PY1103 Revision: Foundations of Critical Thinking and Psychology Research Methods

THE OBSERVE FRAMEWORK: FOUNDATIONS OF CRITICAL THINKING

The OBSERVE framework is a foundational structure for applying critical thinking through a sequence of seven specific steps designed to move from initial awareness to systematic evaluation.

Observe the Phenomenon: This stage involves carefully noticing and describing events or situations to gather relevant details and form an initial understanding. An example is identifying patterns such as how sleep deprivation might affect test performance.
Examine Beliefs and Emotions: This step requires recognizing personal biases and emotions that shape interpretations. Individuals must acknowledge their pre-existing views and how those influences may color their observations.
Cultivate Self-Awareness of Cognitive Biases: It is essential to identify and acknowledge internal patterns of thought, such as confirmation bias or anchoring bias, that hinder rational judgment or distort analysis.
Establish Primary Hypothesis: Practitioners formulate a specific, tentative explanation or prediction that is both measurable and testable to guide the investigation.
Recognize Alternative Hypotheses: To ensure thorough analysis and avoid premature conclusions, multiple potential explanations for the phenomenon must be considered.
Verify the Evidence: This involves gathering and assessing data from credible sources to scrutinize the reliability and relevance of evidence for each proposed hypothesis.
Evaluate the Criteria of Adequacy: Hypotheses are compared using established scientific principles to determine the most supported explanation.

Criteria of Adequacy

When evaluating hypotheses based on evidence, five criteria are used:

Testability: The hypothesis must be measurable and falsifiable (able to be proven false).
Fruitfulness: The hypothesis leads to new insights and valid predictions.
Scope: The applicability of the hypothesis to a wide range of related phenomena.
Simplicity: The explanation avoids unnecessary complexity and remains clear.
Conservatism: The hypothesis aligns well with existing, established knowledge.

DEDUCTIVE AND INDUCTIVE ARGUMENTATION

Arguments are classified into two primary categories based on the relationship between their premises and conclusions.

Deductive Arguments

In a deductive argument, the premises are intended to provide conclusive support for the conclusion. If the premises are true, the conclusion must be true.

Structure: Based on logical structure, these arguments are either valid or invalid.
Soundness: An argument is sound if it has a valid structure and all its premises are true.
Binary nature: Validity is absolute; an argument cannot be "partially" valid.
Example:
- Premise 1: All mice have tails.
- Premise 2: Sniffy is a mouse.
- Conclusion: Therefore, Sniffy has a tail.

Inductive Arguments

Inductive arguments provide probable support for their conclusions. Even with true premises, the conclusion is only likely, not certain.

Structure: These are characterized as strong or weak.
Cogency: An inductive argument is cogent if it is strong and the premises are true.
Degree of Strength: Strength is a matter of degree rather than a binary state. For example, an argument based on a statistic of $96\%$ is stronger than one based on $75\%$ .
Example:
- Premise 1: Most mice have tails.
- Premise 2: Sniffy is a mouse.
- Conclusion: Therefore, Sniffy probably has a tail.

Implicit Premises

Implicit premises are unstated assumptions that support an argument. They are often omitted because they seem obvious; however, they carry the risk of being false or controversial. To evaluate an argument properly, these must be made explicit. For example, if the conclusion is that Sniffy's teeth will never stop growing because he is a mouse and all rodents have teeth that grow forever, the implicit premise is "All mice are rodents."

CATEGORIES OF INDUCTIVE INFERENCE

Enumerative Inference: Drawing conclusions based on observed patterns. The strength of this inference depends on sample size, the proportion of the sample relative to the population, and the variety/randomness of the collection method.
Argument from Sample: Concluding something about a whole population based on a smaller subset. The reliability depends on how representative the sample is.
Argument by Analogy: Comparing two similar cases. If Object A has traits W, X, Y, and Z, and Object B has W, X, and Y, one infers Object B likely has Z. Strength depends on the relevance and number of similarities and the lack of significant differences.
Causal Reasoning: Identifying cause-and-effect relationships. It requires caution because correlation does not equal causation, and alternative explanations like coincidence must be considered.
Inference to the Best Explanation: Selecting the most plausible explanation from several options. Criteria include consistency with accepted theories and the ability to explain the specific phenomenon.

