LOADING THE COMPUTER SOFTWARE

As we already have mentioned, most of the statistical analysis nowadays is being done with some kind of software package. Measurement analysis is no different. Before actually conducting one, the plant's SPC facilitator or computer system coordinator should be contacted for a copy of the GAGE program and for information concerning the specific command needed to gain access to the program. The access codes and procedures are unique to every personal computer and must be obtained from the plant's SPC facilitator, quality engineer, or computer coordinator .

There are many computer software programs on the market that are suitable for conducting measurement studies. Here, we provide a very simple one with minimum outputs of measurement to demonstrate the rationale and the flow of all measurement software.

The following command will load the GAGE.BAS program, which is used in many manufacturing facilities on most IBM personal computers. (Other advanced programs may be MINITAB, SPSS, SAS, and so many others. Obviously, their commands may be different, but the results will be much more comprehensive and more complete than the one presented here.)

Underlined words and commands must be entered by the analyst.

BASIC GAGE

When the software has been successfully loaded, a title screen will be displayed on the computer video screen.

INPUT THE DATA

The program will allow data to be read from a saved file or input from the keyboard. The program expects to read data from a file when the analyst types "F" and from the keyboard when the letter "K" is typed.

When the data are read from a file, the program will try to read a data file that is stored on a floppy disk in Drive A. The correct floppy disk should be in Drive A before trying to read data from a file. The program will ask for a file name .

The name of the data set should be entered without using the three-letter extension. Below are the responses that were used to read data from the file MOD27.DAT that was on a floppy disk in Drive A.

INPUT DATA FROM {F}ILE OR FROM {K}EYBOARD?	F <Enter>
ENTER NAME OF DATA FILE?	Mod27<Enter>
DATA INPUT COMPLETE

When data are entered from the keyboard, the program requests several lines of descriptive information. This information is used to label the output of the program and to control the organization of the data entry prompts. The responses in Table 15.3 were used to label the data.

Table 15.3: Responses from Computer Software

Part name?	Name <Enter>
Part number?	Name <Enter>
Gage description?	Name <Enter>
Characteristic measured?	Name <Enter>
Number of parts (1 “60)?	10
Number of repeated measures by each appraiser (1 “9)?	3
Number of appraisers/gages used (1, 2, 3, 4)?	3

The data file may be built by entering the data from across the rows or down the columns of the data collection sheet (see Figure 15.10). The program asks whether data are to be entered by measurement trial or by part. The program will display the measurement that is expected (that is, what was actually entered).

The program signals the end of the data-building routine by printing DATA INPUT COMPLETE on the computer screen.

The specification limits for the measured characteristics are requested by the program. The specifications should be entered by using the same scale that was used for the data entry. A grossly incorrect analysis will result when the data are coded but the specification limits are not.

If a one-sided (unilateral) specification is defined for the characteristic, two options are available to the analyst. If zero is a reasonable measurement, it may be entered for the lower specification limit. In this case, the specification interval is calculated as the distance between zero and the specification limit (USL - 0). If zero is far from the measured data range, the analyst may respond to the requested specification with a blank return (<Enter> only). In this case, the program calculates the specification interval as twice the distance from the grand average of all the study data to the specification limit [2 x (Xbar - Spec)]. The following responses were used for the data in this example:

Upper spec?	75 <Enter>
Lower spec?	<Enter>

PROOF, EDIT, AND SAVE THE DATA

When all the data are entered into the computer, the following series of prompts is displayed on the screen:

[D]isplay [S]ave [E]dit [C]alculate D (enter)

It is always a good practice to check the data file for errors that may have occurred during the entering process before analyzing or saving the data. The data set will be placed on the screen if the letter D is typed as the response to the prompt. The data for each appraiser are displayed on the screen at one time. If after checking the data everything is OK, then by using the F response at the bottom of the computer screen, the display session will be ended. Next, proceed with the Save command and then the analysis.

Caution  

If you are to save the data on a floppy disk, make sure that a formatted disk is in the drive before attempting to save the data. The program will ask for a name for the data file. Enter a name of eight or fewer characters . The program will automatically add the extension ".DAT" to the file name. Do not add this extension; let the program do it.

CALCULATE MEASUREMENT VARIABILITY

The program will calculate the measurement variability and the P/T ratio after C is typed in response to the prompt. As the computer is calculating the measurement variability, a prompt on the screen will display the message: CALCULATING MEASUREMENT VARIATION.

The output from the program will appear on the screen as in Table 15.4.

Part name:

Part number:

Gage description:

Characteristic measured:

Standard deviation of measurement error:

Due to repeatability :

Due to reproducibility :

Total: Upper spec:

Lower spec:

Spec interval:

Measurement variability:

INTERPRET THE FINDINGS

Measurement error studies are conducted as a means of quantifying and evaluating the magnitude of measurement system error. The proportion of the product tolerance used by measurement system error is described by the P/T ratio. This value compares the spread of error inherent to the measurement system with the product tolerance. The variation due to the measurement system should be smaller than the product tolerances. So what are you looking for in the output of the program? The following decision rules are usually used to evaluate the magnitude of measurement error:

• Measurement system error under 10% of the product tolerance is an acceptable level of measurement error.

• Measurement system error of 10 to 30% of the product tolerance may be considered acceptable. The level of acceptability should be based on the importance of the measured characteristic, cost of incorrect machine adjustment decisions, cost of the gage, or cost and/or time needed for gauge repairs .

• Measurement system error greater than 30% of the product tolerance is unacceptable. Every effort should be made to reduce the level of measurement error.

If the percentage of the product tolerance used by measurement system error is less than 10%, the study of the measurement system should be continued to provide further insight to the nature of measurement error. The basic measurement error study is only a one-point-in-time study. This is a good screening study that will identify poor measuring systems. The basic measurement error study does not provide information concerning the stability of the measuring system. The study should be extended so that additional measurement trials are made. Additional trials are needed for the assessment of the measurement system stability.

Measurement systems that yield unacceptable levels of measurement error must be diagnosed and improved. If the percentage of the product tolerance is unacceptably large for the basic measurement error study, the system will demonstrate even greater levels of error over a longer period of time.

If the error due to repeatability (%RPT error) is significantly larger than that due to reproducibility, the following may be required:

• Maintenance or repair of the measurement hardware (gage, scale, meter, and/or fixtures)

• Training for all appraisers related to the measurement procedure

• Change of the measurement strategy (replace an analog readout with a digital readout as a means to avoid rounding errors)

If the error due to reproducibility (%RPD error) is significantly larger than that due to repeatability, the following may be required:

• Standardization of the calibration technique used by the appraisers

• Training for specific appraisers related to the standard measurement procedure

• Standardization of where on the parts measurements are taken