Citing scopusflow | The developer | Licence | Versioning and archival | Contributing and support

Coefficient (forest) plots | Comparing models | Predicted values and interactions | Frequentist and Bayesian estimates together | Posterior distributions on their own | Random effects | Optimiser checks | A one-figure model report | References

Principal component analysis | Clustering | Choosing the number of clusters | Survival curves | References

A query, built safely | Describing the search as a plan | Why page size is a quota decision | Sizing before spending | Fetching, with caching and resume | Watching progress | Combining separate retrievals | Writing the search up | When the ceiling bites | Handling interruptions

Five datasets to explore | A tour by task | The shared spine: tidy_estimates() | A consistent, accessible look | Auditing the figure you are about to submit | Where to next | References

Take stock first | A clean, deduplicated DOI list | Into a reference manager | Handing off to science mapping | Saving the working set

One variable | Two variables, any types | Many variables at once | Distributions across groups | Estimation plots: effect size rather than a p-value | Comparing two groups across categories | Data quality: outliers and missingness | A descriptive summary table | Customising and extending the plots | Tidying the legend | Moving the legend into the plot | Built-in layout controls

Author keywords from a search | References via Abstract Retrieval | A minimal, cross-tool corpus

Field tags decide where to look | One term, many disciplines | Combining terms with boolean operators | Phrases | Brackets in a query written by hand | From a query to a plan | Searching by affiliation | When a term is empty | Checks before any request | Checking recall against known records

What is in a record set | How a literature grows | Where a niche sits | Reading the fuller record | Beyond five thousand records

Launching the app | Demo mode | Describing and sizing a search | Fetching, with a live terminal | The reproducible code panel | Reading the results | Comparing topics | Exporting | Beyond the app

Plotting one or several series | Decomposition | Autocorrelation | The seasonal pattern up close | Forecasting | References

Regression diagnostics | Diagnostics for a generalised linear model | Classification on an imbalanced outcome | Comparing two models in one figure | Choosing an operating point | Calibration | Uncertainty | Power curves | Composing and saving | References

Citing lexsync | Citing the corpus | The developer | Licence | Versioning and archival | Contributing and support

The trial-event model | Content and field references | Triggers on an event | Timing that varies from trial to trial | The paradigm registry | Balanced list assignment | Practice, fillers and feedback | The three presentation targets | Trigger timing and the flip lock | Adding a paradigm | References

Define and run a design | Anatomy of a design file | The keys | Running one | Inspect the match quality | Generate the experiment scripts | Where next

The lexical dimensions | Computing dimensions on a pool | Pool filters | Supplied item pools | Relational designs, where the item is a pair | Conditions, the anchor and matched sets | The four matching methods | The parity caveat on two methods | Tolerance windows | Continuous designs | Reading the validation output | Pseudowords | Items as a random factor | References

What is guaranteed | How byte-identical output is achieved | Trial order: a seeded shuffle without a generator | The Unicode input contract | The documented exceptions | The parity test | The materials datasheet | The methods paragraph and the pre-registration template | The run log | Reproducing a run | References

Launching it | The sidebar | Describing the design | A factorial word contrast | Lexical decision | Priming and self-paced reading | Running it | Reading the results | What you leave with

The challenge | A shared generator | Determinism in R | The same design in Python | The analysis layer | Why this matters

One specification, three interfaces | Describe a design | Simulate | Reproducible by construction | Export a self-contained script | Where to go next

Two-group Gaussian | Crossed mixed-effects designs | A power curve | The sample size the curve implies | Parallel execution | A bridge to the Bayesian workflow | See also

The idea | A worked design | Precision at a fixed N | Sweeping sample size | See also

Gaussian | Shifted lognormal (reaction times) | Lognormal (positive continuous) | Bernoulli (accuracy) | Poisson (counts) | Ordinal (Likert) | Beta (proportions) | Choosing a family | Writing the specification directly

Two ways to run it | What the app does | One specification, three interfaces

Building a theory | Checking rigour | Screening for redundant constructs | Diagramming | Simulation

What the comparison measures | The shape of the result | A first plot | When lines converge at the right end | Drawing the eye to one topic | Adjusting the labels | Reading the result as a table

Describing a search as a plan | Sizing and fetching | The record schema | Most frequent sources and authors | DOIs and change tracking | Comparing topic trends | Author keywords and references | Export and interoperability | Writing the search up | Handling failures

The baseline | A later retrieval | What changed | Merging without duplicates | Keeping a record of each pull