Experimental and quasi-experimental research - used to test causal hypotheses by deliberately exposing participants or subjects to specific conditions prior to making observations
Correlational research - tests hypotheses about how variables relate to one another or how they co-vary
- Despite the language of predictor variables and response variables, these designs do not allow the researchers to test causal hypotheses
Data - may be from actual observations or from responses to a test or survey, physiological or biological assessment, chemical assay, etc.
The distinction between experimental and quasi-experimental research is based on random assignment of subjects to conditions.
- In a true experiment there must be random assignment
Many variables of interest to researchers cannot be assigned – they are characteristics of a cell line, a species, or a person.
- For example, gender group comparisons require quasi-experimental designs because gender is innate in the subject
- Similarly, in education research it is common to compare one or more teaching approaches using naturally occurring groups such as children in a specific school or students in two parallel sections of the same course