Evidence-Based Materials Research Design Guide

The problem

Many materials research plans fail at the evidence level: claims are not tied to checkable sources, and experimental variables are not mapped to falsifiable mechanism hypotheses. The output may look complete, but key parameters lack provenance, characterization covers only one observation, and novelty is overstated.

How it works

The skill starts from a material system + target performance pair and applies an evidence gate before drafting a design:
- Literature verification: cross-checks DOI, authors, journal, and year, then builds an evidence table with claim, source, boundary, and confidence.
- Experimental design: maps hypotheses to factors, levels, controls, replicates, success criteria, and next steps after failure, with low-cost screening first.
- Characterization matrix: states what each method can support, what it cannot prove, and possible artifact risks.
- Mechanism and paper structure: adds competing hypotheses, falsification experiments, and organizes the manuscript around gap, hypothesis, evidence, and boundary.

Boundaries

It assists with designing experiments, characterization, mechanism hypotheses, and paper structure. It does not replace lab safety approval, equipment SOPs, medical advice, or unvalidated scale-up decisions. Unverified items are marked as pending verification or working assumptions rather than conclusions.