Chapter 1-6 Psychology: Introduction to Research Methods (Vocabulary)
Correlations
A correlation is a statistical measure of the strength of a relationship between two variables.
Two key directions of relationship:
Positive correlation: as one variable increases, the other also increases, or they move in the same direction.
Negative correlation: as one variable increases, the other decreases, or they move in opposite directions.
Zero correlation: no relationship between the variables.
Correlation coefficient (r): a value on a scale from -1 to 1 that quantifies the strength and direction of the relationship.
The symbol used is r with -1 \le r \le 1 .
The closer |r| is to 1, the stronger the relationship; the closer r is to 0, the weaker the relationship.
Data visualization: correlations are often displayed on a scatter plot.
Uses in psychology:
Allows looking for patterns or relationships without manipulating variables (unlike experiments).
Useful when ethical or practical constraints prevent experimentation.
Data collection methods for correlational studies: surveys, naturalistic observation, archival data.
Reasons for using correlational methods:
Some questions cannot be ethically studied with experiments (e.g., effects of divorce on children’s self-esteem).
Some variables cannot be directly manipulated (e.g., height).
Large datasets can make correlation analysis efficient and reveal patterns.
Example patterns:
Positive: more study time → higher grade; fewer absences → fewer discipline referrals.
Negative: more absences → lower grade; less sleep → more car accidents.
Important caution: correlation does not imply causation.
Example where both variables rise together does not mean one causes the other (ice cream sales and violent crime example).
Correlation vs Causation and Related Issues
Illusory correlations: perceiving a relationship where none exists, a bias that can be reinforced by selective observation.
Regression toward the mean: extreme scores tend to move toward the average on subsequent measurements, which can create the illusion of a relationship.
These phenomena can create or exaggerate perceived links between variables.
Experiments: Core Concepts
Experiments are the only research method that can establish a cause-and-effect relationship.
Definition: an experiment is a research method in which an investigator manipulates one or two factors (variables) to observe the effects on behavior or mental processes.
Experimental findings are often generalizable to a larger population, though experiments can be time-consuming and expensive and can be influenced by extraneous factors (confounding variables).
Key terms:
Hypothesis: a prediction about the relationship between variables.
Independent variable (IV): the factor that is manipulated or controlled by the researcher.
Dependent variable (DV): the outcome measured or observed.
Control group: participants who do not receive the manipulated variable.
Experimental group: participants who do receive the manipulated variable.
Example to identify IV and DV:
If testing whether a new medication affects hair growth in men, the DV is the amount of hair growth, and the IV is the medication.
Random assignment vs random sampling:
Random assignment: participants are assigned to control or experimental groups randomly, used in experiments.
Random sampling: participants are randomly selected from a population for a study, used in surveys.
Experimental Design and Control of Biases
Random assignment helps ensure groups are comparable and reduces bias.
Control vs experimental groups enable comparison to determine the effect of the IV.
Blinding:
Single-blind study: participants do not know whether they are in the control or experimental group.
Double-blind study: neither participants nor researchers know who has received the placebo, to reduce expectations influencing results.
Placebo and placebo effect:
Placebo: a fake treatment (e.g., decaf coffee when testing caffeine) used to control for expectations.
Placebo effect: participants’ changes in behavior due to belief they received an active treatment.
Confounding Variables, Bias, and Validity
Extraneous/Confounding variables: factors other than the IV that can influence the DV.
Example: heat as a confounding variable in a study linking ice cream sales to violent crime.
Experimental bias: researcher bias that supports the hypothesis (cherry-picking data or methods).
Validity: the extent to which a test or experiment measures or predicts what it is supposed to.
Hawthorne effect (not in notes, but mentioned): participants alter behavior simply because they know they are being observed.
Barnum effect (not in notes, but mentioned): people believe vague personality descriptions apply uniquely to them (e.g., astrology).
Non-Descriptive (Descriptive/Non-Experimental) Methods
These methods describe behavior and effects but do not manipulate variables.
Examples mentioned: case studies, naturalistic observation.
Purpose: observe and record behavior and its effects without experimental manipulation.
Qualitative vs Quantitative Research
Quantitative research: relies on numerical data (e.g., survey results, test scores).
Qualitative research: relies on in-depth narrative analysis (descriptions, interviews) that can be translated into numerical data.
The course will revisit these distinctions and how both can contribute to understanding.