TYPES OF POSSIBILITIES AND LAWS OF LOGIC

Logical Possibility

Something is logically possible if it does not violate the Laws of Logic (or Laws of Thought):

Law of Identity: A is A. If something exists, it has a single, specific nature.
Law of Noncontradiction: A statement cannot be both A and not A at the same time and in the same sense.
Law of Excluded Middle: A statement is either true or false; there is no middle ground.

Physical Possibility

Something is physically possible if it aligns with the Laws of Science (e.g., Newton's Laws, Thermodynamics, Relativity). A cow jumping over the moon is logically possible (no contradiction) but physically impossible due to gravity and physiology.

Technological Possibility

Something is technologically impossible if it is currently beyond human capability, even if it does not violate scientific laws. For instance, manned intergalactic travel is technologically impossible due to food and energy storage limitations, though it is physically possible.

INFORMAL FALLACIES IN REASONING

Fallacies are errors in reasoning where premises fail to justify the conclusion adequately.

Fallacies of Insufficiency

Hasty Generalization: Drawing broad conclusions from small or unrepresentative samples (e.g., "One student failed, so the test is impossible").
Post Hoc Ergo Propter Hoc (False Cause): Assuming that because event B follows event A, A caused B (e.g., "I wore lucky socks and we won").
Slippery Slope: Predicting a chain of events leading to a disaster without evidence for each step (e.g., "If we allow one late assignment, no one will ever meet a deadline").
Weak or Faulty Analogy: Comparing items that lack meaningful similarities (e.g., "Employees are like nails; they must be hit to work").
Appeal to Authority (Unqualified): Citing an expert in an unrelated field or a celebrity to support a claim.
Appeal to Ignorance: Claiming a statement is true just because it hasn't been proven false (e.g., "You can't prove aliens don't exist, so they do").
Begging the Question: Circular reasoning where the truth of the conclusion is assumed in the premises (e.g., "The diet works because it's effective").
False Dichotomy: Presenting only two choices when more exist (e.g., "You're either with us or against us").

Fallacies of Irrelevance

Argumentum Ad Hominem: Attacking the person's character or motives instead of their argument.
Red Herring: Introducing irrelevant distractions to shift focus from the main issue.
Tu Quoque ("You Too"): Dismissing an argument by accusing the speaker of hypocrisy.
Straw Man: Misrepresenting an opponent's argument to make it easier to attack.
Appeals to Emotion: Using fear, pity, or joy to persuade instead of logical evidence.

Fallacies of Ambiguity

Equivocation: Shifting the meaning of a word during an argument. For example, using "man" to mean both "species" and "gender" to argue women are not rational.
Amphiboly: Ambiguity caused by poor sentence structure (e.g., "I shot an elephant in my pajamas").
Fallacy of Composition: Assuming what is true of the parts must be true of the whole (e.g., "Every player is the best, so the team is the best").
Fallacy of Division: Assuming what is true of the whole must be true of the parts.
Moving the Goalposts: Changing criteria for success after they have been met.

COGNITIVE BIASES

Cognitive biases are internal, systematic deviations from rational judgment arising from mental shortcuts known as heuristics.

Dunning-Kruger Effect: Individuals with low knowledge in a task overestimate their ability, while experts may underestimate their relative competence.
Confirmation Bias: The tendency to seek, interpret, and remember information that supports personal beliefs while filtering out opposing evidence.
Self-Serving Bias: Attributing success to personal traits and failure to external factors.
Optimism/Pessimism Bias: Overestimating positive or negative outcomes respectively.
Sunk Cost Fallacy: Continuing an endeavor because of past investments of time or money, even when future costs outweigh benefits.
Negativity Bias: Negative events have a more significant psychological impact than positive ones of equal intensity.
Backfire Effect: Strengthening one's beliefs when confronted with contradictory evidence, often because those beliefs are tied to identity.
Fundamental Attribution Error: Attributing others' actions to their character while excusing one's own behavior as situational.
In-group Bias: Favoring people within one's own group over outsiders.
Forer Effect (Barnum Effect): Believing that vague, general personality descriptions apply specifically to oneself.

PERCEPTION AND ATTENTION

Perception is not a passive recording of reality but an active construction.

