Questions researchers ask

No. Equivalence is a workflow and documentation platform, not a translation service. Your project team appoints qualified human translators and reviewers. The platform records their independent work, manages handovers and blinding, and keeps the decisions and final wording together.

Yes. Measures in the same project share a target context, language route, team, and workflow, but each keeps its own name, reference or DOI, item target, source items, instructions, and response options. Those measure-specific texts move through translation and reconciliation with the items.

Members are notified when a stage is handed over to them. Each accepted assignment starts with a seven-day deadline, which the assignee can change. If the task is still open on its due date, Equivalence sends a reminder at 9 a.m. Hong Kong time.

Independent translations can reveal different interpretations of the same item. Equivalence keeps versions hidden until all required translations are submitted, preventing early wording convergence (Beaton et al., 2000).

A back-translator who knows the source may reproduce it from memory rather than translate the reconciled version independently. Forward and back-translator roles therefore remain separate. Back-translation is optional in some frameworks and is not part of TRAPD (Beaton et al., 2000; Epstein et al., 2015; Harkness, 2003).

The rationale explains why the team selected one wording over the alternatives. This creates a record of how disagreements were resolved. Equivalence requires a rationale for each reconciled item.

Each stage uses the version completed in the previous stage. For example, expert review should assess the reconciled items, not an earlier draft. The project page lists the requirements for the current stage and shows any work still outstanding.

Experts rate each item’s relevance on a 4-point scale. I-CVI is the proportion of experts rating an item 3 or 4, with 0.78 the conventional floor for panels of six to ten (Lynn, 1986). S-CVI/Ave is the mean I-CVI across items, with 0.90 the accepted scale-level threshold (Polit & Beck, 2006). Modified kappa corrects I-CVI for chance agreement. Equivalence computes all three live as ratings arrive.

A project may contain several measures, but it uses one target context and language route. Gates, blinding counts, and content validity indices are computed for that route; mixing target languages would mix their denominators. For another target language or context, duplicate the project to reuse its source content with a fresh workflow.

Yes. Adaptation can be necessary even when both versions use English, because wording, concepts, examples, or response options may function differently across cultural and ESL contexts. Select English as both languages, choose the intended country or region, and document the contextual changes through the same workflow.

Role restrictions preserve the independence required by the selected framework. Forward and back-translator roles remain separate, while translator participation in review varies by method. Permitted but weaker combinations require a recorded rationale (Beaton et al., 2000; Harkness, 2003; McKenna & Doward, 2005; International Test Commission, 2018).

The certificate records the stage sequence and dates, team roles, content validity results, and a SHA-256 hash of the final items. You can use the hash to check whether a copy matches the version signed off by the team. The certificate documents the process; it does not certify the quality of the translation.

The supported frameworks rely on independent human judgment. A shared machine-generated suggestion could influence wording and weaken translator independence, so Equivalence does not generate translations.