Research Methodology — Glossary

Term	Meaning
Abstract	A short summary of a paper: background, methods, results, and conclusion
Anonymity	Participant identity is never linked to the data
Appraisal	Critical evaluation of the quality of a source (used in reviewing)
Assumption	A condition required for a statistical method to hold
Attrition	Loss of participants during a study (can cause attrition bias)
Audit trail	A complete record of how data were handled, for verification
Bayes' theorem	Formula to update a probability given new evidence
Bayesian inference	Updating prior beliefs with data to obtain a posterior
Beneficence	Ethical duty to maximise benefits and minimise harms
Bias	Systematic error that distorts study results
Blinding	Concealing group allocation from participants, staff, or analysts
CRediT taxonomy	Standard contributor roles (conceptualisation, methodology, etc.)
Cohort study	Follows a group over time to see who develops an outcome
Collinearity	Near-linear dependence among predictor variables (multicollinearity)
Comparator	The control or alternative treatment against which a test is measured
Completeness	Proportion of expected data that is actually present
Comprehensiveness	Thoroughness of a search; how few studies are missed
Confidence interval	A range that, with stated confidence, contains the true parameter
Confounder	A variable causally linked to both exposure and outcome
Construct validity	Whether a measure captures the theoretical construct intended
Content validity	Whether an instrument samples the full domain of interest
Convenience sampling	Non-probability sampling of easy-to-reach participants
Correlation	Strength and direction of a linear relationship between two variables
Cross-over trial	Each participant receives multiple treatments in random order
Cross-sectional study	Measures exposure and outcome at a single point in time
Data cleaning	Checking and correcting data errors/inconsistencies before analysis
Data dredging	Testing many hypotheses on the same data (inflates false positives)
Data imputation	Filling in missing values with estimated values
Data saturation	Point in qualitative work where no new themes emerge
Descriptive study	A study describing characteristics of a population
Descriptive statistics	Summaries of data (counts, mean, SD, etc.)
Degrees of freedom	The number of independent pieces of information in a statistic
Dichotomous	A variable taking only two possible values (yes/no)
Discrete variable	A numeric variable taking countable values
Distribution	The set of values a variable can take and how often
Dummy variable	A 0/1 variable encoding a category in a regression
Effect size	The magnitude of a difference or relationship (not just p-values)
Effect modification	A variable whose effect on the outcome differs by level
Eligibility criteria	Rules defining who may participate in a study
Empirical	Based on observation or experiment rather than pure theory
Ethics committee	IRB/REC body reviewing protocol and participant protections
Ethnography	Qualitative study of culture through fieldwork and observation
Evaluation	Assessing the value, merit, or performance of something
Evidence-based	Decisions supported by the best available evidence
Experimental study	A study that assigns a treatment to infer causation
External validity	Whether findings generalise beyond the study conditions
F-distribution	The probability distribution used for variance-ratio statistics
False discovery rate	Expected proportion of false positives among significant results
FAIR data	Data that is Findable, Accessible, Interoperable, Reusable
Family-wise error rate	Probability of at least one Type I error across many tests
Fishbone diagram	An Ishikawa cause-and-effect diagram for root-cause analysis
Five whys	An iterative root-cause method of asking "why?" repeatedly
G*Power	Software for statistical power and sample-size analysis
Gap (literature)	What is not yet known, which the current study addresses
Grounded theory	A qualitative method building theory directly from data
Grey literature	Unpublished/non-commercial works (reports, theses, preprints)
H0	The null hypothesis (no effect or no difference)
H1	The alternative (research) hypothesis
Hazard ratio	A ratio of hazard rates (used in survival analysis)
Heterogeneity	Variation in results across studies or groups
Histogram	A bar chart of frequency for continuous data (bars touch)
H-index	An author-level metric: h papers each cited at least h times
Human subjects	People participating in research (or their identifiable data)
Hypothesis	A testable statement about the relationship between variables
Inclusion criteria	Characteristics required for a participant to join a study
Informed consent	Voluntary agreement after understanding purpose, risks, benefit
Instrument	A tool for measuring or observing a variable
Instrumentation	The process or tools used to collect/measure data
Integrity	Honesty and truthfulness in researching and reporting
Inter-rater reliability	Agreement between two or more raters (κ, ICC, %)
