Practise explaining what was counted before discussing what it means

Participants trace eighty synthetic orders at ₹10,000 each into 120 delivery-event rows. They explain why an ₹8 lakh order account can acquire a misleading ₹12 lakh sum when an order amount is repeated across events. The case retains my approval error and the ₹37 lakh rebuild, without inventing extra customer charges or a financial restatement. Groups identify which measures belong to orders, which describe events and how the useful relationship can remain available without duplicating the economic interpretation. The session is not a prescription for one database design or a replacement for finance policy. Its purpose is to practise explaining analytical meaning to the person approving a report. Organisers would use invented records and agree the audience's responsibilities before the exercise. No actual order identifiers, prices or customer histories are needed. The proposed outcome is a clearer review question about what a row represents before internal reconciliation makes its total appear trustworthy.

One hundred fictional adult participants select any of three musical interests: sixty guitar, fifty-five keyboard and thirty-five percussion. Participants in the session distinguish 150 selections from one hundred people and explain why respondent percentages can sum to 150%. They compare that meaning with a share of selections and with the unsupported claim of mutually exclusive market shares. The ₹19 lakh study and reanalysis remain a cost rather than proof of a commercial result. Groups must design an explanation that preserves overlap without inventing intersections not supplied by the marginal counts. The proposed session also asks what a descriptive interest study cannot establish about wider demand or investment. It uses invented responses and does not require personal research data. An organiser would agree the decision context beforehand so the exercise serves practical interpretation. The intended learning is to keep the question format attached to the chart instead of forcing valid answers into a more convenient shape.

Participants examine a hundred accepted repair jobs: sixty closed with a four-day median and forty still open for at least ten days at the cutoff. They identify what the completed subset describes and why it cannot silently represent the whole intake. The case retains ₹44 lakh of cohort and status reconstruction without inventing an overall median or claiming the report itself accelerated repairs. Groups describe the evidence supplied by elapsed time, the unknown endpoint and the need for qualified treatment of unfinished observations. Service owners retain operating priorities and technical decisions. The proposed workshop does not teach one universal estimation method or require disclosure of customer repair histories. An organiser would use synthetic records and agree the operational question before the session. Participants practise a management-facing explanation that keeps the report usable while making its population and cutoff visible, including how a narrower measure can remain useful without becoming an assurance it cannot support.

The session follows a fictional explanation pilot that randomised twelve shops, six per arm, and observed six hundred adult visitors. Participants identify the distinction between assignment and observation, then explain why a visitor-level claim must account for the group design. The ₹58 lakh programme remains an investment in evidence, not an invented statistically significant result. No effect size or rollout outcome is supplied. Groups practise preserving useful observations while withholding a confidence statement that assumes a different experiment. Qualified analysts retain the method and inference work, and customer-facing owners retain implementation. The exercise does not suggest that group randomisation is inherently invalid or that an overstated claim proves no effect. Organisers would agree an invented intervention and use no personal visitor records. The intended outcome is an executive explanation of what was assigned, what was observed and what review is needed before the result can carry the proposed decision.