Module 0.4 & 0.5 – Correlation, Experimentation, Research Design & Ethics

Page 1

Module 0.4 – Correlation & Experimentation (Intro + Key Terms)
• Learning Targets 0.4-1 to 0.4-3:
– Explain correlation (+/-).
– Explain illusory correlations & regression toward the mean.
– Describe experimental characteristics that isolate cause & effect.
• Key Vocabulary introduced:
– Correlation = measure of how two factors vary together; how well either predicts the other.
– Correlation coefficient $(r)$ ranges from $-1.00$ to $+1.00$.
– Variable = anything that can vary & is feasible/ethical to measure.
– Scatterplot = graphed cluster of dots (each dot = two-variable value pair); slope ⇒ direction, scatter ⇒ strength.
• Psychologists use multiple research methods:
– Correlation (non-experimental) describes relationships.
– Experiments attempt to establish cause–effect.

Page 2

0.4-1 What does “correlated” mean? Positive vs. Negative Correlations
• Describing behavior is 1st step → prediction.
• Naturalistic observations & surveys often reveal that traits/behaviors ‘coincide’.
• Correlation coefficient quantifies how closely two things vary together → predictive power (e.g., aptitude scores & school success).
• Scatterplots reveal relationships visually:
– Perfect positive $r=+1.00$ (direct proportional increase).
– No relationship $r=0.00$.
– Perfect negative $r=-1.00$ (one increases, other decreases proportionally).
• Positive correlation: two sets rise/fall together (height & weight).
• Negative correlation: inverse relation (height vs. distance head-ceiling).
• Statistics catch patterns casual observation misses. Example: Czech/Slovak sample (N = 2291) rating fear/disgust for 24 animals (Table 0.4-1).

Page 3

Working with Table 0.4-1 – Fear vs. Disgust Example
• Raw table doesn’t reveal relationship easily.
• Scatterplot (Fig 0.4-2) shows upward slope ⇒ positive correlation: $r=+0.72$.
• Key skills: convert raw data → graphical display to ‘see’ relationships.
• AP Science Tips:
– ‘Positive/negative’ refer only to direction, NOT desirability/strength.
– Correlation ≠ causation (directionality & third-variable problems).
• Table 0.4-2: practice classifying example news statements as positive/negative correlations.

Page 4

AP Science Practice – Developing Arguments: Correlation vs. Causation
• Directionality problem illustrated with smoking & mental illness. 3 possible explanations:

1. Smoking → mental illness.

2. Mental illness → smoking.

3. Third variable (e.g., stressful home) → both.
   • Similar logic applied to sexual hook-ups & depression, parental love & teen health behaviors.
   • Take-away: Correlation suggests possible cause–effect but does not prove it.

Page 5

0.4-2 Illusory Correlations & Regression Toward the Mean
• Illusory correlation = perceiving relationship where none exists or overestimating strength.
– Fueled by confirmation bias (noticing confirming instances).
– Example: belief that dreams predict events; gamblers thinking they influence dice.
• Regression toward the mean = tendency for extreme/unusual scores/events to fall back to average on later trials.
– Extreme low test score often followed by higher score because unlucky factors unlikely to co-occur.
– Leads to superstitious attributions (coach’s scolding ‘worked’ because performance regressed).
• Check-Your-Understanding prompts: interpret $r=-0.87$ (strong negative), describe scatterplot, identify media misuses, explain basketball coach example.

Page 6

0.4-3 Experimentation – Isolating Cause & Effect
• Real-world example: smartphone/social-media boom & teen girls’ depression spike. Correlation & longitudinal studies suggest link but experiments needed for causality.
• Experiments isolate factors by:

1. Manipulating IV(s).

2. Holding other factors constant (control).
   • Key design elements:
   – Experimental group (receives treatment).
   – Control group (no / alternative treatment).
   – Random assignment equalizes groups (distinguish from random sampling).
   • Example study: Facebook deactivation (N≈1700, 4 weeks). Results: More TV & F2F socializing; lower depression; ↑ life satisfaction; ↓ post-experiment Facebook use.
   • General consensus: unlimited teen social-media poses modest risk; ongoing research.

Page 7

Procedures & Placebo Effect
• Need to control misleading testimonials (e.g., zinc for colds, historical bloodletting).
• Single-blind: participants unaware of assignment.
• Double-blind: participants & researchers unaware (controls placebo & experimenter bias).
• Placebo effect: expectations alone produce measurable change (pain, depression, athletic performance, caffeine illusion; price increases effect).

Page 8

Independent vs. Dependent Variables; Confounding Variables
• Illustrative classroom-quiz experiment:
– IV = study procedure (review vs. self-test).
– DV = final-exam performance.
• IV varied independently of participant traits; confounding variables = extraneous factors needing control.
• Operational definitions specify how IV is manipulated & DV measured → enables replication.
• Benassi study: testing (75 %) vs. restudy (51 %).
• Facebook experiment diagram (Fig 0.4-3) summarises: random assignment → groups → measure depression after 4 weeks.

Page 9

Understanding & Validity Check
• Rental-housing discrimination experiment:
– IV = implied ethnicity of name (“McDougall”, “Al-Rahman”, “Jackson”).
– DV = % positive replies (89 %, 66 %, 56 %).
• Validity = extent experiment tests what intended.
• Random assignment controls confounds; IV manipulated, DV measured.
• Review match-ups:
i Double-blind ⇒ controls placebo (c).
ii DV ⇒ outcome measured (a).
iii Random assignment ⇒ equalize groups (b).
• Random assignment vs. random sampling difference emphasized.

