HCD.22:5 - Archetypal Grounding
HCD.22:5.1 - A shared diagnosis with individual model selection
This constructed case continues HCD.28’s future-diagnostician product. The task uses safe supplied data and a domain-qualified answer basis; it does not authorize testing a live pump.
The baseline is 100 m³/h, 400 kPa and 20 kW. A later observation reports the same flow and 22.4 kW, with pressure still unknown. The group must recommend a warranted next observation and a bounded diagnostic conclusion. Each participant must learn to distinguish a ratio change from evidence selecting its cause.
After a worked contrast, each records an initial explanation from the common data. The record shows whether “12% more energy per volume” has already become “12% more resistance” without the needed assumptions. References and arithmetic tools remain available; an answer naming the preferred model is withheld during this particular first choice.
For four participants, prepare four accessible contributions:
| Contribution | What its holder must do | Use in the common result |
|---|---|---|
| Measurement evidence | Interpret a repeat power reading of 22.3 ± 0.2 kW and flow of 99.5 ± 1 m³/h against the baseline. | Test whether the stated measurement uncertainty could remove the ratio increase. |
| Resistance model | Explain what pressure change would be expected if flow and efficiency were unchanged. | Supply a conditional prediction, not a diagnosis inferred from power alone. |
| Rival model | Explain how changed efficiency or regulation can produce a different relation and what remains unknown. | Prevent a resistance explanation from being accepted merely because it is familiar. |
| Observation budget and access | Compare the obtainable observations and their effort under the remaining allowance. | Select an informative observation that can actually be obtained. |
The group has four specialist-hours for measurements in this constructed setting. The repeated power and flow work costs one and one-and-a-half hours respectively, leaving one-and-a-half hours for the available pressure observation. Each holder explains their contribution and challenges or answers a consequential question about another contribution. The two three-minute contributions per person and six minutes of reconciliation use thirty minutes. HCD.20:5.1 adds a separate six-minute qualified-help visit after all four initial explanations and before the second personal round. Its staggered starts and named mentor assignments therefore require thirty-six active minutes per group; help does not occupy the time already promised for personal reasoning.
The common comparison can now be reconstructed. Even the smallest ratio allowed by the stated readings, 22.1/100.5, exceeds the baseline 20/100. The stated uncertainty alone therefore does not erase the increase. That does not decide its cause. At unchanged flow and efficiency, the resistance model predicts 448 kPa. A qualified pressure observation is obtainable within the remaining allowance and can distinguish that prediction from the rival condition.
Provide the selected observation only after the participants justify its use. In one branch it is compatible with 448 kPa; the resistance account remains a conditional candidate, not a unique proven cause. In another it is compatible with the 400 kPa baseline; the simple unchanged-efficiency resistance explanation no longer accounts for the power change. The appropriate return is to the qualified efficiency, regulation or measurement question, not to a confident unsupported repair order.
The group result retains the observations, model conditions, chosen test and remaining uncertainty. It identifies which earlier inference changed and why. A facilitator rejects neither uncertainty nor a bounded continuation merely for failing to name one cause.
Then each participant receives a changed case, with the same subject comparison available as a reference but no peer, mentor or AI supplying the model choice. The ten-minute personal attempt asks what the observations support and what the receiving engineer should do next. Success in the group remains practice evidence; the personal response and its later correction are interpreted separately.
Personal case: the pressure result is already available. Reuse HCD.6’s baseline of 100 m³/h, 400 kPa and 20 kW. The supplied independent results are 22.3 ± 0.2 kW, 99.5 ± 1 m³/h and 400 ± 4 kPa, compatible with the pressure baseline. The four specialist-hours have been used on those three observations. No further worthwhile observation is obtainable in the present window, and the day-long shutdown is outside the allowed arrangement. The receiving engineer needs a bounded diagnostic return now, not an unsupported resistance-specific remedy.
The learner may use the quantity definitions, arithmetic tools and HCD.6’s conditional subject comparison. They must interpret the existing pressure result, retain the live explanation or uncertainty, and justify either an obtainable further observation or closure of the present inquiry. Keep the worked responses below with the provider during this attempt; showing them would change the help supplied.
A plausible wrong response is: “The indicator rose by about 12%, so the increased resistance should be corrected.” The ratio increase survives the stated power/flow uncertainty, but the unchanged-efficiency resistance account predicted a pressure near 448 kPa. The supplied pressure does not support that prediction. A second pressure request would also spend effort obtaining a result already available.
A warranted personal return distinguishes the indicator change from its cause: the increase remains, the simple resistance account under its stated assumptions is unsupported, and the question returns to regulation or efficiency while preserving measurement explanations outside the stated uncertainty. The engineer receives the observations, rejected conditional inference and remaining uncertainty; no resistance-specific remedy is selected. The absence of another worthwhile obtainable test permits this bounded closure. It does not establish one unique cause.
If the first response still infers resistance from the ratio, feedback points to the pressure condition and asks the learner to rewrite the diagnostic return. The learner reads the feedback and revises within HCD.28’s fifteen-minute feedback-and-correction period. That corrected return demonstrates work with this assistance; it does not replace the preserved first response or establish unassisted transfer. A subsequent personal claim uses another suitable observation under HCD.11–13 when its question is actually live.
HCD.22:5.2 - Recover a missing contribution without pretending learning occurred
Suppose the measurement-evidence participant is absent. A prepared qualified card can restore those data and their interpretation to the group. Assign a present participant to explain how the uncertainty changes the inference, adjusting the time construction if needed. The group can proceed if the restored contribution is usable.
The absent participant has not performed either explanation or model selection. Offer a viable individual or later shared task that elicits the missed operation. If no such route is available, report that missing learning opportunity. A signed group report cannot repair it.
Suppose instead the card is present but compresses the conclusion to “measurement is not the cause.” Repair its claim: the stated uncertainty does not explain away this ratio increase; that does not eliminate every possible measurement fault. The error belongs to the supplied material and its dependent conclusion. It is not evidence that learners need more motivation.
HCD.22:5.3 - Shared equipment and an explanation without a report
In a safe instrument-training task, one participant operates while others predict, identify a mismatch or select a measurement. The operator uses that contribution and explains the response; roles then change so that required control actions are personally performed. No private first answer is necessary before a novice has seen how to use the instrument. Later interpretation still needs a task that does not supply the decisive choice.
A different task asks one participant to explain why two visually similar technical passages imply different actions. The recipient applies each passage to a small case, exposes an ambiguity and requests a repair. The explainer revises the account and the recipient retries. The learning interaction can be complete without a common written report. The result is the explained distinction and its observed use, with any help and remaining difficulty preserved.