Research Methods and Biases in Psychology

This set of flashcards covers important concepts and terminology in research methods and biases in psychology, providing definitions and explanations for key terms.

Theory

A theory describes general principles about how multiple variables relate to one another and must be supported by data.

Confound

Anything in a study other than the variable being measured that could explain the results.

Availability Heuristic

Relying on what comes to mind easily; if something is easily recalled, it is perceived as more common or true.

Present/Present Bias

Focusing on what is present while forgetting to consider what is unseen or absent.

Confirmation Bias

Asking biased questions or seeking information that confirms pre-existing beliefs.

Bias Blind Spot

The tendency to believe that others are biased while denying one's own biases.

Covariance

For a causal relationship, X and Y must change together; they can increase together (positive) or decrease together (negative).

Temporal Precedence

The cause must occur before the effect.

Internal Validity

The extent to which confounds have been ruled out; high internal validity allows for causal claims.

Association Claim

A claim that two variables are related, typically based on measurement rather than manipulation.

Causal Claim

A claim that one variable causes another, requiring manipulation of one variable and measurement of the other.

Convergent Validity

different measures or sources give similar results, showing they agree with each other.

Criterion Validity

How well a measure predicts a real-world behavior or outcome.

Margin of Error

The potential difference between an estimate and the true value, with a smaller margin indicating more precision.

Random Sampling

The method of selecting participants randomly from a population to create a representative sample.

Random Assignment

The process of randomly assigning participants to groups, helping to ensure that groups start out equal.

Quota Sampling

Choosing a sample that reflects the population traits, e.g., ensuring a percentage of women reflects the population.

Purposive Sampling

Choosing participants based on specific qualities needed for the research, done non-randomly.

Forced-choice questions

Questions where respondents must select from provided options with no open-ended responses allowed.

Question-ordering effects

When the sequence of questions influences how respondents answer.

Acquiescent response set

A tendency for respondents to agree with statements regardless of their content.

Fence-sitting

Choosing a neutral option instead of taking a definitive stance.

Double-barreled questions

Questions that ask about two things at once but allow for only one response.

Likert scale

A rating scale often used to gauge levels of agreement or disagreement.

Reverse-worded items

Questions phrased oppositely to check if respondents are paying attention.

Unobtrusive observation

Observing individuals without their knowledge, allowing for natural behavior.

Obtrusive observation

Observing individuals while they know they are being watched, which may alter their behavior.

Unobtrusive data

Data collected without interacting with subjects, such as archival data or records.

Codebook

A guide that explains how data should be categorized, labeled, or scored.

Masked design

A design where participants or researchers lack knowledge about certain information to avoid bias.

Inter-rater reliability

The degree to which different observers agree when measuring the same phenomenon.

Covert participant observation

A research method where the observer joins a group without revealing their identity.

Quasi-experiment

An experiment lacking random assignment, typically involving pre-existing groups.

Regression to the mean

The phenomenon where extreme scores tend to move closer to the average on subsequent measurements.

Cross-lag (cross-lagged panel design)

A method that evaluates the relationship between two variables over multiple time points.

Interrupted time series design

A research design measuring a variable multiple times before and after a specific event or treatment.

Non-equivalent control group

A comparison group that is not randomly assigned, leading to potential differences from the treatment group.

Longitudinal design

A research design studying the same subjects over time to observe changes.

Small-N design

A type of study utilizing a very small number of participants for an in-depth analysis.

Reversal design (A-B-A or A-B-A-B)

A small-N design where a treatment is introduced and then removed to observe changes in behavior.

Selection effects

When groups differ prior to the study due to a lack of random assignment.

Nonlinear association

A relationship between variables that does not follow a straight line, indicating a curve.

Moderator

A variable that alters the strength or direction of the relationship between two other variables.

Spurious association

A situation where two variables appear related, but the relationship is actually driven by a third variable.