Home / Statistical Tools / Control Charts / Z-MR / Options
Options¶
This page covers what is specific to the Z-MR Chart dialog. The parts shared by every control chart, the Data Selection window, the Data tab, and the analysis pane, are described once on Create Control Charts. What each option means statistically is on the Math Details page.
The Dialog¶
The dialog is titled Z-MR Chart, whatever the ribbon called it, and there is no variant selector above the tabs: this chart always draws both panels, the Z Chart and the Moving Range Chart.
The tabs¶
The dialog is a set of tabs with one set of buttons underneath. Data is always first and always the tab the dialog opens on.
| Tab | What it holds |
|---|---|
| Data | What is charted: the data source, the data types, the columns, and how rows are collected into points |
| Limits & Estimation | How the center line and the control limits are worked out: the sigma method, the limit type and multiplier, historical values, the baseline, any transformation, and the phase splits |
| Rules & Display | The out-of-control tests and what is drawn on the chart |
| Capability | Specification limits and the capability summary. Present only on the charts whose page says so |
Switching tabs never recalculates anything, so you can move between them freely and read your settings back. Only a change you make triggers the rebuild.
Z-MR has three tabs, not four. There is no Capability tab. That is not just a hidden tab: with no Capability tab the capability indices and the histogram are switched off in the engine as well, so no specification limit, target, capability index, or histogram can reach a Z-MR worksheet.
The dialog is also missing several groups that other variable charts carry. None of the following exists anywhere in it: Estimate Limits From, Transformation, Historical Values, the Control Limits group with its limit type radios, its Number of standard deviations (k): box and its manual limits grid, and Split Control Limits. Each one is covered below at the point where you would look for it.
Data Tab¶
The Data tab¶
The Data tab is always the first tab, and it is where you choose what is charted. It holds, from the top:
- The Excel and GroupBy radio buttons, which decide how the rows are divided.
- A red message line that appears when a selection cannot be honored. The text clears itself after about five seconds, so read it when it appears.
- The Data Types list, one row per column.
- The column slot lists, which differ by chart family: what is measured or counted, the optional X-axis labels, and any extra role the chart needs.
- The GroupBy region, visible only in GroupBy mode.
- A metadata line at the bottom describing the column you last clicked.
Every slot list is a small scrolling box of check boxes. Three behaviors are the same in all of them:
- A long column name is shortened with an ellipsis; hover it to see the full name.
- A column that is not available is disabled, not hidden, so you can see that it exists and that something else is using it or that its data type does not fit.
- A disabled check box does not say why it is disabled. The usual reason is that the column is already doing another job on this chart.
The type drop-downs inside these lists ignore the mouse wheel and let the list scroll instead, so scrolling past a drop-down cannot silently change a data type.
On Z-MR the slot lists are Measurement Columns, Part Column, and X-Axis Labels (Optional). See Create Control Charts for what each one does.
There is no Subgroups group on this tab. One worksheet row is one plotted point, so there is no subgroup size, no subgroup identifier, and no subgroups-across-columns mode. Checking more than one measurement column gives you one complete Z-MR chart per checked column, each using the same part column.
Part Column¶
Check the column that identifies which part each row belongs to.
- This selection is required, even though the group header does not say so. With no part column checked, Finish stays unavailable and no message explains why.
- Any data type is accepted, because the values are compared as the text they display on the sheet.
- The list is drawn as check boxes but behaves as a single choice: checking a second column silently unchecks the first.
- One column is checked for you when the dialog opens: the first available candidate, which in practice is the second column of your range, because the first is already taken as the measured column.
- A column used as the part column is not available as the measured column, the X-axis label, or a grouping column.
Two behaviors follow from the part column and will surprise you if you do not expect them:
- Parts are matched on the text you see, not the underlying number. Under a two-decimal cell format, 1.001 and 1.004 both display as 1.00 and are treated as the same part. Widen the cell format if two parts must stay separate.
- The part overrides grid on the options tab is filled from the distinct values of this column. If the column holds more than 100 distinct parts, the grid comes back empty and no override can be entered.
Runs¶
The part column does more than label the points. Quantum XL walks the rows in worksheet order and cuts them into runs, where a run is a maximal block of consecutive rows carrying the same part name. A run ends the moment the part name changes.
The consequence is the one to remember: the same part appearing again later starts a new run rather than joining the earlier one. Sort your sheet so a part's rows sit together if you want that part to be a single run. Runs matter in three places: they are drawn as dividers on the finished chart, they stop the moving range window on the lower panel, and under the By runs sigma mode each one gets its own standard deviation.
