Methodological guide
The Multilinguistics project follows the usual procedures of systematic and scoping reviews. This guide answers, with their justifications, the questions most often asked by editors, authors and reviewers about these procedures. The examples refer to the project's first review, a scoping review of the digital transformations of Spanish, Portuguese and English.
Planning and registration
What is a scoping review, and why does the project begin with one?
A scoping review maps the literature on a broad topic: which phenomena have been studied, with which data and methods, and where the gaps lie. Unlike a systematic review, it does not aim to measure the effect of an intervention or to rank the strength of the evidence. The project begins with one because research on digital language change is scattered across sociolinguistics, corpus and computational linguistics and education; a map of this literature is needed before more delimited studies are designed. The review follows the PRISMA-ScR guidelines (Tricco et al., 2018) and the Joanna Briggs Institute (JBI) guidance for scoping reviews.
Why register the protocol before the searches?
The protocol sets out the questions, the inclusion criteria, the sources, the searches and the way studies are selected and synthesised. Registering it publicly, with a date, on the Open Science Framework (OSF), before the database searches, makes it possible to verify that the criteria were not adjusted after the results were known. It guards against selection bias and is valued by scientific journals.
What if part of the work began before registration?
The registration states exactly what had already been done. In the first review, the screening of the project's bibliographic collection and a pilot of the selection took place before registration, and two rules were added after the pilot; the registered protocol describes these steps, their dates and why the new rules do not change the pilot's decisions. Declaring this history is accepted good practice; omitting it would not be.
What happens if the protocol has to change after registration?
Every later change is recorded as a deviation, with its date, justification and effect on the work already done. The final article reports these deviations.
Sources and searches
Why begin with the project's bibliographic collection?
The project keeps a collection of the works on which its research already rests, and these works are screened first. The collection does not, however, delimit the review: bibliographic databases and grey-literature sources are searched next, and the review's flow diagram shows the records from the collection and from each source separately.
Which sources are searched?
Scopus, Web of Science, SciELO, Redalyc, Dialnet and OpenAlex; for grey literature, the Brazilian Digital Library of Theses and Dissertations (BDTD), the CAPES Catalogue of Theses and Dissertations, Dialnet's doctoral theses and UNESDOC (UNESCO); and the reference lists of the included reviews. The combination aims to cover both international and Ibero-American research.
How are the searches run, and why can they be repeated?
The search strings are set in the protocol, in English, Spanish and Portuguese, for each theme of the review, and applied to titles, abstracts and keywords. Each search is recorded with its date, the database, the exact string used, the filters and the number of results. With this record, any researcher can repeat the search and trace where each record came from.
What is the coverage test of the searches for?
Before the final searches, the strings are tested in Scopus against a set of studies already known to be relevant. If they retrieve fewer than 80% of these studies, the terms are revised and the test is repeated, with each round recorded. The test reduces the risk that a poorly formulated search leaves obvious literature out.
Why does the period start in 2007?
2007 saw the launch of the iPhone and the spread of Twitter and Facebook; from then on, smartphone messaging and social media became the main setting of the phenomena studied. Earlier works are cited as background but not counted. The period ends on the date of the last search, which is reported.
Is grey literature included?
Yes. Theses, dissertations, reports and preprints often hold data that never reach journals, especially in Ibero-American research. These documents are always identified as grey literature and presented separately in the synthesis, so that no conclusion rests on them without saying so.
Selection of studies
Why two independent screeners and a third to settle disagreements?
Each record is assessed by two people who work without knowing each other's decisions; when they disagree, a third reviewer decides. The procedure reduces the effect of a single person's preferences and errors. Agreement between the screeners is measured with Cohen's kappa and reported in the article.
Why are the screeners anonymous?
The screeners act as peer reviewers do: anonymity protects the independence of their judgement. Their contributions are recorded internally, and the article describes the procedure without identifying them.
What happens to records without an abstract or an available full text?
A record without an abstract is assessed from its title and first pages; if doubt remains, it goes on to the next stage. When the full text is not available, it is sought in libraries, open sources, lending services and from the authors; if it still cannot be obtained, the record is excluded for lack of access and listed in the report. Inclusion never depends on the work already being in the project's collection.
Is artificial intelligence used in the selection?
Screeners may use artificial intelligence tools as support, but the decision on each record is theirs. The use of these tools is declared in the article's methods.
Analysis and synthesis
How are data extracted from the studies?
For each included study a fixed set of information is recorded: language, country and region, platform, theme, phenomenon, data, design, method, notion of change, findings and limitations. The definitions are fixed before extraction and tested on ten studies by two people; the rest is extracted by one person and checked by another on a sample.
Why is the Ibero-American evidence presented first?
Studies from any country are included, but the region of each study is recorded, and the synthesis presents first the Ibero-American evidence, the project's immediate focus, and then that of other regions, as comparison.
Why is there no critical appraisal of each study's quality?
As is usual in scoping reviews, the aim is to map the evidence, not to assess the strength of each study. Consequently the review neither ranks results nor supports causal conclusions, which is stated as a limitation.
Authorship and credit
How do co-authors take part, and how are they credited?
Participation is coordinated by the editor through delimited tasks. Contributions are recognised according to the CRediT taxonomy and confirmed by the co-authors before each article is submitted; see Collaborate.
Where can the protocol be consulted?
The protocol of the first review was publicly registered on OSF on 9 November 2026: doi:10.17605/OSF.IO/67EVR. The registration is a frozen version that cannot be edited; any later deviations are reported in the article. The Multilinguistics project gathers the registrations of its studies on OSF.