Research Design, Sampling & Metrics

← Back to Domain-Specific Families

Abstractions about defining units, drawing samples, designing studies, and measuring research performance or bias. They include factorial and repeated designs, sampling frames and errors, poststratification, bibliometrics, prevalence estimation, reporting guidelines, review procedures, and verification bias.

19 abstractions in this family — domain-specific abstractions that sit near one another in structural-signature space (k-means over structural-signature embeddings). Each is shown with its short description.

• Absence rate — The proportion of an employed reference population absent from usual full-time work during a specified measurement period under a declared absence definition.
• Altmetrics — Indicators of scholarly attention derived from online platforms, downloads, news, policy documents, social media and other traces that complement rather than replace citation measures.
• ARRIVE guidelines — A reporting standard for in vivo animal research specifying essential and recommended information needed to assess rigor, reproducibility and ethical context.
• Balanced repeated replication — A replicate-weight variance estimator for complex surveys that repeatedly selects one primary sampling unit from each paired stratum according to a balanced sign matrix.
• Designated Member Review — A committee-review pathway in which authorized members examine a protocol on behalf of the full committee under rules preserving member access and escalation.
• Fractional factorial design — An experimental design using a structured subset of full-factor combinations to estimate selected effects with fewer runs at the cost of aliasing.
• H-index — A bibliometric index equal to the largest integer h for which an author or other publication set has at least h works cited at least h times each.
• Historiometry — The quantitative study of historical people, achievements and change using coded documentary records and statistical analysis.
• Multilevel regression with poststratification — An estimation method fitting a hierarchical outcome model to sample data and averaging cell predictions using known target-population cell counts.
• Multiple baseline design — A single-case experimental design that staggers introduction of an intervention across behaviors, participants or settings to replicate causal change without withdrawing treatment.
• Non-sampling error — Survey or estimation error arising from causes other than random selection of the sample.
• Sampling frame — The operational list or spatial representation from which members or units of a target population can actually be selected.
• Seroprevalence — The proportion of a defined population whose specimens contain antibodies meeting a specified assay criterion for a pathogen, antigen, or exposure at a stated time.
• Statistical unit — The elementary entity or event about which variables are measured and observations, sampling, and inferential claims are defined in a statistical study.
• Stroke count method — A Chinese-character input method that encodes a character by the ordered sequence of a small number of stroke-type categories rather than by pronunciation.
• Systematic sampling — A probability-sampling design that chooses a random start in an ordered frame and then selects units at a fixed interval, with variants for unequal probability and spatial grids.
• Unmatched count — A privacy-preserving survey experiment estimating prevalence of a sensitive trait from differences in mean item counts between randomized lists.
• Verification bias — Bias arising when receipt of a definitive reference assessment depends on the result of the preliminary test being evaluated.
• Wiki survey — An open, adaptive survey method in which participants both evaluate statements and contribute new ones while an aggregation system identifies broadly supported or bridging positions.