Scribd
Upload
26 views

3. Blocking and Confounding System

The document discusses blocking and confounding in two-level factorial designs, emphasizing their role in controlling variability and improving experimental accuracy. Blocking groups similar experimental units to minimize nuisance factors, while confounding mixes the effects of factors with block effects to streamline analysis. The text also covers the application of these concepts in both unreplicated and replicated designs, highlighting their importance in managing error variance.

Uploaded by

shaikhmariya431
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
26 views

3. Blocking and Confounding System

The document discusses blocking and confounding in two-level factorial designs, emphasizing their role in controlling variability and improving experimental accuracy. Blocking groups similar experimental units to minimize nuisance factors, while confounding mixes the effects of factors with block effects to streamline analysis. The text also covers the application of these concepts in both unreplicated and replicated designs, highlighting their importance in managing error variance.

Uploaded by

shaikhmariya431
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 11

Blocking and

Confounding
System for Two-
Level Factorial
- Dr. Khushwant Yadav
Index

● Blocking

● 2k Factorial Design

● Blocking in Unreplicated Designs

● Confounding

● Blocking in Replicated Designs to Reduce Error Variance


Blocking
Blocking is a technique for dealing with nuisance factors.
A nuisance factor is a factor that has some effect on the response, but is of no
interest to the experimenter; however, the variability it transmits to the
response needs to be minimized or explained.

● Blocking helps control unwanted variations (like time, location, or material


differences) that could affect results.
● It groups similar experimental units together to reduce variability caused by
external factors.
● Blocking factors are conditions that can influence results but are not of primary
interest (like temperature, operator, time of day).
● Randomized Complete Block Design (RCBD) is used when one blocking factor
exists.
2k Factorial Designs

Factorial designs are the basis for another important principle besides blocking -
examining several factors simultaneously. We will start by looking at just two factors
and then generalize to more than two factors. Investigating multiple factors in the
same design automatically gives us replication for each of the factors.

The 2𝑘 refers to designs with k factors where each factor has just two levels.
Blocking in Unreplicated Designs
An unreplicated design means that each treatment combination appears only once in the
experiment.
Blocking is used to control variability by grouping similar experimental units into
blocks. A common rule is to sacrifice (confound) an interaction effect (usually the
highest-order interaction) to define the blocks

Treatment A. B AB Block
Table 2: Generalizing to k=3 design
-1 -1 -1 1 1
Treatment I A B C AB AC BC ABC
a 1 -1 -1 2
-1 1 -1 -1 -1 1 1 1 -1
b -1 1 -1 2
a 1 1 -1 -1 -1 -1 1 1
ab 1 1 1 1 b 1 -1 1 -1 -1 1 -1 1
Table 1: Assigning treatments to blocks ab 1 1 1 -1 1 -1 -1 -1
c 1 -1 -1 1 1 -1 -1 1
ac 1 1 -1 1 -1 1 -1 -1
bc 1 -1 1 1 -1 -1 1 -1
abc 1 1 1 1 1 1 1 1
Table 3: Blocking with Block Size = 4 Table 4: Blocking with Block Size = 2

Treatment ABC Block Treatment AB AC BC Block


-1 -1 1 -1 1 1 1 4
ab -1 1 a -1 -1 1 1
ac -1 1
b -1 1 -1 3
bc -1 1
ab 1 -1 -1 2
a 1 2
c 1 -1 -1 2
b 1 2
ac -1 1 -1 3
c 1 2
abc 1 2 bc -1 -1 1 1
abc 1 1 1 4
Confounding
● Confounding occurs when the effect of a factor or interaction is mixed with the
block effect.
● It is intentionally introduced in blocked designs to control variability and reduce
noise.
● In two-level factorial designs, higher-order interactions like ABC are often
confounded with blocks.
● Confounding helps in dividing the experiment into blocks without increasing the
experiment size.
● The effect that is confounded cannot be estimated separately from the block
effect.
● It is generally applied in replicated designs where controlling unwanted variation
is essential.
● Using confounding allows for a more practical and manageable analysis of
factorial experiments.
Blocking in Replicated Designs to Reduce
Error Variance
In replicated designs, we have multiple replicates of the same design, allowing us to extract more
information about the confounded effects.

Replicate
A repeat of the entire experiment to improve accuracy.
(Rep)

A group of experimental runs that share similar conditions to


Block (Blk)
reduce variability.

Confounded An interaction (e.g., ABC) whose variation is mixed with the


Effect block effect, making it impossible to estimate separately.

Random Factors like replicates that are assumed to be drawn from a


Effects larger population and are analyzed differently.
Yijkl​=μ+Repi​+Blk(Rep)ij​+Ak​+Bl​+Cm​+ABCk​lm+ϵijkl​
where:
i → Replicate index
j → Block index within replicate
k,l,m→ Treatment factor indices
ϵijkl→ Random error term

Table 5: ANOVA for Blocking in Replicated Designs


Source of Variation Degrees of Freedom (df) Formula

Replicate (Rep) n−1 3−1=2

Blocks within Replicate


n(b−1) 3(4−1)=9
(Blk(Rep))

Factor A 2−1 1
Factor B 2−1 1
Factor C 2−1 1
Interaction ABC 2−1 1
Error (n−1)×4 8
Total 23
THANK YOU

You might also like