You are on page 1of 35

Important Experimental Designs

Before-and-after without control design


After-only with control design
Before-and-after with control design
Completely randomized design (C.R. design)

Two-group simple randomized design


Completely randomized design (C.R. design)
Random replications design
Completely randomized design (C.R. design)

Illustration of a Completely randomized design

Routes No. of Treatment No. of


passengers passengers
before after
Group 1 of O1 X1 O2
nine routes
Group 2 of O3 X2 O4
nine routes
Group 3 of O3 X3 O6
nine routes
Randomized block design (R.B. design)
Randomized block design (R.B. design)
Illustration of a randomized block design
(Xs indicates various levels of blocking factor Os-No.
of passengers before & after treatment are not
shown, though these measures are taken)
Fare Blocking Factor: Residential areas
reduction Suburbs Crowded Retirement
urban areas areas
5c X1 X1 X1
7c X2 X2 X2
10c X3 X3 X3
Latin square design (L.S. design)
Latin square design (L.S. design)
Illustration of a Latin square design

Residential Day of the week


area
Midweek Weekend Monday/Friday

Suburbs X1 X2 X3

Urban X2 X3 X1
Retirement X3 X1 X2
Factorial designs
Simple factorial designs
Simple factorial designs…
Simple factorial designs…
Simple factorial designs…
Factorial designs
Simple factorial designs
Factorial designs
Complex factorial designs
Factorial designs
Complex factorial designs
Factorial designs
Complex factorial designs
Basic Issues in Experimental Design
Manipulation of the Independent Variable

• Independent Variable & Experimental


treatment
• Experimental group & Control group
• Several experimental treatment levels
• More than one independent variable
Control over extraneous variables
• Matching groups
• Randomization
• Constant error (Bias)
• Demand characteristics
- Experimenter bias
• Establishing control
- Order of presentation bias
- Counterbalancing
- Blinding
- Double blind design
Field Vs. Laboratory Experiment
Laboratory Experiment Field Experiment

Artificial-Low Realism Natural-High Realism

Few Extraneous Many Extraneous


Variables Variables

High control Low control

Low Cost High Cost

Short Duration Long Duration

Subjects Aware of Subjects Unaware of


Participation Participation
Issues of Experimental Validity
• Internal Validity:
- Refers to the confidence we place in the cause and
effect relationship
- Validity determined by whether the experimental
treatment was the sole cause of changes in a
dependent variable.
• External Validity:
- Refers to the extent of generalizability of the results
of a causal study to other settings, people or events
- The ability of an experiment to generalize the results
to the external environment.
Factors Influencing Internal Validity
History - Specific events in the environment between the
Before and After measurement that are beyond the
experimenter’s control
Maturation - Subjects change during the course of the
experiment
Testing - The Before measure alerts or sensitizes subject
to nature of experiment or second measure.
Instrument - Changes in instrument result in response
bias
Selection - Sample selection error because of differential
selection comparison groups
Mortality - Sample attrition; some subjects withdraw from
experiment
Factors Influencing External Validity
• In reactive bias - Participants exhibit abnormal or unusual
behavior simply because they are participating in an
experiment.
• Pretest-manipulation interaction bias - a special form of
reactive bias that is unique to experiments relying on pre-
measurement of consumers before they are exposed to the
experimental manipulation; it arises when the pre-
measurement increases or decreases respondents' sensitivity
to the experimental manipulation.
• Non-representative sample bias - occurs when the units
participating in an experiment are not representative of the
larger body of units to which the experimental results are to
be generalized.
Trade-Off between Internal and External Validity
• Field experiments have greater external validity than
laboratory experiments
• Control is more in laboratory experiments
• Time - Field experimentation is more time consuming than
laboratory
• Cost - Field experimentation is more expensive than
laboratory.
• Exposure to competition - A theory tested in the field may
allow access by the competition to the idea.
• Nature of manipulation - Some things simply cannot be tested
adequately in a laboratory.
Variables in Research
• Anything that may assume different numerical values
or categorical values
• Types:
- Dependent and Independent Variables
- Categorical and Continuous Variables
- Control variable
- Extraneous variable
- Confounding variable

29
Independent and Dependent Variable
• Independent (Experimental, Manipulated, Treatment,
Grouping) Variable:
A variable that is being manipulated in an experiment in
order to observe the effect on a dependent variable.
• Dependent (Outcome) Variable:
That factor which is observed and measured to
determine the effect of the independent variable.
• Eg: Dependent Variable: Test Mark (measured from 0 to
100)
Independent Variables: Revision time (measured in
hours) Intelligence (measured using IQ score)
Categorical Variables (discrete or qualitative variables)
• Result from a selection from categories, any variable that
has a limited number of distinct values
• Further categorized as:
Nominal: variables that have two or more categories but
which do not have an intrinsic order. Eg: The number pinned
on a sports person, A set of countries
Ordinal: variables that have two or more categories just like
nominal variables only the categories can also be ordered or
ranked. we cannot place a "value" to them. Eg: The first, third
and fifth person in a race, Pay bands in an organization, as
denoted by A, B, C and D.
Dichotomous: variables which have only two categories or
levels. Eg: gender as either "male" or "female"
Continuous Variables (quantitative variables)

• Numeric variables that can take any value, such as


weight
• Further categorized as:
Interval: variables for which their central characteristic is that
they can be measured along a continuum and they have a
numerical value. Eg: temperature measured in degrees
Celsius or Fahrenheit 30C to 40C
Ratio: The name "ratio" reflects the fact that you can use the
ratio of measurements (So, for example, a distance of ten
metres is twice the distance of 5 metres). Eg: height, mass,
distance.
Control variable
• A variable that effects the dependent variable.
• When we "control a variable" we wish to balance its
effect across subjects and groups so that we can
ignore it, and just study the relationship between the
independent and the dependent variables.
• You control for a variable by holding it constant, e.g.,
keep humidity the same, and vary temperature, to
study comfort levels.
Extraneous variable
• An undesirable variable other than the independent variable
that may bear any effect on the behavior of the subject being
studied, but it is one that we do not control or manipulate.
• Further categorized as:
Subject variables: characteristics of the individuals being studied
that might affect their actions. Eg: age, gender, health status, mood,
background, etc.
Experimental variables: characteristics of the persons conducting
the experiment which might influence how a person behaves. Eg:
Gender, the presence of racial discrimination, language, etc.
Situational variables: features of the environment in which the
study or research was conducted, which have a bearing on the
outcome of the experiment in a negative way. Eg: the air
temperature, level of activity, lighting, and the time of day.
Confounding variable (third variable or a mediator
variable, confounder or lurking variable)
• A type of extraneous variable.
• A "hidden" variable in a statistical or research model that affects
the variables in question but is not known or acknowledged, and
thus (potentially) distorts the resulting data.
• This hidden third variable causes the two measured variables to
falsely appear to be in a causal relation. Such a relation between
two observed variables is termed a spurious relationship.
• Eg: Ice cream consumption and murder rates are highly correlated. Does
ice cream incite murder or does murder increase the demand for ice
cream?
Neither: they are joint effects of a common cause or lurking variable,
namely, hot weather.
Another look at the sample shows that it failed to account for the time of
year, including the fact that both rates rise in the summertime.

You might also like