EXPERIMENTAL DESIGNS – TYPES

Experimental design refers to the structure or plan of an experiment used to test

hypotheses by manipulating independent variables and observing their effect on dependent
variables under controlled conditions.

1. ONE-GROUP DESIGN

Meaning

A one-group design involves only one group of participants, where the researcher studies
the effect of a treatment without comparing it to another group.

Types of One-Group Designs

1. One-Shot Case Study

 A single group is exposed to a treatment.

 Outcome is measured after treatment.

Structure:
Treatment (X) → Observation (O)

Example:
Giving a training program and testing performance afterward.

2. One-Group Pretest–Posttest Design

 Same group is tested before and after the treatment.

Structure:
Pretest (O₁) → Treatment (X) → Posttest (O₂)

Example:
Measuring anxiety before and after therapy.

Advantages

 Simple and easy to conduct

 Requires fewer participants

  Useful for preliminary or pilot studies

Disadvantages

 No control group → low internal validity

 Effects may be due to:

o History (external events)

o Maturation (natural changes)

o Testing effect (practice)

 Cannot establish strong cause–effect relationship

2. TWO-GROUP DESIGN

Meaning

Involves two groups of participants:

 Experimental group (receives treatment)

 Control group (does not receive treatment)

Types of Two-Group Designs

1. Posttest-Only Control Group Design

 Two groups are formed randomly.

 Only posttest is conducted.

Structure:
Experimental: X → O₁
Control: — → O₂

2. Pretest–Posttest Control Group Design

 Both groups are tested before and after treatment.

Structure:
Experimental: O₁ → X → O₂
Control: O₁ → — → O₂
Key Feature: Random Assignment

 Participants are randomly assigned to groups.

 Helps control extraneous variables.

Advantages

 Better control over variables

 Allows comparison between groups

 Higher internal validity than one-group design

 Stronger cause–effect conclusions

Disadvantages

 Requires more participants

 Ethical issues (withholding treatment)

 Time-consuming and resource-intensive

3. FACTORIAL DESIGN

Meaning

A factorial design studies two or more independent variables (factors) simultaneously to

examine:

 Their individual effects (main effects)

 Their combined effect (interaction effect)

Basic Structure

Represented as:
2 × 2, 2 × 3, 3 × 3, etc.

 First number = levels of first independent variable

 Second number = levels of second independent variable

Example of 2 × 2 Factorial Design

Variables

 Factor A: Type of Therapy (CBT / No Therapy)

 Factor B: Gender (Male / Female)

Groups Formed

1. CBT – Male

2. CBT – Female

3. No Therapy – Male

4. No Therapy – Female

Types of Effects

1. Main Effect

 Effect of one independent variable ignoring the other variable

Example:
Effect of therapy on depression (regardless of gender)

2. Interaction Effect

 Occurs when the effect of one variable depends on the level of another variable

Example:
CBT works better for females than males → interaction

Advantages

 Studies multiple variables simultaneously

 Detects interaction effects (very important in psychology)

 More efficient than multiple separate experiments

 Provides richer and deeper understanding

Disadvantages

 Complex design and analysis

  Requires large sample size

 Difficult interpretation when many variables are involved

COMPARISON SUMMARY

Feature One-Group Two-Group Factorial

Groups One Two Multiple

Variables One IV One IV Two or more IVs

Control Low High Very High

Complexity Simple Moderate Complex

Validity Low High Very High

KEY EXAM POINTS (Quick Revision)

 One-group design → No comparison group → Weak validity

 Two-group design → Experimental vs control → Stronger conclusions

 Factorial design → Multiple IVs → Studies interaction effects

 Main effect vs interaction effect is crucial in factorial design

Here’s your content rewritten into clear, structured exam-ready notes (with headings,
subheadings, and all points preserved) 👇

Quasi-Experimental Design

Definition

 A quasi-experimental design applies an experimental interpretation to results that

do not meet all requirements of a true experiment.

 The independent variable is manipulated, but participants are not randomly

assigned to conditions.

 Since IV is manipulated before measuring the dependent variable, it eliminates the

directionality problem.

 Also called a compromise design.

  Useful when true experiments are not feasible due to ethical or practical reasons.

Types of Quasi-Experimental Design

1. Nonequivalent Groups Design

 Researcher selects existing groups that appear similar.

 Only one group receives the treatment.

 Unlike true experiments:

o Groups are not randomly assigned

o Hence, they may differ → nonequivalent groups

 Researchers try to:

o Control confounding variables

o Choose groups that are as similar as possible

 ✅ Most common quasi-experimental design

2. Time Series Design

 Used when a control group is not possible

 Involves:

o Multiple pretests

o Followed by treatment

o Then multiple posttests

 Difference from single-group pretest-posttest:

o Uses series of observations, not just one

 Features:

o Better control over extraneous variables

o No history effect considered

o Avoids simple statistical comparisons

3. Equivalent Time Series Design

  Extension of time series design

 Uses single group

 Treatment is:

o Introduced and reintroduced repeatedly

 Diagram:

o X₁O₁ X₀O₂ X₁O₃ X₀O₄

 Where:

o X₁ = Treatment

o X₀ = No treatment / alternative condition

 Advantages:

o Controls history effects

o More refined and improved design

4. Counterbalanced Design

 Also called:

o Crossover design

o Switch-over design

o Rotation experiment

 Uses Latin Square arrangement:

o Each treatment appears once in each row and column

 Example:

o Multiple treatments given to different groups in different orders

 Purpose:

o Controls order effects

5. Separate Sample Pretest-Posttest Design

 Used when:

o Treatment cannot be given to all subjects at once

  Process:

o One sample → treatment → measurement

o Another sample → same treatment → measurement

 Different samples used for:

o Pretest and posttest

6. Patched-Up Design

 Starts with a weak/inadequate design

 Researcher adds features to improve it

 Goal:

o Control factors causing invalidity

 Improves:

o Overall validity of results

Quasi-Independent Variable Designs (Time-Based Designs)

