Which tool to use?

Q: Can we use the quality control tools cause-and-effect diagram, tree diagram, affinity diagram, relations diagram and fault tree analysis interchangeably in a problem situation involving multiple root causes? Some quality professionals suggest fault tree analysis is best suited for all such situations.

I’m also wondering whether all root cause analysis approaches can be divided, depending on:

  1. Situations in which an immediate reason is known, so why-why analysis can be carried out by tracing back to the end of the string.
  2. Situations in which the immediate reason for the effect is unknown and there may be multiple causes, so the quality control tools mentioned earlier should be used.

Ramasway Ganesan
Puzhuthivakkam, Chennai, India

A: Great question. To some extent, the answer is undoubtedly yes; however, the tools are not all at the same level of measurement, so depending on the information or data already at hand, you may prefer one tool over another.

In particular, you have divided the analyses into two logical groups. I have found that for situations in which the reason is known, fault tree analysis and tree diagrams are especially good at going back to the root cause.

For situations where the cause is unknown, or earlier in the process of root cause analysis where you may be doing some group brainstorming, cause-and-effect diagrams, and affinity and relations diagrams may be the most useful, as they can guide you to the true cause or causes.

A nice discussion of the tools you mention, as well as some others, can be found in an appendix from a book written by Dean L. Gano, Apollo Root Cause Analysis—A New Way of Thinking.1

I. Elaine Allen
Professor of statistics and
Babson College
Wellesley, MA


  1. Dean L. Gano, "Comparison of Common Root Cause Analysis Tools and Methods," from Apollo Root Cause Analysis—A New Way of Thinking, third edition, Apollonian Publications, 2007, www.realitycharting.com/_public/site/files/pdf/ARCA_Appendix.pdf (case sensitive).

Data for a fee?

Q: I have been an auditor of ISO/ANSI/ASQ 9001:2008—Quality management systems—Requirements since 1992 and recently began consulting hospitals seeking ISO 9001 certification.

My experience with auditing to ISO 9001 is mostly in the manufacturing sector. When I audited against ISO 9001, clause 7.6—Control of monitoring and measuring equipment, I routinely included questions regarding the process for assessing the validity of previous measurement results when equipment did not conform to established limits. I found no real issues with this until lately.

Now, clients say that calibration service providers do not routinely provide "as-found" data in the report that’s sent to clients and customers. I have been told that the "as-found" data only becomes available to the client/customer for an additional charge (and it’s not cheap).

Obviously, organizations cannot comply with the ISO 9001 requirement to perform the aforementioned assessment without this data. Since this has only come to my attention recently, I am wondering about the ethics and legality of withholding specific information in the calibration report—unless an additional fee is paid.

Could you please provide some insight or justification for this business practice?

Celeste Nair

A: It is always a good idea to evaluate one’s suppliers. This requirement is in ISO 9001, clause 7.4—Purchasing. The Measure for Measure column, "Supplier Demand," (May 2010, pp. 48-50)1 provides guidance on evaluating and selecting calibration providers accredited to ISO/IEC 17025:2005—General requirements for the competence of testing and calibration laboratories.

In addition, the ILAC-P14:01/2013 policy document requires ISO/IEC 17025 accredited laboratories to provide measurement uncertainty data with the measurement results as of Dec. 1, 2011.

The customer should specify its requirements in its purchasing documents for calibration. ISO/IEC 17025 has contract review requirements that accredited laboratories must meet to comply with clause 4.4 of ISO/IEC 17025.

For the laboratory to make an out-of-tolerance decision, it has to measure "as-found" data. Even if the laboratory does not report it, it is required to retain it per ISO/IEC 17025, clause, second paragraph:

"When a statement of compliance with a specification is made omitting the measurement results and associated uncertainties, the laboratory shall record those results and maintain them for possible future reference."

So, for a start, it is a good idea to use ISO/IEC 17025 accredited calibration providers and specify the customer’s requirements. Some provide "as-found as-left" data routinely. Others may charge because they may claim it takes extra time. But, if a competing laboratory provides it as part of the service, the other laboratories will follow suit or lose market share.

If the ISO/IEC 17025 accredited providers have to make a compliance decision on an item being calibrated, why would they not record the data? Even if it’s not provided, they are required to retain it for future reference in case of an inquiry.

Calibration providers (whether accredited or not) that do not provide "as-found as-left" data should probably be avoided. You don’t know if they provided a legitimate calibration or if they "stickered" the calibrated item and produced a generic certificate.

Other laboratories complying with ANSI Z540-1 or ANSI Z540.3 requirements are also required to provide "as-found as-left" data. Otherwise, they are not fully complying with Z540 requirements.

The Measure for Measure column, "Calibration Evaluation," (September 2010, pp. 44-45)2 discusses evaluating non-accredited calibration providers and what to look for when assessing them.

Dilip A. Shah
E = mc3 Solutions
Medina, OH


  1. Dilip Shah, "Supplier Demand," Quality Progress, May 2010, pp. 48-50, http://asq.org/quality-progress/2010/05/measure-for-measure/supplier-demand.html.
  2. Dilip Shah, "Calibration Evaluation," Quality Progress, September 2010, pp. 44-45, http://asq.org/quality-progress/2010/09/measure-for-measure/calibration-evaluation.html.

Average Rating


Out of 0 Ratings
Rate this article

Add Comments

View comments
Comments FAQ

Featured advertisers