Causal research design: experimentation

CAUSALITY
The definition of causality for an average person is different from that of a scientist and it
is the scientific definition that is used in marketing research. Causality is not
demonstrable, only a cause and effect relationship can be demonstrated.
CONDITIONS FOR CAUSALITY
Before assuming causality, three conditions must be met:
1. Concomitant variation: it is the way in which a cause and an effect occur or vary
together according to what is stated in the hypothesis. Evidence of this variation can be
obtained quantitatively or qualitatively.
2. Temporal order of the occurrence of the variables: states that the causing event must
occur before or at the same time as the effect.
3. Absence of other causal events: the factor or variable being investigated should be the
only possible causal explanation.
DEFINITIONS AND CONCEPTS
Independent variables: are variables or alternatives that are manipulated and whose
effects are measured and compared.
Test units: individuals, organizations, or other entities whose response to independent
variables or treatments is being studied.
Dependent variables: measure the effect of the independent variables on the test units.
Extraneous variables: are those other than the independent variables that influence the
response of the test units.
Experiment: is the process that manipulates one or more independent variables, and
measures its effect on one or more dependent variables, while controlling extraneous
variables.
Experimental design: set of tests that specify: 1. Test units and sampling procedures 2. The
independent variables 3. The dependent variables 4. How to control extraneous variables.

DEFINITION OF SYMBOLS
X= the exposure of a group to an independent variable, treatment or event, whose effect
must be determined.
O= process of observation or measurement of the dependent variable in the test units.
A = random assignment of test units or groups to separate treatments.
VALIDITY OF THE EXPERIMENTATION
The researcher has two goals when conducting an experiment:
1.- draw valid conclusions about the effects of the independent variables in the study
group (internal validity) and 2.- make valid generalizations to the larger population of
interest (external validity).
Internal validity: measures the precision of the experiment and whether the management
of the independent variables really causes an effect on the dependent variables.
External validity: determines whether it is possible to generalize the causal relationships
found in the experiment.
STRANGE VARIABLES
They are classified into the following categories:
History: refers to specific events that are external to the experiment, but occur at the
same time as it.
Maturation: Extraneous variables attributable to changes in test units that occur over
time.
Test effects: effects on the experiment of making a measurement of the dependent
variable, before and after the presentation of the treatment.
1. Main effect of the test: occurs when a previous observation affects a subsequent
observation.
2. Interactive effect of the test: a previous measurement influences the response of the
test units to the independent variable.
Instrumentation: an extraneous variable that involves changes in the measuring
instrument, observers, or scores.
Statistical regression: Extraneous variable that occurs when test units with extreme scores
approach the average score over the course of the experiment.
Selection bias: extraneous variable attributable to inappropriate assignment of test units
to treatment conditions.
Mortality: extraneous variable attributable to the loss of test units while the experiment is
in progress.
CONTROL OF FOREIGN VARIABLES
Extraneous variables represent alternative explanations of the experimental results. They
are a threat to the internal and external validity of the experiment. If they are not
controlled, they affect the dependent variable and therefore also the results.
There are four ways to control extraneous variables:
Randomization: involves random assignment of test units and test conditions to
experimental groups using random numbers.
Matching: involves comparing test units on a set of key antecedent variables before
assigning them to treatment conditions.
Statistical control: controls extraneous variables, measuring them and adjusting their
effects with statistical methods.
Control design: involves the use of specific experimental designs to control extraneous
variables.
PREEXPERIMENTAL DESIGNS:
They are characterized by the lack of Randomization. Three specific designs are described:
Single case study: A single group of test units is exposed to treatment X and then a single
measurement of the dependent variable is made.
Pretest and posttest design with one group: a group of test units is measured twice. Static
group design: there are two groups, the experimental group that is exposed to the
treatment and the control group. In both groups measurements are made only after
treatment and the test units are not randomly assigned.
TRUE EXPERIMENTAL DESIGNS
The distinguishing feature is randomization. True experimental designs include the
pretest-posttest design with control group, the posttest-only design with control group,
and the Solomon four-group design.
Pretest-posttest design with control group: The test units are randomly assigned to the
experimental group or the control group, and in each group a measurement is made prior
to treatment. Only the experimental group is exposed to the treatment, but post-test
measurements are taken in both groups.
Posttest-only design with control group: It does not involve any prior measurement,
therefore, the effects of the test are eliminated; however, the design is susceptible to
selection bias and mortality. On the other hand, this design has important advantages in
terms of time, cost and sample size requirements. It only includes two groups and a single
measurement per group. Because of its simplicity, the posttest-only control group design
is likely to be the most popular in marketing research.
QUASIEXPERIMENTAL DESIGNS
They are useful because they can be used in cases where true experimentation is not
possible, and because they are faster and less expensive. Time series design: Involves a
series of periodic measurements of the dependent variable on a group of test units. The
treatment is then applied by the researcher or occurs naturally. After treatment, periodic
measurements continue to determine its effect.
Multiple time series design: It is similar to the time series design, except that another
group of test units is added to serve as a control group. Which improves the ability to test
the treatment effect twice.

