Swiss Testing Board

The Swiss Testing Board is the Swiss member organization of the ISTQB and emerged in the early 2000s from a working group of the SAQ IT Section. It accredits training providers, promotes tester certification and, right from the start, introduced didactically measurable learning objectives based on the Bloom taxonomy into the international ISTQB curricula.

Key Takeaways

• The ISTQB emerged in the early 2000s from informal discussions between a few people before there was even a formal organizational structure.
• Learning objectives in the Certified Tester Foundation syllabus only became measurable and testable through Bloom’s taxonomy and Anderson’s extension, because without this system no reliable certification is possible.
• Testers cannot meaningfully test software if they do not understand what it is used for and how it is technically structured.
• Agility requires a high level of discipline: anyone who does not explicitly write down and understand acceptance criteria does not know what they actually have to test.
• Systematic test case creation based on classic techniques remains indispensable, because purely intuitive testing finds errors but does not deliver completeness.

How an anecdote in the Red Light District turned into a certification

The Swiss Testing Board was born out of a specific demand, not a strategic master plan. In the early 2000s, Thomas Müller gave a presentation on use cases in testing at Credit Suisse. Afterwards, Silvio Moser approached him with a simple question: How could testers be certified?

Thomas didn’t have an answer. But he knew someone who did. Through Carol Frühauf, with whom he had already been working for almost ten years in the SAQ’s IT specialist group, contact was made with the British ISEB and the American ASQ, both of which were already working on curricula for testing.

The decisive factor was a meeting on a rainy winter evening in a restaurant in Zurich’s Red Light District. Robert Treffny told him about efforts to merge the ISEB and ASQ curricula and found an international organization. Thomas agreed, without knowing exactly what he was getting himself into.

Why the Swiss Testing Board started as a loose working group

In the beginning, the Swiss Testing Board was not an association, but a working group within the SAQ’s IT specialist group. This pragmatic connection saved the founders from having to set up their own organizational structure.

The international umbrella organization was founded in Scotland. Thomas was unable to attend for family reasons, as his children were still small at the time. Robert Treffny represented Switzerland on site. The following meetings mostly took place in the Rhine region, in Cologne or Düsseldorf, in small groups of six to seven people.

At one of these early meetings, the question arose as to who would coordinate the curriculum work. Thomas volunteered, according to his own description as “nobody” in this group, but with the right background: he had dealt with learning objectives and had already set up courses, including together with Hans Schäfer as part of EU funding programs in Germany.

Learning objectives must be measurable, just like requirements

Thomas’ first substantive intervention in the curriculum work concerned the measurability of learning objectives. The original formulations were not testable, and for him this was a contradiction to his own subject.

His argument to the group was that if you know how to specify requirements for software in a testable way, you can apply the same principle to learning objectives. Bloom’s taxonomy models and Anderson’s extension, of which Thomas was a follower, served as the methodological basis. These cognitive levels, the K-Levels, became the basis for the learning objectives in the Certified Tester.

The background to this went back further. In the early 1990s, during his time at Schweizerische Kreditanstalt, Thomas had studied didactics at the ETH. Professor Frey made the question “How do you measure?” a central concern, and Thomas carried this influence into his curriculum work.

Convincing the group of the measurability was not the difficult part. Finding the right level of detail in the texts, on the other hand, led to massive disputes. Native speakers such as Dorothy Graham from the UK Board and Erik van Veenendaal from the Netherlands were involved in the writing process. Finland and Sweden brought in representatives from the universities right from the start.

Switzerland was an early leader in certifications

In terms of population, Switzerland was at times one of the top nations in terms of the number of certified testers. Demand came primarily from the banking sector, with Credit Suisse as co-initiator and Silvio Moser as part of the board.

Three training providers were accredited first; they had a strong vested interest because the courses sold well. Marketing was not handled by the Swiss Testing Board, but by the training providers themselves. Adrian Zwingle is mentioned here as a particularly active driver. At the time, a certificate cost around CHF 400, which was an unproblematic price for the market.

The 2007 General Assembly in Zurich-Oerlikon provided a boost. Many nations attended, the Indians were there for the first time. Switzerland as a conference venue acted as an additional incentive. This momentum later flattened out. Saturation and critical questions about the concrete benefits of certification set in.

Despite all his efforts, one sector remained closed to Thomas: the pharmaceutical industry, his main field of work at the time. The only thing that mattered there was whether something was FDA and GXP certified. In terms of content, a lot of things were included, but no one there was interested in testing certification.

The concepts of the pioneers still apply

The basic test plans have retained their importance, even though the environment has changed considerably. According to Thomas, the concepts formulated by pioneers such as Glenford Myers and Bill Hetzel are still relevant to the question of what to test, how to test it and what to think about.

One concrete example is the combinatorial approach to system integration. This involves checking at a high level which systems communicate with each other and how, and which data elements are affected.

Sampling is a separate topic in the pharmaceutical industry. Many people involved transfer the thinking from hardware production to software and ask about permissible deviations as if nails were being manufactured. Thomas counters this with a clear statement: “Software is not a nail. Sampling logic from production cannot be transferred one-to-one to software.

Speed is the central challenge in testing today

The biggest change in software testing is speed. Testing must be fully integrated into the development cycle from the beginning, not as a downstream step.

This starts with the requirements. Anyone working with acceptance criteria, user requirements or user stories should ask themselves early on whether a requirement can be tested at all. The old principle that requirements must be testable remains valid.

Test automation brings with it its own question, especially in regulated environments: Is what the test automations output also qualified and validated? How do you ensure that the results are correct?

With artificial intelligence and machine learning, the basis is shifting further. It is no longer possible to take a purely deterministic approach here, and different test criteria apply than with traditional software.

You can’t test without understanding the product

Thomas’ most important advice to testers is: understand what you have in front of you, what it is used for and how it is built. If you don’t know this, you can’t test software or an embedded system properly.

He illustrates this with an example from the shipbuilding industry. A company was asked to test the software for the drive motors of container ships, machines the size of a kitchen, in ships that cost around 150 million. The biggest hurdle at first was to understand what it was all about. Only this understanding made testing and simulation possible, precisely because such a drive has to be adapted to new environmental specifications over 20 years.

For testers in the low-level area, technical skills are also required: optimal processes, building deployment models, working with tools such as GitLab. Testers should also be familiar with related IT specialist areas.

This is where experienced testers have an advantage. They already know a lot from experience, and a look at the real product often acts as an eye-opener, for developers and testers alike.

Systematics beats gut feeling

Modeling is underrepresented in testing, although it has high value. If you can’t model, you test according to feeling or whatever comes to mind.

Feelings are important, I have nothing against that. But systematics is something extreme. Thomas Müller

Intuitive testing certainly produces good tests, and sometimes something valuable emerges. But something is missing when the systematic approach to a requirement is omitted. The basic techniques for this were defined by Myers and other early representatives of the field.

Agility does not release anyone from this discipline either. Agile does not mean “just go for it”, but requires a high level of discipline. Those who have written down and understood their acceptance criteria also know what they need to check. Early, early, early is how Thomas sums up the principle.