# Experimental Design

Psychology I · Research Techniques · https://tryals.app/learn/psychology-i/experimental-design

## Establishing Causality

A true **experiment** is the only design that establishes **causality**: showing that changing one factor genuinely causes a change in another. The researcher achieves this through **manipulation**.

| Term | Role |
| --- | --- |
| **Independent variable (IV)** | The factor directly manipulated by the researcher |
| **Dependent variable (DV)** | The outcome measured to see the effect |
| **Control group** | The baseline group receiving no treatment |
| **Confounding variable** | An uncontrolled factor that might explain results |

**Random assignment** gives every participant an equal chance of entering the **experimental group** or control group. This randomization spreads hidden differences evenly, ruling out confounding variables.

$$\text{IV manipulated} \to \text{DV measured}$$

## Experimental Rigor

Several safeguards protect findings from bias and error. **Operational definitions** state precisely how abstract concepts are measured to ensure replicability.

A **placebo** controls for psychological belief in a treatment. A **double-blind** procedure hides group assignments from both participants and staff, protecting against **experimenter bias** where researcher expectations subtly alter data.

| Safeguard | What it Guards Against |
| --- | --- |
| **Placebo** | Psychological expectancy effects |
| **Double-blind** | Participant expectations and experimenter bias |
| **Operational definition** | Ambiguity and poor replicability |

> **Common pitfall:** Assuming any two-group comparison is a true experiment. Without researcher-controlled **random assignment**, a study is merely a quasi-experiment and cannot establish strong causality.

## Practice questions

4 of this lesson's 11 practice questions, with answers. The full set is in the app.

### 1. Select every statement that is TRUE about true experiments.

A. Manipulating the IV without random assignment still counts as a true experiment.
B. Random assignment is what allows an experiment to rule out confounds on average.
C. The independent variable is the outcome the researcher measures.
D. Operational definitions make a study replicable by specifying how concepts are measured.
E. Double-blind guards against experimenter bias and participant expectations.

**Answer:** B. Random assignment is what allows an experiment to rule out confounds on average.; D. Operational definitions make a study replicable by specifying how concepts are measured.; E. Double-blind guards against experimenter bias and participant expectations.

**Why:** Random assignment ruling out confounds, the double-blind guarding both sides, and operational definitions enabling replication are all true. The IV is manipulated (not measured), and manipulation without random assignment is only quasi-experimental - those two are false.

Page: https://tryals.app/practice/psychology-i/experimental-design/select-every-statement-that-is-true-about-true-experiments

### 2. A trial gives one group a drug and the other a placebo, and neither the participants nor the staff measuring the outcome know who got which. Why is keeping the STAFF blind necessary, given that the participants are already blind?

A. Because the placebo effect operates on the person administering the treatment as strongly as on the person receiving it.
B. Because an assessor who knows the allocation can shade judgements.
C. Because blinding the staff prevents them from accidentally revealing the allocation to the participants, which is the only real risk.
D. Because the staff might deliberately falsify the results if they knew the allocation.

**Answer:** B. Because an assessor who knows the allocation can shade judgements.

**Why:** Single blinding handles the participant's expectations and leaves the assessor's untouched, and ratings of pain or mood are full of borderline calls a believer resolves in one direction, without cheating, because that is how expectation acts on judgement. There is a second route too: an assessor may probe one group more encouragingly, so the procedure stops being identical across arms.

Page: https://tryals.app/practice/psychology-i/experimental-design/a-trial-gives-one-group-a-drug-and-the-other-a-placebo-and-neither

### 3. Match each safeguard to the specific threat it is designed to counter.

**Answer:**

- Random assignment → Confounding variables differing between groups
- Placebo control → The effect of merely believing you are treated
- Double-blind procedure → Participant expectation AND experimenter bias
- Operational definition → Vagueness that would make the study non-replicable

**Why:** Random assignment counters confounds, the placebo counters belief effects, the double-blind counters both participant and experimenter expectations, and operational definitions counter vagueness so the study can be replicated. Each safeguard targets one specific threat to validity.

Page: https://tryals.app/practice/psychology-i/experimental-design/match-each-safeguard-to-the-specific-threat-it-is-designed-to-counter

### 4. A researcher could **match** two groups carefully on age, sex and income, or assign at random. Randomisation usually wins. Why prefer chance over deliberate care?

A. Because randomisation produces larger samples than matching
B. Because matching only equates the variables you thought of
C. Because matching introduces experimenter bias in how pairs are selected
D. Because matched groups are harder to recruit, making randomisation more practical

**Answer:** B. Because matching only equates the variables you thought of

**Why:** Matching handles your list; randomisation handles the list you could not write. That is the whole of its power, it neutralises unmeasured and unimagined differences at once, which no amount of care about age and income can do.

Page: https://tryals.app/practice/psychology-i/experimental-design/a-researcher-could-match-two-groups-carefully-on-age-sex-and-income