Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Experimental Research Designs

Introduction to the various designs in experimental research in Marketing.

Priya Sana

on 22 August 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Experimental Research Designs

Experimental Research Designs
Process of Experimentation
Uses of Experimentation
Product Design
Package Design
Pricing Policies
Promotion Policies
Distribution Policies
Limitations of Experimentation
Experimental Designs
Informal Experimental Design
Formal Experimental Design
These are the various designs in marketing research with respect to experimenting.

While experimenting it is necessary to analyse the requirement and choose an appropriate design for effective outcomes.
Variables used :
Formulate Hypothesis
Define variables
Design Experimental Procedure
the Experiment
the Results
the Report
Process Of
Experimental Research
the Problem
Uses of Experimentation
Product Design
Basic characters, colour, shape, size are studied.
Which product customers tend to choose when the above conditions vary, will be experimented.
Package Design
Though related to production function of marketing, various experiments have been conducted to study the package design with respect to product sales.
What colour, material, shape etc. of the package should be used to increase the product sales, can be experimented in marketing research.
Pricing Policies
Price elasticity of the demand is better tested using experiments.
Magazine sellers experiment by subscribing at different prices to various buyers.
Retailers use the same concept with "multiple unit discount" etc.
Promotion Policies
Widely explored than other areas.
Testing of advertising effectiveness, media effectiveness, appeals, efficiency, copy and size effectiveness, frequency of repeating an ad, the effect of colour, optimum frequency of call by a salesman.
Besides, the effect of various sales techniques is measured by experiments.
Distribution Policies
The distribution channel management, physical distribution management, and distribution system management uses experimentation as a technique of research.
Though other methods of research are available, experimentation provides accurate results.
Limitations of Experimentation
The information to formulate hypothesis is not easily available since there is a deficiency of a firm’s theoretical base in marketing.
The cost involved in the experimentation research is more. When data is to be collected for large market, it is tremendously expensive.
Close supervision and administration of the experiments are needed to carry on research effectively. People sometimes may not co-operate or intentionally manipulate results which may give misleading conclusions.
It is not clear to anyone that after what time period the real impact of the ad will be felt on the sales. Hence it may be time consuming.
Experimentation is defined as a process where events occur in a setting at the discretion of the experimenter, controls are used to identify the sources of variation in subject’s response.
It is the process of manipulating one variable in a controlled environment while holding all other variables constant in order to establish a causal relationship.
Basic components:
1. The variable manipulated – independent variables or test units.
2. The actual alteration – treatment.
3. The actual variables chosen to measure the subjects response –
dependent variables.
The treatment on independent variables gives the result or outcome.
Experimental Designs
Experimental designs are blueprints - provide guide lines to conduct the final experiments.
Specifies what steps would be followed in giving treatment to the experimental or test unit.
Informal Experimental Design
Effects can be quantified.
No randomized selection of test units and treatment.
The loose structure of the experiments does not allow the changes caused by the treatment to be singled out.
Widely used in marketing.
Before-After without Control
Simplest design of all.
Measure the test unit before and after it has been subjected to the treatment.
Does not involve any control group.
Merits : Reduced cost and time.
Demerits : The extraneous factors present in real time are not taken into account.
Sales before
exposure to
Sales after
exposure to
Before-After with Control
Experimental unit is selected and studied during the period.
The units selected must be of same size and located almost in same localities.
Sales before exposure to
experimental variable in city1
(Before Clearance Sale)
Sales after exposure to
experimental variables in city1
(After Clearance Sale)
Control Group
Sales before experiment
Sales after experiment
Sales in city2
(Before Clearance Sale)
Sales in city2
(After Clearance Sale)
After-Only with Controls
Most frequently used.
Using two equivalent groups – experimental and control group - eliminates the pre-testing of the groups.
There is no before measurement involved.
The treatment is provided to experimental group.
The design neutralize the impact of pre-tests in both the groups.
Does not allow any measurement of the total change that might occur during the experiment.
Sales after exposure to experimental
variables in experimental territory
Control group
Sales in control terrritory
Formal Experimental Design
Randomly selected test units and treatment methods.
Statistical tests (Analysis of variance) can be applied to results.
Analysis of Variance: Helps researchers to determine whether the observed association is by chance or some factor.
Completely Randomised Design
Simplest of formal designs.
Treatments are assigned to test units randomly, before the analysis of variance.
Can measure only one type of variance between treatments.
Demerit: All extraneous factors are assumed to have equal impact on units. Hence hardly taken into account.
Randomised Block Design
The extraneous factor is separated from the total environment.
This factor is considered along with influence of treatment on test unit.
It should be identified properly and be measured during the experiment.
The design gives truer results.
Latin Square Design
More effective than previous methods with complex design.
Needs expert experimenter.
Two major variations of extraneous factors are considered along with the treatment influence.
Additional effectiveness measured against cost and time.
Assumption: Interaction among the variables is not considered.
Demerit: Can work on only two extraneous variables regardless of the number of treatment.
Factorial Design
Measures the influence of interacting variables on the dependent variables.
Assumption: Consider the interaction amongst the factors.
Design used: Randomised block or Latin Square.
Analysis of variance applied to the results becomes more complex with the design.
Accurate results will be produced.
Ex-post Facto Design
It is a variation of the 'after-only design'.
None of the groups is selected before the experiment.
Groups are identified only after they are exposed to the treatment.
Has merit over 'after-only' because it avoids influence of advance setting the experimental group in otherwise natural setting.
The experimental group is set to receive the treatment rather than exposing it to the treatment by its choice.
Week Rs. 7.00 Rs. 6.00 Rs. 6.50

1 500 400 300
2 450 450 350
3 300 350 450
4 350 350 550

Total 1600 1550 1650

Block store size Rs. 6.00 Rs. 6.50 Rs. 7.00
based on turnover

> Rs. 1000 500 550 450
Rs. 600 - Rs. 1000 600 300 250
< Rs. 600 400 350 150

Total 1500 1200 850
Chain > Rs. 1000 Rs. 600 - Rs. 1000 < Rs. 600

X Price 2 Price 3 Price 1
Y Price 1 Price 2 Price 3
Z Price 3 Price 1 Price 2
A Presentation

Full transcript