PSYC 100 Week 2 Review: Research Methods

Vocabulary practice flashcards covering research methods, operational definitions, validity, reliability, distributions, descriptive research, correlations, and experimental designs.

Descriptive Research

A type of research whose main purpose is to describe what people think, feel, or do, establishing what is happening, but not why.

Correlational Research

A type of research whose main purpose is to measure the association between variables, establishing whether variables are related and how strongly.

Experimental Research

A type of research whose main purpose is to manipulate a possible cause and observe an outcome, establishing cause and effect when the study is well designed.

Variable

Anything that can take different values across people, situations, or time.

Qualitative Variable

A category or kind, such as major, occupation, or favorite food.

Quantitative Variable

A numerical amount, such as age, income, test score, or minutes on social media.

Operational Definition

A definition of an abstract concept in terms of exactly how it is measured or manipulated.

Reliability

Consistency of a measurement, such as when a personality test produces similar scores over time.

Test-Retest Reliability

Consistency of a measure across two occasions, such as when the same person receives similar scores this week and next week.

Inter-Rater Reliability

Agreement between observers, such as when two researchers give similar ratings of the same behavior.

Validity

The extent to which a measure captures what it claims to measure, such as a depression scale measuring depression rather than temporary sadness.

Reliability and Validity Relationship

A measure can be reliable but invalid (like a broken clock that always reads 1:00), but generally cannot be valid if it is not reliable because inconsistent measurements cannot accurately represent the intended construct.

Normal Distribution

A roughly symmetrical distribution in which most scores cluster near the center and fewer scores occur at the extremes; the mean, median, and mode are equal, with about $68\%$ falling within $1\text{ SD}$ and about $95\%$ falling within $2\text{ SD}$ of the mean.

Mean

The sum of the scores divided by the number of scores.

Median

The middle score after scores are arranged in order.

Mode

The most frequent score.

Range

The largest value minus the smallest value.

Standard Deviation

A measure of how spread out scores are around the mean, where a small SD means scores cluster closely and a large SD means greater variability.

Effect Size

A measure of how large a difference or relationship is; $d$ standardizes the mean difference as $d = \frac{\text{Mean 1} - \text{Mean 2}}{\text{SD}}$.

Naturalistic Observation

Observing behavior without interfering, including real world observation, online behavior, or experience sampling.

Case Study

Examining one person or a very small number of unusual cases in depth to provide rich detail.

Survey

Asking people to report attitudes, experiences, or behavior.

Population

The entire group the researcher wants to understand.

Sample

The people who actually participate in the study.

Representative Sample

A sample that reflects important characteristics of the population.

Random Sample

A sample selected so every population member has an equal chance of inclusion.

Sampling Bias

A problem that occurs if some people are more likely to enter the sample, causing results to potentially not generalize to the population (e.g., self-selection).

Correlation

A statistical association between two quantitative variables where neither variable is manipulated; coefficient $r$ ranges from $-1.00$ to $+1.00$.

Positive Correlation

A correlation where the variables tend to move in the same direction.

Negative Correlation

A correlation where the variables tend to move in opposite directions.

Scatterplot

A graph where each dot represents one observation; an upward pattern indicates a positive correlation, a downward pattern indicates a negative correlation, and a diffuse pattern indicates a weaker association.

Causation Explanations for Correlation

If $X$ and $Y$ are correlated, three possible explanations remain: 1. $X$ causes $Y$, 2. $Y$ causes $X$, or 3. A third variable causes or influences both $X$ and $Y$.

Appropriate Correlational Language

Non-causal phrasing such as 'is associated with', 'is related to', 'covaries with', or 'is more likely'; causal words like 'causes', 'makes', 'increases', 'decreases', 'prevents', or 'leads to' should be avoided unless design supports causation.

Experiment

A study with two essential features: the researcher manipulates an independent variable and randomly assigns participants to conditions.

Independent Variable

The variable the researcher manipulates.

Dependent Variable

The outcome the researcher observes or measures.

Experimental Group

The group that receives the treatment or focal condition.

Control Group

The group that provides a comparison and does not receive the focal treatment, or receives a placebo or alternative condition.

Random Assignment

Placing participants into conditions by chance to make groups similar before the manipulation and reduce preexisting group differences.

Confound

Any systematic difference between conditions other than the independent variable, making it unclear whether the IV or the extra difference caused the outcome.

Experimenter Expectations

A threat where researchers unintentionally treat groups differently; controlled using standardized procedures and blinding researchers to condition or hypothesis.

Placebo Effect

Improvement in participants because they expect improvement; controlled using a placebo comparison condition.

Participant Demand

When participants guess the hypothesis and alter their behavior; controlled by reducing cues, using blinding, and concealing the exact hypothesis when ethical.

Double-Blind Design

A study design where neither participants nor interacting experimenters know condition assignments, reducing both participant and experimenter expectancy effects.

Quasi-Experiment

Also known as a natural experiment; a study of existing groups or naturally occurring conditions when random assignment is impossible or unethical.

External Validity

The extent to which findings generalize to other people, settings, and real world conditions.