1.

SOME BASIC CONCEPTS:

1.1. Factorial design:

 Factorial experiment is an experiment whose design consist of two or more factor each with
different possible values or levels.

 To evaluate the combined effect of two or more factors when they are used simultaneously.

 𝑍𝑍𝐾𝐾 = K factors, Z levels.

23 = 3factors, 2levels.

32= 2factors, 3 levels.

 Factorial design is divided into two types:

i. Full factorial design (FD).

ii. Fractional factorial design (FFD).

i. Full factorial design (FD):

 It consists of all possible factor combinations in an experiment and factors varies at

simultaneously rather than one factor at given time.
 Number of runs (N)=𝑍𝑍𝐾𝐾

where, Z = number of levels,

𝐾𝐾 = number of factors

E.g.- 3 factors, 2 levels each, N = 23 = 8 runs.

ii. Fractional factorial design (FFD):

 In full factorial design as a number of factor or level increases, the number of experiments
required exceeds to unmanageable levels.
 In such cases, the number of experiments can be reduced systemically and resulting design is
called as fractional factorial design.
 Applied when number of factors are > 5.

 In general, 2𝑘𝑘−𝑝𝑝 design is a ( 12)𝑝𝑝 fraction of a 2𝑘𝑘 design using 2𝑘𝑘−𝑝𝑝 runs.

 For example: 25−2 design is a (12)2 of a 25 design using 25−2 = 23 =8 runs.

1
2. INTRODUCTION:

2.1. PLACKETT BURMAN DESIGN (PBD):

In a classic 1946 published research paper in the journal of Biometrika, the Plackett and Burman
design showed how to construct 2-level orthogonal designs when the number of runs ‘N’ is a
multiple of 4 (𝑁𝑁 =4,8,12,16,20,24, and so on). If the run size is a power of R=2(for example, 𝑁𝑁 =
8,16,32, ...), these designs are identical to the fractional factorial designs. Asper the literature, this
is called as popular screening design. These designs are very efficient screening designs when only
the main effects are of interest to evaluate in complex way. These are useful for detecting large
main effects economically, assuming all interactions are negligible when compared with important
main effects. These are two-level fractional factorial designs for studying k =n -1 variables or
factors in n runs, where n is a multiple of 4 (n = 4,8,12,16,20,24, and so on). It is a two-level
design, each variable represented in two levels i.e., High (+) and Low (-). Each horizontal row
represents a trial or runs and each vertical column represents the either of 2 levels (High or Low
level).
Table 1: COMPARISON

Full Factorial Designs Fractional factorial Designs Plackett-Burman design

Considers all possible Considers only a fraction of Suitable for screening

combinations of levels all possible combinations, experiments when the number
for each factor. reducing the number of of factors is large.
runs.

Requires a large number Requires a smaller number Requires very few runs for
of runs. of runs compared with the a large number of variables.
factorial design.

Represented as 2𝑘𝑘 , where k Represented as 2𝑘𝑘−𝑝𝑝 where p Multiples of 4.

is the number of factors. is the fraction of the full
factorial design.

Used when a Mainly used for screening Primarily used for screening
comprehensive experiments when the experiments, especially
understanding of all factors number of factors is large. when resources are limited.
and interactions is required.

Inefficient for large numbers More efficient than a full Efficient for screening a large
of factors would be factorial design, allowing for number of factors with a
exponentially distributed. a reduction in the number of minimal number of runs
runs.

2
3. DESIGN RESOLUTION:
Plackett Burman Design are Resolution III.

Resolution III designs:

 Designs in which no main effects are aliased (chain) with any other main effects.
 Main effects are aliased with at least two-factors interactions.
 Some two-factor interactions may be aliased with each other.