Runs are not the same thing as parts. The mean subtracted from every measurement is the part's mean, pooled across all of that part's rows wherever they are in the sheet, so two runs of one part are always centered on the same number.
One column, one job¶
A column can hold only one role on the chart, and up to six roles compete for it:
| Role | Where you choose it | Present on |
|---|---|---|
| The measured column | Measurement Columns, Data tab | every variable chart |
| X-axis label | X-Axis Labels (Optional), Data tab | every variable chart |
| Split column | Split Control Limits, on the second tab | every variable chart except Z-MR |
| Grouping column | GroupBy Order, Data tab | every variable chart |
| Subgroup ID | Subgroup ID column:, Data tab | the charts that offer the Subgroups group |
| Part | Part Column, Data tab | Z-MR only |
Claiming a column for one role disables it everywhere else. It does not move and it is not unchecked: you have to release the first role before the column becomes available again. That is why a column can appear greyed out in a list for no visible reason.
GroupBy is the one exception. Moving a column into GroupBy Order actively takes it away from the other roles: if it was the measured column, the X-axis label, the subgroup ID, the part column, or the split column, that selection is cleared. Everywhere else the rule is disable, here it is evict.
Two messages can appear in the red line when a claim conflicts:
- "'{0}' is already in GroupBy. Remove it from GroupBy first."
- "'{0}' is already selected as a Data Column. Deselect it from Data Columns first."
Releasing a role does not always make the column usable again in the measurement list. The data type filter still applies, so a column that has been retyped to Nominal stays greyed there even after nothing else is using it.
Finish¶
Z-MR is the one variable chart that will not let you press Finish with nothing selected. Finish is available only when at least one measurement column and a part column are checked. No message explains which of the two is missing.
Limits and Estimation Tab¶
This tab carries two groups on Z-MR, and only two: Sigma Estimation and Part Overrides.
Sigma Estimation¶
The group is headed Sigma Estimation and it opens with the label Estimate sigma: above four radio buttons. Exactly one is selected at a time, and the whole group applies to the standard deviation that each measurement is divided by. It never changes the mean that is subtracted, which is always the part's own mean.
| Control | Type | Default | What it estimates |
|---|---|---|---|
| Constant | radio button | selected | One standard deviation for the entire chart, estimated across every row |
| Relative to size | radio button | not selected | The same single chart-wide estimate, but computed after taking the natural logarithm of the data |
| By parts | radio button | not selected | One standard deviation per part, so every row of a part is divided by that part's own estimate |
| By runs | radio button | not selected | One standard deviation per run, so a part that appears twice can be divided by two different estimates |
An italic note under the four radios states that Relative to size takes the natural logarithm of the data, so all values must be positive. That is a hard requirement, not a suggestion: a zero or negative measurement stops the analysis under this mode. Marking the offending value as an outlier does not rescue it, because the positivity check is made before the outlier exclusions are read.
Under Relative to size everything downstream lives on the logarithmic scale: the part means, the deviations, the estimated standard deviation, and the plotted Z values are all computed from the transformed measurements. It is not the same thing as the Transformation group on other charts, and it is not offered as one.
Next comes the moving range statistic, which decides how the set of moving ranges behind the estimate is reduced to a single number. It is labelled Moving-range statistic:.
| Control | Type | Default | What it uses |
|---|---|---|---|
| Average moving range | radio button | selected | the mean of the moving ranges, divided by the \(d_2\) constant |
| Median moving range | radio button | not selected | the median of the moving ranges, divided by the \(d_4\) constant, which is much less disturbed by one large jump |
This choice also changes the Moving Range Chart panel's own center line and limits, because those lines are the constants for the method you picked. Switching from average to median lowers the whole band on that panel even when the data have not changed.
Last is the window length, labelled Moving-range length (L): with an italic hint under it reading "2 to 100; default 2".
| Control | Type | Default | Accepts |
|---|---|---|---|
| Moving-range length (L): | text box | 2 | a whole number from 2 to 100 |
Validation: a value below 2 or above 100 is refused and the box snaps back to the value it held before, when it loses focus. There is no message, so it is easy to believe an entry was accepted when it was discarded. There is no Nelson moving-range correction checkbox on this chart, and no unbiased-constants checkboxes: the unbiasing constant for the method you chose is always applied.
Estimate Limits From is not offered on this chart¶
There is no baseline control in this dialog and no way to freeze the limits on an early period. The limits always come from every point in the chart. If you need a frozen baseline, use one of the charts that offers it.
