Replication Studies

Highly specialised requests are first checked for feasibility from this information. Please do not send confidential data through the public form; describe it first and agree secure transfer afterwards.

1. 01Enquiry and requirement assessment
2. 02Technical scoping and proposal
3. 03Execution
4. 04Quality review and delivery, with an agreed round of revisions

Read the full project workflow.

In depth: how the work is done

1. Why replicate and what kind

   The manuscript explains why the original finding merits replication, such as importance, practical use or inconsistency with other work, and states whether the replication is direct, close or conceptual. These categories shape the interpretation. We help authors choose and justify the label.

2. Comparing with the original protocol

   A table that lists the original and replication procedures side by side, with differences marked, is the heart of a replication report. It covers materials, conditions, equipment, sample sizes, analysis methods and software versions. We help authors build it and ensure it is accurate.

3. Documenting deviations

   Deviations from the original are unavoidable, for example different suppliers or equipment. The manuscript explains each and considers whether it could influence the outcome. Honest documentation allows readers to judge whether differences in results are meaningful.

4. Statistical comparison of results

   Replication success should be assessed with defined criteria, such as whether effect estimates are consistent, whether confidence intervals overlap and whether the replication has sufficient power. Reporting both effect sizes and intervals avoids simplistic success or failure labels.

5. Interpreting consistent results

   Agreement strengthens confidence but does not guarantee generality. The discussion states the range of conditions covered by original and replication and notes what remains untested.

6. Interpreting inconsistent results

   Disagreement can arise from differences in conditions, hidden moderators, errors or chance. The manuscript should explore these without accusing the original authors. Contacting the original authors for clarifications and inviting comment is good practice.

7. Respectful tone and collaboration

   Constructive tone and acknowledgement of the original work support the research community. Where possible, contact the original authors, obtain materials and discuss findings before submission.

8. Venues for replication

   Some journals explicitly publish replications, others do not. Registered report formats can reduce publication bias. We help authors identify suitable venues, which they verify.

9. Planning a replication before starting

   A replication is easier to interpret if planned like any study: precise hypotheses, a protocol agreed in advance, a power calculation based on the original effect and criteria for success. Pre-registration protects against flexible analysis. Authors should also collect materials and information from the original paper and, where possible, from the original authors. We help authors turn this plan into a transparent Stage 1 style description.

10. Computational and data-based replications

    Replicating computational work requires code, data and environment. When the original code is unavailable, an independent reimplementation tests whether the description is sufficient. Differences in results may come from undocumented choices, library versions or random seeds. Reporting what was needed to obtain agreement helps the community improve documentation standards.

11. Sharing materials and data from replications

    Replication work is more valuable when others can build on it. Sharing raw data, analysis scripts, protocol documents and notes on practical difficulties allows further replication and meta-analysis. Authors should also record the communication with original authors where appropriate. We help prepare a clear data and materials statement and a repository structure that mirrors the report.

12. Reporting power and sensitivity for replications

    A replication needs enough power to detect an effect of the size originally reported, or even a smaller one, because original effects are often overestimated. Authors should report the power calculation and the smallest effect the design could detect. If the study is underpowered, the report should say that failure to find an effect is not strong evidence against the original claim. We help authors write this assessment accurately.

Common pitfalls this service helps avoid

• Undocumented differences. Readers cannot interpret disagreement.
• Simple pass/fail judgement. Effect sizes and intervals tell more.
• Adversarial tone. Respect advances discussion.
• Underpowered replication. Small samples cannot confirm or refute reliably.

What good reporting looks like

Support for a replication report is documented with an original-versus-replication comparison table, a deviation log, a statistical comparison note and a tone review. Authors perform the replication and report the results as they are. We do not alter results to show agreement or disagreement, and acceptance is not promised.

An illustrative scenario

Hypothetical example written for explanation. It is not a client case, a result or a testimonial.

Imagine a group replicating a reported strength improvement from a new additive. Using the original recipe but a different cement supplier, they find a smaller improvement. The report tabulates differences, shows both effects with intervals, notes that the cement composition differed and recommends testing across suppliers. The original authors were contacted and their comments are acknowledged.

Questions to settle before you begin

• What type of replication is this?
• How do procedures differ?
• What criteria define consistency?
• Is the study adequately powered?
• Have I contacted the original authors?

Key terms

Direct replication

A repeat of the original procedure as closely as possible.

Conceptual replication

A test of the same idea with different methods.

Effect size

The magnitude of an effect.

Moderator

A factor that changes the effect.

Statistical power

The probability of detecting a true effect.

Working together on Specialised Scholarly Article Formats projects

Specialised article engagements begin by confirming which format suits your material, since short communications, methods, data, software, protocol and replication papers each have conventions and length limits. We ask for the results, code or datasets that the article describes, the target journal's guidelines and your deadline. Drafting follows the format's structure, and reporting checklists are applied where they exist. Statements about data and code availability reflect what you can actually share. We do not fabricate results, and we do not promise acceptance.

Few projects need a single service in isolation. These are the services most often combined with this one.

• Negative Results Articles

  Support for articles reporting results that did not confirm expectations, framed so that other researchers can learn from them.

• Registered Reports

  Support for registered-report formats in which hypotheses and methods are peer-reviewed before data collection.

• Methods Papers

  Support for methods papers that present a new or improved method, with clear algorithm or protocol description, validation and comparison to existing methods.