# Concept of Causality

Concept of Causality

Ü  Causality may be thought of as understanding a phenomenon in terms of conditional statements of the form “if x, then y”

Some simple examples

If:

n    We increase the temperature on the thermostat

n    We reduce price

Then:

n    The room will get warmer

n    Sales may increase

n    Sales may increase

Scientific meaning of Causality

Ü  Marketing effects are caused by multiple variables

Ü  The relationship between cause and effect tends to be probabilistic

Ü  Causality can never be proved

Conditions for Causality

Concomitant Variation

Ü  Extent to which a cause (x) and an effect (y) occur together or vary together in the way predicted by the hypothesis under consideration.

Ü  Association or relationship between variables

eg. level of education “causes” people to play golf??

Watch out for spurious correlation

Conditions for Causality cont.

Time Order of Occurrence of Variables

Ü  Causing events (x) must occur either before or simultaneously with the effect (y); can not occur afterwards

eg. If we increase promotional expenditure, sales may increase (not the other way around)

Ü  An event in a relationship could be both a cause and effect

eg. Which is the cause?

Store loyalty or owning a store credit card.

Conditions for Causality cont.

Absence of other Probabilistic Causal Factors

Ü  The variable being investigated should be the only possible casual explanation

Ü  Difficult to rule out all other casual factors

If all the evidence is strong and consistent, it may be reasonable to conclude that there is a causal relationship

Definition

Experiment

Ü manipulating an independent variable to see how it affects the dependent variable while also controlling the effect of additional extraneous variables

Examples of…

Independent variables

Dependent variables

Ü  sales, market share, level of customer satisfaction

Extraneous variables

Ü  store location, competition, questionnaire used, respondents

Validity in Experimentation

Internal Validity

Ü  Extent to which the research design accurately identifies causal relationships

Ü  The researcher can rule out rival explanations [other variables] for the observed results and conclusions about the relationship

External Validity

Ü  Extent to which the cause-and-effect relationships found in the experiment can be generalised

Extraneous Variables

History

Ü  Specific events that are external to the experiment but occur at the same time as the experiment

Ü  These events may affect the dependent variable.

[economic downturn, competitors’ price reductions]

O1 X1 O2

If there is no difference between O1 and O2, we may conclude that X1 (treatment) was ineffective.  However, consider if the effects might have been compounded by history.

Extraneous Variables cont.

Maturation

Ü  Changes in the test units over time

[changes in the people who are part of the experiment, changes to stores (change in layout, décor, traffic)]

Extraneous Variables cont.

Test Effects

Ü  The effect on the experiment of taking a measure on the dependent variable before and after the presentation of the treatment

[The actual act of ‘testing’ may contaminate the results of the experiment]

Extraneous Variables cont.

Instrumentation

Ü  Changes in the measuring instrument, observers or scores

[Changing the questionnaire, observers, or scores during the course of the experiment may contaminate the results of the experiment]

Extraneous Variables cont.

Statistical Regression

Ü  When test units with extreme scores move closer to the average during the course of the experiment

[Respondents’ opinions move to the average because they are part of the experiment, rather than as a result to the treatment]

Extraneous Variables cont.

Selection Bias

Ü  Improper assignment of test units to treatment conditions

[sampling error]

Extraneous Variables cont.

Mortality

Ü  Loss of test units while the experiment is in progress

[respondents selected no longer wish to participate]

Controlling Extraneous Variables

Randomisation

Ü   Randomly assigning test units to experimental groups by using random numbers

Matching

Ü   Comparing test units on a set of key background variables before assigning them to the treatment

Statistical Control

Ü   Measuring the extraneous variables and adjusting for their effects through statistical analysis

Design Control

Ü   Use of experiments designed to control specific extraneous variables

Case: Beauty is in the eye of the beholder

Ü  Scientists have found a link between drinking alcohol and perceptions of beauty

Ü  80 students were shown colour photographs of 120 male and female students and were asked to rate the aesthetic properties on a 7-point scale from high unattractive to highly attractive

Ü  Half the students had drunk up to four units of alcohol, the other half had no alcohol.

Ü  The students who had consumed alcohol rated the people in the photographs as more attractive than the student who did not consume alcohol.

Next slide fig 6.1

Source: http://www.theage.com.au/articles/2002/09/091031115991721.html

Laboratory versus Field Experiments

Laboratory

Ü  The independent variable is manipulated and measures of the dependent variable are taken in a contrived, artificial setting for the purpose of controlling the many possible extraneous variables that may affect the dependent variable

Field

Ü  Independent variables are manipulated and measures of the dependent variable are made on test units in their natural setting

Table 6.3 Laboratory versus Field Experiments

Limitations of Experimentation

Ü  Time

Ü  Cost

Test Marketing

Ü  An application of a controlled experiment, done in limited but carefully selected parts of the marketplace called test markets.

Classifications

Ü  Standard market test

Ü  Controlled test market

Ü  Simulated test market

Test Marketing cont.

Criteria for the selection of test markets

Ü   Be large enough to produce meaningful projections

(ie.at least 2% of the potential actual population).

Ü   Be representative:

Ü  demographically

Ü  with respect to product consumption behavior

Ü  with respect to media usage

Ü  with respect to competition

Ü   Be relatively isolated in terms of media and physical distribution.

Ü   Have normal historical development in the product class

Ü   Have marketing research and auditing services available

Ü   Not be over-tested