1. Longitudinal Design

 Same group studied over time

 Purpose:

o Observe changes due to passage of time

 Limitations:

o Affected by extraneous events

o Limited generalizability over time

2. Cross-Sectional Design

 Between-subjects design

 Compares:

o Different groups at different ages or time points

 Faster than longitudinal

  But:

o Cannot track individual change over time

3. Cohort Design

 Combines:

o Longitudinal + multiple groups

 Studies:

o Different generations (cohorts) over time

 Shares:

o Similar advantages and disadvantages as longitudinal design

Advantages of Quasi-Experimental Design

 ✅ Higher external validity

o Conducted in real-world settings

 ✅ Better than non-experimental research

o More control over confounding variables

Disadvantages of Quasi-Experimental Design

 ❌ Lower internal validity than true experiments

o No randomization → harder to control all variables

 ❌ Retrospective data issues

o Data may be:

 Inaccurate

 Incomplete

 Difficult to access

When to Use Quasi-Experimental Design

1. Ethical Reasons
  When random assignment is unethical

 Example:

o Oregon Health Study

o It would be unethical to:

 Give health insurance to some

 Deny others intentionally

 Instead:

o Study naturally occurring situations (like lottery-based selection)

2. Practical Reasons

 When true experiments are:

o Too expensive

o Difficult to implement

o Require too many resources

 Quasi-experiments:

o Use existing or previously collected data

o Often rely on government or secondary data

Here are your quasi-experimental designs with clear, simple examples for each type (exam-
friendly and easy to remember) 👇

Quasi-Experimental Design (with Examples)

Definition (with example idea)

 IV is manipulated, but no random assignment

 Example:

o Studying effect of a new teaching method on one class vs another already

existing class (no random assignment)

Types with Examples

1. Nonequivalent Groups Design

Concept

 Two existing groups

 One gets treatment, the other does not

Example

 A researcher studies effect of CBT on anxiety

 Group 1: Patients from Hospital A → receive CBT

 Group 2: Patients from Hospital B → no CBT

 Groups may differ (age, severity) → nonequivalent

2. Time Series Design

Concept

 Repeated measurements before and after treatment on same group

Example

 Measuring students’ test scores

o Pretests: 3 monthly tests before new teaching method

o Treatment: New teaching method introduced

o Posttests: 3 tests after implementation

 Compare trends before vs after

3. Equivalent Time Series Design

Concept

 Treatment is repeatedly introduced and removed

Example

 Studying effect of music on productivity

o Week 1: Music (X₁) → measure output

o Week 2: No music (X₀) → measure

o Week 3: Music again → measure

 o Week 4: No music → measure

 If productivity increases only during music weeks → stronger evidence

4. Counterbalanced Design (Crossover Design)

Concept

 Same participants receive all treatments in different orders

Example

 Studying effect of 3 diets on weight loss

 Group 1: Diet A → B → C

 Group 2: Diet B → C → A

 Group 3: Diet C → A → B

 Controls order effects

5. Separate Sample Pretest-Posttest Design

Concept

 Different samples for pretest and posttest

Example

 Study effect of awareness program on drug use

o Sample 1: Measure drug awareness (before program)

o Sample 2: After program → measure awareness

 Same treatment, but different participants

6. Patched-Up Design

Concept

 Start with weak design → add controls later

Example

 Researcher studies online learning effectiveness

 Initially:
 o Only posttest scores taken (weak design)

 Later adds:

o Pretest scores

o Control variables (IQ, prior knowledge)

 Improves validity

Quasi-Independent Variable Designs (Time-Based)

1. Longitudinal Design

Concept

 Same group studied over time

Example

 Follow children from age 5 to 15

 Measure:

o Cognitive development each year

2. Cross-Sectional Design

Concept

 Different age groups studied at same time

Example

 Compare:

o 5-year-olds, 10-year-olds, 15-year-olds

 Measure cognitive ability at one point in time

3. Cohort Design

Concept

 Multiple groups from different generations followed over time

Example

 Study:
 o People born in 2000, 2010, 2020

 Track their development over years

 Compare generational differences

When to Use (with Examples)

1. Ethical

 Example:

o Cannot deny life-saving treatment to control group

o So use existing groups instead

2. Practical

 Example:

o Government already collected education data

o Researcher uses it to study policy impact instead of conducting costly

experiment

Quick Memory Trick 🧠

 Nonequivalent → Existing groups

 Time series → Many tests before & after

 Equivalent time series → Treatment ON-OFF

 Counterbalanced → All groups get all treatments

 Separate sample → Different people pre & post

 Patched-up → Fix weak design

If you want, I can next convert this into MCQs or case-based questions (very useful for your
exam).