Transformation¶
A transformation replaces each measurement with a rescaled version of itself, chosen so the rescaled values sit closer to a normal distribution. Control charts assume a normal distribution and never fit any other one, so a transformation is the only remedy the product offers for skewed measurement data on a control chart.
| Charts | Transformation group box |
|---|---|
| Xbar-R, Xbar-S, I-MR, I-MR-R/S, EWMA, CUSUM | Offered |
| Levey-Jennings, Z-MR | Not offered. There is no control, and no transformation is applied. |
| p, np, Laney P', c, u, Laney U', G, T | Not offered. These charts plot counts, proportions, or waiting times and have their own distributions built in. |
There is no Historical Values group either. You cannot supply a known mean or a known standard deviation for the chart as a whole. What you can supply is a mean or a standard deviation for one named part, in the grid below.
Part Overrides¶
The group is headed Part Overrides and holds a grid with three columns, Part, Override Mean, and Override Sigma. The Part column is read-only text. The other two are text boxes, one row per distinct value found in the part column, and every box starts blank.
Blank means estimate it from the data, which is what the whole grid does until you type in it.
The two override fields are independent. Filling in Override Mean and leaving Override Sigma blank uses your mean and the computed standard deviation for that part. Filling in Override Sigma and leaving Override Mean blank does the reverse. There is no requirement to supply both, and a row where both are blank behaves exactly as if the grid were not there.
| What a row contains | What that part does |
|---|---|
| Both boxes blank | Both values are computed from the data |
| Override Mean only | Your mean is subtracted; the standard deviation is computed |
| Override Sigma only | The part's computed mean is subtracted; your standard deviation is divided by |
| Both filled | Neither value is estimated for that part |
Validation: an Override Sigma must be a finite number greater than zero. A blank, a non-numeric entry, zero, a negative number, or a non-finite value is refused and the box snaps back to what it held before, with no message. An Override Mean must be finite, and is otherwise unrestricted, including zero and negative values.
Two behaviors are worth knowing before you rely on the grid:
- Overrides are allowed under Relative to size, and they are read on the logarithmic scale like everything else under that mode. An Override Mean you type is the untransformed center for the part and Quantum XL takes its logarithm for you. An Override Sigma you type is used exactly as entered, as a spread on the logarithmic scale.
- The grid is filled from the distinct values of the part column, so it is empty until a part column is selected, and it repopulates when you change that selection. The empty-state message reads "No parts found. Select a part column on the Data tab." Values you have typed are kept by part name across a refresh. If the part column holds more than 100 distinct parts the grid comes back empty and no override can be entered, which is worth checking early, because many parts is the normal case for a short-run chart.
Split Control Limits is not on this tab¶
The Split Control Limits group does not exist in the Z-MR dialog. There is no split mode, no break point list, and no split-by-column drop-down.
Dividers on a Z-MR chart are not split control limits
A Z-MR chart draws dividers with labels beside them that look exactly like the ones a split produces, and they are not this feature. They mark part runs, a run being a block of consecutive rows with the same part name, and they are derived from the part column in your data. They are not editable, they follow no setting, and the Z-MR dialog has no Split Control Limits group box at all. The same part appearing again later starts another run and gets another divider. The limits do not step at these dividers.
Rules and Display Tab¶
Out-of-Control Rules¶
The group box headed Out-of-Control Rules holds one row per rule. Each row is a check box carrying the rule's description, followed by up to two small numbered boxes labelled N = and M =. N is how many points the rule needs, and M, where a rule has one, is the length of the window those points must fall inside.
Ticking a rule adds its test to the analysis. Nothing else about the chart changes: the center line, the limits, and the plotted points are the same whether a rule is on or off.
Thresholds. Every box arrives pre-filled with that rule's standard value rather than blank, so you can see what the rule will do before you change it. Clearing a box means use the standard value; it does not mean zero. Unticking a rule leaves its numbers visible but locks the boxes, so the value you set is still there when you tick the rule again.
There is no message when a threshold is refused. On the attribute and rare-event dialogs a count below 2 is silently raised to 2 as soon as you leave the box. On the variable chart dialogs a value that is blank, zero, or negative is treated as the standard value.
