🎯⭐ INTERACTIVE LESSON

Research Methods & Study Design

Learn step-by-step with interactive practice!

← Back to Standard Lesson

Research Methods & Study Design - Complete Interactive Lesson

Part 1: Variables, Sampling & Study Types

Research Methods & Study Design

Part 1 of 4 — Variables, Sampling & Study Types

Types of Variables

Variable Type	Definition	Example
Independent Variable (IV)	What the researcher manipulates/measures	Drug dose, light exposure, temperature
Dependent Variable (DV)	What the researcher measures as an outcome	Patient recovery time, test score, enzyme activity
Confound Variable	Unmeasured/uncontrolled variable affecting DV	Age, baseline health status, observer bias

Sampling Methods

Method	Description	Bias Risk
Random Sampling	Every participant has equal chance	Low bias; representative
Convenience Sampling	Easiest to access (first n patients)	High bias; may not represent population
Stratified Sampling	Divide population into groups, sample proportionally	Lower bias than convenience
Matched Sampling	Match participants on key variables	Controls specific confounds; less effective than randomization

Study Types (by Causation Inference)

Type	Design	Causation Evidence
Experimental (RCT)	Researcher manipulates IV, randomly assigns	Strongest
Quasi-Experimental	Researcher manipulates IV, no randomization	Moderate
Correlational	Researcher measures variables, finds association	Weak
Observational	Passive observation; no manipulation	Weak

Key: Only random assignment (RCT) can establish causation by balancing confounds.

Variables & Sampling 🎯

Key Takeaways — Part 1

IV = Independent Variable (what's manipulated/measured); DV = Outcome (what's measured)
Confound = Unmeasured variable that could influence DV
Random Assignment ⟹ RCT ⟹ Strongest causation inference
Convenience/Stratified Sampling ⟹ Observational ⟹ Weaker causation inference
Matched Sampling controls specific confounds but not unknown ones (inferior to randomization)

Worked Examples — Variables, Sampling & Study Types

<details> <summary>Example 1: Separate IV and DV cleanly</summary>

Question: Participants receive either 0 mg, 50 mg, or 100 mg caffeine, then complete a reaction-time test.

Manipulated factor is caffeine dose.
Measured outcome is reaction time.

IV: caffeine dose. DV: reaction time.

</details> <details> <summary>Example 2: Spot sampling bias</summary>

Question: A stress survey recruits only pre-med students from one campus.

Recruitment is convenience-based and narrow.
Sample likely differs from general population.

Main issue: selection/sampling bias, reducing external validity.

</details> <details> <summary>Example 3: Distinguish observational from experimental</summary>

Question: Researchers record average sleep and exam scores without assigning sleep schedules.

No manipulation of sleep duration.
No random assignment.

Design type: observational/correlational, not experimental.

</details>

Part 2: Validity & Threats to Validity

Research Methods & Study Design

Part 2 of 4 — Validity & Threats to Validity

Types of Validity

Validity Type	Definition	Example
Internal Validity	Causal inference: Did IV cause the DV?	Did drug (not age/diet) cause recovery?
External Validity	Generalizability: Do results apply beyond the study?	Do lab findings translate to real patients?
Construct Validity	Does the measurement actually measure what it claims?	Does IQ test truly measure "intelligence"?
Statistical Conclusion Validity	Are conclusions about relationships statistically sound?	Is the sample large enough to detect effect?

Threats to Internal Validity

Threat	What it is	Example
Confounding	Unmeasured variable causes apparent effect	Older patients recover faster (age ← good health status)

Part 3: Blinding & Experimental Controls

Research Methods & Study Design

Part 3 of 4 — Blinding & Experimental Controls

Blinding in Studies

Type	Who is Blinded?	Effect
No Blinding	Participants & researchers know treatment	Maximum placebo effect + researcher bias
Single-Blind	Participants don't know (researcher knows)	Reduces placebo effect; researcher may bias
Double-Blind	Both participants & researchers don't know	Gold standard; minimizes bias

Mechanism of Placebo Effect: Brain expectation → neurotransmitter release → real physiological changes (30-40% of patients benefit from placebo alone)

Placebo & Control Groups

Group	Purpose
Placebo Control	Isolates IV effect from placebo effect
Active Control	Compares new drug to gold-standard treatment (ethical)
No-Treatment Control

Part 4: Sample Size, Ethics & Meta-Analysis

Research Methods & Study Design

Part 4 of 4 — Sample Size, Ethics & Meta-Analysis

Sample Size & Power

Larger sample → More power to detect true effects (↓β, Type II error)

Factor	Effect on Power
Larger sample size	↑ Power
Larger effect size	↑ Power
Lower variability	↑ Power
Higher α (0.05 vs 0.01)	↑ Power

Rule of thumb: Aim for 80%+ power (allow 20% ≤β).

