Experiment
An empirical research method used to examine a hypothesized causal relationship between
independent and dependent variables. The antecedent event in a proposed causal sequence
is called the independent variable. The measured effect in the causal sequence is called
the dependent variable. The main methodological rule of experimentation is that the
experimenter must have precise control over the experimental situation. Control
involves the creation and variation (manipulation) of the independent variables.
The values of the independent variable(s) define the experimental conditions or
the design of the experiment. At a minimum, the design involves the application
of a treatment to one group of participants and the withholding of the treatment
from a comparison or control group. Aside this manipulation the situation in both
groups must be hold identical. Accordingly, the dependent variable(s) has (have) to
be assessed by consistent measures. In order to render the group samples comparable,
participants must be randomly assigned to conditions. Differences that then occur on
the measures of the dependent variable can be attributed to the factor which differentiates
the groups systematically, the presence, absence or level of the independent variable.
Without randomization the method is quasi-experimental (e.g. if gender is used as a factor
in the design). Experiments can be conducted in the laboratory or in natural settings
(field). Because it is easier to precisely control the experimental situation in the
laboratory, this kind of experiments allow the experimenter to achieve a higher level
of internal validity than in field experiments. However, if it is possible to sufficently
control the experimental situation in natural conditions, field experiments are more
likely to be externally valid.
See also:
experimental design,
objectivity,
reliability,
validity
Literature:
Carlsmith, Ellsworth & Aronson (1976)
