Causal Design

8/13/2019 Causal Design

1/35

Chapter 3

Causal Research Design:Experimentation


2/35

7-2

Chapter Outline

1) Overview

2) Concept of Causality

3) Conditions for Causality

4) Definition of Concepts5) Definition of Symbols

6) Validity in Experimentation

7) Extraneous Variables8) Controlling Extraneous Variables


3/35

7-3

Chapter Outline

9) A Classification of Experimental Designs

10) Pre-experimental Designs

11) True Experimental Designs

12) Quasi Experimental Designs13) Statistical Designs

14) Laboratory vs. Field Experiments

15) Experimental vs. Non-experimental Designs16) Limitations of Experimentation

17) Application: Test Marketing


4/35

7-4

Chapter Outline

18) Determining a Test Marketing Strategy

19) International Marketing Research

20) Ethics in Marketing Research

21) Internet and Computer Applications22) Focus on Burke

23) Summary

24) Key Terms and Concepts


5/35

7-5

Concept of Causality

A statement such as "Xcauses

Y" will have thefollowing meaning to an ordinary person and to a

scientist.

____________________________________________________

Ordinary Meaning Scientific Meaning____________________________________________________

Xis the only cause of Y. Xis only one of a number of

possible causes of Y.

Xmust always lead to

Y The occurrence of

Xmakes the(Xis a deterministic occurrence of Ymore probable

cause of Y). (Xis a probabilistic cause of Y).

It is possible to prove We can never prove that Xis a

that Xis a cause of Y. cause of Y. At best, we can

infer that Xis a cause of Y.


6/35

7-6

Conditions for Causality

Concomitant variationis the extent towhich a cause, X, and an effect, Y, occurtogether or vary together in the waypredicted by the hypothesis underconsideration.

The time order of occurrenceconditionstates that the causing event must occureither before or simultaneously with theeffect; it cannot occur afterwards.

The absence of other possible causalfactorsmeans that the factor or variablebeing investigated should be the only possiblecausal explanation.


7/35

7-7

Evidence of Concomitant Variation betweenPurchase of Fashion Clothing and Education

High

High Low

363 (73%) 137 (27%)

322 (64%) 178 (36%)

Purchase of Fashion Clothing, Y

Table 7.1

500 (100%)

500 (100%)LowEducation,

X


8/35

7-8Purchase of Fashion Clothing ByIncome and Education

Low Income

Purchase

High Low

High

LowEducation 200 (100%)

300 (100%)

300

200

122 (61%)

171 (57%)

78 (39%)

129 (43%)

High Income

Purchase

High

High

Low

Low

241 (80%)

151 (76%)

59 (20%)

49 (24%)Education


9/35

7-9

Definitions and Concepts

Independent variablesare variables oralternatives that are manipulated and whose effectsare measured and compared, e.g., price levels.

Test unitsare individuals, organizations, or otherentities whose response to the independent variables

or treatments is being examined, e.g., consumers orstores.

Dependent variablesare the variables whichmeasure the effect of the independent variables onthe test units, e.g., sales, profits, and market shares.

Extraneous variablesare all variables other thanthe independent variables that affect the response ofthe test units, e.g., store size, store location, andcompetitive effort.


10/35

7-10

Experimental Design

An experimental designis a set ofprocedures specifying

the test units and how these units are to

be divided into homogeneous subsamples, what independent variables or treatments

are to be manipulated,

what dependent variables are to bemeasured, and

how the extraneous variables are to becontrolled.


11/35

7-11

Validity in Experimentation

Internal validityrefers to whether themanipulation of the independent variables ortreatments actually caused the observedeffects on the dependent variables. Controlof extraneous variables is a necessarycondition for establishing internal validity.

External validityrefers to whether thecause-and-effect relationships found in theexperiment can be generalized. To whatpopulations, settings, times, independentvariables and dependent variables can theresults be projected?


12/35

7-12

Extraneous Variables

Historyrefers to specific events that are external tothe experiment but occur at the same time as theexperiment.

Maturation(MA) refers to changes in the test unitsthemselves that occur with the passage of time.

Testing effectsare caused by the process ofexperimentation. Typically, these are the effects onthe experiment of taking a measure on thedependent variable before and after the presentation

of the treatment. The main testing effect(MT) occurs when a prior

observation affects a latter observation.

