Last time we traced how a request becomes a finished piece. This time we stop looking at that piece as a single sheet of paper and recode it onto two dials: "does it stay true to the facts" and "does it actually reach the reader." Split this way, what looks like skill or its absence comes into focus.
Reading one finished piece on two separate dials
Think of a medical check-up. A doctor does not stop at "you look healthy enough." Height, weight, blood pressure, blood sugar — each is measured on its own dial. A single impression is translated into several numbers. Reading a promotional piece works the same way. The offhand verdict "nice piece of work" is here translated onto two dials.
The first is "faithfulness to the facts." Here it means whether the claims on the page stay true to their source — the trial results, the package insert, the evidence behind them. When you go back to the source, does it say the same thing? We call this "grounding": are the feet on solid ground. The second is "craft of delivery": the skill of arranging the same facts so the reader does not misunderstand and can reach the information they need. We call this "design." Because it is about reaching the reader, we also call it "arrival."
A finished piece looks like a single sheet, but two separate abilities are carved into it: "does it stay true to the facts (grounding)" and "does it reach the reader (arrival)." Reading these two apart is the work of installment three.
Why split them? Lump them together and good looks will pull your eye away from a slipped fact. The reverse happens too: plain but accurate work gets dismissed as "boring." Only once you split the two can you say which side is strong and which is weak.
Four combinations — the same "looks skilled" hides different things
Consider a driver's license. Driving safely without accidents and guiding passengers to the destination without getting lost are different abilities. Some people do both; some are safe but lose their way; some speed along but seem about to crash. A piece, too, splits into four types when you cross faithfulness and design as high or low.
| Type | Faithfulness (grounding) | Design (arrival) | Sign in the piece |
|---|---|---|---|
| Dangerous amateur | Low | Low | Vague evidence, hard to read. Cannot return to the source, and the message is scattered |
| Correct but unreachable | High | Low | Facts match the source, but text is cramped and the point is buried; the reader cannot pick it up |
| Dangerous sales pitch | Low | High | Polished and persuasive, but back at the source there is overstatement or substitution. The most dangerous |
| The main road | High | High | True to the source and reaches the reader's point without misunderstanding |
The scariest of these is the "dangerous sales pitch." Because it looks skilled, at a glance it is hard to tell from "the main road." For example, a chart of benefit is shown large, while the share of non-responders or the safety cautions are shrunk and pushed to the edge. No single word is a lie, yet the reader comes away firmly convinced "this drug works." That is a "persuasive misreading." The higher the design skill, the higher the power to hide a slip. So design skill alone must never be the score.
Faithfulness sets the ceiling for design
Think of cooking. However masterful the chef, rotten ingredients yield only a good-looking, dangerous dish. The skill of plating (design) only matters within the range where the ingredients (the facts) are safe. A piece is the same: if the facts have slipped, the better the craft of delivery, the greater the harm. Faithfulness sets the ceiling under which design can do its work.
Grounding sets the ceiling. However much skilled design you stack on a claim that has slipped from its source, the score will not rise. If anything, the more skilled it is, the wider the misreading spreads, so the danger grows.
That is why the order of reading matters. Do not let the polish of the design catch your eye first. First confirm "can this claim return to its source," and only then look at "is the delivery appropriate." Separating the floor (the condition that must be met) from the total score (how excellent it is) — the groundwork for installments five and six — sits right here. Reverse the order and you will hand a high mark to the "dangerous sales pitch."
Learning the signs of low faithfulness from reported deviations
Think of airport baggage screening. The officer does not pick out suspicious people by their faces. They stop things because they have memorized the "typical shadow" of a dangerous object on the X-ray. Signs of low faithfulness work the same way: learn the typical shapes first, and you can spot them in a real piece. Let us line up, as typical shapes, cases actually flagged by Japan's monitoring project. Company names are withheld in every report, so here too we cite them in general terms, "in a reported case."
In a reported case, one chart kept a normal vertical axis in the product information summary (the official explanatory document), yet on the presentation slide alone part of the axis was stretched to make the gap between two drugs look larger than it was. In another, a curve showing the surviving proportion started its vertical axis at 0 when it should have started at 0.8, making the two drugs look indistinguishable. In another, no material was even prepared for the primary endpoint (the foremost measure promised first in that trial), and only a secondary result that happened to show a significant difference (a gap hard to explain by chance) was presented. In a Japanese subgroup with no difference, the speaker said "a difference is appearing," and when challenged fled to authority: "the professor says it is fine too." One case argued an effect from a chart of just nine subjects (four versus five) with no statistical handling at all. None of these is craft of delivery; each is a trace of faithfulness shaved away.
Behind such slips, the maker's psychology is at work. The scariest part is that the person has no sense of lying. The table below lines up the reported trace, the psychology working behind it, and the strength that stops it.
| Trace left in the piece | Psychology behind it | Strength that stops it |
|---|---|---|
| Stretch part of the axis / start it at 0 | Motivated reasoning (the way data is shown is pulled toward the conclusion one wants to sell) | Source-grounding — can you return to the original chart |
| Omit the primary endpoint, explain only the secondary | Sin of omission (defending oneself by not speaking) | Self-review — check it yourself before release |
| "The professor says it is fine too" | Externalizing responsibility (handing judgment to authority) | Source-grounding — return the basis to documents, not people |
| Claim an effect from nine cases with no statistics | Local rationalization (just this once, this is enough) | Balance design — limit the claim to match the strength of the data |
Look at the rightmost column. A sign of low faithfulness can never be cancelled by skill of design. What stops it is always the grounding strength of "can you return to the source." So when the one measuring finds a trace, they do not blame the maker's character; they record the vertical axis (faithfulness) one step lower. These are not the deeds of special villains but circuits an ordinary maker falls into under pressure — which is exactly why we need this encoding, to watch for the same circuit inside ourselves.
