Name: Process Measurements Sample Dataset
Creator: Stratum
License: https://creativecommons.org/publicdomain/zero/1.0/

A time-ordered manufacturing log with deliberate drift and defect counts — the dataset behind the control-chart, normality, and Pareto lessons.

Columns

Column	Type	Description
`Timestamp`	DateTime	Measurement time, YYYY-MM-DD HH:MM, in time order.
`Lot`	Categorical	Production lot id (e.g. L0001).
`Wafer`	Integer	Wafer number within the lot.
`Tool`	Categorical	Processing tool (Tool-A/Tool-B…).
`Recipe`	Categorical	Process recipe (Fine/Coarse…).
`Shift`	Categorical	Day/Swing/Night.
`Thickness_nm`	Numeric	Film thickness in nanometers (the key measured variable, drifts over time).
`Conductivity`	Numeric	Measured conductivity.
`Resistivity`	Numeric	Measured resistivity.
`UnitsInspected`	Integer	Units inspected in the sample.
`Fails`	Integer	Number failing inspection.
`ParticleCount`	Integer	Particle defects counted.
`Disposition`	Categorical	Pass/Rework/Scrap.

Things to try

Catch the drift. Plot Thickness_nm on an I-MR or X-bar control chart, with automatic Nelson-rule flags. → Shewhart control charts lesson
Chart the fail rate. Put Fails over UnitsInspected on a p-chart. → Attribute control charts lesson
Check for normality. See whether Thickness_nm is normal with a Q-Q plot. → Q-Q plots lesson · Normal distribution
Rank the defect causes. A Pareto chart sorts the biggest contributors first. → Dot-strip & Pareto lesson · relate spread to standard deviation

How to open it. Download the CSV, then drag it onto Stratum (or use File ▸ Import). Columns are typed automatically — numbers, the date, and the categories all detected — so you're charting within seconds. Get Stratum →

Process measurements dataset

At a glance

Columns

Things to try

See also