Direct answer
How do students conduct a responsible business investigation?
Start with a decision, not a preferred answer. Convert the problem into a narrow question that identifies the setting, one factor to change, measurable outcomes, a comparison, and a defined scope. Before collecting results, document the method and the claim each source could legitimately support.
Run the comparison consistently, preserve inconvenient results, and separate observation from interpretation. The final brief should answer only the research question, name uncertainty and model limits, and identify what real-world evidence would be needed before acting.
Investigation menu
Six questions students can test
| Simulation | Research question | Change only | Measure | Worksheet |
|---|
| Lemonade stand | How does one price change affect a fixed-day result? | Price | Units, revenue, profit | Open |
|---|
| Restaurant | How does staffing affect service and profit? | One staffing setting | Wait, quality, profit | Open |
|---|
| Coffee shop | Does extra rush capacity improve the outcome? | Rush staffing | Queue, sales, profit | Open |
|---|
| Bakery | How does production quantity change availability and waste? | Batch quantity | Stockouts, waste, profit | Open |
|---|
| Motel | What changes when the room rate moves? | Room rate | Occupancy, revenue, profit | Open |
|---|
| Bookstore | How does inventory emphasis affect sales and slow stock? | One inventory setting | Sales, inventory, margin | Open |
|---|
A simulator is a controlled learning environment, not a representative sample of customers or businesses. Frame conclusions as findings about the model.
Ready-to-use sequence
50-minute question–audit–test–report lesson
Learning goal: plan a reproducible investigation, evaluate evidence fitness, and make a qualified claim without confusing simulated results with real-market proof.
- Frame the decision — 5 minutes. State who needs to decide what, the relevant outcome, and one guardrail such as cash, safety, quality, waste, or capacity.
- Write the question — 6 minutes. Use: “Within [scenario and period], how does changing [one factor] from [A] to [B] affect [primary outcome] while monitoring [guardrail]?”
- Audit three sources — 8 minutes. Classify the simulator instructions, a relevant strategy guide, and the worksheet as model documentation, explanatory source, or collection tool. Record currency, authority, relevance, method, and limits.
- Preregister the method — 6 minutes. Write the baseline, treatment, fixed settings, trial count, outcome units, stopping rule, and expected pattern before viewing results.
- Run the comparison — 10 minutes. Complete A and B under matching conditions. Copy results exactly and note any departure from the plan.
- Analyze rival explanations — 7 minutes. Calculate absolute and percent change where meaningful. Identify randomness, capacity constraints, setting differences, or model rules that could also explain the result.
- Publish a bounded brief — 8 minutes. Answer the question, cite the evidence log, state confidence and limits, and propose either another model test or an ethical real-world research step.
Printable student record
Research plan and evidence log
Decision and decision-maker: __________________________________________
Testable question: _________________________________________________
Primary outcome and unit: __________________ Guardrail: __________________
Expected pattern and reason: _________________________________________
| Source or evidence | Type | Who / when | Method or basis | Useful for | Not proof of |
|---|
| Simulator instructions | | | | | |
|---|
| Strategy guide | | | | | |
|---|
| Run A / baseline | Generated | | | | |
|---|
| Run B / treatment | Generated | | | | |
|---|
| Method record | A — baseline | B — treatment | Difference / note |
|---|
| Changed factor | | | Only planned change? |
|---|
| Fixed settings | | | Matched? |
|---|
| Primary outcome | | | B − A = ___ |
|---|
| Guardrail outcome | | | Tradeoff? |
|---|
| Unexpected condition | | | Possible influence? |
|---|
| Repeat result | | | Consistent? |
|---|
Audit whether a source fits the claim
- Currency: is it current enough for the question?
- Authority: who created it, and with what expertise or responsibility?
- Relevance: does it address the same setting, population, and outcome?
- Method: how was the information generated and checked?
- Purpose: is it explaining, persuading, selling, or documenting?
- Limits: what conclusion would exceed the evidence?
A polished page is not automatically strong evidence. Prefer primary documentation for model rules and current authoritative sources for real legal, safety, or regulatory questions.
Separate five layers of a finding
- Question: what was defined before the run?
- Method: what changed and what stayed fixed?
- Observation: what values appeared?
- Interpretation: what explanation fits those values?
- Decision implication: what limited next step follows?
Do not turn “profit increased in this run” into “customers prefer this price.” The second claim requires customer or market evidence the simulation did not collect.
Write the one-page research brief
- Decision context: identify the decision and guardrail.
- Research question: reproduce the preregistered wording.
- Method: name A, B, fixed settings, trials, outcomes, and deviations.
- Finding: report values and differences before interpretation.
- Assessment: give the leading explanation and at least one rival explanation.
- Conclusion: answer only within the simulated setting and assign low, medium, or high confidence.
- Next evidence: propose a repeat, a different model condition, or a lawful, voluntary, minimal-data real-world method.
Sentence frame: In [model and period], changing ___ from ___ to ___ was associated with ___ while ___ changed by ___. Confidence is ___ because ___. This does not establish ___. Before a real decision, we would verify ___ using ___.
16-point research methods rubric
| Criterion | 4 — strong | 3 — capable | 2 — developing | 1 — beginning |
|---|
| Question and plan | Decision-linked, testable, scoped, preregistered, and reproducible. | Clear with one minor method gap. | Question or controls partly defined. | Preferred answer replaces a plan. |
|---|
| Source audit | Types, authority, method, relevance, currency, purpose, and limits evaluated. | Sources evaluated with most checks. | Checks are mostly descriptive. | Sources accepted without evaluation. |
|---|
| Evidence record | Exact units, matched settings, deviations, repeats, and rival explanations documented. | Accurate comparison with minor omissions. | Incomplete or partly confounded record. | Selective or irreproducible evidence. |
|---|
| Research brief | Bounded finding, confidence, limits, ethical next evidence, and no overclaim. | Supported conclusion with most limits. | Claim exceeds some evidence. | Unsupported real-world conclusion. |
|---|
Reward transparent methods and preserved contrary evidence. Use the full simulation assessment rubric for collaboration and presentation criteria.
Teacher notes, supports, and extension
Make preregistration visible
Have pairs exchange and initial plans before running. Changes remain allowed, but students label and explain them instead of silently rewriting the method.
Support learners
Provide one approved question, highlight the independent and dependent variables, and let students complete the brief orally using the sentence frame.
Extend the study
Ask teams to replicate another team, compare results, and decide whether disagreement comes from random variation, method drift, or a different model condition.
Research legally, privately, and honestly
This activity does not require human participants or personal data. Do not ask classmates or the public for names, contact information, precise locations, health details, financial details, protected characteristics, or other sensitive information. Do not scrape sites, evade access controls, copy proprietary datasets, or present copyrighted material as your own.
Real surveys, interviews, observation, experiments, and customer research can require informed consent, accessibility, privacy safeguards, institutional approval, sampling expertise, and compliance with local rules. Simulation output is educational model evidence—not financial, legal, market, safety, or regulatory advice. Cite sources, preserve contrary results, and never fabricate participants, observations, citations, or testimonials.
Frequently asked questions
How can a business simulation teach research methods?
Students define a testable question, document a method, hold settings constant, compare results, and qualify the claim.
Is simulation output primary or secondary research?
It is generated model evidence. It supports claims about that model, not claims about real customers or markets.
Do students need to collect personal information?
No. Use public non-personal sources and anonymous simulation results.
What makes a question testable?
It names the setting, one deliberate change, measurable outcomes, comparison, and scope.
Can results prove what a real business should do?
No. They identify a bounded model finding and the real-world evidence still needed.