Ethics in Psychological Research (APA Guidelines)
APA: American Psychological Association; provides guidelines for ethical research.
Voluntary participation: participants freely choose to participate.
Informed consent: participants are given enough information about the study to decide whether to participate.
Passive or partial disclosure is possible regarding purpose, but participants should understand procedures.
If participants are under 18, parental consent is required.
Confidentiality: participant information must be kept private and secure.
Debriefing: after participation, explain the study’s purpose and any deception used.
Deception and its disclosure: if deception occurred, debriefing must reveal it and justify its use.
No long-term harm: researchers must ensure no lasting physical or psychological harm to participants; animals must be well cared for.
IRB approval: research institutions require Institutional Review Board approval before performing experiments; this involves submitting a full proposal with purpose, methods, questions, and participant details.
Scientific integrity: reporting accurate, verified data and avoiding misrepresentation or fake science; citing reputable sources.
Citations: acknowledge others’ ideas when used; helps protect against errors and maintains scholarly integrity.
Practical note: computers may not be used during certain activities; prepare to engage with correlational activity as described in the notes and Canvas.
Relationships Between Methods and Applications
When planning research, choose method according to ethical considerations, feasibility, and what you aim to determine (relationship vs causation).
Correlational studies are valuable for identifying relationships, patterns, and hypotheses that can later be tested with controlled experiments.
Experiments provide stronger evidence for causation but require careful design to minimize biases and confounding factors.
Quick Reference: Key Formulas and Concepts
Correlation coefficient: r = \frac{\mathrm{cov}(X,Y)}{\sigmaX \sigmaY} = \frac{\sum (xi-\bar{x})(yi-\bar{y})}{\sqrt{\sum (xi-\bar{x})^2}\sqrt{\sum (yi-\bar{y})^2}}
r is bounded: -1 \le r \le 1
Interpretations (rules of thumb):
|r| close to 1: strong relationship; |r| close to 0: weak relationship.
Positive r: variables move in the same direction.
Negative r: variables move in opposite directions.
Examples from the lecture:
Positive: more study → higher grade; fewer absences → fewer referrals.
Negative: more absences → lower grade; less sleep → more car accidents.
Ice cream sales and violent crime: both rise with temperature but one does not cause the other.
Experimental structure: IV, DV, control and experimental groups, random assignment.
Blinding types: single-blind vs double-blind.
Placebo effect and placebo control.
Confounding variables and experimental bias must be minimized for valid results.
Hawthorne effect and Barnum effect as additional considerations in interpretation and communication of findings.
Ethics and integrity: voluntary participation, informed consent, confidentiality, debriefing, minimization of harm, IRB approval, animal welfare, and accurate reporting with proper citations.
Notes on Upcoming Activities
The course will return to experiments and their components in a couple of days.
There is an assignment to practice with correlational methods on Canvas.
Students will identify independent and dependent variables in practice questions.
Connections to Foundational Principles and Real-World Relevance
Correlational research supports hypothesis generation and ethical study of relationships when experiments are not feasible.
Understanding causation requires controlled manipulation and random assignment to rule out confounding variables.
Ethical guidelines (APA) protect participants and ensure the reliability and integrity of scientific findings.
Knowledge of biases and effects (Placebo, Hawthorne, etc.) improves experimental design and interpretation of results in real-world psychology research.
Common Misconceptions and Practical Tips
Do not infer causation from correlation alone; always consider possible confounds and alternative explanations.
Be mindful of regression toward the mean when interpreting extreme scores in repeated measurements.
Use random assignment to reduce selection bias and improve internal validity.
Use blinding when possible to reduce expectancy effects on participants and researchers.
Maintain transparency: preregister hypotheses, report methods, and cite sources to uphold scientific integrity.
Quick Summary for Exam
Correlation measures relationships; r ∈ [-1,1]; strength grows with |r|; correlation does not imply causation.
Experiments establish causation via manipulation of IV, measurement of DV, random assignment, control groups, and blinding.
Ethical guidelines (APA) govern voluntary participation, informed consent, confidentiality, debriefing, deception handling, and IRB approval.
Descriptive and non-experimental methods (case studies, naturalistic observation) describe phenomena but do not establish causation.
Distinguish quantitative (numerical data) from qualitative (narrative data translated to numbers).
Be aware of biases and effects (confounding variables, experimental bias, placebo effect, Hawthorne effect, Barnum effect) when designing and interpreting research.