Methodology guide
Which analysis should I use?
Lacuna is three engines over one research memory, plus synthesis bridges that re-read a single frame through another family. Start from your question; the right family follows.
Start from your question
Who shapes the field, and along which axes?
Bibliometrics
What does the literature say — and what is left unsaid?
Computational thematic
How strong is the pooled evidence, and how defensible the decision?
Meta-analysis
I have one frame and want to read it through another lens.
Synthesis bridges
The three engines and the synthesis bridge
- Bibliometrics6 methods
Turns citation, co-authorship, keyword and temporal flows into a research map — where the field is dense, who its centres of influence are, and where the bridges and disconnects sit.
Use it when your question is about the structure of a field rather than a single effect.
- Computational thematic6 methods
Surfaces conceptual proximity, repetition and zones of silence across the corpus text. It is computational — not Braun–Clarke reflexive thematic analysis.
Use it to map themes, trace how they evolve, and mark candidate gaps with textual evidence.
- Meta-analysis12 methods
Takes meta-analysis out of a calculator output and binds it to a decision trail: effect size, heterogeneity, model selection and provenance.
Use it when you have an effects table and need a defensible pooled estimate.
- Synthesis bridges4 methods
Cross-head readings of the SAME frame: each bridge re-reads the attached effects table through another family's run — time, theme, dependency or citation.
Use it after a base run, to read one body of evidence through a second lens.
The full method registry
Below is a public summary of every method, grouped by family. The full rule registry — with each rule's defense text and provenance — lives inside the workspace.
Bibliometrics
Bibliometric Overview
Headline descriptives: production, growth, top sources, authors and keywords.
Engine: Lacuna native (Python)
Sources Analysis
Source productivity and core journals via Bradford’s Law zones.
Engine: Lacuna native (Python)
Authors Analysis
Author productivity (Lotka’s Law), impact and dominance over time.
Engine: Lacuna native (Python)
Documents Analysis
Most-cited documents and the words that travel with them.
Engine: Lacuna native (Python)
Intellectual Structure
Co-citation and bibliographic coupling — the intellectual base and research fronts.
Engine: Lacuna native (Python)
Social Structure
Collaboration networks across authors, institutions and countries.
Engine: Lacuna native (Python)
Computational thematic
Co-word Analysis & Thematic Map
Keyword co-occurrence clustered into a strategic diagram (centrality × density).
Engine: Lacuna native (Python)
Topic Modeling (LDA / NMF)
Probabilistic or matrix-factorisation topics over the full-text corpus.
Engine: Lacuna native (Python)
Embedding-based Topic Clustering
Documents embedded (TF-IDF by default; transformer/SBERT optional) and clustered into semantically coherent topics.
Engine: Lacuna native (Python)
Factorial Analysis
MCA / correspondence analysis projecting terms into a conceptual map.
Engine: Lacuna native (Python)
Thematic Evolution
How themes split, merge and flow across time slices (Sankey).
Engine: Lacuna native (Python)
Trending Topics & Burst Detection
Emerging terms and citation bursts over the publication timeline.
Engine: Lacuna native (Python)
Meta-analysis
Univariate Meta-Analysis
Pool one effect size per study into a single summary estimate with heterogeneity diagnostics.
Engine: metafor 5.0.1 (R)
Multilevel Meta-Analysis
Two- or three-level models for dependent effect sizes nested within studies or labs.
Engine: metafor 5.0.1 (R)
GLMM Meta-Analysis
Generalised linear mixed models for binary/count outcomes without normal approximation.
Engine: metafor 5.0.1 (R)
Network Meta-Analysis
Mixed treatment comparison across ≥3 interventions with a connected evidence network.
Engine: netmeta (R)
Diagnostic Test Accuracy
Bivariate sensitivity/specificity model with a summary ROC curve.
Engine: mada (R)
Dose–Response Meta-Analysis
Model the shape of an exposure–outcome relationship across dose levels.
Engine: dosresmeta 2.2.0 (R)
P-Uniform*
Bias-corrected effect estimate robust to selective publication.
Engine: Lacuna native (Python)
P-Curve Analysis
Test the significant p-values for evidential value vs p-hacking via the p-curve's right-skew (Simonsohn et al. 2014).
Engine: Lacuna native (Python)
Trial Sequential Analysis
Adjust cumulative meta-analysis for repeated significance testing — is the evidence conclusive, or is more information needed (Wetterslev et al. 2008)?
Engine: Lacuna native (Python)
MA Power Analysis
Prospective / retrospective power for a meta-analytic design specification.
Engine: Lacuna native (Python)
Synthesis bridges
Cumulative Reading (Time Bridge)
Re-estimate the pooled effect after each study in year order — when the evidence first became conclusive.
Engine: Lacuna native (Python)
Theme Subgroup Reading (Topic-Model Bridge)
Re-pool as a subgroup analysis whose grouping variable is each study's theme, taken from a topic-model run.
Engine: Lacuna native (Python)
Dependency Reading (Author-Network Bridge)
Re-pool with cluster-robust variance for studies produced by the same research team, from an author-network run.
Engine: Lacuna native (Python)
Evidence-Base Diagnosis (Citation Bridge)
Read the body of evidence behind a meta-analysis descriptively, matched to the project's bibliographic records.
Engine: Lacuna native (Python)
Sign in to browse the live rule registry in the Library.
Open the full registryReady to build the evidence?
Pick a family, attach your corpus, and let Lacuna turn the gap into a defensible report.