Home / Statistical Tools / MSA / Create MSA Template

Create MSA Template¶

Create MSA Template builds a blank worksheet laid out for a measurement system analysis: the columns in the right places, the gage and part information fields, space for the specification limits, and printed instructions for whoever takes the measurements.

Find it at QXL Stat Tools > MSA / Gage R&R > Create MSA Template (Optional). A workbook must be open; nothing needs to be selected.

Use it before you collect data. Once the sheet is filled in, run the analysis on it directly.

When to use it¶

Use it when you are planning a study and want the data collection sheet laid out correctly, rather than building the columns by hand. It is optional: the analysis tools read any worksheet with the right columns, however it was made.

The advantage of a template is that the analysis can be launched straight from it. The template records the specification limits and gage information under defined names, and the analysis reads them back, so those do not have to be re-entered.

A template written by the previous version of Quantum XL runs unchanged. The first template on a sheet uses the same defined names the older version used, so a one-template sheet made by either version is the same sheet, and the v18 analyses read it without conversion. A second and any later template on the same sheet is numbered, which is necessary because a worksheet can hold each name once.

What it produces¶

Finish adds a new worksheet named MSA Template, placed after the sheet you were on (a second one is named MSA Template (2), and so on), and writes one template block on it for each tab in the dialog. Each block contains:

a title cell naming the analysis, for example MSA Template 1: Gage R&R: XbarR, EMS, and REML
a Gage Information area: Gage name: (filled in as My gage), Gage No.:, Gage type:, Part name:, Part No.:, Characteristics: and Specifications:, and beside them, on a quantitative template, USL:, LSL:, Date: (filled in with today's date) and Performed by:
printed instructions under the heading How to use this template, describing what to do in numbered steps
a data table sized to the number of operators, parts and trials you asked for, with one column per operator trial

With more than one tab, each block is written below the previous one on the same new sheet.

The instructions it writes¶

The instructions are printed on the sheet next to the table, so the person taking the measurements does not need the help system. For a quantitative study they read:

Enter the gage and part information.

Enter the USL and LSL if known (needed for %Tolerance).

Have each operator (appraiser) measure each part in random order; type each result under that operator's trial column.

When the table is full, run the MSA analysis on this template.

Running the analysis from a template¶

Crossed and Nested skip the analysis dialog entirely. Launch either one with a template on the sheet and a small MSA Analysis Method dialog appears in its place, because the template has already supplied everything the Data tab would ask for. It is shown once for the whole launch, however many templates are on the sheet.

An Extended analysis opens its full dialog instead, prefilled from the template. The Model tab is where additional factors and nesting are declared, and a template cannot declare them.

The MSA Analysis Method dialog¶

Estimation Method offers the same four choices as the analysis dialog's own group, spelled the same way, and described on Crossed > Options:

Automatic (balanced: EMS; unbalanced: REML)
Restricted Maximum Likelihood (REML)
Expected Mean Squares (EMS) - Equivalent to ANOVA when balanced
Xbar and Range (XbarR)

Xbar and Range (XbarR) is offered only when it can run on every template about to run. Where it cannot, the radio button is disabled and the reason is shown beneath it.

Specification Limits is a grid with one row per template about to run, under the headings Template, LSL and USL. Every row arrives prefilled from that template's own limit cells and can be edited here. A blank limit means the template has no limit on that side, and where both are given the lower must be strictly below the upper. A limit you change is written back into the template's cell, so the sheet records what actually ran; a limit you leave alone is not touched.

Display Confidence Intervals is ticked by default. Clearing it drops the confidence interval table and the interval that otherwise follows the distinct-categories count.

Display gage performance curve is cleared by default, so a template run draws the curve only when you ask for it. It is also disabled, with Not available when part variation is not representative of the population. beneath it, while the box below is ticked: the curve is drawn from the part to part standard deviation, and declaring that variation unrepresentative leaves nothing to draw it from. Your tick is kept while the box is disabled, so clearing the box below gives the choice back.