Internal consistency	How well items of a scale hang together (Cronbach's α)
Internal validity	Confidence that the observed effect is a true causal effect
Interpretivism	Epistemology that meaning is interpreted, not measured objectively
Intention-to-treat	Analysis including all randomised participants in original groups
Interquartile range	The middle 50% of data (Q3 − Q1); robust spread measure
Intervention	The treatment or exposure under study
Interview	A data-collection method using open-ended questions
Kappa (κ)	Statistic for agreement beyond chance between raters
Keyword	A term indexing/searching literature
Literature review	A systematic survey of published work on a topic
Mean	The arithmetic average; sensitive to outliers
Measurement	Assigning numbers or labels to attributes by defined rules
Median	The middle value of ordered data; robust to outliers
Method	A planned way of addressing a research question
Mode	The most frequently occurring value
Model	An idealised description of how data were generated
Multiple imputation	Creating several datasets by imputing missing values
Nominal scale	Categorical labels with no order (e.g., blood type)
Normal distribution	A symmetric bell-shaped distribution (mean, SD)
Null hypothesis	The "no effect" statement tested against the data
Observational study	A study observing exposure/outcome without assigning treatment
Odds ratio	Ratio of the odds of an event in two groups
Operational definition	A precise, concrete procedure defining how a variable is measured
Outlier	A value far from the rest; may be error or a true extreme
p-value	Probability of the data (or more extreme) if the null hypothesis is true
Parameter	A numerical characteristic of a population
Pearson correlation	A coefficient (r) ranging −1 to +1 for linear association
Pilot study	A small run of the method to refine procedures before the main study
Plagiarism	Using others' words or ideas without credit
Population	The entire group about which you want to generalise
Power	The probability of detecting a true effect (1 − β)
Pre-registration	Publicly registering the analysis plan before collecting data
Precision	Closeness of repeated estimates (low variance means high precision)
Primary data	Data collected firsthand for the current study
Primary source	The original report of an event or study
Probability	A number between 0 and 1 quantifying uncertainty
Propensity score	Estimated probability of treatment, used to balance confounders
Protocol	A written plan for a study
Publication bias	Tendency to publish significant results more readily
Quantitative	Data expressed numerically, amenable to statistical analysis
Quartile	A value dividing ranked data into four equal groups
Randomisation	Assigning units to groups by a chance mechanism
Range	The difference between the largest and smallest values
Relative risk	(Risk ratio) ratio of risk in exposed vs unexposed groups
Reliability	Consistency of a measure across time, items, or raters
Replication	Repeating a study to check whether findings hold
Reproducibility	Whether the same analysis on the same data gives the same result
Residual	The difference between an observed and a predicted value
Risk of bias	The risk that errors influenced a study's results
Sampling	Selecting a subset of the population
Sampling bias	Systematic error from how participants are selected
Sampling distribution	The probability distribution of a statistic over all samples
Sampling error	Natural variability between a sample statistic and the population
Sampling frame	A list used for sampling (should match the population)
Scale	Any measuring instrument; also a scale of measurement (NOIR)
Scatter plot	A plot of paired observations to show a relationship
Secondary source	A source interpreting or summarising primary sources
Significance level	The threshold α for rejecting H₀ (usually 0.05)
Skew	Asymmetry of a distribution (positive or negative)
Snowball sampling	A chain-referral method for hidden populations
Standard deviation	The average distance of values from the mean
Standard error	The standard deviation of a sampling distribution
Statistical significance	A result unlikely to be due to chance (small p-value)
Statistic	A number describing a sample characteristic
Stratum	A subgroup used in stratified sampling (plural strata)
Systematic review	A rigorous, protocol-driven literature review
T-distribution	The bell-shaped distribution used for small-sample inference
t-test	A test comparing means (one-sample, paired, or two-sample)
Test-retest reliability	Whether a measure gives the same result when repeated
Triangulation	Using multiple methods/data sources to study one question
Type I error	Rejecting a true null hypothesis (a false positive)
Type II error	Failing to reject a false null hypothesis (a false negative)
Uncertainty	The inherent unknownness in observations and estimates
Variable	A characteristic that can take different values
Variance	The average squared deviation from the mean
Visual analogue scale	A line respondents mark to rate an attribute (e.g., pain)