A Testing Dojo starts with the facilitator introducing the rules. The facilitator provides a mission and clarifies the structure of the dojo. If you run a Testing Dojo for the first time, you could consider providing a first focused test that gets the group going, rather than leaving it too broad.
The session facilitator should change so that everyone gets the opportunity to lead a small group of people. Any testing can be done by a single tester in front of the computer or in a paired setup.
The missions vary between testing as product, evaluating the usage of the following tools or using a new approach to check if we could incorporate it into our testing process.
When the team has run many Testing Dojos, the introduction tends to get shorter since the facilitator has to explain less of the standard dynamics.
In a single tester session, the person with access to the keyboard takes on the role of the tester and the others fulfill the role of observers. On a previously agreed upon time the tester is replaced by another participant from the audience. Five minutes seem to be just enough for this, but you can try different timings based on the size and skills of your group. The new tester continues to follow the mission tackling the product under test. When the individual tester gets stuck, he can ask for support from the observers. Otherwise observers have to remain silent while watching the performance.
For instance the facilitator has picked the latest application from the company and decided to explore it. A tester unfamiliar with the software is asked to begin for ten minutes. Right after starting the application, she faces a problem because she does not know what it does. After a few hints from the observers about the documentation, she continues to explore the software, but does not quite understand it. After the ten minutes have expired, a tester more familiar with the application starts, coming from the group of observers. He shows some of the features and explores areas of the product. He provides the first tester with some new insights about the application, as well as the approaches used to test it. A third tester then takes charge and explores the product from a completely new perspective.
The tester must explain every step of her thoughts for the observers to follow the individual actions. As a rule of thumb, the more the tester speaks, the better it is for novice testers in the audience. More experienced testers need less explanations of a particular motivation or applied heuristic.
In a paired session, two participants sit in front of the computer. The tester is working on the keyboard, while the recorder writes down the test ideas and discovered bugs. After a previously agreed timeframe, between 5 and 10 minutes), the tester goes back to the group of observers. The recorder takes over the role of the tester and one of the observers becomes the new recorder.
During a paired session the tester and the recorder in front of the computer need to clarify their steps so that everyone from the observers understands what they're doing, and more importantly why they are doing it. They should at least talk as much as they test and probably talk more than test. The pair is allowed to involve other participants only when it gets stuck just as in the single tester setting.
The paired setup is similar to a Randori Kata in a Coding Dojo. The people in front of the keyboard need to make their steps clear to the audience. Over the course of the time-box, they explain the model they got from testing the application. This knowledge is then transferred when the pairing partners switch. Since all observers watched and listened to the particular steps performed, the next partners will have similar knowledge about the product. Over the course of the session, each participant takes notes and tests.
Pairs switching can be decided using a round robin style or by picking the next volunteer for a larger group. In round robin switching, each participant should get into the role of the tester and the recorder at least once. The size of the group will influence how much time you schedule for the session. If you have one or two testers who feel uncomfortable to expose their testing steps to others, you can make testing and recording optional.
To make the activity more fun, you can include a ceremony activity for pair switching like a handshake when the tester leaves or a Karate-style bow when a new recorder joins. You can also bring in special things if the team found a bug, like a high five.
Suppose that John and Susi are the first pair. John decides to pick up the keyboard and Susi begins with recording notes. They take the application and discuss which areas to focus on. The session facilitator starts a timer for five minutes, informing the pair when they have two and one minutes left. After getting a brief charter up, John and Susi start to test the application with the first item on their list. They make pretty good progress. The time also passes very quickly. John goes back to the observers after the facilitator announced that the time is up. Susi now takes the keys and Paul joins her. Paul raises a question about one area that he just saw while John and Susi were performing. Susi gets back to this area and Paul asks some questions that eventually result in Susi doing some follow-up testing.