7 13


13/35

7-13

Extraneous Variables

In the interactive testing effect(IT), a priormeasurement affects the test unit's response to theindependent variable.

Instrumentation(I) refers to changes in themeasuring instrument, in the observers or in the

scores themselves. Statistical regressioneffects (SR) occur when test

units with extreme scores move closer to the averagescore during the course of the experiment.

Selection bias(SB) refers to the improper

assignment of test units to treatment conditions. Mortality(MO) refers to the loss of test units while

the experiment is in progress.

7 14


14/35

7-14

Controlling Extraneous Variables

Randomizationrefers to the random assignment oftest units to experimental groups by using randomnumbers. Treatment conditions are also randomlyassigned to experimental groups.

Matchinginvolves comparing test units on a set of

key background variables before assigning them tothe treatment conditions.

Statistical controlinvolves measuring theextraneous variables and adjusting for their effects

through statistical analysis. Design controlinvolves the use of experiments

designed to control specific extraneous variables.

7 15


15/35

7-15

A Classification of Experimental Designs

Pre-experimental designsdo not employrandomization procedures to control forextraneous factors: the one-shot case study,the one-group pretest-posttest design, andthe static-group.

In true experimental designs, theresearcher can randomly assign test units toexperimental groups and treatments toexperimental groups: the pretest-posttestcontrol group design, the posttest-onlycontrol group design, and the Solomon four-group design.

7 16


16/35

7-16


Quasi-experimental designsresult whenthe researcher is unable to achieve fullmanipulation of scheduling or allocation oftreatments to test units but can still applypart of the apparatus of true

experimentation: time series and multipletime series designs.

A statistical designis a series of basicexperiments that allows for statistical controland analysis of external variables:randomized block design, Latin squaredesign, and factorial designs.

7 17


17/35

7-17


Pre-experimental

One-Shot CaseStudy

One GroupPretest-Posttest

Static Group

TrueExperimental

Pretest-PosttestControl Group

Posttest: OnlyControl Group

Solomon Four-Group

QuasiExperimental

Time Series

Multiple TimeSeries

Statistical

RandomizedBlocks

Latin Square

FactorialDesign

Figure 7.1

Experimental Designs

7 18


18/35

7-18

One-Shot Case Study

X 01

A single group of test units is exposed to atreatment X.

A single measurement on the dependentvariable is taken (01).

There is no random assignment of test units.

The one-shot case study is more appropriatefor exploratory than for conclusive research.

7 19


19/35

7-19

One-Group Pretest-Posttest Design

01

X 02

A group of test units is measured twice.

There is no control group. The treatment effect is computed as

0201.

The validity of this conclusion isquestionable since extraneous variablesare largely uncontrolled.

7-20


20/35

7-20

Static Group Design

EG: X 01

CG: 02

A two-group experimental design.

The experimental group (EG) is exposed tothe treatment, and the control group (CG) isnot.

Measurements on both groups are made only

after the treatment. Test units are not assigned at random.

The treatment effect would be measured as01- 02.

7-21T E i t l D i


21/35

7 21True Experimental Designs:Pretest-Posttest Control Group Design

EG: R 01 X 02

CG: R 03 04

Test units are randomly assigned to either the experimental orthe control group.

A pretreatment measure is taken on each group.

The treatment effect (TE) is measured as:(02- 01) - (04- 03). Selection bias is eliminated by randomization.

The other extraneous effects are controlled as follows:

0201= TE+ H+ MA+ MT+ IT+ I+ SR+ MO

0403= H+ MA+ MT+ I+ SR+ MO

= EV(Extraneous Variables)

The experimental result is obtained by:

(02- 01) - (04- 03) = TE+ IT

Interactive testing effect is not controlled.

7-22


22/35

7 22

Posttest-Only Control Group Design

EG :R

X

01

CG : R 02

The treatment effect is obtained byTE= 01- 02

Except for pre-measurement, the

implementation of this design is verysimilar to that of the pretest-posttestcontrol group design.

7-23Q i E i t l D i


23/35

7 23Quasi-Experimental Designs:Time Series Design

01

02

03

04

05

X

0

60

70

80

90

10

There is no randomization of test units

to treatments. The timing of treatment presentation,

as well as which test units are exposed

to the treatment, may not be within theresearcher's control.