Perceptual Processes

Perceptual Constancies: The tendency to perceive objects as unchanging despite sensory input variations (e.g., color and size constancy).
Bottom-Up Processing: Data-driven processing starting with raw sensory input sent to the brain.
Top-Down Processing: Concept-driven processing where the brain uses prior knowledge and expectations to interpret data.

THE GESTALT PRINCIPLES

Figure-ground: Structuring input to see an image against a background.
Similarity: Grouping stimuli that look alike.
Proximity: Grouping stimuli that are physically close to each other.
Continuity: Perceiving stimuli in smooth paths rather than discontinuous ones.
Closure: Filling in gaps to create a whole object.

Attention Mechanisms

Selective Attention: Acting as a filter to focus on specific inputs while ignoring others. This creates the "User Illusion" that we perceive everything.
Cocktail Party Effect: The ability to tune into one voice in a noisy environment but immediately shift attention if a significant word (like one's name) is heard.
Divided Attention: Splitting mental resources between tasks (e.g., driving while talking).
Inattentional Blindness: Failing to see visible objects because attention is focused elsewhere.

MEMORY, BELIEFS, AND KNOWLEDGE

Memory and Priming

Subtle cues can shape behavior and recall. In research by Bargh, Chen, and Burrows (1996), participants primed with elderly stereotypes walked more slowly. In eyewitness studies by Loftus and Palmer (1974), the language used to describe a car accident (e.g., "smashed" vs. "hit") significantly altered participants' speed estimates.

Belief Systems

Beliefs are interconnected systems rather than isolated facts. Quine describes a "protective belt" of auxiliary beliefs that shield core beliefs from contradictory evidence. People often adjust peripheral beliefs to protect their core identity-related beliefs.

Defining Knowledge

Philosophically, knowledge is often defined as Justified True Belief (JTB). This requires:

Belief: Personal commitment to the statement.
Truth: The statement must correspond to reality.
Justification: Evidence and reasoning supporting the belief.

Types of Knowledge

Procedural: Knowing how to perform a task.
Acquaintance: Familiarity with people or places.
Propositional: Knowledge that something is true (e.g., facts and science).

THE SCIENTIFIC METHOD

Scientific inquiry integrates intuition, authority, rationalism, and empiricism into a systematic evidence-based approach.

Methods of Acquiring Knowledge

Intuition: Sudden gut feelings based on past experience.
Authority: Trusting experts (necessary because no one can verify everything).
Rationalism: Using logical reasoning and deduction (e.g., Descartes).
Empiricism: Knowledge through observation and sensory experience.

The Research Cycle

Observation or Question.
Hypothesis Formation.
Experimentation.
Data Collection and Analysis.
Conclusion and Revision.
Replication and Peer Review.

Goals of Science

Describe: Systematically record events to identify patterns.
Predict: Use patterns to anticipate future events.
Explain: Uncover causal mechanisms to understand why phenomena occur.

PSYCHOLOGICAL MEASUREMENT AND CONSTRUCTS

Psychological measurement, or psychometrics, involves quantifying abstract mental traits.

Constructs and Definitions

Psychological Constructs: Variables that cannot be directly observed, such as intelligence or self-esteem.
Conceptual Definition: Outlines what a construct includes (e.g., defining neuroticism as the tendency to experience negative emotions).
Operational Definition: Specifies the method of measurement in a study.

Types of Measurement

Self-Report: Participants describe themselves (e.g., Rosenberg Self-Esteem Scale). Limitation: Social desirability bias.
Behavioral: Observing actions (e.g., Bandura’s Bobo doll study). Limitation: Participant reactivity.
Physiological: Measuring biological markers (e.g., heart rate, cortisol levels, fMRI scans).

Levels of Measurement (Stevens, 1946)

Nominal: Categorical labels without ranking (e.g., nationality).
Ordinal: Ranked order but unknown intervals (e.g., race placement).
Interval: Equal intervals but no true zero (e.g., temperature in Celsius).
Ratio: Equal intervals with a true zero (e.g., time, height).