4. GENERATOR OF ROWS FOR CONSTRUCTING PB DESIGN.

Table 2: Plus and Minus Signs for the Plackett–Burman Designs

𝒌𝒌=Factors 𝑵𝑵=Runs Plus, and minus signs

8-11 12 ++-+++---+-

16-19 20 ++--++++-+-+----++-

20-23 24 +++++-+-++--++--+-+----

32-35 36 -+-+++---+++++-+++--+----+-+-++--+-

Table 3: PBD FOR UP TO 𝟏𝟏𝟏𝟏 FACTORS IN N=𝟏𝟏𝟏𝟏 RUNS

FACTORS
Runs A B C D E F G H I J K
1 + - + - - - + + + - +
2 + + - + - - - + + + -
3 - + + - + - - - + + +
4 + - + + - + - - - + +
5 + + - + + - + - - - +
6 + + + - + + - + - - -
7 - + + + - + + - + - -
8 - - + + + - + + - + -
9 - - - + + + - + + - +
10 + - - - + + + - + + -
11 - + - - - + + + - + +
12 - - - - - - - - - - -

3
5. METHODOLOGY:
1. Using F- values.
2. Using alias chain.
3. Using R- programme.

5.1. USING F- VALUES:

a. Determine the difference between the + (high) and - (low) responses for each independent and
dummy variable.
Difference = ∑ 𝐇𝐇𝐇𝐇𝐇𝐇𝐇𝐇 𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥 (+) - ∑ 𝐋𝐋𝐋𝐋𝐋𝐋 𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥(−)
b. Determine effect of each variable
(∑ 𝐇𝐇𝐇𝐇𝐇𝐇𝐇𝐇 𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥(+)−∑ 𝐋𝐋𝐋𝐋𝐋𝐋 𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥(−))
c. Effect of Factors =
𝐍𝐍

Where, N = number of runs.

d. Estimate the mean square of each variable (the variance of effect).
(∑ 𝐇𝐇𝐇𝐇𝐇𝐇𝐇𝐇 𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥(+)−∑ 𝐋𝐋𝐋𝐋𝐋𝐋 𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥𝐥(−))2
Factor Mean square =
𝐍𝐍

e. The experimental error can be calculated by averaging the mean squares of the dummy
effects.
∑ (𝐌𝐌𝐌𝐌𝐌𝐌𝐌𝐌 𝐬𝐬𝐬𝐬𝐬𝐬𝐬𝐬𝐬𝐬𝐬𝐬 𝐨𝐨𝐨𝐨 𝐝𝐝𝐝𝐝𝐝𝐝𝐝𝐝𝐝𝐝 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯)
EMS =
𝐍𝐍𝐍𝐍𝐍𝐍𝐍𝐍𝐍𝐍𝐍𝐍 𝐨𝐨𝐨𝐨 𝐝𝐝𝐝𝐝𝐝𝐝𝐝𝐝𝐝𝐝 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯

f. The final stage is to identify the factors which are showing large effects and this was done
using an F-test.
𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅𝐅 𝐌𝐌𝐌𝐌𝐌𝐌𝐌𝐌 𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒
F=
𝐄𝐄𝐄𝐄𝐄𝐄𝐄𝐄𝐄𝐄 𝐌𝐌𝐌𝐌𝐌𝐌𝐌𝐌 𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒𝐒

g. Larger F-ratios and smaller p- values suggest more significant effects.

Table 4: ANOVA
Df SS MSS F-ratio p-value
Source of Variation

Model (Factors) 𝑘𝑘 SSF SSF/k MSF/MSE XXX

Residual (Error) 𝑁𝑁 - 𝑘𝑘 - 1 SSE SSE/(N-k-1) - -

Total 𝑁𝑁 - 1 SST - - -

4
 𝑘𝑘=Number of factors and 𝑁𝑁= Total number of runs.

 The ANOVA table can guide decisions on which factors are likely to have a substantial
impact.
 A low p-value for a factor indicates that the factor has a significant effect on the response
variable.

