Home / Statistical Tools / Control Charts / Control Charts Data Sources

Control Chart Data Source¶

From Excel click...

QXL Stat Tools Tab > Control Charts > any chart

Every control chart reads a rectangular block of cells that you give it in the Data Selection window, and turns that block into named columns. Worked examples of each layout, including subgroups formed three different ways, are on Create Control Charts.

This page covers what each family of charts needs from that data.

What each family reads¶

Family	Charts	What you select	Accepted types
Variable	Xbar-R, Xbar-S, Xbar, R, S, I-MR, I-MR-R/S, Individuals, Moving Range, Levey-Jennings, EWMA, CUSUM, Z-MR	Measurement Columns, the measured values	Continuous, Integer, Count
Attribute	p, np, Laney P', c, u, Laney U'	A count column, plus a denominator	Count only
Rare event	G, T	Data Columns (required), the value for each event	Stated on each chart's page

A Nominal or DateTime column is offered in a measurement list but disabled. If a column holds numbers that were read as text, change its type in the Data Types list on the Data tab rather than editing the worksheet.

Subgroups¶

Only some charts collect rows into subgroups. Where a chart does, its dialog shows a Subgroups group with three ways to form them: a constant number of consecutive rows, a column whose repeated value names the subgroup, or one subgroup per row spread across several checked columns. All three are shown with example data on Create Control Charts.

The smallest subgroup a chart accepts differs by chart, because it depends on what the lower panel plots. A chart that draws a range or a standard deviation cannot use a subgroup of one. Each chart's Options page states its own floor, and a value below it snaps back with no message.

Subgroups do not have to be the same size. Unequal sizes are supported everywhere they are accepted, and they change the appearance of the chart: the control limits step in where a subgroup is large and out where it is small, rather than running as straight lines.

Denominators on the attribute charts¶

The attribute charts plot a proportion or a rate, so they need to know how large each sample was. That denominator is entered in the Sample Size group on the p dialog or the Inspection Size group on the u dialog, and either can be read from a column or fixed as one constant for every point.

A defectives count counts whole units, so a unit with three defects on it counts once, and the count can never exceed the sample size for that row.
A defects count counts defects, so one item may contribute several, and the count may exceed the number of items inspected.

As with subgroups, a denominator that varies from point to point makes the control limits step rather than run straight.

X-axis labels¶

Every chart accepts one optional X-Axis Labels column of any data type. It relabels the horizontal axis and nothing else: it never changes what is plotted, how subgroups are formed, or how the limits are estimated. Left empty, the points are numbered from 1.