Confirm the identity of the source
Record the supplier or publisher, document title, product or grade identifier, and revision or publication date where available. Check that the source refers to the form and condition relevant to the candidate. A broad material-family summary and a supplier's product-specific document serve different purposes.
Keep a link or an authorized copy so another person can revisit the information. If a source has changed, preserve the version used for the comparison where permission allows. An unexplained number copied into a spreadsheet becomes difficult to evaluate once its original context is lost.
Read what kind of statement the value makes
A typical value, a minimum, a maximum, and a declared range are different statements. Note which one the document provides. Do not convert a typical value into a guaranteed limit simply because a decision table needs one number.
Look for the units, test method, specimen details, and conditioning or environmental information relevant to the reported property. Missing context does not automatically invalidate a source, but it limits the conclusions that can safely be drawn from a comparison.
Compare like with like where possible
Two values with the same property name may have been obtained under different methods or conditions. A unit conversion can align the numerical units without resolving that methodological difference. Keep a note when the values are useful only for broad screening.
Separate a directly comparable result from an indicative value and from a gap requiring clarification. Ask the supplier or a qualified specialist when a difference affects the decision. Avoid averaging incompatible values merely to produce a complete-looking table.
Treat uncertainty as information
A measured result is connected to a method and its limitations. When a source reports measurement uncertainty, retain that information. When it does not, do not invent a precision or assume that additional decimal places imply stronger evidence.
There are several kinds of uncertainty in an early material review: incomplete source details, variation between products, uncertain use conditions, and the limits of the measurement itself. Keep them distinguishable so the next action addresses the relevant question rather than applying a generic confidence score.
Make a traceable comparison entry
A practical entry includes the property name, value or range, unit, statement type, source reference, relevant conditions, and the application requirement it informs. Add an explicit status such as comparable, screening only, clarification needed, or not applicable.
The purpose is not to make every source equally complete. It is to make the remaining differences readable. NIST's introduction to measurement uncertainty provides useful background on why a result should be accompanied by information about its uncertainty; supplier-specific interpretation still depends on the actual document.