2.2 Approaches to Research

Research Methods in Psychology

Psychology research methods are the tools scientists use to study behavior and mental processes. Each method has its own strengths and limitations, and understanding them helps you evaluate whether a study's conclusions are actually supported by its evidence. That skill, thinking critically about research, is central to this course.

Research Methods Psychologists Use to Study Behavior and Mental Processes

There are three broad categories of research in psychology. The biggest distinction between them is whether the researcher manipulates anything or simply observes and measures.

Experimental research is the only method that can establish cause and effect. The researcher deliberately changes one variable (the independent variable, or IV) and measures what happens to another variable (the dependent variable, or DV). To make sure the IV is actually responsible for any change in the DV, experimenters use random assignment to place participants into experimental and control groups. This controls for extraneous variables that might otherwise muddy the results.

• Example: A researcher assigns one group to get 4 hours of sleep and another to get 8 hours, then tests both groups' reaction times. Sleep amount is the IV; reaction time is the DV.

Correlational research measures the relationship between two variables without manipulating either one. The result is a correlation coefficient, a number from $-1$ to $+1$ that tells you the direction and strength of the relationship.

• A positive correlation means both variables move in the same direction (e.g., more hours studying tends to go with higher exam grades).
• A negative correlation means they move in opposite directions (e.g., higher stress tends to go with lower job satisfaction).
• A coefficient near $0$ means there's little to no relationship.

The critical limitation: correlational research cannot tell you that one variable causes the other.

Descriptive research aims to systematically observe and describe behavior. It includes surveys, naturalistic observation, case studies, and archival research. Descriptive methods are great for generating hypotheses and providing a foundation of data, but like correlational research, they can't establish causation.

• Example: Surveying 1,000 college students about their attitudes toward climate change, or observing parent-child interactions on a playground.
• Researchers use various sampling methods to select participants who are representative of the larger population they want to generalize to.

Advantages and Limitations of Observational Research

Observational research lets you study behavior in natural settings, free from the artificial constraints of a lab. This produces rich, detailed data about how people (or animals) actually behave in real-world contexts. For instance, a researcher might observe customer behavior in a store to understand purchasing decisions, or study primate social hierarchies in the wild.

There are notable limitations, though:

• Observer bias: Researchers may interpret what they see through the lens of their own expectations, skewing the data.
• Hawthorne effect: People often change their behavior when they know they're being watched. A child who notices a researcher with a clipboard might act differently than they would otherwise, which reduces the authenticity of the observations.

Longitudinal vs. Cross-Sectional Designs

These two designs take fundamentally different approaches to studying how people change or differ.

Longitudinal designs follow the same individuals over an extended period, sometimes years or even decades. This lets researchers track how things like cognitive abilities, personality, or mental health develop within the same people over time.

• Panel study: follows the exact same sample at multiple time points (e.g., testing the same group of students every two years from age 6 to age 18).
• Cohort study: follows a group that shares a common characteristic, such as birth year, through time (e.g., tracking health outcomes for everyone born in 1990).
• The tradeoff: longitudinal studies are expensive, time-consuming, and participants often drop out over the years (this is called attrition).

Cross-sectional designs study different groups of people at a single point in time. For example, a researcher might compare memory performance in 20-year-olds, 40-year-olds, and 60-year-olds all tested on the same day.

• The advantage is speed and lower cost.
• The limitation is that you're comparing different people, not tracking the same individuals. Any differences you find between groups could be due to age, but they could also reflect generational differences (called cohort effects) rather than actual developmental change.

Correlation and Causation in Research

This distinction is one of the most important ideas in the entire course. Just because two things are related does not mean one causes the other.

Correlation describes the statistical relationship between two variables. The correlation coefficient tells you two things: direction (positive or negative) and strength (how close to $+1$ or $-1$). A coefficient of $+0.85$ indicates a strong positive relationship; $-0.30$ indicates a weak negative one.

Causation means one variable directly influences another. To establish causation, three conditions must be met:

1. Temporal precedence: The cause must come before the effect in time.
2. Covariation: Changes in the cause must be associated with changes in the effect.
3. Elimination of alternative explanations: You must rule out other variables that could account for the relationship.

Why does correlation fall short? Because of third variables (also called confounding variables). The classic example: ice cream sales and shark attacks are positively correlated, but ice cream doesn't cause shark attacks. The third variable is summer weather, which increases both ice cream consumption and time spent in the ocean. Only experimental research, where you manipulate the IV and control for confounds, can establish true causation.

Research Quality and Ethics

Two concepts help you evaluate whether a study's findings are trustworthy:

• Validity is the extent to which a study actually measures what it claims to measure. If a "creativity test" really just measures vocabulary, it has low validity.
• Reliability is the consistency of results. If you repeat the study under similar conditions and get similar findings, the measure is reliable. A study can be reliable without being valid (you get the same wrong answer every time), but it can't be valid without being reliable.

Research ethics protect participants' rights, safety, and well-being. Before any study begins, researchers must address several key principles:

• Informed consent: Participants must be told what the study involves and agree to participate voluntarily.
• Confidentiality: Personal data must be kept private.
• Minimizing harm: Researchers must take steps to avoid physical or psychological harm to participants.
• Debriefing: After the study, participants should be told its true purpose, especially if any deception was used.

These ethical standards are enforced by Institutional Review Boards (IRBs), which review and approve research proposals before data collection can begin.