article / GVI

A question gives curiosity a shape.

Students start wide. GVI-style project work narrows that interest into a question a later module can actually answer.

Field noteUpdated 9 Sep 2026gvi.cc

Start broad

Industry, company type, or trend.

Narrow the decision

What does someone need to decide?

Define evidence

What information would make the answer stronger?

A funnel that does not take a month

Interest: consumer brands. Topic: entering a new city. Decision: this year or not, with public data. Evidence: category growth, competitor density, a consumer trade-off, one operational constraint. Stop. You have a module-one question.

Tests of a good question

It is possible to be wrong. It is possible to finish. A table or a customer segment could change the answer. A stranger understands it. If all four fail, you have a theme, not a question.

High-school specific

Do not require original market interviews unless the school can supervise ethics. Public sources plus a clear decision are enough for a virtual internship brief.

When the question needs data module two

If the answer cannot move without a calculation, write that in the brief. Do not pretend a narrative is analysis.

Kill prestige questions

“Should this famous firm do a merger?” is usually unfinishable for a student. Smaller, local, or hypothetical constrained firms make better internships because the student can actually see the edges.

Filters that kill bad questions quickly

Can I finish with public data? Could a table change the answer? Is there a person who would use the answer? Is the noun smaller than an industry? If you fail two filters, you have a theme. High-school students should fail filters on purpose until the question is almost embarrassingly small. Small questions produce real internships. Giant questions produce posters.

A useful trick: write the recommendation you might already believe, then ask what evidence could kill it. The research question is that killing test. If nothing could kill it, you are writing an opinion essay.

More questions that work from a laptop

Should a student club raise prices this term given last year’s attendance table? Should a local cafe add delivery given public competitor density and a simple margin sketch? Should a school cohort run one track or two given staff hours? These are not glamorous. They are markable. Glimmer can come later, after the student can finish.

Killing a question is allowed

If two days of search show the decision cannot be answered with lawful public data, rewrite the decision. Record the kill in the log. That is research skill, not failure. Students who refuse to kill a prestige question spend the pathway performing impossibility.

Mentors should praise a clean kill and a smaller replacement. Families sometimes hear “the intern gave up.” Teach them the difference in the parent letter. GVI’s honesty about finishability depends on this cultural point.

A replacement question should reuse some of the sources already gathered so the days are not wasted. Transfer the log. Do not start a new identity. Long pathways that restart from zero every time a question dies never reach module four.

Continue

Keep exploring GVI.

Each page covers a different part of the remote internship pathway so families can compare structure, work, and evidence.