7-24


24/35

7 24

Multiple Time Series Design

EG : 01 02 03 04 05 X 06 07 08 09 010CG : 01 02 03 04 05 06 07 08 09 010

If the control group is carefully

selected, this design can be animprovement over the simple timeseries experiment.

Can test the treatment effect twice:

against the pretreatment measurementsin the experimental group and againstthe control group.

7-25


25/35

7 25

Statistical Designs

Statistical designsconsist of a series of basic

experiments that allow for statistical control and analysisof external variables and offer the following advantages:

The effects of more than one independent variable canbe measured.

Specific extraneous variables can be statisticallycontrolled.

Economical designs can be formulated when each testunit is measured more than once.

The most common statistical designs are the randomizedblock design, the Latin square design, and the factorialdesign.

7-26


26/35

Randomized Block Design

Is useful when there is only one majorexternal variable, such as store size,that might influence the dependentvariable.

The test units are blocked, or grouped,on the basis of the external variable.

By blocking, the researcher ensures

that the various experimental andcontrol groups are matched closely onthe external variable.

7-27


27/35

Randomized Block Design

Treatment GroupsBlock Store Commercial Commercial CommercialNumber Patronage A B C

1 Heavy A B C2 Medium A B C3 Low A B C4 None A B C

Table 7.4

7-28


28/35

Latin Square Design

Allows the researcher to statistically control two noninteracting

external variables as well as to manipulate the independentvariable.

Each external or blocking variable is divided into an equalnumber of blocks, or levels.

The independent variable is also divided into the same number

of levels.

A Latin square is conceptualized as a table (see Table 7.5), withthe rows and columns representing the blocks in the twoexternal variables.

The levels of the independent variable are assigned to the cellsin the table.

The assignment rule is that each level of the independentvariable should appear only once in each row and each column,as shown in Table 7.5.

7-29


29/35

Latin Square Design

Table 7.5

Interest in the StoreStore Patronage High Medium Low

Heavy B A CMedium C B ALow and none A C B

7-30


30/35

Factorial Design

Is used to measure the effects of two ormore independent variables at variouslevels.

A factorial design may also beconceptualized as a table.

In a two-factor design, each level ofone variable represents a row and eachlevel of another variable represents acolumn.

7-31


31/35

Factorial Design

Table 7.6

Amount of HumorAmount of Store No Medium HighInformation Humor Humor Humor

Low A B C

Medium D E FHigh G H I

7-32


32/35

Laboratory versus Field Experiments

Factor Laboratory Field

Environment Artificial RealisticControl High LowReactive Error High Low

Demand Artifacts High LowInternal Validity High LowExternal Validity Low HighTime Short LongNumber of Units Small Large

Ease of Implementation High LowCost Low High

Table 7.7

7-33


33/35

Limitations of Experimentation

Experiments can be time consuming,particularly if the researcher is interested inmeasuring the long-term effects.

Experiments are often expensive. Therequirements of experimental group, controlgroup, and multiple measurementssignificantly add to the cost of research.

Experiments can be difficult to administer. Itmay be impossible to control for the effectsof the extraneous variables, particularly in afield environment.

Competitors may deliberately contaminatethe results of a field experiment.

7-34


34/35

Competition

Overall Marketing Strategy

Socio-CulturalEnvir

onment

NeedforS

ecrecy

New Product DevelopmentResearch on Existing ProductsResearch on other Elements

Simulated Test Marketing

Controlled Test Marketing

Standard Test Marketing

National Introduction

Stop

and

Reevaluate

-ve

-ve

-ve

-ve

Very +ve

Other Factors

Very +veOther Factors

Very +ve

Other Factors

Selecting a Test-Marketing Strategy

7-35


35/35

Criteria for the Selection of Test Markets

Test Markets should have the following qualities:1) Be large enough to produce meaningful projections. They

should contain at least 2% of the potential actual population.

2) Be representative demographically.

3) Be representative with respect to product consumption behavior.

4) Be representative with respect to media usage.

5) Be representative with respect to competition.

6) Be relatively isolated in terms of media and physical distribution.

7) Have normal historical development in the product class

8) Have marketing research and auditing services available

9) Not be over-tested

Documents

Causal Design