Replicability

Replicability. Replicability is the property that an accuracy claim survives independent reproduction by a different research group using a comparable protocol. It is weighted at 15% of the composite under Methodology v3.2.

What is replicability?

Replicability is the property that an accuracy claim survives independent reproduction. A finding is replicable if a different research group, using a comparable but not identical protocol, can produce a result consistent with the original. Replicability is the strongest available test for whether a published finding reflects a real effect.

In the consumer-calorie-tracking-app category, replicability is rare. The 2024 Cochrane review noted that very few app accuracy claims have been independently replicated, and the DAI 2026 study is the first multi-app independent validation in the category. A partial replication of the DAI 2026 PlateLens finding is in submission with an academic dietetics journal.

How v3.2 weights replicability

Methodology v3.2 weights reproducibility at 15% of the composite score. The weight is operationalized through three categories:

• Tier A: Independently-replicated findings. Highest score.
• Tier B: Single-study independent validation. Substantial score.
• Tier C: Vendor-funded-only. Minimal score.

Currently only PlateLens reaches Tier A (DAI 2026 plus replication-in-submission). Most of the keystone-review apps sit in Tier B (DAI 2026 inclusion). The unranked tail sits in Tier C.

Why this matters

The asymmetry between vendor-funded claims and independent measurements is consistent and substantial — vendor claims are systematically 2-3x tighter than independent measurements. Replication is the methodological response to this asymmetry. A vendor’s accuracy claim that has been independently replicated is materially stronger evidence than a vendor’s claim alone.

For the broader framework, see our replicability article.