5.2. USING ALIAS CHAIN:

Example: A human performance analyst is conducting an experiment to study eye focus time
and has built an apparatus in which several factors can be controlled during the test.
Seven factors are

A= Sharpness of vision
B= distance from target to eye
C= target shape
D = illumination level
E = target size
F = target density
G = subject
Table 5: Design for the eye focus time experiment (𝒏𝒏=8, 𝒌𝒌=7).
Factors

Runs A B C D=AB E=AC F=BC G=ABC Time

1 - - - + + + - 85.5

2 + - - - - + + 75.1

3 - + - - + - + 93.2

4 + + - + - - - 145.4

5 - - + + - - + 83.7

6 + - + - + - - 77.6

7 - + + - - + - 95.0

8 + + + + + + + 141.8

5
Seven main effects and their aliases may be estimated from these data, the effects and their aliases
are:

 [A]=20.63 A+BD+CE+FG

 [B]=38.38 B+AD+CF+EG

 [C]=-0.28 C+AE+BF+DG

 [D]=28.88 D+AB+CG+EF

 [E]=-0.28 E+AC+BG+DF

 [F]=0.63 F+BC+AG+DE

 [G]=-2.43 G+CD+BE+AF

5.3. USING R PROGRAMME:

> Library (BsMD)

> data (PB12Des, package = “BsMd”)

>colnames (PB12Des) <- c ("c11", "c10", "c9", "c8", "G", "F", "E”, “D", "C”, “B","A")

>castf <- PB12Des [c (11,10,9,8,7,6,5,4,3,2,1)]

>y<c(4.733, 4.625, 5.899, 7.0, 5.752, 5.682, 6.607, 5.818, + 5.917, 5.863, 6.058, 4.8)

> castf <- cbind (castf, y)

> modpb <- lm (y ~ (.), data = castf)

> library(daewr)

> cfs <- coef(modpb) [2,12]

> names<-names(cfs)

>model<-aov (y~(.), data=castf)

>Summary(model)

> halfnorm (cfs, names, alpha =.35, refline=FALSE

6. MERITS AND DEMERITS:

6.1. MERITS:

 This design helps us to explore which factor is found to be highly important in an

6
  agricultural field trial, multi centric drug trial experimentation, DNA sequcing analysis.
 To reduce experimental runs with different experiment block.
 This design screens out less scope factors from design.

 This design acts as navigation tool for enabling a quick reduction in the number of potential
factors whichever the factor researcher affixed.

6.2. DEMERITS:

 The selection of factors is very crucial, and if key factors are omitted from the design, the
results may be incomplete.

 The researcher will allow careful consideration and research domain knowledge will ignore
higher order interaction.
 enquired during selection of factors on predominant basis.
 It requires prior knowledge.

7. APPLICATIONS:

7.1. Pharmaceutical industry:

 Screening drug formulations in different age groups when the drug is effective or
ineffective.
 Optimizing manufacturing processes.

 To identify the critical factors affecting the product quality.

7.2. Chemical manufacturing:

 Different concentrations were screened to know factors influencing yield, purity, reaction
kinetics and chemical structure of substances.
7.3. Food and beverage industry:

 Assessing factors attributed taste at different persons with different time intervals, texture,
and shelf life.
7.4. Biotechnology:

 Identifying the key variables in the bioprocess optimization.

 Screening factors influencing cell culture, cell biology, fermentation, and product yield.

7.5. Agricultural research:

 Assessing factors influencing crop yield and quality, meteorological parameters.

7
8. CASE STUDIES:

8.1. CASE STUDY 1:

This study focuses on Screening of process components and their effects on production of lactase
by newly isolated Bacillus sp. VUVD101 strain from dairy effluent by using PBD.

8.1.1. OBJECTIVES:

 Lactase production increases.

 Screening the factors affecting of the production process.

 Isolated bacterium increases more profit.

 The Plackett- Burman statistical design is very frequently used to study the effects of broth

components on lactase production.