Part variation isn't representative is the same display-only choice the analysis dialog's Options tab offers as Part Variation is Not Representative of the Population, described on Crossed > Options.

Cancel abandons the whole launch and no template is run.

Where the launch leaves you. Every template on the sheet runs in the one launch, and it ends on the last results sheet it wrote, so that is the sheet in front of you when the summary box closes. When no template ran at all you are left on the sheet you started from.

The part count is read from the sheet, not remembered. The analysis counts the parts from the rows the template's grid actually covers rather than from the number of parts you asked for when the template was made, so a grid that has grown is read at its new size.

A template anchored on another worksheet is ignored. Its defined names are treated as absent, so a launch reads only the templates on the sheet you ran it from.

A sheet with no template opens the ordinary analysis dialog, exactly as on any other worksheet.

When a template cannot run¶

Situation	What happens
Fewer than 5 measurements filled in	Template 1 does not have enough filled measurements yet (at least 5 are needed; 3 found). With several templates on the sheet, that one is skipped and the rest run
Nested, and fewer than two operators have any data	Template 1: a nested analysis needs at least two operators (appraisers) with data; this template has fewer. Skipped the same way when other templates can run
A blank measurement cell	It is a missing measurement, not a zero. The rest of the template still runs

Messages about skipped templates are collected and shown together in one MSA Template box when the launch ends.

Attribute templates¶

The Data Format choice offers Qualitative (Attribute) as well as Quantitative (Variable). An attribute template collects ratings rather than measured values, and it is read by the attribute tools further down the MSA / Gage R&R menu rather than by Crossed, Nested or Extended.

Which of the two attribute templates you get depends on your reference values, and the choice is made for you rather than asked:

Data Format	Reference values	The template built
Qualitative (Attribute)	none, or in qualitative format	for Attribute MSA: Crosstabulations
Qualitative (Attribute)	in Quantitative (Variable) format	for Attribute MSA - Analytic Method

The title cell names which one you got, reading MSA Template 1: Attribute MSA: Crosstabulations Method for the first of them.

The Crosstabulations template¶

Alongside the usual gage details it carries a Data type: cell, a dropdown offering Binomial, Nominal and Ordinal, which starts at Binomial. The analysis reads that cell, so the data type is decided on the sheet rather than in the dialog.

The grid is headed Part down the left, then one Operator (Appraiser) 1, Operator (Appraiser) 2 and so on across the top, each spanning its Trial 1, Trial 2 columns, the same headings a quantitative template uses. Typing a real name over a heading is honoured: the report uses the name you typed, and falls back to Operator (Appraiser) 1 and so on only when the heading is left blank.

The instructions block, headed How to use this template, reads:

Enter the gage and part information.
Choose the data type: Binomial for two levels such as pass and fail, Nominal for three or more unordered levels, Ordinal for three or more ordered levels. A sentence follows when the template has a reference column: Reference entries use the same labels as the ratings.
Have each appraiser rate each part in random order; type each rating under that appraiser's trial column.
When the table is full, run Attribute MSA - Crosstabulations Method on this template. The order of Ordinal levels and the conforming level are set in its dialog.

There are no Accept and Reject cells on this template. The levels are simply the distinct values you type, and which of them counts as conforming is chosen in the analysis dialog, where it is also guessed from the words you used.

Running the wrong kind of template¶

The two template kinds are not interchangeable. Running Attribute MSA - Crosstabulations Method on a sheet whose only template is a quantitative one stops with one message box rather than opening a column picker: The MSA template on this sheet is for continuous (measurement) data. Use Crossed, Nested or Extended MSA on it, or create an attribute template with Create MSA Template.

The other way round is refused the same way. Running Crossed, Nested or Extended on a sheet whose only template is an attribute one shows: The MSA template on this sheet is for attribute data. Use Attribute MSA - Crosstabulations Method on it, or create a continuous template with Create MSA Template.