The International Organization for Standardization (ISO) has been updating the present ISO 9126 international standard on software product quality measurement. The update has been ongoing since 2005. The present standard will be replaced by the quality measurement division of the forthcoming ISO 25000 series of international standards on software quality requirements and evaluation (SQuaRE). This paper focuses on the weaknesses of the current ISO 9126 by collecting information from various researchers and interested parties.
The ISO 9126 was established in 1991 to check on the quality characteristics for software product assessment. From 2001 to 204, ISO published an expanded version of the guidelines and the current version comprises one international standard and three technical reports. The updates on the current ISO 9126 version have been necessitated by advances in the field of information technologies and changes in environment.
Weaknesses in the current ISO 9126 version are divided into two categories: self-discovered weaknesses and weaknesses discovered by other researchers. In the first category, the following weaknesses have been noted.
- The language use din the paper needs to be matched with the ISO International Vocabulary of Basic and General Terms in Metrology;
- The metrics in parts 2, 3, and 4 of the current version are fully merged and need to be categorized into base and derived metrics to be aligned with ISO 15939;
- Many results metrics in the current version range between 0 and 1. Therefore, converting them into percentages is simple. However, these results would be simpler to understand if they were arranged in qualitative values, such as “excellent” for 100% efficiency and “very good ” for 80%, and so forth;
- When mapping the ISO 9126 to ISO 12207, most of the quality metrics referred to ISO 12207 processes. However, each process in the ISO 12207contains a number dissimilar processes. Hence, it is more practical for the ISO 9126 users to refer t the activities of the ISO 12207.
The following weaknesses of the ISO 9126 fall under the weaknesses discovered by other researchers.
- Jung et al observed that including the six compliance sub-characteristics leads to two possible outcomes. The first is that the compliance sub-characteristic for each characteristic belongs to its characteristic as indicated in ISO 9126. The second possible result is that the six compliance sub-characteristics comprise a separate dimension. Besides, their outcomes expose vagueness in the way that ISO 9126 is structured in terms of characteristics and sub-characteristics;
- In their study, Chua and Dyson discovered some inherent weaknesses in the ISO 9126 quality model, especially in relation to the Usability feature. To make the model easier to use for educators, who may lack the required expertise, they suggest that this feature should be extended to include more precise factors such as constancy, effortlessness, legibility, and color use. They also propose that a Help sub-characteristic be included as part of Usability, primarily to ensure that this important factor is not ignored. Besides, they suggest the inclusion of user satisfaction as a universal characteristic to sum up the general impact of the system on the user in their specific requirements. These changes will ensure that the newer version is an effective tool for assessing the quality of e-learning systems.
- Based on the ISO 9126 standard, on top of other models proposed by the five documents under the standard, Abran et al stated the usability model should be revised to include two other characteristics: learnability and security.
- Bøegh stated that the ISO 9126 metrics could not be applied directly as they are inexact. Besides, the intricate features based on the standard require a thorough examination.
- Koscianski and Costa stated that normalization of metrics and scores have to satisfy the following: every metric used must be mapped to the same reference scale.
- Behkamal et al stated that the current standard is a general one, in addition, it needs to be customized.
- Erazo stated that each characteristic should be ranked with their respective valued aspects.
The paper presented the weaknesses of the current ISO 9126 international standard as found out by researchers and practitioners. In order to resolve the weaknesses, the following ought to be done:
- Ensure the language used in software product quality measurement is fully in line with the classic measurement terminologies in the sciences and engineering;
- Identify and categorize the ISO 9126 metrics into base and derived metrics;
- Transform the result of the standards to percentage values;
- To eliminate the problem on the mapping between the ISO 9126 and ISO 12207, a cross-reference numbers from ISO 12207 can be used;
- Disregard the compliance sub-characteristic for each characteristic in ISO 9126 quality model;
- Additional experimental work is required to clarify and validate the ISO 9126 quality model structure in terms of characteristics and sub-characteristics;
- The usability characteristic should be widened to include more particular sub-characteristics;
- Two baseline characteristics should be added to the current standard: learnability and security;
- The ISO 9126 metrics should be redefined to a similar reference scale;
- The ISO 9126 should be made ready for customization so it can be used any domain;
- The ISO 9126 quality model characteristics need to be ranked by their usability and significance.