 It is a two factorial (i.e. -1 and +1) design that locates significant variables for the production
by “n” variables.
 14 factors chosen in the present investigation were tested at these two levels, based on the
Plackett- Burman matrix design.
 The main effect was calculated basically as a difference between the average measurements
of each variable made at a high level (+) and low level (-).

8.1.2. RESULTS AND DISCUSSION:

Table 6: The variables and levels used in statistical design for screening medium
components affecting lactase production.

Variable Code Variable Name Minimum Value Maximum Value

A Incubation time 10 40

B Temperature 25 40

C pH 6 10

D RPM 100 200

E DO 1 3

F Inoculum size 0.25 1

G Inoculum age 24 72

8
 H MgSO4 0.5 1

J L-Cysteine 0.1 1

K KH2PO4 0.01 0.05

L CaCl2 0.001 0.005

M K2PHO4 0.01 0.05

N Corn steep liquor 0.5 1

O Lactose 0.5 2

Table 7: Effect of variables on lactase production by bacillus sp. VUVD101 using Plackett- Burman
design
Runs A B C D E F G H J K L M N O Lactase
activity
U/ml
1 40 40 10 100 3 0.25 24 0.5 0.1 0.05 0.001 0.05 0.5 2 0.82

2 10 40 10 100 3 1 72 0.5 0.1 0.05 0.005 0.01 1 2 1.85

3 40 40 6 100 3 1 24 1 1 0.01 0.001 0.01 0.5 2 6.10
4 10 40 10 200 3 0.25 24 1 1 0.01 0.005 0.05 0.5 0.5 5.06
5 10 25 6 200 3 0.25 72 1 1 0.05 0.001 0.01 1 2 15.85
6 10 40 6 200 1 0.25 24 1 0.1 0.05 0.001 0.05 1 2 6.82
7 40 40 6 100 1 0.25 72 0.5 1 0.01 0.005 0.05 1 2 2.00
8 10 25 10 100 1 1 24 1 1 0.05 0.005 0.01 0.5 2 5.75
9 40 40 6 200 1 1 72 0.5 1 0.05 0.001 0.01 0.5 0.5 4.28
10 10 25 10 200 1 1 72 0.5 0.1 0.01 0.001 0.05 0.5 2 8.60
11 40 40 10 200 1 0.25 72 1 0.1 0.05 0.005 0.01 0.5 0.5 2.01
12 40 25 6 100 1 1 24 1 0.1 0.05 0.005 0.05 1 0.5 5.35
13 40 25 10 200 1 0.25 24 0.5 1 0.01 0.005 0.01 1 2 5.18
14 40 25 10 100 3 0.25 72 1 0.1 0.01 0.001 0.01 1 0.5 2.50
15 10 25 6 100 3 0.25 72 0.5 1 0.05 0.005 0.05 0.5 0.5 9.01
16 40 25 10 200 3 1 24 0.5 1 0.05 0.001 0.05 1 0.5 2.00
17 10 25 6 100 1 0.25 24 0.5 0.1 0.01 0.001 0.01 0.5 0.5 7.92

9
18 40 25 6 200 3 1 72 1 0.1 0.01 0.005 0.05 0.5 2 16.25
19 10 40 6 200 3 1 24 0.5 0.1 0.01 0.005 0.01 1 0.5 2.99
20 10 40 10 100 1 1 72 1 1 0.01 0.001 0.05 1 0.5 1.02

