Measuring Software Product Quality: A Survey of ISO/IEC 9126

From Maisqual Wiki

Jump to: navigation, search

Ho-Won Jung and Seung-Gweon Kim, Korea University.

Chang-Shin Chung, Telecommunications Technology Association.



  1. ISO/IEC 9126
  2. Background and method
  3. The survey
    1. Rating scale
    2. Data collection
    3. Dimensionality's role
  4. Results
    1. Summary
    2. Internal-consistency reliability
  5. Discussions


 author = {Ho-Won Jung and Seung-Gweon Kim and Chang-Shin Chung},
 title = {Measuring Software Product Quality: A Survey of ISO/IEC 9126},
 journal = {IEEE Software},
 volume = {21},
 issn = {0740-7459},
 year = {2004},
 pages = {88-92},
 doi = {},
 publisher = {IEEE Computer Society},
 address = {Los Alamitos, CA, USA},



The international standard ISO/IEC 9126 defines a quality model for software products. Based on a user's survey, this study of the standard helps clarify quality attributes and provides guidance for revising the standard.


Some have expressed concerns about a lack of evidences to support such standards.

Shari Lawrence Pfleeger et al.[1] "Standards have codified approaches whose effectiveness has not been rigorously and scientifically demonstrated. Rather, we have too often relied on anecdotes, gut feeling, opinions of experts, or even flawed research rather than careful, rigorous software engineering experiment.

Quality [2] is "The totality of characteristics of an entity that bear on its ability to satisfy stated and implied needs".

Characteristics are a composite (aggregate) of their subcharacteristics. But if a subcharacteristic doesn't correlate to its characteristic, the characteristic fails to show its intended definition. In such a case, the subcharacteristic should be moved to a better, more appropriate characteristic.

Background and Method

This study compares users satisfaction to ISO 9126 quality characteristics.

On 75 users, both end-users and developers, questions were asked with 5 answers plus a "I don't know" box. Statistical methods (Imputation) were used to fulfill missing values and the "I don't know" answers. PCA (Principal Component Analysis) was used to correlate the 18 sucharacteristics.

4 of them were not correlated (i.e. the correlation coefficient was below 0.6): testability, installability, replaceability, coexistence were not correlated to any dimension. Of the 18 remaining subcharacteristics, the correlation coefficients are shown below:

Dimension 1 Dimension 2 Dimension 3 Dimension 4 Dimension 5
Analyzability (0.616) Understandability (0.769) Time Behavior (0.805) Suitability (0.818) Security (0.856)
Changeability (0.653) Learnability (0.827) Resource utilization (0.766) Accuracy (0.648)
Stability (0.814) Operability (0.848) Interoperability (0.796)
Adaptability (0.669) Attractiveness (0.616)
Cronbach's alpha 0.817 0.850 0.758 0.750 --
Percent of variance 18.67 17.42 11.90 11.82 7.80
Cumulative percent of variance 18.67 36.09 47.99 59.81 67.61

Most of the ISO/IEC 9126 structure is demonstrated as being valid, but some subcharacteristics may be moved, revised or dismissed.

Limitations of this study are:

  • Cultural differences with other countries than Korea regarding product quality judgement.
  • Study based on a single packaged software.
  • Other statistical methods could be used for correlation.


  1. S.L. Pfleeger, N. Fenton, and S. Page, "Evaluating Software Engineering Standards", Computer, vol. 27, no. 9, 1994, pp. 71-79.
  2. ISO 8402: Quality management and quality assurance - Vocabulary.

See also


Personal tools