Define BIBD with its all Parameters.

A Balanced Incomplete Block Design (BIBD) is a statistical tool used in the design of experiments to structure experimental trials efficiently. BIBDs are particularly useful in situations where testing every possible combination of treatments (or items) is either impractical or too costly, and where we want to ensure that every treatment is compared against others as fairly as possible.

In BIBDs, treatments are arranged in blocks, but not all treatments are included in each block. The "incomplete" nature of BIBD refers to the fact that not all treatments are included in each block, while "balanced" refers to the fact that each treatment appears in an equal number of blocks and each pair of treatments occurs together in a block the same number of times. The design ensures that comparisons between treatments are conducted uniformly across the experiment.

Definition of BIBD

A Balanced Incomplete Block Design (BIBD) is a combinatorial design defined by a set of elements and blocks. It is denoted as a $(v,b,r,k,\lambda )$-BIBD, where:

• $v$ represents the total number of treatments or varieties (also called points or elements),
• $b$ represents the total number of blocks,
• $r$ is the number of blocks in which each treatment appears,
• $k$ is the number of treatments in each block (size of the block),
• $\lambda$ is the number of times each pair of distinct treatments occurs together in the same block.

A BIBD satisfies the following conditions:

1. Each block contains exactly $k$ distinct treatments.
2. Every treatment appears in exactly $r$ blocks.
3. Any pair of distinct treatments appears together in exactly $\lambda$ blocks.

Parameters of a BIBD

A BIBD is completely characterized by the five parameters: $v$, $b$, $r$, $k$, and $\lambda$, and these parameters must satisfy certain relationships to ensure that the design is balanced and valid.

1. $v$ (Number of treatments)

• $v$ is the total number of distinct treatments (or objects) under study in the experiment.
• For example, in an agricultural experiment, these could represent different crop varieties, fertilizer types, or irrigation methods.

2. $b$ (Number of blocks)

• $b$ is the total number of blocks or experimental units in which the treatments are assigned.
• A block represents a group of treatments that are tested or compared together. For example, in a crop field experiment, a block might represent a single plot of land where some subset of the treatments (e.g., crop varieties) is grown.

3. $r$ (Replications per treatment)

• $r$ is the number of blocks in which each treatment appears. In other words, each treatment appears in exactly $r$ different blocks.
• This ensures that every treatment is tested multiple times, improving the reliability of the experiment’s conclusions.

4. $k$ (Size of each block)

• $k$ is the number of treatments in each block.
• In practice, this means that instead of testing all treatments in every block (as in a complete block design), we only test a subset of $k$ treatments in each block. This allows for more efficient designs when the number of treatments $v$ is large.

5. $\lambda$ (Number of times each pair of treatments appears together)

• $\lambda$ is the number of times each pair of distinct treatments appears together in the same block.
• This parameter ensures the "balance" in the design. Each pair of treatments appears in exactly the same number of blocks, ensuring that no treatment pair is favored over others when making comparisons.

Relationships Among Parameters

The parameters , $b$, $r$, , and $\mathrm{\lambda }$in a BIBD are not independent of one another. They are linked by the following relationships:

1. Total number of treatment occurrences: Each of the $v$ treatments appears in exactly $r$ blocks, and each block contains $k$ treatments. Hence, the total number of treatment occurrences across all blocks must be the same whether counted by blocks or by treatments:

vr=bk 

This equation shows that the total number of occurrences of treatments in blocks, $vr$, is equal to the total number of occurrences of blocks multiplied by the block size, $b$.

2. Pairwise occurrence of treatments: Each treatment appears in $r$ blocks, and every pair of treatments appears together in $\$ blocks. There are $\binom{v}{2}$ distinct pairs of treatments, and in each of the $b$ blocks, there are $\binom{k}{2}$ pairs of treatments. Therefore, the total number of times pairs of treatments appear together must be the same whether counted by blocks or by pairs of treatments:

$$\lambda (v-1)=r(k-1)$$

This is the most crucial relationship in a BIBD. It ensures that each pair of treatments appears together the same number of times across all blocks, contributing to the "balance" of the design.

Example of a BIBD

Consider a BIBD with parameters $(v,b,r,k,\lambda )=(7,7,3,3,1).\; Here:$

• $v=\mathrm{7 }$: There are 7 treatments (let's label them as $T_1,T_2,\dots ,T_{\mathrm{7 }}$).
• $b=7$: There are 7 blocks.
• $r=3$: Each treatment appears in 3 blocks.
• $k=3$: Each block contains 3 treatments.
• $\lambda =1$: Each pair of treatments appears together in exactly 1 block.

An example of the block structure might look like this:

• Block 1: $\{T_1,T_2,T_3\}$
  
• Block 2: $\{T_1,T_4,T_5\}$
  
• Block 3: $\{T_1,T_6,T_7\}$
  
• Block 4: $\{T_2,T_4,T_6\}$
• Block 5: $\{T_2,T_5,T_7\}$
  
• Block 6: $\{T_3,T_4,T_7\}$
  
• Block 7: $\{T_3,T_5,T_6\}$
  

In this design:

• Each treatment appears in exactly 3 blocks (e.g., $T_1$ appears in Blocks 1, 2, and 3).
• Each pair of treatments appears together in exactly 1 block (e.g., $T_1$ and $T_{\mathrm{2 }}$ appear together only in Block 1).

Applications of BIBD

BIBDs are widely used in agricultural experiments, industrial testing, and other experimental settings where a full factorial design is impractical. Some specific applications include:

• Agriculture: Testing crop varieties, fertilizer treatments, or irrigation methods in field plots, where resources or space are limited.
• Pharmaceutical research: Testing combinations of drugs or treatments in clinical trials.
• Industrial experiments: Testing combinations of materials, machines, or production processes.

By using BIBDs, experimenters can ensure that every treatment and pair of treatments are tested fairly, even when a complete test of all combinations is infeasible. The balance inherent in these designs leads to more reliable and interpretable results, while the incomplete structure reduces the resources required to conduct the experiment.

Subscribe on YouTube - NotesWorld

For PDF copy of Solved Assignment

Any University Assignment Solution

WhatsApp - 9113311883 (Paid)