Table 8: ANOVA
Source Df Adj. SS Adj. MS F-Value p-Value

Model 14 347.14 24.796 8.60 0.013

Incubation time 1 16.85 16.85 5.84 0.06

Temperature 1 103.345 103.345 35.83 0.002

pH 1 87.196 87.196 30.23 0.003

RPM 1 35.725 35.725 12.39 0.017

DO 1 9.09 9.09 3.15 0.136

Inoculum size 1 0.443 0.443 0.15 0.711

Inoculum age 1 11.833 11.833 4.10 0.099

MgSO4 1 24.358 24.358 8.44 0.034

L-Cysteine 1 0.064 0.064 0.02 0.887

KH2PO4 1 0.753 0.753 0.26 0.631

CaCl2 1 0.010 0.010 0.00 0.954

K2PHO4 1 0.3130 0.3130 0.11 0.755

Corn steep liquor 1 20.462 20.462 7.09 0.045

Lactose 1 36.697 36.697 12.72 0.016

Error 5 14.422 2.884 - -

Total 19 361.562 - - -

Main effect of variables on production by Bacillus sp. VUVD101.

a. Normal plot of standardized effects (Fig.1) and
b. Pareto chart of standard effects (Fig.

10
 Fig. 1 Fig. 2

8.1.3. CONCLUSION:

This study had led to the isolation of bacterium, Bacillus sp. VUVD101 strain with high activity
for lactose hydrolysis. The components namely temperature, pH, RPM, MgSo4, Corn steep
liquor and lactose were found to be highly significant to achieve the maximum production of
lactose (18.31U/ml). These results revealed that the isolated bacterium could be used as a good
source for industrial production of lactase.

8.2. CASE STUDY 2:

This study focuses on application of Plackett-Burman Design for screening the media components
for tannase production from red gram husk using submerged fermentation.

8.2.1. OBJECTIVE:

To identify which ingredients of medium, have significant effect on tannase enzyme production.
 Placket-Burman design used in screening experiment as the number of experiments run
required are very few, leading to saving of time, chemicals and man power.
 Placket-Burman design was used for screening the media components to enhance tannase
enzyme production.

 This design does not consider the interaction effects between the variables and is used to
screen the important variables affecting tannase production. It can represent by first-order
polynomial Equation: 𝑌𝑌= 𝛽𝛽0+ ∑𝛽𝛽𝑖𝑖𝑋𝑋𝑖𝑖 .
Where, 𝑌𝑌 represents the response, β0 is the model coefficient, βi is the linear coefficient,

11
 𝑋𝑋𝑋𝑋 is the level of the independent variable.
 The statistical software package ‘Design expert 7.1.5’, was used for analyzing the
experimental data.
 The 12 variables (Table 5) and inducer tannic acid were screened in 20 experimental trials.

 The low level (–1) and high level (+1) of each factor are listed in (Table 5).

8.2.2. RESULTS AND DISCUSSION:

Table 9: The effect of Variables on Tannase Enzyme Production by A. foetidus using PBD

Run A B C D E F G H I J K L Tannase Activity (U/ml)

No.
Experimental Predicted

1 1 1 1 1 -1 -1 1 1 -1 1 1 -1 44.35 33.35
2 -1 1 -1 1 -1 1 1 1 1 -1 -1 1 89.09 86.97
3 -1 -1 1 -1 1 -1 1 1 1 1 -1 -1 87.98 85.79
4 -1 1 1 -1 -1 -1 -1 1 -1 1 -1 1 48.49 56.01
5 1 -1 1 -1 1 1 1 1 -1 -1 1 1 78.74 88.91
6 1 -1 -1 -1 -1 1 -1 1 -1 1 1 1 88.11 80.05
7 -1 -1 -1 1 -1 1 -1 1 1 1 1 -1 70.53 77.26
8 1 -1 1 1 -1 -1 -1 -1 1 -1 1 -1 49.96 55.37
9 -1 1 1 -1 1 1 -1 -1 -1 -1 1 -1 120.40 108.95
10 1 1 -1 -1 1 1 -1 1 1 -1 -1 -1 105.80 103.64
11 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 100.43 103.92
12 -1 1 1 1 1 -1 -1 1 1 -1 1 1 64.70 64.94
13 -1 -1 1 1 -1 1 1 -1 -1 -1 -1 1 87.40 81.16
14 -1 -1 -1 -1 1 -1 1 -1 1 1 1 1 99.65 98.06
15 1 -1 1 1 1 1 -1 -1 1 1 -1 1 47.56 38.96
16 -1 1 -1 1 1 1 1 -1 -1 1 1 -1 77.65 85.43
17 1 1 1 -1 -1 1 1 -1 1 1 -1 -1 53.20 55.71
18 1 1 -1 -1 -1 -1 1 -1 1 -1 1 1 89.60 91.38
19 1 -1 -1 1 1 -1 1 1 -1 -1 -1 -1 69.34 70.24
20 1 1 -1 1 1 -1 -1 -1 -1 1 -1 1 36.76 43.67

