Problems, then questions, then knowledge—again
PQK Lab is an inquiry workshop, not a poster factory. Teams often treat a finished trial as a reason to abandon the setting and invent a new theme that photographs better. That restart erases the trail. Problems, questions, knowledge is a sequence: the next question should grow from the same friction, the same notebook, and the rival explanations still alive in the model.
This page teaches how to choose that next question without pretending the previous weeks did not happen. It does not invent mentors or results. It teaches a move until it can be attempted in a real doorway, lunch line, planter, or sensor the team can reach.
Why restarting feels easier
Restarting feels clean. A new title hides misses. A new board hides an ugly log. Inquiry-based STEM fails quietly when display outruns the trail. The failure is usually not a lack of enthusiasm. It is a missing problem continuity, an untestable next question, or a knowledge claim that arrived before the evidence log.
Repair is always a return to the trail. Redesigning the poster comes last. PQK would rather keep an ugly notebook with a live question than a beautiful board with a dead one.
A short checklist for the next question
Use this as a critique script, not as filler bureaucracy. After each item, write a sentence in the notebook. If you cannot walk to a setting, the item is not finished.
- Where, exactly, was the original problem noticed—and is that setting still reachable?
- What did the last evidence log confirm, miss, or leave rival?
- Which uncertainty in the knowledge model is still binding?
- Write five candidate next questions; strike any that cannot be embarrassed in two weeks.
- Compare the survivors by evidence cost: what can you count, photograph, measure, or trial with tools you already have?
- Pick the cheaper honest question. Open a new dated log page before any prototype glue.
Reuse the problem, do not reinvent the theme
Problem framing for students separates a grievance, a wish, and a friction you can observe. When choosing the next question, keep the same friction unless the team has evidence that the friction was misidentified. A grievance that is still too large does not become researchable by changing fonts. A wish that is still untestable does not become researchable by adding a sensor for show.
From the same framed problem, question design produces variants. Some are too broad. Some hide the answer. Some require a laboratory the team does not have. The surviving question is researchable because evidence could confirm or embarrass it. That is the PQK meaning of researchable questions when the work continues rather than restarts.
Evidence cost decides the winner
Evidence habits take over after candidates exist. Teams decide what they will count, photograph, measure, or trial, and how they will record a miss. The cheaper honest question wins because a two-week log that can fail beats a semester slogan that cannot. Knowledge is not a paragraph of certainty. It is a model that can be revised when the next trial disagrees.
An evidence log for the next question is dated, located, and unit-aware. A knowledge model names causes, constraints, and uncertainties carried forward from the previous cycle. An inquiry build is a test, a prototype, a protocol, or a data story that tries one cause—not all causes at once.
A field-scale worked example
Worked at PQK, choosing the next inquiry question always returns to a setting the team can actually reach. The mentor asks what evidence could embarrass the current sentence. If the room cannot answer, the essay’s advice has not yet become a habit. That is inquiry-based STEM: not a larger theme, a harder local question.
They then write five next questions about the same friction and strike three as untestable. The remaining two are compared by evidence cost. The cheaper honest question wins. An evidence log is opened before any new showcase title is drafted. After a week of observations the knowledge model stays ugly and useful: two likely causes, one constraint, one uncertainty. The inquiry build tests one cause. The showcase draft is forced to include the miss.
What to keep from the last cycle
Keep the noticing note, the problem frame, the failed questions, the dated log, and the rival explanations. Keep consent and safety notes if people or shared spaces were involved. Keep the limit sentence that said what the last model could not claim. Throw away only the decorative title that outran the log.
Showcase communication remains last because communication should carry knowledge, not conceal its absence. A poster that cannot point to continuity is advertising a restart. A talk that cannot admit a failed trial is theatre.
Critique questions the lab keeps repeating
- Where, exactly, was the problem noticed?
- Which next question could a two-week evidence log actually embarrass?
- What did you count, photograph, measure, or trial last time, and what will you skip now?
- What rival explanation still fits the same observations?
- Which sentence on the next poster is still a hope rather than a finding?
Who uses this page
Students use it when a cycle ends and the urge to switch themes appears. Families use it to support continuity without replacing the noticing. Schools use it to design inquiry clubs that do not collapse into craft hour. Inquiry mentors use it as a shared critique language: problem, question, evidence, model, build, communication. Learning partners use it to see how a challenge brief can stay student-scale.
PQK does not publish rankings, invent faculty celebrity, or quote unnamed testimonials. The lab’s proof is the quality of questions and the inspectability of evidence.
What to do after reading
Open the notebook to the last model. Run the checklist. Write the cheaper honest next question. Enter the matching track in the curriculum if you need longer practice. An inquiry note should include the setting, the continued problem, the candidate questions, and constraints of time, tools, and school context. The lab will answer with a smaller question more often than with a larger plan. That is a compliment to the work.