing value in the latin square experiment can be estimated as follows
 http://www.unaab.edu.ng

Where the R0, C0, T0 are the total of the row, column and treatment respectively that
contain the missing observation. Again one degree of freedom is subtracted from both
total and error degrees of freedom in the case of one missing value.
ADVANTAGES AND DISADVANTAGES
1. The elimination of two sources of variation often lead to a smaller error mean
square than would be obtained by use of CRD and RCBD.
2. ANOVA is simple
3. Missing values can easily be handled
Disadvantages.
1. Assumption of no interaction between different factors may not hold.
2. Unlike the CRD and RCBD, the number of treatments is restricted to the number
of replications. Hence it is limited in application.
3. For large number of treatments such as t > 12, the square becomes too large and
does not remain homogeneous.
4. For small number of treatments such as t < 3, degrees of freedom for the error is
usually too small for any meaningful comparison or conclusion.
5. A square field is often required for the design and this may not be practicable.

Example: The following show the field layout and yield of a 5 x 5 latin square
experiment on the effect of spacing on yield of millet, the spacing are: A(2cm), B(4cm),
C(6cm), D(8cm) and E(10cm)

Column
Row 1 2 3 4 5 Total
1 B: 257 E: 230 A: 279 C: 287 D: 202 1255
2 D: 245 A: 283 E: 245 B: 280 C: 260 1313
3 E: 182 B: 252 C: 280 D: 246 A: 250 1210
4 A: 203 C: 204 D: 227 E: 193 B: 259 1086
5 C: 231 D: 271 B: 266 A: 334 E:339 1440
 http://www.unaab.edu.ng

Total 1118 1240 1297 1340 1309 6304

GRAECO LATIN SQUARE DESIGN (GLSD)

Introduction
In the Graeco latin square design, the treatments are grouped into replicates in three
different ways: this triple grouping is to eliminate from the error, three sources of
blocking of blocking variation. Recall the earlier design:
CRD - No blocking variation
RCBD - Single blocking variation
Latin Square - Double blocking variation
Graeco latin - Tripple blocking variation
Thus GLSD provide more opportunity than the other designs in the reduction of error
through skillful planning. The experimental unit should be arranged and the experiment
conducted so that differences in the three directions represent major sources of variation.
RANDOMIZATION
Arrange the rows and the columns independently at random including the treatment.
Assign the subscript at random to their respective classification. The treatment and
subscript must appear once in a column and once in a row, and must appear together only
once. The layout can be difficult to design but like in the latin square design, once we
obtain a standard form then we can randomize between the columns and between the
rows.
I II III I II III
I A1 B3 C2 C3 A2 B1
II B2 C1 A3 A1 B3 C2
III C3 A2 B1 B2 C1 A3
Randomize between rows

I II III
A2 B1 C3
 http://www.unaab.edu.ng

B3 C2 A1
C1 A3 B2
Randomize between columns

T=4 A, B, C. D
I II III IV
I A1 C4 B3 D2
II B2 D3 A4 C1
III C3 A2 D1 B4
IV D4 B1 C2 A3

T=5 A, B, C, D, E
I II III IV V
A1 B3 C5 D2 E4
B2 C4 D1 E3 A5
C3 D5 E2 A4 B1
D4 E1 A3 B5 C2
E5 A2 B4 C1 D3

MODEL
yijkl = μ + αi + βj + δk + τl + eijkl
where i, j, k, l = 1, 2, … , t
where is the observed value from row j and column k receiving treatment i.

is the grand mean

is the effect of treatment i
 http://www.unaab.edu.ng

is the effect of row j

is the effect of column k

τl is the effect of subscript factor l.
is the random error component

Assumptions:
The model is completely additive i.e. no interaction between the row, the column, the
subscript factor and the treatment

N (0, σ2)

, and ∑τl = 0

Estimation of Parameters
A procedure similar to that used in CRD can be utilized here to obtain the desired
estimates.

STATISTICAL ANALYSIS
The total variation is petition into five components i.e. the row, column, subscript factor,
treatment and error.

The sum of square are obtain in the usual form i.e.

SSR =
∑y 2
. j ..
−
y 2....
t N
 http://www.unaab.edu.ng

ANOVA TABLE
Source Df SS MS F
MSR
Rows t-1 SSR MSR /MSE
MSC
Column t-1 SSC MSC /MSE
MS
Subscript t-1 SSSubscript MSSubscript Subscript/MSE
MS
Treatments t-1 SSTrt MSTrt Trt/MSE

Error (t-1) (t-3) SSE MSE

Total t2-1 SST

Note the followings:

1. The number of replications equals the number of treatments hence for large
number of treatments, the design is not practicable (t > 12).
2. The experimental error is likely to increase with the size of the square.
3. Small square provide only a few degree of freedom for the error.
4. Experimental units are difficult to balance conveniently in all the three groupings.
Example
Consider the following data which was obtain from an experiment to study the sampling
error, the data consist of five samplers (A – E) being controlled for order of sampling,
area of sampling and qualification of sampler. Analyse the data.