12
  The effect of 12 variables namely concentrations of eleven nutrients and tannic acid as
inducer on tannase enzyme production by submerged fermentation by A. Foetidus were
analyzed.
 Table 9 shows the Plackett–Burman experimental design of experiments and the results
obtained from the experiments which are generated by the MINITAB 15 software.
 From the table 5, it was observed that the variation in tannase activity was 36.76–120.40
U/ml.
 In run no.9, the maximum tannase activity was obtained with the medium.

 In run no.20, the minimum tannase activity was obtained with the medium.

Table 10: Statistical analysis of medium optimization using Plackett-Burman design for
tannase production using A. foetidus.
Nutrient Nutrient Minimum Maximum t-value p-value
Code Value Value
A Tannic acid 1 5 3.40 0.011
B Yeast extract 0.1 1 1.19 0.273
C Magnesium sulphate 0.1 1 3.31 0.013
D Ferrous sulphate 0.1 1 3.85 0.006
E Ammonium nitrate 0.1 1 0.47 0.655
F Ammonium chloride 0.1 1 1.89 0.101
G Urea 0.1 1 1.08 0.317
H Potassium chloride 0.1 1 0.88 0.407
I Sodium nitrate 0.1 1 0.43 0.678
J Potassium di-hydrogen 0.1 1 3.96 0.005
phosphate
K Ammonium sulphate 0.1 1 1.75 0.124
L Peptone 0.1 1 0.11 0.918

 On analysis of regression coefficient (t -value) of 12 medium components in (Table 6),

Ammonium sulphate, ammonium nitrate, ammonium chloride, urea, showed positive
effect on tannase production, Whereas the remaining components showed negative effect
on tannase production.
 There is a close agreement between the experimental values of tannase production and
theoretical values predicted by PB design model equation for all the medium components.
 The Pareto chart as shown in (Fig.3) offers a convenient way to view the results obtained
by P.B. Design and the order of significance of the variable affecting tannase production.

13
 Fig.3: Pareto Plot for Plackett-Burman Design of Experiments for Tannase
Production Using A. foetidus.

8.2.3. CONCLUSION:
The Plackett-burman design was effectively applied for screening of nutrients for the
production of tannase from aspergillus foetidus (MTCC 3557) using redgram husk as a
substrate in submerged fermentation. From standard Plackett-burman data analysis it was
conformed that, concentration of tannic acid, concentration of potassium di-hydrogen
phosphate, concentration of magnesium sulphate and concentration of ferrous sulphate
were found to be the most significant for tannase enzyme production.

8.3. CASE STUDY 3:

 The researcher conducted the experimentation at in vitro condition to know the plasma
concentration level of newly administered drug significance, he has included the 31
geographical areas with the following factors:
a. CD4: X1 f. SGOT: X6

b. Gender: X2 g. SGPT: X7

c. BMI: X3 h. Blood urea: X8

d. PPBS:X4 i. TC: X9

e. WBC: X5

 He wishes to determine the experimental reliability in PB design.

14
8.3.1. RESULTS AND DISCUSSION:

Table 11: Statistical analysis using Plackett-Burman design.

