Methodological second opinion

Because when you have done the analysis yourself, its weak points are hard to see: you know the data too well and view the objections that may come from outside with too little distance. It applies before an analysis is presented to a journal, a committee or a funder, while there is still room to strengthen it. Anyone who works with a database for months ends up taking for granted decisions that an outside reader will not. An independent review brings that critical distance and reads the analysis the way an assessor would, asking about what the team has come to treat as self-evident. The best time to identify a methodological weakness is before the analysis goes out, not in the reviewer's report, when addressing it tends to be costlier and slower. What the service offers is not distrust of your analysis, but the outside scrutiny it is going to receive anyway, at a point when you can still act on it.

We review the key components of a statistical analysis that has already been carried out: the fit between method and question, the models' assumptions, the correspondence between results and conclusions, and the tables and figures. It applies to a finished or nearly finished analysis that is about to be assessed. Specifically, we assess whether the chosen model is consistent with the design, the type of data and the question; whether the model's assumptions are checked and how they have been handled when they are not met; whether the conclusions stay within what the data support; and whether the tables and figures show what the text claims, with uncertainty and effect sizes properly represented. When the work is a meta-analysis, the review can also cover the suitability of the synthesis model, the handling of heterogeneity and the reading of the pooled results, within the agreed scope. What you get is a report with recommended adjustments and points to strengthen, not a rewrite of the analysis.

We review, identify and recommend; the corrections and the decisions stay with the research team. This division is worth spelling out because it defines the nature of the service: it is a second, expert reading, not a re-run of the work. The service examines whether the method is appropriate, whether the assumptions are valid and whether results and conclusions correspond, and returns a report with the recommended adjustments, the points to strengthen and, depending on scope, proposals for complementary analyses. The changes are applied, and the work signed, by the team that did it: the analysis remains theirs, as do the authorship and the final decision on what to incorporate. This line is deliberate and it protects the client. A second opinion provides outside judgement on which to base a decision, not an alternative version of the analysis that has to be adopted. What you get is an actionable diagnosis, not an analysis redone and signed by someone else.

Yes: support with the response to reviewers is one of the most specific ways into the service. It applies when a manuscript comes back with methodological comments that need a technical, well-grounded response. The review separates the objections that require a substantive methodological answer from those that a clarification will settle, helps frame clear technical responses to the comments received and identifies what additional analysis could strengthen the work against the objection raised. It also works before the manuscript goes out: anticipating the methodological objections common in peer review makes it possible to strengthen those points before a reviewer asks for it. The service provides the technical judgement and the argument; the research team decides what to answer and signs the response. What you get is support in building a methodologically sound response, not a response written and sent on the team's behalf.

No: a methodological second opinion is a scoped engagement by design, meant to settle a specific methodological question without the need for a full project. It applies when what you need is a review of a well-defined piece, such as a model, a meta-analysis, a set of figures or a methods section, before you send, defend or redo the work. That bounded format is part of its value: a focused, high-value intervention on the point that raises the question, not a broad commitment. The scope is agreed at the outset according to what is to be reviewed, so that the engagement fits the real question rather than a fixed package. If the review shows that broader work would be worthwhile, we raise it, but it is not the starting point. What you get is an expert answer to a bounded question, in a format proportionate to that question.

No. A methodological second opinion is a peer-to-peer scientific review of the method, models, assumptions, conclusions, figures and response to reviewers, aimed at publication, committees and funders in academic and research settings. It is not a formal validation of the analysis and its reproducibility for a regulatory dossier: it does not include double programming, code auditing, quality control to regulatory standards or the documentation a regulatory dossier requires. This limit is worth spelling out because "reviewing an analysis" can be understood either way, and they are different jobs with different frameworks. What is included: the critical methodological reading that helps an analysis reach a journal or a committee in stronger shape. What is not included: formal validation for regulatory purposes, which is a separate competence with its own procedure. Anyone who needs that formal validation will find a clear line here, not a vague promise.

Figures and tables are reviewed as part of the second opinion, from a methodological angle: whether what the text claims matches what they actually show, whether uncertainty and effect sizes are properly represented and whether they are clear to their audience. It applies when the question is whether a manuscript's figures support what the text says and will stand up to a reviewer's reading. This should be kept apart from figure design. Here the existing figure is critically assessed and adjustments are recommended; designing new visualisations, building the visual narrative or preparing publication figures from scratch belongs to the scientific visualisation service. The line is in the verb: this review assesses and recommends on what already exists; the visualisation service designs or reworks the visual piece and its narrative. If what you need is a critical look at the figures before they go out, it belongs here; if you need them designed or reworked, it's visualisation.

You receive a methodological review report with the recommended adjustments, the points to strengthen and the aspects a reviewer or a committee might question, organised so that the team can act on them. It applies to anyone who wants to reach assessment with a more solid analysis, or to respond to a review already received. Depending on the agreed scope, the report covers the fit between method and question, the validity of the assumptions, the correspondence between results and conclusions, the reading of tables and figures and, where relevant, proposals for complementary analyses and support with the response to reviewers. The service reviews, identifies and recommends; the research team decides what to incorporate, applies the changes and signs the work. What you get is actionable outside judgement on an analysis that remains the client's, not a redone analysis or a validation with a formal seal.