Scientific Process & Experimental Design – Bullet-Point Revision Notes

Chapter 1 – The Scientific Method

1.1 Introduction

• Everyday stereotype: psychology = Sigmund Freud, Dr Phil, clinical work.
• Reality: > $50$ APA divisions (e.g., Military, Sport, Organisational, Law, Addictions).
• Unifying feature across this diversity = commitment to the scientific method.

1.2 Psychology & the Scientific Process

• Epistemology asks, “How do we know what we know?” Five common sources:
– Personal experience / intuition ➜ prone to selective perception & over-generalisation (jogging–knee myth).
– Authority ➜ efficient but credentials & evidence often unquestioned.
– Rationalism (reason) ➜ syllogisms valid only if premises sound (e.g., “All brown dogs are friendly” vs “All men are mortal”).
– Empiricism (observation) ➜ observations filtered by senses, culture, mood, intoxication.
– Science ➜ combines reason & empirical evidence within a self-correcting method.

1.2.5 Knowledge from Science – 5 Core Attitudes

• Objectivity
• Skepticism
• Openness / open-mindedness
• Tentativeness
• Independence from authority

1.3 Principles of the Scientific Method

• Objectivity – evidence observable by anyone; use physiological indices (heart-rate, GSR) instead of self-report.
• Skepticism – “Show me the data”; Galileo dropping objects in a vacuum disproved Aristotle.
• Openness – full method disclosure, inter-observer reliability; willingness to revise interpretations.
• Tentativeness – findings always provisional (global-warming debate; Pavlov’s conditioning replicated $\sim$ $1{,}000s$ of times).
• Independence from authority – no “because I said so”.

1.4 Assumptions of Science

• Nature is lawfully organised.
• Determinism – lawful relations allow prediction (forgetting curve example).
• Concern with solvable problems – rephrase metaphysical questions into testable forms (bystander effect derived from “Are people good?”).

1.5 Goals of Psychological Science

1. Describe behaviour (pilot altitude under-estimation at night).

2. Explain behaviour (lack of visual cues, pseudo-explanations to avoid).

3. Predict behaviour (night ↔ worse estimates).

4. Control behaviour (add ground lights ⇒ improved landings).

1.6 Theories & Hypotheses

• Theory = logically organised propositions that summarise, organise, explain & predict.
– Example: Piaget’s $4$ developmental stages (sensorimotor $0!–!2$ yrs … formal operations $11+$ yrs).
– Operational definitions needed (e.g., “object permanence”).
• Hypothesis = specific, testable, tentative prediction derived from theory.
– Null ($H<em>0$): no difference; Alternate ($H</em>1$): predicted difference.
• Sources: literature gaps, replication (exact / conceptual), serendipity.
• Falsifiability – must be possible to refute (gravity drop test analogy).
• Pitfalls: circularity, supernatural forces, ill-defined terms, lack of parsimony (Lloyd Morgan’s Canon).

1.7 Scientific vs Non-Scientific Evidence

• Criteria: empirical, objective, systematic, controlled.
• Randomised, repeatable observations resolve disputes.

1.8 Critical Evaluation Checklist

• Did study uphold objectivity, skepticism, openness, tentativeness, independence?
• Evidence test: empirical? objective? systematic? controlled?

1.9 The Scientific Process (Flow-Chart)

Theory → Hypothesis → Design → Data (describe & analyse) →
– Supported → report → replication → scientific “truth”.
– Not supported → refine / discard → new cycle.

Chapter 2 – Research & Experimental Design

2.1 Experiments & Statistics

• Experiment = systematic, objective observations under controlled conditions.
• When true experiments infeasible ➜ observational or quasi-experimental designs.

2.2 General Types of Research

2.2.1 Basic Research

• Knowledge-driven “What happens if…?” (dichotic listening → channel theory).

2.2.2 Applied Research

• Problem-solving (cell-phone use impairs simulated driving even hands-free; Strayer & Johnston, $2001$).

2.2.3 Sub-types

• Qualitative – rich descriptions (interviews, case studies).
• Quantitative – numerical data analysed statistically (focus of course).

2.2.4 Experimental Variables

• Independent Variable (I.V.) – manipulated / grouping factor.
• Dependent Variable (D.V.) – measured outcome.

2.2.5 Unwanted Variables (Random Error)

• Situational (room temp, noise).
• Individual differences (IQ, motivation).
• Measurement error (experimenter lapse).
– Analogy: static on TV obscures signal.

2.2.6 Confounding Variables

• Vary systematically with I.V. ⇒ alternative explanation (drug A am vs drug B pm; fix by holding constant or counterbalancing).

2.2.7 Control Groups / Conditions

• Baseline for comparison; avoid empty controls; use placebo in clinical trials (magnetic blanket example).

2.3 Three Main Research Approaches

2.3.1 True Randomised Experiments

• Mill’s Joint Method: Agreement (If $X$ then $Y$) + Difference (If not $X$ then not $Y$).
• 3 requirements: ≥ 2 I.V. levels; random assignment; control confounds.
• Designs:
– Independent Groups (between-subjects). Example: robot-tickle vs self-tickle; $N=32$ → $16$ per group.
– Repeated Measures (within-subjects) – same Ps all conditions; counterbalance order; reduces individual-difference error; inappropriate when carry-over (learning) permanent.

2.3.2 Correlational Studies

• Observe natural covariance; scatterplot shows direction/strength; cannot infer causality (smoking ↔ doctor visits $N=200$).

2.3.3 Quasi-Experiments

• Groups formed by pre-existing traits (light vs heavy smokers, age groups); no random assignment ⇒ limited causal claims.

2.4 Relationships & Causality

• 3 conditions (Shaughnessy & Zechmeister): covariation, time-order, elimination of alternatives.
• Ultrasound–birth-weight case: met covariation but failed time-order & alternative-cause tests.

2.5 Measurement

2.5.1 Scales

• Nominal (labels only: $male=0$, $female=1$).
• Ordinal (rank order: 1st, 2nd, 3rd in race).
• Interval (equal units, no true zero: °C, °F).
• Ratio (equal units + absolute zero: Kelvin, reaction time in $ms$).

2.5.2 Quality of Measurement

• Reliability – consistency (rigid ruler vs floppy ruler).
• Validity
– Face validity (looks like it measures construct).
– Predictive validity (OP score predicts uni success).
– Construct validity (IQ correlates with life outcomes in line with theory).

Illustrative Examples & Analogies

• Jogging‐knee anecdote → dangers of over-generalising personal experience.
• Galileo’s feather & boulder in vacuum → skepticism + objectivity.
• Pavlov’s dogs → replication builds confidence/law.
• Kitty Genovese → bystander effect derived from reformulating ethical question.
• Magnetic blankets infomercial → need for proper placebo control.
• Saturday-morning cartoon antenna story → random variability as TV static.

Practical / Ethical / Philosophical Implications

• Tentative nature of science demands willingness to revise policy (e.g., global warming, health guidelines).
• Ethical limits restrict true experimentation (cannot randomise people to smoke). Observational designs fill gap but curb causal claims.
• Independence from authority guards against misinformation from media/political leaders.
• Parsimony cautions against anthropomorphising animal behaviour without necessity.