Dear Mr.Breyfogle,
I came across with a thread in another board where the user has used mean, standard deviation to analyze the data on Supplier Performance assessment values (Likert Scale 1-10). This triggered my thought and hence the following question.
In your book "Implementing Six sigma" on Page 653, you had suggested to use Sixsigma tools like Normal probability plotting and ANOM to obtain insight on variability for data like Employee Opinion Survey. As indicated in the text, these are "Likert Scales". Hence the data is discrete.
Would you please explain how statistical procedures relevant to Continuous distribution can be used for Likert Scale data?
I incidentally came across with a webpage
http://www.icbl.hw.ac.uk/ltdi/cookbook/info_likert_scale/
that suggest median or mode to measure central tendency and range, intrerquartile range for variability.
Also comparison of groups through Non Parametric methods.
What are the applicable statistical procedures/analysis for Likert Scale Data?
Regards,
GR
Sign In to get started!
The visual standards are
The visual standards are just that the standards. You cannot place the part up to the standard and have the standard tell you if the part is good or bad the person actually is the measuring device.
--------------------------------
deals
epb 5/21/2004
You need to give visual inspectors a Snellin or Jaeger eye test on a yearly basis (near vision test). This should be part of your inspection procedures. Here is an example:
2.0 ABC Company Visual Inspection Requirements
2.1 Vision Test - ABC Company employees that perform Visual Inspections for the Final acceptance of product must pass a one-time-only color vision test and a 12-month near vision test. Vision tests will be performed by a medically qualified personnel. Records of these examination will be retained on file according to [Procedure 4.2] "Documentation Requirements" Individuals must meet the following minimum standards in at least one eye, corrected or uncorrected:
2.1.1 Near Vision: Snellen 14/18 or better, 20/25 or better, Jaeger Type 1 at 14 inches or greater, Ortho-Rated 8 or equivalent.
2.1.2 Rater Color Vision: Average or Normal.
paulirvin 5/18/2004
Do you use a visual standards for acceptance/reject? Is the inspector using those standards? Also, are the inspectors given yearly eye exams, including Jaeger I test? If so, this is your documentation.
In addition, be careful on how you word your procedures. You might want to say something like, "no nicks, scratches, etc. that are readily apparent to the naked eye."
Regardless, this auditor went a little overboard, in my opinion, and I would ask on what element of the standard he based his decision, and then challenge him.
Paul Irvin - President, Irvin Quality Solutions
rw1 5/18/2004
Have had the same problem but not with an auditor. What I recomend is an ongoing solution but it should work.
1. Have the "ompany experts"evaluate a number of parts independantly. I suggest over 30.
2. Compile the results and make sure there is aggreement. If not resolve the discrepancies on all the parts so there is aggreement.
3. Put a training program together on how to identify and rate cosmetic defects. Visual Aids, Pictues, sample parts, Customer Visual Standard, etc (Don't use the original 30+ parts)
4. After training have the trainees rate 5 parts. Give immediate feedback to the trainee.
5. Bring them back in a week or so and go through five more parts, again with immediate feedback. When a trainee aggrees with 5 out of 5 with the experts, they have been calibrated. Give them a T Shirt or Cap with "DD" on it for Defect Detector.
Repeat the process a couple of times a year and you will still have cosmetic defects but not as many!
tisanewt 5/18/2004
laugh. . .and hope to God they were kidding! If not, think about finding a new auditor.
jaimezepeda 5/18/2004
Once one of our external auditors began his opening meeting by stating that "External auditors are not calibrated..."
It kinda makes you wonder how accurate any external auditor's findings can be if they are not calibrated.
shelly 5/18/2004
um, gosh....she isn't measuring she is making a judgement call. Do you do training? Do operators have some sort of examples of good vs bad? I would write up a CA that says that this is not a calibration issue, but a training issue and then state whatever you do to train operators on what to reject, as well as a statement that you will also train people on answering questions about calibration of body parts. sheesh!
bobdoering 4/15/2004
I would agree that the auditor is being pretty rough giving this a minor - and you may want to push back.
But, consider this: visual inspection crosses the line of training and metrology. All audit questions can be boiled down to "How do you know?" questions. In this case, the question can be how does the inspector know they are passing good parts and failing bad parts? Your evidence should include training records - with evidence of effectiveness (ISO standard requirement). Your evidence of effectiveness could very well be "Calibration records". Remember, calibration is determining if the measurement device (human or otherwise) accurately detects the value of a standard. If your visual inspectors can correctly pick out of the lineup of visual standards the specimens that pass and fail, then they are calibrated. If they cannot, then they are "out of calibration". How do you adjust the human gage? Retrain, retest. The results of the retest is your "after" calibration data. Not that hard to figure.
Does this seem extreme? Well, the correct answer for any question - especially metrology - is "it depends". If your records show many rejections from your customer for visual problems, then this could very well be part of your corrective action. On the other hand, of you are just looking for missing parts or features that are readily seen by the naked eye, then it could be more than necessary. Your rejections will be a good gage of the validity of your gaging system – although a little too late.
The bottom line is you need to evaluate the gaging system. The people are a part of the system. The visual standards are just that – the standards. You cannot place the part up to the standard and have the standard tell you if the part is good or bad – the person actually is the measuring device. The standards need calibration, too. Are they worn or dirty from handling? Changes the visual appearance.
Bob Doering
dwade 4/15/2004
I worked in job shop that did a lot of military and aerospace work and we had to have our eyes "certified" for color, depth perception and one or two other things. Some companies who work with color will also have inspection people certified for color contrasting(??).
This could be what the auditor was looking for but I agree with the other posts. You don't calibrate people for visual inspection. You show what reference standards you have for the inspection people to use to make the judgement between and bad product.
Duane Wade
hershal 4/15/2004
1. Push back on the auditor, have him remove the minor. If he refuses, then.....
2. File a formal complaint with both the Registrar and the Registrar's Accrediting Body. The Registrar is REQUIRED to have a formal complaints and appeals procedure, and to let you know what that procedure is upon request.
Oh, and ask for the auditor's qualifications to make a judgment call regarding Metrology. If the auditor is not a trained Metrologist with optics in his background, then he is not qualified (in the accreditation world the term used is competent) to make that judgment.
Hope this helps.
Hershal
ryan 4/15/2004
This sounds like the auditor trying to find a nonconformance for the report. We do extensive manual visual inspection. We only use visual aids with text describing the defect. The other plan we have is to collect different rejects and during training use the rejects for certifying the operator.
baku 4/14/2004
This sounds crazy! We have a vision exam that is given to MPI inspectors, to verify they are not color blind, but as far as 'normal' visual inspection (I am in auto manufacturing, wheel hubs and such, and we deal with a lot of nicks and scratches as well) we just use visual aids as far as what is acceptable and unacceptable. Did the auditor say specifically what requirement you did not follow?