Psychology - Research

Common Questions, Beliefs, and Myths

• What people often think vs. what scientific evidence suggests

• Page 3: The belief that we use 10% of our brain (myth or misconception to be tested scientifically)

• Page 4: Intuition on MC exams—initial gut responses vs. evidence-based strategies

• Page 5: The belief that opposites attract in relationships

Science and Skepticism in Psychology

• Evaluating beliefs requires scientific research to avoid relying on intuition, authority, luck, or personal experience alone

• The scientific method uses: careful evidence collection, formulation of hypotheses, and theory-based evaluation

• Key attitude: be skeptical and demand evidence

Key Concepts: Theory and Hypothesis

• Theory: An organized system of statements that explains when, how, and why some phenomena occur

• Hypothesis: A clear predictive statement/general premise/idea; can be based on intuition, observation, or better, theory

• Example: Hypothesis — Watching violent TV shows makes people aggressive

  • Question: How do you test this? (introduces empirical testing)

The Cycle of Science

• Hypothesis/idea → empirical observations → Results → Supported/Not supported

Reasoning in Psychology: Deductive vs Inductive

• Deductive reasoning

  • Start with an idea or hypothesis and test it against the real world

  • Largely used in psychology; conclusions flow from theories

• Inductive reasoning

  • Start with observations and generalize from them

  • Example given: dog barks before dinner → assumes all dogs bark when hungry (a generalization that may be flawed)

Fallacies in Inductive Reasoning

• Just because we observe a pattern doesn’t guarantee the conclusion is valid

• Example: You loved The Matrix → Keanu Reeves is in The Matrix → Keanu Reeves is in Speed → Conclusion: You will like Speed

• What’s wrong: Drawing a conclusion from a chain of observations without a solid logical link or sufficient data

Affirming the Consequent (Fallacy)

• If you live in Brooklyn, then you live in New York

• You live in New York

• Therefore, you must live in Brooklyn

• Note: This is a logical fallacy; correlation or co-occurrence does not imply causation or the exact conditional relationship

Testing Our Theories: Complexity and Multiple Explanations

• Behaviors are complex; more than one theory can explain results

• Sometimes more than one theory can predict the same outcomes

Multiple Theories and Phenomena

• Phenomena can be consistent with Theory 1 and Theory 2 simultaneously

• Illustration: Theory 1 vs Theory 2 with shared predictions (Prediction 1)

• Testing: Confirming Prediction 1 corroborates both theories

Falsification and Revision in Theory Testing

• Falsification: If a prediction is contradicted by evidence, the theory can be falsified

• Example structure: If Prediction 1 is false, Theory 1 is falsified; Theory 2 remains viable if its predictions hold

• Revision: Update wrong theories with new assumptions to better predict new evidence

Newer Tests and Generating New Predictions

• Look for differences in the assumptions between theories to generate new predictions

• Combine theories or introduce new ideas to create Prediction 1, Prediction 2, Prediction 3, etc.

Survival of the Fittest: Competing Theories

• When several theories can explain phenomena, evaluate by generating new, testable predictions

• More than one theory may survive if they withstand empirical testing

Parsimony: Simpler Explanations First

• Parsimony principle: Prefer explanations with fewer, simpler, or more coherent assumptions when facts fit multiple explanations

• Sometimes the simplest explanation is the best

Can You Read My Mind? Cognitive Biases in Prediction

• Activity: A number between 1 and 50 with constraints (both digits odd, distinct)

• Demonstrates how people infer patterns and search for structure with limited information

• Highlights human tendency to impose patterns even with minimal data

Did You Read My Mind? Interpreting Ambiguity in Choices

• You have many options (1–50) but constraints steer choices

• People often infer probable choices (e.g., 37 is most likely, 35 follows) based on heuristics

What’s Going On Here? Mind Reading and Beating the Odds

• Link to Derren Brown video: https://www.youtube.com/watch?v=sEmCQzueyEQ&ab channel=Derren Brown

• Demonstrates psychology of prediction, suggestion, and cognitive biases

Paranormal Activity and Skepticism

• No laboratory evidence, non-replicable results for paranormal claims

• When unusual events occur, consider alternative explanations; controlled studies are needed to advance science

Research Methods: Overview and Options

• Naturalistic observation

• Surveys

• Archival research

• Longitudinal & Cross-Sectional designs

• Correlational studies

• Experiments (lab and field)

• All are tools to study behavior and mental processes

Naturalistic Observation

• Observing behavior in natural settings (e.g., do people wave and thank when cars let them cross the street?)

Pros of Naturalistic Observation

• Captures behavior in real-world settings

• Ecological validity can be higher than lab settings

• Less artificial; cheaper; potentially greater external validity

• Fewer ethical concerns in some contexts

Cons of Naturalistic Observation

• Lack of control over extraneous variables

• Risk of observer bias

• Time-consuming

• Cannot prove causality

Surveys: What They Do

• Questionnaires to collect attitudes, beliefs, preferences, behaviors in a population

• Examples: voting, customer experience, polling

Issues with Surveys

• Representativeness and sample size affect conclusions

• Social desirability bias: respondents give desirable answers

• Inattention: use attention checks to ensure data quality (Aust et al., 2013)

Attention Checks in Surveys

• Where to place attention check? Beginning, end, or throughout

• Referenced work: Oppenheimer et al., 2009

• Example items include intrusive or clearly off-topic questions to detect inattention

Examples of Survey Items and Biases

• Example 1: I oppose raising taxes (1 = strongly disagree, 7 = strongly agree)

• Example 2: I make it a practice to never lie (1 = strongly disagree, 7 = strongly agree)

• Example 3: I would be willing to pay extra taxes for high-quality education (1 = strongly disagree, 7 = strongly agree)

• Example 4: Like all humans, I occasionally tell a lie (1 = strongly disagree, 7 = strongly agree)

• Exercise: Raise your hand if your score on all four questions is greater than 4

Leading Questions and Biases in Surveys

• Biased phrasing can influence responses

• Examples: "How satisfied are you with our product?" vs. "How do you think the new policies are?"