Where to look in the piece, and how to recode it onto the two axes
Picture marking up a proof sheet (the galley you check before printing) with a red pen. The proofreader does not write impressions. They point at specific spots: "this differs from the original," "without a comma here it will be misread." Recoding onto the two axes is the same: decide by pointing at clues inside the piece, not by impression.
The clue for faithfulness (grounding) is the distance between claim and source. Do the charts and numbers match the original data? Is the basis for "improves" actually shown in that trial? Has it dropped the conditions (which patients, how long, what dose) and overstated the reach? Are there notes or references that let you return to the source? Conversely, a spot where the assertion is strong but the basis is thin is a sign that the grounding is floating.
The clue for design (arrival) is how easily the reader can move. Does the main point come first? Are technical terms paraphrased or explained? Is the safety and risk information at a size and position in balance with the body text? Do the figures use emphasis that breeds misunderstanding? The table below sorts how the same spot reads on each axis.
| Spot to look at in the piece | Read as faithfulness (grounding) | Read as design (arrival) |
|---|---|---|
| Chart of benefit | Do values and conditions match the original data | Is only the favorable side stressed; is the axis chosen fairly |
| Side-effect / caution text | Does it conflict with the package insert and such | Is it sized and placed in balance with the body so the reader notices |
| Headline copy | Does it overstep the source by overstating or dropping conditions | Does it convey the point briefly and accurately without inviting misreading |
The key is to view the very same spot with two eyes. One emphatic phrase, seen through the grounding eye, is "overstatement (a slip)"; seen through the arrival eye, it is "a skilful presentation" — each judged separately. Only when you can read this double way can you catch, without missing it, the most dangerous type: "skilled yet slipped."
Measuring Skill from Work and Behavior ── Map of all 10 episodes
- Vol. 1: Measure by the Materials Actually Made, Not by Impressions or Self-Report ── A material maker's skill is measured from the actual deliverables and observable conduct, not from self-report or others' impressions.
- Vol. 2: Tracing the Brief, the Choices, and the Result — In Order ── Read a creator's skill from evidence by walking through one real project in order: the brief, the thinking, the actions, and the result.
- Vol. 3 (this episode): Reading "Faithfulness to the Facts" and "Craft of Delivery" Out of the Work Itself ── This installment shows how to recode a finished piece into two axes — faithfulness to the facts and the craft of getting it across — by reading concrete clues, not impressions.
- Vol. 4: The Rules That Keep Measurement Honest ── Six ground rules that keep the evaluator from drifting when measuring an author's real skill.
- Vol. 5: Three Rulers: Accuracy, Clarity, and Balance ── Defines three rulers for grading material-making skill and scores each on a four-step scale: accuracy as the floor, clarity as the reach, and balance as the adjustment between too much and too little.
- Vol. 6: How to Decide the Level — Returning to the Source Sets the Ceiling ── Work that cannot be traced back to its source cannot earn a higher level, however polished it looks. Grounding sets the ceiling.
- Vol. 7: What Deliverables Signal Which Level ── An anchor table that reads a creator's level (L1-L4) from visible deliverables and behavior patterns.
- Vol. 8: How Far Can We Trust a Judgment? ── How sure a level judgment is depends on how visible the evidence is; less observable skills produce shakier judgments, so we attach a confidence to each verdict.
- Vol. 9: Combine More Than Self-Assessment: Add the Reviewer's and Requester's View ── Layering four viewpoints — self, reviewer, requester, and AI — surfaces the deviations of omission that a single pair of eyes cannot see.
- Vol. 10 (final): Connecting the Measurement to Pass/Fail and a Development Plan ── The finale links the score to the pass floor and a plan for what to grow next.
Splitting a piece onto two dials is the first step from impression to measurement. Break "looks skilled" into faithfulness (can it return to the source) and design (does it reach the reader), and place it in one of four types. The thing to guard against most is letting good looks hide a slipped fact.
Next time we take up the agreements that make this reading come out the same no matter who does it. So that the same piece does not earn scattered scores, we decide what counts as a clue and what we agree not to look at. We firm up the shared dial of two axes into a shared measuring stick.
- Split one piece into two axes. Translate a piece's "looks skilled" onto two dials: faithfulness to the facts (grounding — can it return to the source) and craft of delivery (arrival — does it reach the reader).
- The most dangerous of the four types is high design, low faithfulness. The better the looks, the more a slip can be hidden — a "persuasive misreading." Never score on design alone.
- Grounding sets the ceiling. If the facts have slipped, more skilled craft only spreads more harm. First confirm the return to source, and read the floor apart from the total score.
- Ministry of Health, Labour and Welfare, Compliance and Narcotics Division (commissioned project). Report on the Monitoring of Promotional Information Activities for Prescription Drugs (March 2024 and prior years). Flagged cases are published with company names anonymized; the deviation patterns cited here are generalized from this report.
- Japan's Ministry of Health, Labour and Welfare, "Standards for Fair Advertising of Drugs and Related Products" — the general principle that claims of efficacy and safety must rest on fact and avoid exaggerated or misleading expression.
- Japan Pharmaceutical Manufacturers Association, "Promotion Code" — the industry's general standard that information on prescription drugs should be accurate, fair, and objective.
- General accounts of the Behavioral Event Interview (BEI) and the STAR method — the approach of reading ability from concrete behavior and work products rather than from impressions.
- Textbook literature on competency assessment — the general framework of judging ability in stages using observable behavioral indicators.