Research Ethics (MCAT focus: Informed Consent, IRB)

Principle	Requirement
Informed Consent	Subjects understand risks/benefits; voluntary participation
Beneficence	Maximize benefits; minimize harms
Justice	Fair distribution of risks/benefits; equitable access
IRB (Institutional Review Board)	Ethical review before study starts

Extra protections for children, prisoners, cognitively impaired (cannot give true consent)

Threats to External Validity

Threat	Impact
Limited Sample	Results may not generalize to other populations
Lab Conditions	Artificial environment ≠ real-world effects
Volunteer Bias	Volunteers differ from general population
Interaction with IV	Effect depends on specific conditions (works in winter, not summer)

Solution: Randomization (internal validity) + Large diverse sample (external validity)

Validity & Threats 🎯

Key Takeaways — Part 2

Internal Validity: Random assignment controls confounds (high IV = good study)
External Validity: Representative sampling + diverse conditions enable generalization
Construct Validity: Measurement must truly measure the construct intended
Key Threats: Confounding, selection bias, attrition (internal); volunteer bias, lab conditions (external)
Ideal Study: Randomized (strong internal) + Large diverse sample (strong external)

Worked Examples — Validity & Threats

<details> <summary>Example 1: Internal vs external validity</summary>

Question: A tightly controlled inpatient trial shows strong results, but participants are all healthy young adults.

Strong protocol control supports internal validity.
Narrow participant profile limits generalization.

Result: high internal validity, weaker external validity.

</details> <details> <summary>Example 2: Identify attrition bias</summary>

Question: In a weight-loss trial, dropout is 35% in treatment vs 8% in control.

Unequal dropout changes who remains in each group.
Final comparison can be biased.

Threat: attrition bias.

</details> <details> <summary>Example 3: Construct validity check</summary>

Question: "Empathy" is measured only by number of social media comments posted.

Metric may not map to the intended construct.
Behavior proxy is weak and noisy.

Main concern: construct validity.

</details>

Example: Depression RCT

Placebo group: 30% improve (placebo effect alone)
Antidepressant group: 60% improve
True drug effect = 60% − 30% = 30%

Types of Experimental Designs

Design	Structure	Use Case
Within-Subjects	Same subjects in all conditions	Measures change pre-post (power maximized)
Between-Subjects	Different subjects per condition	No carryover effects; larger N needed
Factorial	Multiple IVs tested simultaneously	Efficiency; can detect interactions
Longitudinal	Follow subjects over time	Tracks development, long-term effects

Carryover Effect: Practice in Condition A affects performance in Condition B. Randomize order or use between-subjects design.

Blinding & Controls 🎯

Key Takeaways — Part 3

Double-Blind > Single-Blind > No Blinding (minimizes placebo effect + researcher bias)
Placebo Control isolates IV effect by subtracting baseline placebo response
Within-Subjects design increases power (subjects are own controls); risk of carryover effects
Between-Subjects design avoids carryover; requires larger N
MCAT Tip: Always check: Is the study blinded? Is there a placebo control? Could carryover effects bias results?

Worked Examples — Blinding & Experimental Controls

<details> <summary>Example 1: Why double-blind helps</summary>

Question: In a pain trial, nurses who know treatment assignment rate pain outcomes more favorably for the drug group.

Knowledge of assignment can influence ratings.
Blinding raters removes expectancy effects.

Fix: use a double-blind design.

</details> <details> <summary>Example 2: Compute true treatment effect from placebo baseline</summary>

Question: Placebo improves 25%; drug improves 55%.

Total observed drug-group response includes placebo component.
Subtract placebo baseline.

Estimated drug-specific effect: 30 percentage points.

</details> <details> <summary>Example 3: Carryover in within-subjects design</summary>

Question: Subjects take memory test after caffeine in week 1 and after placebo in week 2. Scores improve in week 2.

Practice effects can increase later scores independent of treatment.
Order effects threaten interpretation.
Counterbalancing or between-subjects design can reduce this issue.

Threat identified: carryover/order effect.

</details>

Special Populations:

Vulnerable Populations: Cannot be excluded from research solely for "protection"; must justify any exclusion

Combines data from multiple studies to increase statistical power.

Advantage: Large sample → more generalizable conclusions
Risk: Publication bias (only positive findings published)

Issue	Impact
Heterogeneity	Studies differ in methods; may not combine
Publication Bias	Null findings ↓ published; overestimates effect
Quality Variation	Poor-quality studies can bias pooled analysis

Forest Plot Interpretation: Vertical line at 1.0 (OR) means no effect; if confidence intervals cross it, effect not significant overall.

Sample Size, Ethics & Meta-Analysis 🎯

Key Takeaways — Part 4

Sample Size: Larger N → More power (can detect small true effects)
Power Goal: 80%+ (give ≤20% Type II error allowance)
Informed Consent: Disclosure of risks/benefits; voluntary; ability to withdraw
Vulnerable Populations: Include with extra protections, don't exclude
Meta-Analysis: Combines studies for power, but publication bias can overestimate effects
Forest Plots: If CI crosses 1.0 (OR), effect is not statistically significant
MCAT Tip: Ethics questions focus on consent, IRB review, protection of vulnerable groups

Worked Examples — Power, Ethics & Meta-Analysis

<details> <summary>Example 1: Increase power in a weak study</summary>

Question: Trial with n=40 per arm yields p=0.09 for a moderate effect.

Small sample can underpower the test.
Increasing n reduces standard error.
Better power improves chance of detecting a true effect.

Most direct improvement: larger sample size.

</details> <details> <summary>Example 2: Informed consent essentials</summary>

Question: A participant signs a form but was never told key side effects.

Signature alone is insufficient.
Valid consent requires meaningful disclosure and understanding.

Ethics issue: informed consent was inadequate.

</details> <details> <summary>Example 3: Read a forest plot threshold</summary>

Question: Pooled odds ratio is 0.85 with 95% CI 0.62 to 1.10.

For OR, null value is 1.0.
CI crosses 1.0.
Pooled effect is not statistically significant.

Interpretation: possible benefit, but inconclusive overall.

</details>