Reliability and Validity

Reliability (Consistency)

Test-retest: Stability of scores over time. Correlation of $+0.80$ or higher is generally good.
Internal Consistency: Consistency across test items. Measured by Cronbach’s alpha ( $\alpha$ ).
Inter-rater: Agreement between observers. Measured by Cohen’s kappa ( $\kappa$ ).

Validity (Accuracy)

Face Validity: Does it look right surface-level?
Content Validity: Does it cover all aspects of the construct?
Criterion: Does it correlate with real-world outcomes (Concurrent or Predictive)?
Convergent: Does it align with other measures of the same trait?
Discriminant: Does it avoid correlating with unrelated traits?

EXPERIMENTAL RESEARCH DESIGNS

Variables and Sampling

Independent Variable (IV): The factor manipulated to observe its effect.
Dependent Variable (DV): The outcome being measured.
Random Sampling: Every population member has an equal chance of being in the study.
Random Assignment: Every participant has an equal chance of being in any condition.

Between-Subjects Design

Each participant is exposed to only one level of the IV.

Strength: No carryover/order effects.
Weakness: Requires more participants; vulnerability to individual differences.
Matched-groups: Participants are matched on traits (like health) before assignment to reduce variability.

Within-Subjects Design

Each participant experiences all conditions.

Strength: Controls for individual differences; higher statistical power with fewer people.
Weakness: Carryover effects (practice, fatigue, contrast).
Counterbalancing: Varying the order of conditions to control for these effects. Latin Square designs ensure each condition appears in each position once.

Quasi-Experimental Research

Resembles experiments but lacks random assignment. It can eliminate the directionality problem by manipulating the IV before measuring the DV but remains vulnerable to confounding variables.

One-group Pretest-Posttest: Testing before and after treatment.
Interrupted Time Series: Repeated measurements before and after an intervention to detect trends over time.

Factorial Designs

Used when studying multiple independent variables concurrently.

2 x 2 Design: Two IVs, each with two levels (4 conditions).
Main Effects: The individual effect of each independent variable.
Interaction Effect: When the effect of one IV depends on the level of another IV (e.g., disgust affecting moral judgment only in people with high body consciousness).

DESCRIPTIVE AND INFERENTIAL STATISTICS

Descriptive Statistics

Mean ( $M$ ): Arithmetic average ( $\text{Mean} = \frac{\sum X}{N}$ ). Sensitive to outliers.
Median: The middle score. Preferred for skewed data.
Mode: The most frequent value.
Standard Deviation ( $SD$ ): Average distance of scores from the mean.
Range: The difference between the highest and lowest scores.
Z-score: Measures how many standard deviations a score is from the mean ( $z = \frac{X - M}{SD}$ ). Outliers are typically identified as z < -3 or z > +3.
Correlation ( $r$ ): Measures the relationship strength and direction from $-1.00$ to $+1.00$ .

Inferential Statistics

Statistical Significance: Determined by the p-value. Standard alpha level is p < 0.05, meaning there is less than a $5\%$ chance the result is due to random error.
Cohen’s d: Measures effect size in standard deviation units ( $0.20$ =small, $0.50$ =medium, $0.80$ =large).

RESEARCH ETHICS

Core Ethical Principles

Weighing Risks against Benefits: Benefits to participants, science, and society must outweigh the risks of harm or distress.
Acting Responsibly and with Integrity: Honesty in reporting and professionalism in dealing with participants.
Seeking Justice: Fair distribution of the burdens and benefits of research.
Respecting Rights and Dignity: Protecting privacy and autonomy.

Major Ethical Considerations

Deception: Sometimes necessary for scientific validly but creates an ethical dilemma regarding transparency. Milgram’s (1963) obedience study is a primary example of this conflict.
Minimal Risk: Risks no greater than those in daily life or routine exams.
Animal Research (Standard 8.09): Requires humane treatment and care. Justified when benefits to human/animal health outweigh risks and no alternatives exist.
Scholarly Integrity: Zero tolerance for data fabrication, falsification, or plagiarism. Self-plagiarism (recycling one's own work) is also unethical.
Authorship: Must accurately reflect significant contributions of researchers.

Reporting and Engagement

Scientific findings are communicated through peer-reviewed journals, book chapters, and academic conferences. Peer review acts as the "gold standard" for quality control. Public engagement bridges the gap between academia and society to influence policy and real-world outcomes.