Capstone: a complete evidence-based recommendation
Final integration. You are given symptoms, constraints, three product pitches and trial summaries. Make an ethical decision, justify the strongest evidence, flag uncertainty and define a safe follow-up.
The fictional brief
A 32-year-old has recurring inflammatory acne, older brown marks, frequent stinging, moderate sun exposure, a limited budget and no consistent sunscreen. Product A is a high-strength peeling serum with a 14-person uncontrolled satisfaction survey. Product B is a comfortable moisturiser studied in 120 users against a matched vehicle for dryness. Product C is a pigment serum whose independent controlled trial showed a modest validated spot-score benefit after twelve weeks.
Capstone: a complete evidence-based recommendation — a published visual reference for the topic, not SESH product-test data.
Open full-size image ↗Evidence is not a prescription
B's evidence addresses dryness, while C's trial is more relevant to pigment, not necessarily the irritation or acne. A's attractive headline is weak for causal claims and its potential irritation matters. The buyer's current stinging, sun-exposure habits and scarring risk may dominate immediate priorities even if C has good data.
An expert decision includes a 'not yet'
A reasonable first stage is to address ongoing irritation, avoid the suspected triggers, establish a tolerable basic routine and reliable sun protection. Assess acne severity and refer deep, scarring or resistant inflammatory disease. Defer optional brightening or anti-aging actives until tolerability and priorities are clearer. This is a provisional cosmetic plan, not treatment of a diagnosed disorder.
Document a reproducible verdict
Write your final recommendation in five lines: observations, priority, evidence used, uncertainty and follow-up trigger. Specify what would change your mind, such as appearance of nodules, persistent burning, improved adherence or valid head-to-head evidence. The licensed forest plot here is a methodological example, not a meta-analysis of these invented products.
These curves show how changing mean and variance affects a statistical model. They are theoretical teaching data, not skin or fragrance-panel results.
Open full-size image ↗Your lab · decide before revealing answer
CAPSTONE: Three product options, two goals, irritation, budget, and incomplete trials. Choose a first-stage plan; say what cannot be proven; name two measurements and one medical referral condition. Do not recommend every product simply because each has a positive claim.
Reveal analysis after making your decision
Model reasoning and safety decision
Stage one: simplify suspected irritants, establish tolerable support and sun protection, and assess scarring acne for clinical care. Do not infer A causes improvement or that B treats pigment. Track new inflammatory lesions and stinging; assess spots later with consistent imaging. Refer nodules, new scars or significant persistent inflammation.
Further research
Related SESH references
Links do not establish treatment of the hypothetical condition.
Cases are fictional training exercises, not clinical diagnoses.