AREA
Order 1 2 3 4 5 Total
1 A1 (3.5) B3 (4.2) C5 (6.7) D2 (6.6) E4 (4.1) 25.1
2 B2 (8.9) C4 (1.9) D1 (5.8) B3 (4.5) A5 (2.4) 23.5
3 C3 (9.6) D5 (1.7) E2 (2.7) A4 (3.7) B1 (6.0) 25.7
4 D4 (10.6) E1 (10.2) A3 (4.6) B5 (3.7) C2 (5.1) 34.1
5 E5 (3.1) A2 (7.2) B4 (4.0) C1 (3.3) D3 (3.5) 21.1
Total 35.6 27.2 23.8 21.8 21.1 129.5
 http://www.unaab.edu.ng

ASSIGNMENT
The data below is obtained from an experiment using Graeco Latin Square Design with
four diet, (A, B, C, D). breed I, II, III, IV weight group {1, 2, 3, 4} feed concentration {i
ii iii iv}. Is there any significant difference between the diets. If any, compare them and
make necessary recommendation. Also comment on the management of the experiment.
Breeds
I II III IV
1 A1 (5.9) B3 (4.2) C4 (10.2) D2 (6.6)
2 B2 (8.9) A4 (4.5) D3 (6.0) C1 (3.0)
3 C3 (9.6) D1 (5.8) A2 (7.2) B4 (4.6)
4 D4 (10.5) C2 (4.1) B1 (3.5) A3 (6.7)
SIMPLE FACTORIAL EXPERIMENT
Introduction
Factorial experiments are used in the study of the effects of two or more factors. In
factorial experiments, all the possible combinations of the level of the factors make up
the treatments. For example, if there are two factor A, B each with ‘a’ and ‘b’ levels
respectively, then we have ‘ab’ treatment combinations.
A B
Maize variety Fertilizer Rate

There are 10 treatment combinations. Also factorial experiments allow us to investigate

the interaction between the factors. That is, how the levels of a factor perform in the
presence of the levels of another factor. We are able to answer the question on how the
responses to one factor were affected by another.
A B
Maize Variety Fertilizer Rate
a0 White b0 0 kg/ha
a1 Yellow b1 30 kg/ha
 http://www.unaab.edu.ng

In analyzing data from a factorial experiment, we would be interested in the main effect
and the interaction effect of a factor. The main effect of a factor is a measure of the
change in response in the level of a factor averaged over all levels of the other factors.
For example, let two factors A and B be at two level a0, a1 and b0, b1 respectively with
treatment combinations a0b0, a0b1, a1b0, a1b1. The main effect of A is a measure of change
in A from a0 to a1 averaged over the two levels of B.
i.e. At level b0 of B: the simple effect of A is a1b0 – a0b0.
Similarly, at level b1 of B: the simple effect of A is a1b1 – a0b1..
Main effect of A =

The r represent the replication, where each treatment total response is from r units.
Also at level a0 of A: simple effect of B is a0b1 – a0b0.
Similarly, at level a1 of A simple effect of B is
Thus
Main effect of B =

Each effect of a factor at a given level of the other factor is known as simple effect. The
interaction effect is the differential response to one factor in combination with varying
levels of a second factor. That is, an additional effect due to the combined influence of
two or more factors. For example, interaction between A and B (AB) is estimated as the
difference between two simple effects. and

i.e. the interaction effect AB =

BA =

Consider the following result

Mean Responses
b0 b1 b0 b1 b0 b1
a0 20 30 20 30 20 40
a1 40 50 30 50 50 30
(i) (ii) (iii)
 http://www.unaab.edu.ng

From above illustration (i) shows the case of no interaction, (ii) shows the case of mild
interaction and (iii) shows the case of strong interaction.

LAYOUT
Any of the earlier design discussed can be used, in particular, the RCBD. The treatment
combinations are assigned to each block randomly
For example, consider the case of two factors A and B, each at two levels a0, a1 and b0, b1
respectively. The treatment combinations are a0b0, a0b1, a1b0, a1b1.
a0b1
 http://www.unaab.edu.ng

RCBD a1b1 CRD

MODEL CRD

where is the observed value from the kth unit corresponding to level i of factor A and

the jth level of B

is the grand mean
is the effect of level i of factor A
is the effect of the jth of factor B

is the effect of interaction between and .

is the random error component.

MODEL RCBD

Where Tk is the effect of the kth block

H0:
H1: atleast one
H20:
H21: atleast one

H30:

H31: {there is interaction between A and B}

ANALYSIS OF VARIANCE
 http://www.unaab.edu.ng

The total variation is partitioned into that due to factor A, factor B, the interaction AB
and the error i.e.

where the SST, SSA, SSB, SSAB and SSE are obtain in the usual form i.e.

-SSA- SSB

ANOVA TABLE
Source Df SS MS F
MSA
A a-1 SSA MSA /MSE
MSB
B b-1 SSB MSB /MSE
MSAB
AB (a-1) (b-1) SSAB MSAB /MSE
ERROR ab(r-1) SSE MSE
Total abr-1 SST
Example
An engineer designing a battery for use in a device that would be subjected to some
extreme variation in temperature has three types of plate materials to use. He decided to
test the plate materials under three temperature settings (150F, 700F, 1250F) to see their
effect on the life of a battery. Four test runs are to be made at each treatment
combination. Test the effect of temperature and plate material and their possible
interaction on the battery life.
 http://www.unaab.edu.ng

Temperature
Type 150F 700F 1250F
A

539 229 230 998

B

623 469 198 1290

C

576 583 342 1501

1738 1281 770 3789
ASSIGNMENT
Three types of tyres are to be compared using four different brand of cars. The threading
on the tyres are measured after a period of use. Below is the rescaled data. Are the tyres
significantly different? Also does the performance of the tyres depend on the brand of the
car?

Tyres
Car brand A B C
1

3
 http://www.unaab.edu.ng