Home / Statistical Tools / MSA / Type 1 Gage Study / Options

Options¶

Every control on the Type 1 Gage Study dialog. The dialog has two tabs, Data and Options.

Worked example

The How-To page runs a complete study from sample data and prints the numbers you should see. Start there if you would rather read a finished report than a list of settings.

Data tab¶

Two radio buttons at the top of the tab choose where the data come from:

Excel reads the range you selected on the worksheet.
GroupBy reads the same range but splits it into groups first, and runs a separate complete study for each group.

Data Types¶

Each column in the selection carries a data type, shown beside its name. A Type 1 study will take a measurement column typed Continuous, Integer or Count. Count is offered so that a whole number readout is selectable without your having to retype its data type first. A column of any other type is not offered as a measurement column.

Data Columns¶

Tick the column holding the repeated measurements of the one part. You may tick more than one, and each ticked column becomes its own separate study with its own chart, its own table and its own reference value and limits. The studies do not share an axis and are never overlaid.

This list is headed Data Columns rather than Measurement Columns, which is what the Crossed, Nested and Extended dialogs call theirs. It is the same idea under a different heading.

GroupBy¶

In GroupBy mode, the columns you move into GroupBy Order split the measurements into groups, and each group becomes a study of its own. The group name appears in the study title and in the run chart title.

Options tab¶

Per Measurement Column¶

A grid with one row for each measurement column you ticked, headed Column, Reference Value, LSL, USL and Resolution. The first column is the column's name and cannot be edited; the other four are yours to fill in.

Everything here is per column, which is the point of the grid: two measurement columns can be two different characteristics with two different specifications, and each gets its own row.

Box	Leave empty and	Fill it in and
Reference Value	no bias, no T, no P value and no Cgk	those four are computed, and the run chart centres on it
LSL and USL	no tolerance, so no Cg, no Cgk, neither percent of tolerance figure, no percent of resolution and no limit lines	all of those are computed
Resolution	no percent of resolution and no acceptance verdict	both are computed

Every box may be left empty. None of them is required, and an empty box is not an error.

Both specification limits or neither

A tolerance is the upper limit minus the lower limit, and it is computed only when you have given both. One limit on its own buys nothing, and nothing is substituted for the missing one. The sheet says so with a note under User Supplied Data.

Unticking a measurement column discards that row's typed numbers. Tick the column again and the row comes back empty, so retype the reference value and the limits rather than assuming they survived.

Study Options¶

Three boxes, each of which starts at a default the calculation engine supplies.

Percentage of Tolerance:

: The fraction of the tolerance the study is judged against, 0.2 by default. It is a fraction on both sides of the exchange: the box reads 0.2, not 20. It sets the reference span on the run chart and is the numerator of Cg and of % Repeatability and Bias.

Number of Standard Deviations:

: The multiplier that turns the standard deviation into the study variation, 6 by default. It also names the row on the report, which reads 6*StDev at the default and follows whatever you type.

Alpha for the bias test:

: The significance level the bias test is judged against, 0.05 by default. It decides when the P value is printed in red, and it is the number quoted in the note under the Bias Calculations block.

Two behaviours are worth knowing about these three boxes.

Clearing a box restores its default. Empty one, leave it, and the engine default comes back and is repainted into the box. You never send a blank.

Nothing is clamped. A percentage of tolerance of 1.5, or a multiplier of 3, reaches the engine exactly as typed. Values outside the ranges the mathematics allows are not silently corrected; they stop the study and print the sentence that says why (see When a study is refused).

Gage Information¶

Seven free text boxes recording what the study was of and who ran it: Gage Name:, Gage No.:, Gage Type:, Part Name:, Part No.:, Date: and Performed By:.

Nothing here is parsed or checked, and nothing here changes a number. Date: is free text on purpose, so "week of 3 March" is as acceptable as a date. All seven are written to the Gage Information block on the report, and all seven rows are printed whether or not you filled them in.

Save as Default Settings¶

Save as Default Settings stores three settings only: the percentage of tolerance, the number of standard deviations and alpha. They are read back the next time you open the dialog, on this machine, for every Type 1 study.

Nothing else is saved, and that is deliberate. The reference value, the specification limits, the resolution and the seven gage details belong to one part on one gage, not to the next study. They are stored on the worksheet instead, so Modify brings them back for that study.

On a machine where nothing has been saved, all three settings come from the calculation engine's own defaults.

When a study is refused¶

Some conditions stop a study before any number is produced. The sheet prints the heading This study could not be computed in place of the five sections, followed by one of these sentences. Other studies in the same run are unaffected.

Sentence on the sheet	What caused it
This study needs at least 2 usable measurements. Select a range with more measurements of the same part.	Fewer than two readings could be used
Every measurement in this study is the same value, so there is no variation to analyze.	Every reading is identical, so the standard deviation is zero
The percentage of tolerance must be greater than zero.	Percentage of Tolerance: is zero, negative or not a number
The number of standard deviations must be greater than zero.	Number of Standard Deviations: is zero, negative or not a number
Alpha must be greater than zero and less than one.	Alpha for the bias test: is outside that range
The resolution must be greater than zero. Clear the box if the gage resolution is not known.	A Resolution was typed but is zero or negative
These measurements are too large or too close together for a standard deviation to be computed at double precision.	The readings differ, but not by enough for the arithmetic to hold them apart

Two more messages replace the whole worksheet rather than one study:

No data was found for analysis. when the run produced no studies at all.
Too many results. Please reduce the number of groups. when GroupBy would make more than 100 studies. Refused studies count towards that total.