Effect Coefficient t-test P-value

𝑪𝑪𝑪𝑪𝟒𝟒 count: X1 – 25.52 339.96 0.000

Gender: X2 0.492 0.27 3.27 0.047

BMI: X3 6.311 3.15 41.95 0.000

PPBS:X4 −4.85 −2.4 32.28 0.000

WBC: X5 −5.27 −2.65 35.03 0.000

SGOT: X6 0.82 0.410 5.45 0.012

SGPT: X8 1.01 0.507 6.75 0.000

Blood urea: X6 1.84 0.921 12.24 0.000

TC: X7 0.75 0.372 5.01 0.015

8.3.2. CONCLUSION:

The PBD is well suited for massive data sets of life-threatening diseases (HIV, Cancer, other
neurological disorders) if the patient repeatedly undervent treatment with periodically change
of hematological parameters and biochemical parameters.

9. SUMMARY:
Plackett-Burman design is a statistical technique used to identify important factors that affect
a process. It is a type of screening design that allows researchers to efficiently explore a large
number of factors with a small number of experimental runs. This seminar provides valuable
insights into PBD and their construction, methodologies and emphasizing their significance
in screening most important factors. Additionally, three case studies are presented: one focuses
on screening of process components and their effects on production of lactase, while second
case study involves screening the media components for tannase production from red gram
husk and third case study presents the plasma concentration level of newly administered drug
significance.

15
It is great tool for optimizing processes and reducing costs in various fields, including
manufacturing and agriculture.

10. REFERENCES:
ABRAHAM, P. KARLAPUDI, KRUPANIDHI, S., ERVA, R., M. INDIRA, MD. N. BOBBY,
AND VEKATESWARULU, T.C., 2018, Plackett-burman design for screening of process
components and their effects on production of lactase by newly isolated bacillus sp.
Vuvd101 strain from dairy effluent. Beni-Suef. univ. j. basic appl.Sci.,7: 543-546.

BASAVARAJAIAH, D. M. AND BHAMIDIPATI NARASIMHA MURTHY, 2020, Design of

Experiments and Advanced Statistical Techniques in Clinical Research. Elsevier, pp.:
117-120.
GEORGE, E. P. BOX, J. STUART HUNTER AND WILLIAM, G. HUNTER, 2015, Statistics
for Experimental Design, Innovation and Discovery. John Wiley & Sons, Inc., Hoboken,
New Jersey, pp.: 281-282.

JOHN LAWSON, 2015, Design and Analysis of Experiments with R. Taylor & Francis Group,
New York, pp.:231-232.

JOHANNES LEDOLTER AND ARTHUR J. SWERSEY, 2007, Testing 1–2–3 experimental

design with applications in marketing and service operations. Stanford University Press,
Stanford, California, pp.:150.
MONTGOMERY, D.C., 2012, Design and Analysis of Experiments. John Wiley &Sons, Inc.,
Hoboken, New Jersey, pp.:351-357.
MOHAN, S. K., VIRUTHAGIRI, T. AND ARUNKUMAR, C., 2013, Application of Plackett-
Burman design for screening the media components for tannase production from red gram
husk using submerged fermentation. Int. J. Pharma res. Rev., 2 (9): 24-29.

PETER F. STANBURY, ALLAN WHITAKER, AND STEPHEN J. HALL, 2013, Principles of

16
fermentation technology. Elsevier, pp.:110-112.

17
 UNIVERSITY OF AGRICULTURAL SCIENCES, GKVK, BENGALURU.
DEPARTMENT OF AGRICULTURAL STATISTICS,
APPLIED MATHEMATICS AND COMPUTER SCIENCE.

NAME: Nandhini Laxman Naik DATE: 13-01-2024

ID NO.: PAMB 2147 TIME: 9:30 AM
CLASS: Sr. M.Sc. (Agril. Statistics)

Seminar Synopsis
Plackett-Burman Design.

