Understand the reporting choice
Short guides to the decisions a formula cannot make for you.
Keep coupon quotes, settlement cash and duration in their own units →
Follow a regular coupon schedule from a clean quotation to accrued interest, an implied yield and a fixed-date rate scenario.
Choose the sample design before calculating a p value →
Separate individual variation, uncertainty of a mean, matched differences and independent-group comparisons.
Write the event before choosing a distribution →
Distinguish exact counts, inclusive boundaries, replacement assumptions and observation-window units.
Read a fitted model before using its estimate →
Match coefficients to predictors, keep the intercept and compare residuals on the scale that was actually fitted.
Separate an annual charge, a reporting window and carrying value →
Read salvage, fiscal stubs and declining-balance switches before combining financial functions.
Read the cash-flow timeline before comparing returns →
Keep initial costs, numeric zero periods, actual dates and finance/reinvestment assumptions explicit.
Read the criteria table before calculating →
Separate selected records from numeric observations, interpret AND/OR rows and keep DGET cardinality errors visible.
Choose which formula errors to handle →
Read actual error types, keep unexpected failures visible and preserve the denominator of a numeric summary.
Choose a LAMBDA loop from the result you need →
Pair corresponding records, carry one final total or calculate every row-and-column combination.
Keep record identity and order when reshaping arrays →
Align exported fields, distinguish numeric zero from absence and declare the order before flattening or taking rows.
State the clock and endpoint rules before calculating →
Keep date/time values numeric, validate normalized imports and explain what an overnight or inclusive interval means.
Choose the population behind a grouped report →
Filter original records, recompute mean totals and name the denominator before interpreting a cross-tabulated share.
Choose identity and row grain before reconciling exports →
Separate a missing record, a split payment and a grouped amount difference before choosing a matching formula.
Keep loan units, signs and observation times consistent →
Explain what a constant-payment formula measures before interpreting its payment, interest or balance.
Keep field boundaries and identifiers before converting text →
Decide which separators, empty records and character rules define an imported text report before changing its values.
Read the sample before acting on its summary →
Keep sample size, percentile conventions and review decisions separate when interpreting measurement data.
Move a deadline or count an interval? →
Choose the calendar rule before combining dates, weekly closures and holidays.
Decide what one row means →
Separate order totals from line items before counting or adding anything.
An average needs its count →
Keep the denominator when combining groups, and distinguish a zero from no records.
Keep the sign in the picture →
Show refunds, order the horizontal axis and inspect the numbers behind a chart.
When a label contains an asterisk →
Decide whether a symbol is a pattern or part of an exact label before writing criteria.