Internal validity
From EvaluationWiki
An experiment has internal validity if it logically demonstrates a causal relationship between two variables. To demonstrate a causal relationship, three criteria must be met:
- the cause precedes the effect
- the cause and the effect covary
- there are no other plausible explanations for the covariation.
In conducting an experiment, researchers manipulate a condition (independent variable) to see what effect it has on a second variable (dependent variable). For example, a new technique for teaching reading (independent variable) is implemented into a classroom to determine its effect on students' scores on reading tests (dependent variable). If the design of the experiment logically demonstrates that the new technique for teaching reading caused a change in the reading scores of students (see the criteria above), then the study possesses internal validity.
In summary, an experiment demonstrates internal validity to the extent to which changes in the dependent variable are logically attributable solely to the independent variable. A central purpose of effective research design is to assure that an experiment or study eliminates or controls for factors that could provide alternative, plausible explanations for observed effects other than the independent variable. Factors that limit internal validity are commonly referred to as threats to internal validity.