1. INTRODUCTION
In a classic 1946 published research paper in the journal of Biometrika, the Plackett and Burman
design showed how to construct 2-level orthogonal designs when the number of runs ‘𝑁𝑁’ is a
multiple of 4 ( 𝑁𝑁 = 4,8,12,16,20,24, and so on). If the run size is a power of R=2(for
example, 𝑁𝑁 = 8,16,32, ...), these designs are identical to the fractional factorial designs. Asper
the literature, this is called as popular screening design. These designs are very efficient
screening designs when only the main effects are of interest to evaluate in complex way. These
are useful for detecting large main effects economically, assuming all interactions are negligible
when compared with important main effects. These are two-level fractional factorial designs
for studying k =n -1 variables or factors in n runs, where n is a multiple of 4 (n =
4,8,12,16,20,24, and so on). It is a two-level design, each variable represented in two levels i.e.,
High (+) and Low (-). Each horizontal row represents a trial or runs and each vertical column
represents the either of 2 levels (High or Low level).

2. DESIGN RESOLUTION
Plackett-Burman designs are Resolution III. Designs in which no main effects are aliased
(chain) with any other main effects. Main effects are aliased with at least two-factors
interactions. Some two-factor interactions may be aliased with each other.

3. GENERATOR OF ROWS FOR CONSTRUCTING PB DESIGN

𝒌𝒌=Factors 𝑵𝑵=Runs Plus, and minus signs

8-11 12 ++-+++---+-
16-19 20 ++--++++-+-+----++-
20-23 24 +++++-+-++--++--+-+----
32-35 36 -+-+++---+++++-+++--+----+-+-++--+-
4. MERITS AND DEMERITS
Merits:
 To reduce experimental runs with different experiment block.
 This design helps us to explore which factor is found to be highly important in an
agricultural field trial, multi centric drug trial experimentation, DNA sequencing
analysis.
Demerits:
 Ignore higher order interaction.
 The selection of factors is very crucial, and if key factors are omitted from the design, the
results may be incomplete.

5. CONCLUSION
Plackett-Burman design is a statistical technique used to identify important factors that affect a
process. It is a type of screening design that allows researchers to efficiently explore a large
number of factors with a small number of experimental runs.

6. REFERENCES:
ABRAHAM, P. KARLAPUDI, KRUPANIDHI, S., ERVA, R., M. INDIRA, MD. N. BOBBY,
AND VEKATESWARULU, T.C., 2018, Plackett-burman design for screening of process
components and their effects on production of lactase by newly isolated bacillus sp.
Vuvd101 strain from dairy effluent. Beni-Suef. univ. j. basic appl.Sci.,7: 543-546.

BASAVARAJAIAH, D. M. AND BHAMIDIPATI NARASIMHA MURTHY, 2020, Design of

Experiments and Advanced Statistical Techniques in Clinical Research. Elsevier, pp.:
117-120.
GEORGE, E. P. BOX, J. STUART HUNTER AND WILLIAM, G. HUNTER, 2015, Statistics
for Experimenteral Design, Innovation and Discovery. John Wiley & Sons, Inc.,
Hoboken, New Jersey, pp.: 281-282.

JOHN LAWSON, 2015, Design and Analysis of Experiments with R. Taylor & Francis Group,
New York, pp.:231-232.

JOHANNES LEDOLTER AND ARTHUR J. SWERSEY, 2007, Testing 1–2–3 experimental

design with applications in marketing and service operations. Stanford University
Press, Stanford, California, pp.:150.
MONTGOMERY, D.C., 2012, Design and Analysis of Experiments. John Wiley &Sons, Inc.,
Hoboken, New Jersey, pp.:351-357.
MOHAN, S. K., VIRUTHAGIRI, T. AND ARUNKUMAR, C., 2013, Application of Plackett-
Burman design for screening the media components for tannase production from redgram
husk using submerged fermentation. Int. J. Pharma res. Rev., 2 (9): 24-29.

PETER F. STANBURY, ALLAN WHITAKER, AND STEPHEN J. HALL, 2013, Principles of

fermentation technology. Elsevier, pp.:110-112.