What are Experimental Designs?
Any design that is based on random allocation of participants allows manipulation of the Independent Variable (IV) (a variable that will be manipulated/controlled by the experimenter), and allows control of extraneous and intervening variables. Because everything is controlled, I also get to increase predictability and find a clear cause-and-effect relationship.
Example–
To study the effect of lighting (IV) on reading ability (DV), the researcher:
- Manipulates IV → chooses different lighting conditions
- Randomly assigns participants to lighting conditions
- Controls extraneous variables like noise, font size, and distance from text
- Conducts an experiment in a laboratory or natural setting
Now that we understand the concept of an experiment, let’s explore the different types of experimental designs, focusing on quasi-experiments.
Types of Experimental Designs
True Experimental Designs & Quasi-Experimental Design
True Experimental Designs – are where you can randomly choose the candidates, nothing in specific giving every individual an equal chance to get selected, you can control extraneous variables or confounding, like light sound, and also the surroundings in which you want to conduct this experiment, giving more control. means you can find a cause and effect, and are good at predicting.
whereas,
Quasi-Experimental Design– It is used when random assignment of participants is not possible, meaning the researcher works with pre-existing groups such as school classes, depressed individuals, or different age groups. In these designs, the Independent Variable (IV) may be manipulated (like introducing a new teaching method to only one class) or may be naturally occurring (like gender or clinical status), but because complete control over confounding and extraneous variables is lacking, the internal validity and causal conclusions are weaker than true experiments. However, quasi-experiments are very valuable in real-life field settings, giving high ecological and external validity, and are widely used in applied research where randomization is not practical or ethical.
Two main types of quasi-experiments.
Non-equivalent control group design- A Non-Equivalent Control Group Design is a type of quasi-experimental design where the groups used in the study are already formed before the experiment begins, meaning no random assignment is done. Both groups are given a pre-test to measure their initial status. Then, one group receives the treatment (IV) while the other group does not and acts as a control group. After the intervention, both groups are post-tested to compare changes in their performance. Because the groups differ naturally and the researcher cannot fully control confounding variables like motivation or environment, the internal validity is lower. However, this design is highly practical, has high ecological validity, and is widely used in field research, especially in real educational or clinical settings where random selection is not possible.
- A quasi-experimental design (no random assignment)
- Uses pre-existing groups (e.g., two different school classes)
- Pre-test → Treatment to one group → Post-test for both
- The control group does not receive the intervention
- Comparison of pre-test & post-test scores shows the effect of treatment
- IV may be manipulated (planned intervention)
- Lower internal validity due to uncontrolled confounding variables
- Higher external & ecological validity because it occurs in natural settings
- Very useful in applied research (education, clinical psychology, counselling)
Time series – Time Series Design is a quasi-experimental design used to study the effect of a treatment by repeatedly measuring the same group multiple times, both before and after introducing the intervention. The researcher first collects baseline data through several pre-tests to understand the natural trend or pattern of the behaviour. Then the IV (treatment/intervention) is introduced, and again multiple post-tests are conducted to observe whether there is a significant change in the trend. Since random assignment is not used and only one group is studied, this design may have lower internal validity, but it is very useful in real-life settings like policy implementation, school interventions, or clinical improvements over time. It has high ecological validity because measurement occurs in natural conditions.
- A quasi-experimental design (no random assignment)
- The same group is measured repeatedly over time
- Multiple pre-tests → Intervention → Multiple post-tests
- Focuses on trends and changes over time
- Helps detect whether the change is due to the intervention or natural fluctuations
- Only one group is usually used (sometimes + control in “Interrupted Time Series”)
- Lower internal validity because other factors may influence change
- High ecological & external validity — real-world applications
- Useful in education, healthcare, public policy, and clinical psychology
| Feature | Non-Equivalent Control Group Design | Time Series Design |
|---|---|---|
| Type | Quasi-experimental | Quasi-experimental |
| Groups Used | Two or more pre-existing groups | One group (sometimes + control in Interrupted Time Series) |
| Random Assignment | Not used | Not used |
| Pre-testing | Pre-test for both groups | Multiple pre-tests for same group |
| Post-testing | Both groups post-tested | Multiple post-tests |
| IV Manipulation | Often manipulated for treatment group | Introduced once after baseline measurements |
| Focus | Comparison between groups | Trend changes within same group |
| Internal Validity | Moderate (Confounding variables exist) | Low–moderate (History/Maturation threats) |
| External Validity | High due to real-life settings | High due to field application |
| Best Use | Applied field research, educational & clinical group comparisons | Evaluating effects over time like policies, therapy effects |
MCQs on Experimental & Quasi-Experimental Designs
In true experimental designs, the most essential element is:
A. Field setting
B. Random assignment
C. Pre-testing
D. Non-equivalent groups
Which variable is manipulated by the experimenter?
A. Dependent Variable
B. Predictor Variable
C. Independent Variable
D. Control Variable
Quasi-experimental designs are most commonly used when:
A. Random assignment is unethical or not possible
B. IV cannot be measured
C. Lab setting is mandatory
D. Sample size is too small
Which design uses pre-existing groups and compares them before and after treatment?
A. Solomon Four Group Design
B. Non-Equivalent Control Group Design
C. One Shot Case Study
D. Correlational Design
Time Series Design mainly focuses on:
A. Differences between two groups
B. Changes over multiple time points
C. IQ changes between genders
D. Random allocation of participants
Which validity is stronger in quasi-experimental designs due to real-life settings?
A. Internal validity
B. Construct validity
C. External/Ecological validity
D. Face validity
In Non-equivalent Control Group Design, which step comes first?
A. Post-test for both groups
B. Introducing intervention
C. Random assignment
D. Pre-test for both groups
A major threat to internal validity in Time Series Design is:
A. Maturation & history effects
B. Fatigue effects only
C. Sampling bias
D. Demand characteristics only
A design that uses ‘Multiple pre-tests → Intervention → Multiple post-tests’ is called:
A. True experimental design
B. Interrupted Time Series Design
C. Case study
D. Factorial design
Which of the following best describes True Experiments?
A. No control over IV
B. Focuses on naturally occurring variables
C. Allows strong cause-effect conclusions
D. Always uses only one group
Answers with Explanations
| Q No. | Answer | Explanation |
|---|---|---|
| 1 | B | Random assignment differentiates true from quasi experiments |
| 2 | C | IV is the manipulated variable in experiments |
| 3 | A | Quasi designs are used when randomization cannot be done |
| 4 | B | Pre-existing groups + pre-test + post-test = NECG Design |
| 5 | B | Time Series focuses on trend changes over time |
| 6 | C | Field settings increase ecological/external validity |
| 7 | D | Pre-test is done before treatment to measure baseline |
| 8 | A | Time passage changes behavior independent of IV |
| 9 | B | Interrupted Time Series studies trends before & after intervention |
| 10 | C | High control → high internal validity → cause-effect |
Share this article by Power Within Psychology (PWP):
Discover more from Power Within Psychology | BLOGS
Subscribe to get the latest posts sent to your email.
