Experimental Design: Quick Reference Notes

Participants ideally unaware of group assignment to avoid behavior change due to awareness.
Hawthorne effect: simply being part of a study can alter outcomes.
Use random assignment to groups to reduce biases from knowledge of condition.
In drug studies, use a placebo for control group to compare against the active treatment.
Independent variable (IV) is the drug; the observed outcome is the dependent variable (DV).
Placebo can lead to perceived improvement even without an active drug (placebo effect).
Example: antidepressants often take 4–6 weeks to enter the bloodstream; early improvements may be placebo-driven.

IV: the drug (the treatment being tested) $\text{IV} = \text{drug}$
DV: the outcome measured (e.g., headache relief) $\text{DV} = \text{headache relief}$
Operational definitions are required to specify exactly how variables are measured.

Define what you mean by each variable in real-life terms so others can replicate.
Example hypothesis: chocolate causes violent behavior.
Operationalize chocolate: e.g., use a specific chocolate bar (e.g., Hershey) and a defined amount.
Operationalize violent behavior: a specified set of actions (e.g., kicking, fighting, scratching).
Clear operational definitions improve validity and reliability by enabling replication.

For a hypothesis to become a theory, results must be replicated across studies with the same methods and results.
Define variables precisely to ensure other researchers can replicate the experiment.
Validity and reliability depend on clear definitions and consistent procedures.

A confounding variable is anything that causes the DV other than the IV.
Confounds threaten the internal validity of an experiment.
Examples of potential confounds:
- An argument with a parent before the study.
- Alcohol consumption.
- Not eating breakfast (affects mood/behavior).
In a smoking-and-heart-issues study, confounds include genetics, diet, lifestyle, and overall health.

Hypothesis: chocolate causes violent behavior.
Operationalize:
- Chocolate = a specified type/brand (e.g., Hershey bar).
- Violent behavior = defined set of actions (e.g., kicking, fighting, scratching).
Why: ensures different researchers use the same measurement and can replicate results.

Random assignment helps minimize Hawthorne effects and bias.
Use a placebo when a drug is involved to isolate the drug effect (IV) on the DV.
Clearly define operational definitions to enable replication and assess validity.
Watch for confounding variables that could explain the DV.