Quality control sheet: recording measurements so the data is usable later

A quality control sheet is the record of measurements and checks taken on a product or process: which characteristic was checked, what the specification allows, what was actually measured, and who took the reading. It is simpler than a full inspection report and more frequent — often one per batch, shift or machine. Its value is almost entirely in the data it leaves behind, which is why a sheet that is quick to fill in but impossible to analyse later has done only half its job.

What a quality control sheet should contain

• Header: product or part number, revision, batch or lot, machine or line, date and shift.
• One row per characteristic: dimension, weight, temperature, visual defect — whatever the control plan specifies.
• The specification for each row: nominal value and tolerance, or the acceptance criterion in words.
• The measuring instrument used, with its ID, so calibration status can be traced.
• Measured values in their own columns, not a tick for "OK".
• A pass or fail result per row and overall.
• Initials and time for each set of readings, and a sign-off by whoever is responsible.

Record values, not ticks

A column of ticks tells you the part passed; a column of numbers tells you it is drifting towards the limit. Wherever the characteristic is measurable, record the value. That single design decision turns the quality control sheet from a compliance artefact into an early warning: trends, shifts between operators, and a gauge that has started reading high all become visible in the numbers long before they produce a reject.

Put the tolerance on the sheet next to the value. Operators should never need to look up the limit elsewhere to decide whether a reading passes.

Designing the sheet so it gets filled in correctly

1. Follow the order in which checks are physically performed at the station.
2. Keep one characteristic per row and one reading per cell.
3. Pre-fill everything that does not change: part number, specification, instrument type.
4. Make out-of-tolerance results impossible to miss — a separate column for the reaction taken.
5. State what happens on a fail: stop, quarantine, notify, and raise a non-conformance report.
6. Version-control the sheet and record the revision on every completed copy.

From paper to a spreadsheet you can analyse

Paper sheets are fine at the station and useless for trend analysis. A shared spreadsheet with the same layout — specification in fixed cells, readings in columns, a formula flagging out-of-tolerance values — keeps the familiar format and makes the data usable. Ettex Sheets works well for this: each line or product gets a tab, conditional formatting highlights values near the limits, and a simple chart per characteristic shows drift across batches. Readings from the floor can also be captured on a form and land in the same sheet.

What to do with the data

• Chart key characteristics over time and look for trends, not just failures.
• Compare operators, shifts and machines for the same characteristic.
• Link failures to corrective action and check that the trend changes afterwards.
• Use recurring near-limit readings to justify gauge calibration or process adjustment.
• Feed supplier-related failures into the supplier scorecard.

Where dedicated statistical process control software is justified — automatic data capture from gauges, control charts with run rules across hundreds of characteristics — it replaces the sheet. For most small and mid-size operations, a well-designed quality control sheet in a spreadsheet delivers the insight at a fraction of the cost.