Understanding Scientific Method and Research Validity

Confirmation Bias

Tendency to favor confirming evidence over contradictory information.

Scientific Discovery

Process of uncovering causes of natural phenomena.

Laws

Universal statements predicting future events reliably.

Hypothesis

Specific prediction derived from theories.

Validation

Searching for evidence supporting a hypothesis.

Falsification

Searching for evidence contradicting a hypothesis.

Qualification

Determining conditions for hypothesis validity.

Empirical Findings

Results based on measurable, observable data.

Falsifiable

Hypotheses must be testable and refutable.

Temporal Precedence

Cause must precede effect in time.

Covariation

Correlation between cause and effect variables.

Alternative Explanation

Other factors that could influence results.

Scientific Method

Systematic approach to inquiry and evidence gathering.

Variables

Attributes that can vary within a population.

Operationalizing

Defining variables for measurement in research.

Measured Variable

Observed and recorded variable in research.

Manipulated Variable

Variable changed by the researcher in experiments.

Nominal Variable

Categorical variable with no intrinsic order.

Ordinal Variable

Ranked variable indicating relative position.

Interval Variable

Variable with equal distances but no true zero.

Ratio Variable

Variable with a meaningful zero point.

Goals of Psychological Science

Description, prediction, explanation, and practical application.

Theory

Set of principles explaining observed facts.

Frequency Claims

Claims involving one variable's occurrence.

Association Claims

Claims involving relationships between two variables.

Causal Claims

Claims asserting a directional relationship between variables.

Correlations

Statistical relationships indicating variable associations.

Predictions from Associations

Stronger associations lead to more accurate predictions.

Correlation

Statistical relationship between two variables.

Causation

One variable directly affects another variable.

Reverse Causation

Situation where effect is mistaken for cause.

Third Variable Problem

An unaccounted variable influences both studied variables.

Alternative Explanations

Other factors that could account for findings.

Spurious Correlations

False relationships due to a third variable.

Predictor Variable

Variable that forecasts the outcome variable.

Outcome Variable

Variable being predicted or measured.

Independent Variable

Manipulated variable in an experiment.

Dependent Variable

Measured variable affected by the independent variable.

Construct Validity

Accuracy of operational definitions of variables.

Statistical Validity

Data adequately supports the claims made.

External Validity

Study's relevance to other contexts or populations.

Internal Validity

Study's ability to establish causal relationships.

Survey Question Formats

Different styles of questions used in surveys.

Open-ended Questions

Questions allowing free-form responses.

Forced-choice Format

Questions requiring a specific response option.

Likert Scale

Rating scale for attitudes or opinions.

Semantic Differential Format

Rating scale measuring attitudes across dimensions.

Leading Questions

Questions that suggest a particular answer.

Double-barreled Questions

Questions combining two inquiries into one.

Socially Desirable Responding

Answering in a way to appear favorable.

Unbiased Sample

Every member has equal chance of selection.

Biased Samples

Samples that do not represent the population.

Convenience Sampling

Sampling those who are easiest to reach.

Self-selection

Participants volunteer for inclusion in a study.

Probability Sampling

Sampling method ensuring equal selection chance.

Cluster Sampling

Randomly selecting groups from a population.

Multistage Sampling

Two-stage sampling process for population selection.

Stratified random sampling

Selects demographic categories, then samples individuals randomly.

Oversampling

Over-represents one or more groups in sampling.

Systematic sampling

Selects samples using a fixed interval method.

Random sampling

Increases external validity of research findings.

Random assignment

Assigns participants randomly to groups in experiments.

Purposive sampling

Targets specific types of participants for study.

Snowball sampling

Participants recommend others for inclusion in study.

Quota sampling

Sets target numbers for specific population subsets.

Central Limit Theorem

Sample means approximate normal distribution with large samples.

Naturalistic Observation

Observes variables in natural settings without interference.

Observer effects

Participants alter behavior to meet observer expectations.

Masked design

Blends in to prevent observer bias in studies.

Self-report measure

Participants provide their own responses to questions.

Observational measure

Researcher records behaviors without participant input.

Physiological measure

Measures biological responses related to behavior.

Categorical variables

Variables that represent distinct categories or groups.

Quantitative variables

Variables measured numerically for statistical analysis.

Test-retest reliability

Consistent scores across multiple uses of a measure.

Interrater reliability

Consistency of scores regardless of the measurer.

Internal reliability

Consistent responses from participants across questions.

Cronbach's alpha

Measures internal consistency, ranges from 0 to 1.

Content Validity

Captures all aspects of the construct of interest.

Predictive Validity

Score predicts related criterion in the future.

Concurrent Validity

Score relates to criterion at the same time.

Convergent Validity

Score relates to theoretically relevant variables.

Discriminant Validity

Score unrelated to theoretically different measures.

Face Validity

Measure appears to assess what it claims.

Criterion Validity

Evidence supporting the measure's predictive ability.

Reliability

Consistency of a measure over time.

Effect Size

Strength of an association between variables.

Cohen's Guidelines

Small: 0.2, Medium: 0.5, Large: 0.8.

Moderating Variable

Influences strength/direction of independent-dependent relationship.

Multivariate Design

Controls for multiple variables in analysis.

Longitudinal Designs

Measure same variables over time in same individuals.

Cross-Sectional Correlations

Evaluate variables at the same time.

Autocorrelations

Evaluate variable over time separately.

Cross-Lag Correlation

Evaluate variables across different time points.

Quasi-Experimental Designs

Measures and manipulates independent variables.

Natural Experiments

Observes effects of arbitrary treatments in real life.

Instrumental Variable Designs

Uses proxy variable affecting the variable of interest.

Reverse Causality

Determines which variable causes the other.

Nonlinear structures

Data indistinguishable in complex relationships.

Omitted variable problems

Missing variables affecting study outcomes.