Page 10

AP Practice MC (Correlation & Experimentation)
• Sample Qs: Identify negative correlation (exercise ↑, depression ↓), IV in temp-performance study (room temp), strongest $r$ value $-0.85$, purpose of random assignment (reduce confounds), recognize double-blind, role of control group.

Page 11

Module 0.5 – Research Design & Ethics (Intro + Key Terms)
• Learning Targets 0.5-1 to 0.5-4:
– Choose appropriate research design.
– Value of simplified lab conditions.
– Why study animals & ethical guidelines.
– How psychologists’ values influence work.
• Table 0.5-1 contrasts main methods:
– Non-experimental (Case, Naturalistic, Surveys): observe/record → no manipulation; weakness = no causal inference.
– Correlational: detect relationships; cannot specify cause-effect.
– Experimental: explore causation; manipulate IV; weaknesses = feasibility, generalizability, ethics.

Page 12

0.5-1 Choosing Research Design & Measurement
• Question selection limited by testability & ethics (free will, evil, afterlife not directly testable).
• Design options: experimental, correlational, case, naturalistic, twin, longitudinal, cross-sectional.
• Constraints: money, time, ethics.
• Quantitative research = numerical (e.g., Likert scales).
• Qualitative research = narrative/ in-depth (interviews).
• Researchers guard against confounding variables & strive for diversity, equity, inclusion.

Page 13

0.5-2 Predicting Everyday Behavior: Lab ⇄ Life
• Lab env. = simplified reality; simulates psychological forces under control (like wind-tunnel analogy).
• Goal = test theoretical principles, not replicate every life detail.
• Principles usually generalize (e.g., aggression button-press ↔ real aggression; signal detection ↔ night flying).

Page 14

0.5-3 Why Psychologists Study Animals & Ethical Safeguards
• Reasons: fascination, understanding learning/cognition, shared biology; leads to medical & psychological advances (insulin, vaccines, transplant methods).
• Ethical debate:
– Is human welfare > animal welfare?
– What safeguards?
• Guidelines:
– BPS: natural housing, social companions.
– APA: humane care, minimize discomfort.
– EU Parliament & institutional Animal Care committees monitor.
• Benefits to animals themselves (dog-shelter stress reduction, zoo enrichment).

Page 15

Studying & Protecting Humans
• Most studies mild; some require deception/stress; confederates used when needed.
• Early controversial studies wouldn’t pass modern ethics (monkey deprivation, Little Albert, WWII starvation).
• APA & BPS four-part ethics code:

1. Obtain informed consent/assent pre-study.

2. Protect from >-usual harm/discomfort.

3. Maintain confidentiality.

4. Debrief fully post-study.
   • Institutional Review Boards (IRB): ≥5 members (scientist, non-scientist, community).

Page 16

Scientific Integrity & Values (0.5-4)
• Honesty is paramount; fraud (plagiarism, data fabrication) ends careers (e.g., retracted 1998 MMR-autism hoax → decreased vaccinations, measles resurgence).
• Values influence: choice of topics (e.g., worker morale vs. productivity), framing/labels (rigid vs. consistent, faith vs. fanaticism), interpretation.
• Psychology’s power can be used for good (education, compassion) or harm (manipulation); ethical responsibility to enlighten.
• Classic application: Clark & Clark doll studies cited in 1954 Brown v. Board school-desegregation ruling.

Page 17

Module 0.5 Review Highlights
• Design: generate testable Qs → choose design → measure variables → interpret with eye on confounds.
• Lab principles explain everyday behavior.
• Animal research: guided by legal & professional standards; aims at human & animal welfare.
• Human research: informed consent, debriefing, confidentiality, minimized harm.
• Values shape research but applications largely serve humanity.

Page 18

AP Practice MC (Research Design & Ethics)
• Informed consent = pre-study info to allow rational participation choice.
• Debriefing = post-study explanation, incl. deception.
• Lab aims to simulate psych forces under controlled conditions.
• IRB-approvable animal study example: teaching dolphins simple language (minimal harm, potential benefit).
• Scientific process order: create hypotheses → design & measure variables → interpret results.

Essential Equations & Numerical References
• Correlation coefficient range: $-1.00 \le r \le +1.00$.
• Fear–Disgust example: $r=+0.72$.
• Example strong negative: $r=-0.87$.
• Benassi quiz experiment result: $75\%$ vs. $51\%$ correct.
• Rental study invitation rates: $89\%$, $66\%$, $56\%$.

Common Pitfalls & Exam Reminders
• Positive/negative describe direction, not strength or desirability.
• Correlation does NOT imply causation; remember directionality & third-variable issues.
• Random assignment ≠ random sampling.
• IV = manipulated; DV = measured outcome.
• Placebo & double-blind control expectancy & experimenter biases.
• Validity & reliability are separate concepts.

Ethical Acronyms to Memorize
• APA, BPS, IRB, IACUC (Institutional Animal Care & Use Committee).

Study Strategies
• Practice identifying IV/DV/confounds in real & hypothetical studies.
• Convert raw data tables into scatterplots to visually gauge $r$.
• Explain regression toward mean & placebo effect using personal examples.
• Review ethical principles & be able to contrast informed consent vs. debriefing.