Evaluating a source

Five criteria and a verdict

A found source is not yet evidence. Before a fact from it goes into a memo, the source is scored on the credibility criteria the open sources agree on — R-source, a Memra rubric that mirrors them:

1. Author — credentials; is this person or body competent on this question?
2. Currency — is the date recent for this field? Equipment and software age in three years; accounting standards in one; a definition of a fallacy in fifty.
3. Objectivity — does the author have an interest in the conclusion? Is the other side represented?
4. Publisher / peer review — who put it out, and did anyone independent check it?
5. References — does it show where its own facts came from, so they can be followed?

Each criterion gets a score (this course uses 0–2) and a line of justification; then a verdict: use, use with caution (say what for), or do not use. Smith adds writing quality; Olds's online checklist adds the extras — domain and ownership, last-updated date, design, data practices, community reputation. They refine the five; they do not replace them.

Worked example — three sources for Lakeshore

A press vendor's white paper, "Total cost of ownership of production presses", 2025. Author: the vendor's engineering group — competent (2). Currency: this year (2). Objectivity: the vendor sells the press it costs; the comparison is with its own older models, not a competitor's (0). Publisher: self-published; no review (0). References: cites its own service data, not independent studies (1). Verdict: use with caution — for the maintenance schedule of its own machine, which it knows; not for the lease-versus-buy conclusion, which it has an interest in.

A trade-magazine article, "Is it time to lease your next press?", 2011. Author: a trade journalist (1). Currency: fifteen years old, in a field where the equipment and the financing terms have both changed (0). Objectivity: no product to sell (2). Publisher: an industry magazine with an editor (1). References: quotes two dealers, no data (0). Verdict: do not use for figures; the framing of the lease-versus-buy question may still be worth a sentence, attributed and dated.

A Statistics Canada table on printing-industry sales in Ontario, current release. Author: the national statistical agency (2). Currency: monthly, current (2). Objectivity: no interest in the conclusion (2). Publisher: government, with a published methodology (2). References: methodology and sampling stated (2). Verdict: use — for what it measures (industry sales), which is context; it says nothing about one press.

Notice that the highest-scoring source answers the smallest question. Scoring tells you what a source can be used for, not only whether.

Score two for the clinic

A booking-software company's blog post, "How reminders cut no-shows by 40 %" — author competent on its product, objectivity 0 (the 40 % is its own customers, selected), references 0; use with caution, for the feature list only. A peer-reviewed review article on reminder systems and non-attendance in outpatient clinics, in a health-services journal, four years old — author 2, currency 1 (recent enough for behaviour, not for the software), objectivity 2, peer review 2, references 2; use, and cite the range it reports rather than one number.

Score one cold

A Reddit thread in which three print-shop owners describe their lease terms. Author 0–1 (anonymous; possibly experienced); currency 2; objectivity 1 (no product to sell, but self-selected complainers); publisher 0 (no review); references 0. Verdict: do not use as evidence — but as a starting index it may name the vendors and clauses to look up in a source that scores. That is the right use of Wikipedia, too: cite what it cites, never it.