Experimental

Research
Designs
Contents
1 CRD
2RCBD
 COMPLETE
RANDOMIZED DESIGN
The CRD is best suited for experiments
(CRD)
with a small number of treatments.

Did u Did u Did u

know?? know?? know??
• Design can • Because of the
• Simplest
be used when homogeneity
design to experimental requirement, it
use units are may be difficult to
essentially use this design for
homogeneous. field experiments
 RANDOMIZATION
DID U
PROCEDURE
Did u
KNOW? know??
• Treatments •
are assigned Randomization
to is performed
experimental using a
units random
completely at number table,
Did u
random.
know?? computer,
• Every program, etc.
experimental
unit has the
same
probability of
receiving any
Example of
Randomization
Given you have 4
treatments (A, B, C,
and D) and 5
replicates, how many
experimental units
would you have?
Every experimental Note that there is no
unit has the same “blocking” of
probability of experimental units
receiving any into replicates.
treatment.
ADVANTAGE
S OF CRD
 Did u
know??
Statistical
analysis is
Very flexible design simple Loss of information
(i.e. number of due to missing data
compared to
treatments and is small compared to
replicates is only other
other designs due to
limited by the designs. the larger number of
available number of degrees of freedom
experimental units). for the error source
of variation.
 DISADVANTAG
Did u
know??
ES OF CRD Did u
know??
If experimental
units are not Not suited
homogeneous for a large
and you fail to
minimize this number of
variation using treatments.
blocking, there
may
D i d be
u a loss of
k nprecision.
ow??
Usually the
least efficient
design unless
experimental
units are
homogeneous.
EXAMPLE:
Experiment: Investigating the effect of different
fertilizer types on corn yield.

Treatments:Three different fertilizer types (A, B, and C)

with 5 replicates.
Experimental units: Individual plots of land.

Replication: Each treatment (fertilizer type) will be

applied to n identical plots (e.g., n = 5). This creates
n groups of plots, with each group receiving the same
fertilizer type.
EXAMPLE:
PROCEDURE:

Randomization: Divide the available land into 3n

identical plots. Randomly assign each plot to receive
one of the three fertilizer types. This ensures each
treatment has an equal chance of being assigned to
any plot within the experiment.
Replication implementation: Within each group of
n plots assigned the same fertilizer type, further
randomize the plot order (optional). This helps to
control for potential positional effects within the field.
Treatment application: Apply the designated
SEATWORK
:1. Comparing the effectiveness of different pain relief
medications.
Treatment: Three different pain relief medications (A,
B, and C).
Experimental units: Individual participants
experiencing pain.
Replication: 10
Procedure:
Randomly assign participants to receive one of the three
medications.
Administer the assigned medication to each participant.
Measure pain level using a standardized scale before and
after medication administration.
SEATWORK
:1. Evaluating the impact of different learning methods on
student performance:
Treatment: Three different learning methods (traditional
lecture, group discussion, and online learning.
Experimental units: Individual students in a specific course.
Replication: 15
Procedure:
Randomly assign students to one of the three learning
methods.
Conduct the course using the assigned method for each
group.
Administer an identical exam to all students at the end of
the course.
Thank you
Randomized Complete
Block Design
 What is it?

• Probably the most used and useful of the

experimental designs.
• Takes advantage of grouping similar
experimental units into blocks or replicates.
• The blocks of experimental units should be
as uniform as possible.
• The purpose of grouping experimental units
is to have the units in a block as uniform as
possible so that the observed differences
between treatments will be largely due to
“true” differences between treatments.
Randomization Procedure
• Each replicate is randomized
separately.
• Each treatment has the same probability
of being assigned to a given experimental
unit within a replicate.
• Each treatment must appear at least
once per replicate.
 Advantages
No restriction on
1
the number of
treatments or
Generally more replicates.
precise than the 2
CRD.

Some treatments may

3
be replicated more
times than others.
Missing plots are
easily estimated. 4
 Advantages
Whole treatments or
1
entire replicates
may be deleted from
If experimental the analysis.
error is 2
heterogeneous,
valid comparisons
can still be made. 3

4
 Disadvantages
Error df is smaller
1 than that for the
CRD (problem with a
If there is a large small number of
variation between 2 treatments).
experimental units
within a block, a If there are
large error term 3
missing data, an
may result RCBD experiment may
(this may be due to be less efficient
too many than a CRD.
treatments).
 Example

1. The effect of light intensity on plant growth.

Treatments: Three different light intensity
levels (low, medium, high).
Experimental units: Individual plant
pots.
Blocking factor: Sunlight exposure
(sunny vs. shaded location).
Replication: Each treatment-block
combination is applied to n plant pots
(e.g., n = 3).
 Example

1. The effect of light intensity on plant growth.

Rationale:

Sunlight exposure can significantly

impact plant growth, potentially
masking the true effect of different
light intensity levels.
RCBD allows us to control for this
potential confounding factor by
blocking.
 Example
1. The effect of light intensity on plant growth.
Procedure:

• Divide the available space:

• Identify two distinct areas: one with
consistent sunny exposure and
another with consistent shade.
• Consider factors like potential wind
direction and building shadows to
minimize microclimate variations
within each area.
 Example
1. The effect of light intensity on plant growth.
Randomize within blocks:

• Within each area (sunny and

shaded), randomly assign each
treatment (light intensity) to n plant
pots.
• This ensures each treatment has an
equal chance of being assigned to
any pot within a specific sunlight
exposure block.
 Example
1. The effect of light intensity on plant growth.
Analysis:

• Analyze the data by comparing the

average growth parameters within
each treatment-block combination.
• Statistical tests can be used to
assess if there are significant
differences in average growth
between the light intensity levels
within each block (sunny vs. shaded).
 Seat Work
1. Comparing the effects of different
antibiotics on bacterial growth (Microbiology)
• Treatments: Three different antibiotics
(A, B, and C).
• Experimental units: Bacterial cultures
grown in individual Petri dishes.
• Blocking factor: Bacterial species (E. coli
vs. Staphylococcus aureus).
• Rationale: Different bacterial species
might exhibit varying susceptibility to
antibiotics. Blocking on the species
 Seat Work
1. Comparing the effects of different
antibiotics on bacterial growth (Microbiology)
Procedure:
• Culture E. coli and S. aureus bacteria in
separate batches.
• Divide each bacterial culture into three groups
(blocks).
• Within each block (species), randomly assign
each antibiotic (A, B, or C) to one group of
Petri dishes containing the respective
bacteria.
• Apply the designated antibiotic to each group
 Seat Work
1. Comparing the effects of different
antibiotics on bacterial growth (Microbiology)
Analysis:

Compare the average growth inhibition

caused by each antibiotic within each
bacterial species (block). This allows for a
more accurate evaluation of the antibiotic's
effectiveness against each specific bacteria.
THANK YOU !