STATISTICAL DESIGNS
They consist of a series of basic experiments that allow statistical analysis and control of
external variables, that is, several basic experiments are conducted at the same time.
Randomized block design: This is useful when there is only one important external
variable that can influence the dependent variable. The researcher must be able to
identify and measure the variable used in the formation of the block.
Latin square design: Allows the researcher to statistically control two external variables
that do not interact, in addition to the management of the independent variable. Each
external or block-forming variable is divided into an equal number of blocks or levels. The
independent variable is also divided into the same number of levels.
Factorial design: Used to measure the effects of two or more independent variables at
multiple levels. An interaction is said to occur when the simultaneous effect of two or
more variables is different from the sum of their separate effects.
LABORATORY EXPERIMENTS AND FIELD EXPERIMENTS
The experiments can be carried out in two environments. One is the laboratory
environment, an artificial setting that the researcher builds with whatever he or she wants
for the experiment. The term field environment is synonymous with actual market
conditions. Laboratory experiments have some advantages over field experiments.
It offers a high degree of control because it isolates the experiment in a carefully
supervised environment, it also usually produces the same results if repeated with similar
subjects, which produces high internal validity. Compared to the field experiment, the
laboratory experiment involves disadvantages such as the artificiality of the environment
and also causing different demand scenarios, where the participants try to guess the
purpose of the experiment. In short, the researcher must consider all these factors when
deciding on the type of experiment.

LIMITATIONS OF EXPERIMENTATION
Time: Experiments should last long enough for post-treatment measurements to include
most or all of the effects of the independent variables.
Cost: The experimental group, control group, and multiple measurements requirements
significantly increase the cost of research.
Administration: Administration of experiments can be difficult. It may be impossible to
control the effects of extraneous variables, especially in a field environment.
TEST MARKET APP
It is an application of a controlled experiment, conducted in limited and carefully selected
parts of the market called test markets. The main objectives are:
1. Determine the acceptance of the product in the market.
2. Test alternative levels of marketing mix variables. Standard Test Market: Designing a
standard test market involves deciding what criteria should be used in selecting test
markets, how many test markets to use, and the duration of the test.
This duration depends on the product's repurchase cycle, the likelihood of a competitive
response, cost considerations, initial consumer response, and company philosophy.
Controlled market test: Ensures distribution of the product in retail stores that represent
a predetermined percentage of the market. It includes small store testing and the smaller
controlled store panels. This service is provided by several research companies such as
ACNielsen.
Simulated market testing: Provides mathematical estimates of market share, which are
based on consumers' initial reaction to a new product. Respondents are generally
intercepted in a busy place and are subjected to an initial inquiry about the use of the
product. The information they generate is confidential and competitors cannot obtain it.
They are also relatively inexpensive.

ETHICS IN MARKET RESEARCH

It is often necessary to conceal the purpose of the experiment in order to produce valid
results. This concealment should be done in such a way that it does not violate the rights
of the participants. One way to handle this ethical dilemma is to inform them, at the
beginning, that the purpose of the experiment will remain hidden. After the data have
been obtained, the true purpose of the study and the nature of the concealment should
be explained to respondents, and they should be given an opportunity to withdraw their
information.