The rule set on the variable charts¶
Eight rows, and all eight are ticked by default. This is the opposite of the attribute charts, so a chart of the same data can flag very differently depending on which family of chart you chose.
| Row label as shown | Parameters and defaults | Default |
|---|---|---|
| Rule 1 - Beyond K sigma from center | none, see the note below | On |
| Rule 2 - N consecutive points on same side of center | N = 9 | On |
| Rule 3 - N consecutive increasing or decreasing | N = 6 | On |
| Rule 4 - N consecutive alternating up/down | N = 14 | On |
| Rule 5 - N of M beyond 2 sigma (same side) | N = 2, M = 3 | On |
| Rule 6 - N of M beyond 1 sigma (same side) | N = 4, M = 5 | On |
| Rule 7 - N consecutive within 1 sigma | N = 15 | On |
| Rule 8 - N consecutive beyond 1 sigma (either side) | N = 8 | On |
Rule 1 has no parameter box of its own. It tests against the drawn control limits, so its width is the sigma multiplier from the Control Limits section.
What each rule actually tests, and which points of a run get marked, is defined on Out-of-Control Rules.
Z-MR shows all eight rows, and all eight are on when the dialog opens.
Rule 1 on this chart means "outside plus or minus 3 on the Z panel", because the Z panel's limits are fixed at plus and minus 3 and there is no Number of standard deviations (k): box in this dialog to change them. That is also why the caution about the sigma multiplier moving the zone lines, which applies on the other variable charts, has nothing to act on here.
Which panels the rules run on¶
Every panel is checked, the dispersion panels included. On a two-panel or three-panel chart the range, standard deviation, and moving range panels each run their rules over their own plotted values, against their own center line and their own limits.
Rules 5 through 8 are the exception. They read the zone lines and assume a symmetric spread around the center line, which a range or a standard deviation does not have, so they run on the location panel only.
| Panel | Rules that run there |
|---|---|
| Xbar, Individuals, or EWMA (the location panel) | Every rule the chart offers |
| Range, standard deviation, moving range | 1, 2, 3, 4 |
| Z-MR moving range panel | 1, 2, 3 |
| Single-panel charts | Every rule the chart offers |
What the group box does about it depends on which panels the chart is drawing:
- On a variant that draws a dispersion panel only, such as R only, S only, or Moving Range only, the rows for rules 5 through 8 are greyed out along with their threshold boxes. Greying them does not clear them: whatever you had ticked is still ticked if you switch the panel view back.
- On a chart that draws the location panel next to a companion panel, those rows stay available and an italic note appears under the list reading "Rules 5-8 only apply to" followed by the name of the location panel.
On Z-MR the location panel is the Z panel and the companion is the Moving Range panel, and the note under the rule list reads "Rules 5-8 only apply to Z".
The moving range panel runs rules 1 to 3, not 1 to 4
The note under the list warns about rules 5 through 8. The moving range panel of a Z-MR chart is narrower than that: the engine runs rules 1, 2, and 3 there and nothing else. Rule 4, the alternating up and down test, stays ticked and is not greyed out, and it never fires on that panel. It still runs on the Z panel. If you want the alternating test on the moving ranges, there is no setting that will give it to you.
The chart does not tell you which rule fired
Every rule you tick is tested, and the results are combined into a single verdict per point: this point broke at least one enabled rule, or it did not. Points that broke one are drawn in a different color from the rest, and that is the whole of the report. There is no per-rule label, no per-rule marker, no per-rule column, and no list of violations anywhere in the chart or the worksheet. A point that breaks four rules at once looks exactly like a point that breaks one, and the output never names either.
So when you turn several rules on, read a marked point as an instruction to go and look at the chart, not as a diagnosis. If you need to know which pattern caused a particular signal, the practical method is to switch the other rules off and rebuild the chart.
Two further behaviors apply to the rules on this chart. Splitting the run counters is automatic rather than optional: every part run is a phase, so a run in progress when a new part starts is abandoned and cannot combine with the run that follows. And a blank point breaks a count rather than bridging it, which on the moving range panel includes the unplotted points at the start of every run.
Display Options¶
The Display Options group box holds two features: the sigma zone lines, and reference lines of your own at Y values you choose. It is the same group box on every control chart. The variable charts take it from a shared control, and the p, np, Laney P', c, u, Laney U', G, and T dialogs repeat the same block with the same labels.
- Show 1-sigma / 2-sigma zone lines: draws four light grey lines inside the control band, two above the center line and two below, dividing each side into three zones. The checkbox is cleared by default.
The zone lines are always built into the chart. This checkbox only decides whether they start out visible, so they can be shown or hidden later from the Analysis Pane without running the analysis again.
Where the lines are placed is not the same on every chart, and on a variable chart drawn with a sigma multiplier other than 3 the names 1-sigma and 2-sigma do not describe where the lines actually sit. Since several of the out-of-control tests measure a point against these zones, this changes what those tests count. The placement rule for each chart family is on Control Limits and Zones.
