KAAP 400 - Lecture 11

What is a pre-experimental design?

A pre-experimental design introduces an intervention but lacks random assignment or a true control group, making it the weakest form of experimental design for establishing causation. It is often used for pilot studies, feasibility testing, and preliminary investigations.

Example: One group has their VO₂ max measured before → after a 3-week HIIT program.

What is a true experimental design?

A true experimental design is a study in which researchers:

• Manipulate one or more independent variables (IVs).

• Randomly assign participants to different groups.

• Measure one or more dependent variables (DVs).

• Control confounding variables as much as possible.

It is considered the strongest research design for testing cause-and-effect relationships.

What is a pretest-posttest control group design?

Participants are randomly assigned to an intervention group or control group and measured before (pretest) and after (posttest) the intervention.

Why it is strong: It allows researchers to compare changes between groups and account for baseline differences, strengthening causal inference.

Limitation: The pretest may create a testing effect, and the design requires more time and resources.

Example: Older adults are randomly assigned to a 4-week balance program or usual activity. Postural sway is measured before and after the program.

What is a posttest-only control group design?

Participants are randomly assigned to an intervention or control group, but no pretest is conducted. Both groups are measured only after the intervention (posttest).

Advantages:

• Randomization helps make groups similar at baseline.

• Avoids potential testing effects caused by a pretest.

Limitation:

• Cannot measure individual change over time.

• Assumes randomization successfully made the groups comparable at baseline.

Example: Older adults are randomly assigned to a balance program or usual activity, and postural sway is measured only after 4 weeks

What is a factorial design?

A factorial design investigates two or more independent variables (IVs) simultaneously and examines their individual effects and whether they interact with each other.

Example:

• IV 1 = Exercise type: Resistance vs. Aerobic

• IV 2 = Intensity: Moderate vs. High

• DV = VO₂ max

Advantages: Can test interactions between independent variables.

Limitation: The statistical analysis can be more complex after data collection

What is a repeated measures design?

A repeated measures design measures the same participants multiple times, either across different conditions or at different time points.

Key idea: Each participant serves as their own comparison/control, which reduces variability between individuals.

Advantages:

• Reduces variability because participants are compared to themselves.

• Useful for studying changes over time.

• Can compare different treatments or conditions.

Example: Measure the same athletes’ heart rates during low-, moderate-, and high-intensity exercise, or measure balance before, during, and after a 6-week exercise program.

What is a crossover design?

A crossover design is a type of repeated measures design in which each participant receives two or more interventions sequentially, with a washout period between them.

Key features:

• Each participant serves as their own control.

• Participants are often randomized to the order of treatments.

• A washout period helps prevent carryover effects from the first intervention.

• Researchers compare each participant’s response to the different interventions.

Example: Compare treadmill vs. cycling training on VO₂ max:

• Group 1: Treadmill → Washout → Cycling

• Group 2: Cycling → Washout → Treadmill

What is a Randomized Controlled Trial (RCT)?

A Randomized Controlled Trial (RCT) is a study that compares different treatments or interventions by randomly assigning participants to an intervention group or a control group.

• Experimental group: Receives the treatment/intervention.

• Control group: Receives a comparison treatment, placebo, or no treatment.

• Random assignment: Helps make groups similar at the beginning of the study.

• Outcome: Researchers measure an outcome, such as blood pressure, pain, or mobility.

Key idea: RCTs are considered a strong research design for testing treatment effectiveness and cause-and-effect relationships.

Example: 100 people with knee pain are randomly assigned to a new exercise program or usual care. Researchers then compare changes in pain and mobility between the groups.

What is blinding, and what are the different levels?

Blinding means keeping participants or researchers unaware of group assignments to reduce bias.

• Single-blind: Participants do not know which group they are in.

• Double-blind: Participants and researchers do not know group assignments.

• Triple-blind: Participants, researchers, and data analysts do not know group assignments.

Example: In a migraine drug trial, participants may not know whether they received the drug or placebo. In a double-blind study, the researchers administering the treatment also do not know. In a triple-blind study, the statistician analyzing the data is also unaware of which group is the treatment group.

What are placebo and sham conditions, and why are they used?

A placebo is an inactive treatment designed to look like the real intervention but does not contain the active ingredient. A sham is similar but usually refers to a procedure or intervention that is performed without the active component.

Purpose: Both help control for participants’ expectations and psychological effects of receiving a treatment.

Example:

• Placebo: A sugar pill that looks like the real medication.

• Sham: A neurostimulation device that looks and feels like the real device but does not actually provide stimulation.

Key idea: Participants should have difficulty telling the real and placebo/sham conditions apart.

Why are placebo/sham conditions necessary in research?

Placebo and sham conditions are necessary because participants can experience real changes simply because they expect a treatment to work. This is called the placebo effect.

Key examples:

• Henry Beecher (WWII): After morphine ran out, saline was given to an injured soldier while being described as a powerful painkiller. The soldier reportedly experienced pain relief. Beecher later highlighted the importance of placebo controls in clinical research.

• Bruce Moseley (2002): In a study of knee osteoarthritis, patients receiving sham knee surgery reported improvements in pain and mobility similar to those receiving the actual procedure.

Key idea: A placebo/sham condition helps researchers determine whether improvements are caused by the active treatment itself or by expectations and the experience of receiving treatment.

What is a Solomon 4-Group Design?

The Solomon 4-Group Design is an experimental design that uses four randomly assigned groups to determine both the effect of an intervention and whether taking a pretest influences the results.

The 4 groups:

1. Pretest → Intervention → Posttest

2. Pretest → Control → Posttest

3. No pretest → Intervention → Posttest

4. No pretest → Control → Posttest

Why use it? It allows researchers to determine whether the pretest itself changes participant behavior or outcomes, in addition to testing the intervention.

Example: Athletes who see their baseline VO₂ max might train harder before the intervention begins. This is called a pretest effect.

What is a quasi-experimental design?

A quasi-experimental design is a study where the researcher manipulates an independent variable (IV) but does not randomly assign participants to groups.

Key idea:

• Quasi-experimental: Intervention + no random assignment

• True experimental: Intervention + random assignment

Because groups are often pre-existing, differences between groups may influence the results, making it harder to establish cause and effect.

Example: One school implements a new physical activity program while another school does not. Researchers compare students’ fitness outcomes, but differences in baseline fitness, motivation, or socioeconomic status could affect the results.

How can a quasi-experimental study be converted into a true experimental design?

Convert it by adding random assignment and strengthening control:

1. Random assignment: Randomly assign students/classes to an intervention or control group.

2. Standardize the intervention: Keep the program duration, frequency, and intensity consistent.

3. Pretest-posttest: Measure outcomes before and after the 12-week program.

4. Control confounding variables: Use randomization and consistent procedures to balance factors such as age, sex, and baseline fitness.

5. Blind assessors when possible: Keep outcome assessors unaware of group assignments to reduce observer bias.

Key idea: The major difference is random assignment.
Quasi-experimental = intervention + no random assignment
True experimental = intervention + random assignment