• Encourages respondents to think in a particular direction

Correlational Research: What It Measures

• Correlation = measure of the association between two variables

• Examples: height and weight; personality score and number of friends

• Correlational research examines association without manipulating variables

• The correlation coefficient r ranges between -1 and +1 and indicates strength and direction

• Mathematical expression (concept): $r = rac{cov(X,Y)}{sd(X) \, sd(Y)}$

Scatterplots: Visualizing Correlations

• Illustrates the relationship between two variables (e.g., study performance vs. prior test score; absences vs. scores)

• Helps identify direction (positive/negative) and strength of association

Correlation Does Not Imply Causation

• Example: Nobel Laureates per 10M people vs. chocolate consumption shows a correlation but does not imply causation

• Key idea: A third variable can drive the observed association

• Common third-variable examples: GDP, temperature (as a third variable influencing crime, ice cream sales, etc.)

Try This: Sleep and Mortality Association

• Finding: People sleeping ~7 hours/night have lower mortality risk than those sleeping more or less

• Important caveat: Does sleep cause lower mortality? Consider reverse causation (existing illnesses increase sleep) and third-variable explanations

Alternative Explanations and Causality Challenges

• Example: Sleep and health may be confounded by existing illnesses

• Example: Spanking and child misbehavior — correlation does not imply spanking causes misbehavior; alternative explanations include reverse causation and hereditary factors (Larzelere, Kuhn, & Johnson, 2004)

Illusory Correlation

• Tendency to perceive a relationship between two variables where none exists when data are unsystematic

• Example: Sugar intake causing hyperactivity in children

Explanations for Illusory Correlations

• Early studies suggested sugar reduces behavior, but biases and expectations affected observations (Milich, Wolraich, & Lindgren, 1986; Wolraich et al., 1994)

• Mothers’ expectations led to biased ratings (Hoover & Milich, 1994)

Experiments: The Gold Standard for Causality

• Definition: A study where you manipulate at least one variable and measure at least another

• Examples: Angry people lie more than non-angry people; performance differences in lit vs. dark room

Key Experimental Terms

• Independent Variable (IV): what you manipulate

• Dependent Variable (DV): what you measure; changes due to IV

• Control Group vs. Experimental Group: no treatment vs. received treatment

• Random Assignment: each participant has equal probability of assignment to groups

• Operational Definition: how a variable is measured or manipulated

Biases in Experiments

• Experimenter bias: researcher expectations influence the outcome

• Demand characteristics: clues about the study shape participant behavior

• Good participant/Good subject role: participants respond in a way that pleases the experimenter

• Negative subject role: participants respond in a way that disappoints the expectations

Reducing Biases in Research

• Single-blind design: participants do not know which group they are in, but researchers do

• Double-blind design: neither researchers nor participants know group assignments

• These designs control for expectancy effects and reduce demand characteristics

Techniques to Reduce Bias: Summary

• Implement single-blind or double-blind protocols where possible

• Use standardized procedures and objective measures

• Pre-register hypotheses and analysis plans to reduce flexible post hoc reasoning

• Use random assignment to equalize groups on confounding variables

• Employ control groups to establish a baseline

Practical Takeaways for Research in Psychology

• Be skeptical of intuitive or surface-level patterns; seek empirical validation

• Recognize that multiple theories can explain a phenomenon; test predictions to differentiate them

• Use parsimony to prefer simpler, well-supported explanations when evidence is equivalent

• Distinguish correlation from causation; consider third variables and experimental manipulation to establish causality

• Understand and mitigate biases through rigorous design (blind designs, randomization) and robust measurement

• Choose appropriate methods for the research question (naturalistic observation, surveys, archival data, longitudinal vs cross-sectional, correlational, experiments)

Notable Examples and References from the Slides

• 10% brain myth (Page 3) – a classic misconception often challenged by neuroscience research

• Initial intuition in multiple-choice exams (Page 4) – caveat against sticking to first gut answer without evidence

• Opposites attract (Page 5) – common relationship belief; empirical evidence often shows complexity

• Derren Brown video reference (Page 24) – media example illustrating mind-reading tricks and cognitive biases

• Kripke et al. sleep study (Page 42) – sleep duration and mortality association; emphasizes critical interpretation

• Oppenheimer et al. (2009) on attention checks (Page 32) – importance of attention checks in surveys

• Milich, Wolraich, & Lindgren (1986); Wolraich et al. (1994); Hoover & Milich (1994) – sugar and hyperactivity studies and bias considerations

• Larzelere, Kuhn, & Johnson (2004) – alternative explanations for spanking and child behavior

• Messerli (2012) – Nobel laureates vs. chocolate consumption correlation (Figure 1)

Key Formulas and Concepts (Summary)

• Correlation coefficient: $r = rac{cov(X,Y)}{sd(X)\,sd(Y)} ext{ with } r ext{ in } [-1,1]$

• Experimental design basics:

  • Independent variable: manipulated

  • Dependent variable: measured

  • Random assignment: equal probability to groups

  • Operational definition: how variables are measured

• Causation vs correlation: presence of correlation does not imply causation; third variables can explain associations

• Parsimony: prefer simpler explanations when data fit multiple theories

• Bias reduction: single-blind, double-blind techniques to control for expectations and demand characteristics