On a Z-MR chart the placement is the straightforward one. The Z panel's band is fixed at plus and minus 3 around a center line of 0, and the one-sigma width is the upper half band divided by that same 3, so the line labelled 1-sigma really does sit at one full standard deviation, at \(Z = \pm 1\), and the 2-sigma line at \(Z = \pm 2\). Rules 5 through 8 therefore keep their textbook meaning here, with no multiplier to distort them.
Reference lines¶
A reference line is a horizontal line at a Y value you supply: a specification limit, a target, a contract threshold, anything you want the points compared against. Until you add one the editor reads "No reference lines configured. Use the row editor below to add Y-value reference lines."
Each reference line is one row with four editable cells and a remove button:
| Cell | What it sets | Default |
|---|---|---|
| Value | the Y position of the line | 0 |
| Label | the legend entry for the line | blank |
| Color | the line color, picked from a color button | red |
| Thickness | the line weight, chosen from 0.5, 1.0, 1.5, and 2.0 | 1.0 |
| X | removes that row. Its tooltip reads "Remove this reference line" | n/a |
Value is not validated. Type something that is not a number and the cell keeps the value it had. A blank Label keeps the line out of the chart legend; the line is still drawn, under a generated name built from its value. The editor scrolls on its own once the list of rows is taller than the space allowed for it.
- Add Reference Line: appends a blank row and refreshes the chart preview by itself. Removing a row with its X button does the same.
- Apply: commits edits to Value, Label, Color, and Thickness, which do not refresh the preview on their own. Its tooltip reads "Refresh the chart preview to reflect any pending reference-line edits". The button appears only once at least one reference line exists, and does nothing when there is nothing outstanding to commit.
So adding and removing lines is immediate, while editing an existing line needs Apply.
Reference lines are drawn on the location panel only
Reference lines appear on the panel that plots the measurement itself: the Xbar panel, the Individuals panel, the EWMA panel, or the Z panel of a Z-MR chart. The dispersion panels (Range, Standard Deviation, Moving Range) are in different units, so they never carry reference lines. The location panel's Y axis is stretched to include every reference line value, so a line outside the data still shows.
This has a consequence on the single-panel dispersion charts. R only, S only, and Moving Range only draw no location panel, so a reference line entered in the editor never appears anywhere on the chart. The editor still accepts it and still lists it. If you need a reference line against a range or a standard deviation, there is no dialog option that will draw one.
CUSUM is the exception: it is a single overlay chart rather than stacked panels, and it draws reference lines on that chart.
On Z-MR the location panel is the Z panel, so a reference line is drawn there and never on the Moving Range panel. Remember that the Z panel is in standardized units: a reference line at a specification limit expressed in the measurement's own units will land in the wrong place, because the plotted values are not in those units.
Layout you do not choose¶
Nothing else about the appearance of the chart is a dialog option. These are fixed:
- Which panels are drawn comes from the ribbon entry you started from, not from a checkbox in the dialog. Choosing R rather than Xbar-R is how you get the range panel on its own.
- Panel titles are composed for you from the panel type, the group, and the measurement column name, with empty parts dropped.
- Plot order is fixed, and it is also the front-to-back order: the data line, then the zone lines, then the upper and lower control limits, then the center line, then the reference lines. The zone lines are added before the limits deliberately, so they stay in the background.
- Axes are set for you. The X axis is numbered by plotted point and titled Subgroup. The Y axis is scaled to fit every plotted value, including the zone lines and, on the location panel, the reference lines; dispersion panels are held at or above zero. When a chart has more than one panel the panels are aligned with each other so their points line up vertically.
Z-MR has one ribbon entry and no variants, so the two panels, the Z Chart and the Moving Range Chart, are always both drawn. Their Y axes are titled Z Score and Moving Range of Z.
Once the chart exists, the zone lines can be shown or hidden from the Analysis Pane without running anything again. Reference lines cannot: they are not in the pane, so to add, edit, or remove one you modify the analysis and run it again.
Task Pane¶
The Control Chart pane is the same on every chart and is described on Create Control Charts. On a Z-MR sheet it binds to the worksheet named ZMR, and both panels are redrawn together by every pane setting.
Split Control Limits has no effect on this chart
The pane's Split Control Limits section is available on a Z-MR chart and Apply Splits will run, but break rows you enter do not change the chart. Z-MR works out its own phase boundaries from the data, one phase per run of the same part, and those replace anything entered here. The Z-MR dialog has no split section for the same reason.