Between-group (independent) sample
Comparisons of two or more different groups of participants (e.g. comparing CGPA of tutorial class A and tutorial class B).
Within-group (paired) sample
Comparisons of same group of participants over two or more time frames (e.g. comparing weight loss of group A before and after attending weight management programme).
- Only applicable to between group/independent group.
- Purpose: to eliminate confounding variables (e.g. motivations, personality traits).
Variable that you manipulate in your research (e.g. types of medication; antidepressant A, antidepressant B).
Outcome of your research (e.g. depression level after medication).
Measurement of dependent variable/outcome before treatment.
Measurement of dependent variable/outcome after treatment.
Group that receives treatment during an experiment.
Group that does not receive treatment.
Researchers pretending to be participants.
Occurs when the independent variable, but nothing else, influences the dependent variable.
Threats to internal validity
1. Selection bias: participants in both groups have unequal experiences (e.g. those who are health conscious joined treatment group that offers medication).
2. History: event that is not related to the treatment occurs during experiment and affects the ou
Generalizability of the research findings.
The ability to generalize accurately results from experiment to natural settings.
Consistency of experimental conditions with natural settings.