This paper tries to find out ‘what software developers ask’ during their work as a collocated team. To answer this, they firstly conducted an observation where they took notes of what developers were doing minute by minute in a 90-minute session. Summarizing their notes, they came up with a task structure and what questions are frequently asked in each task. Based on that, they also conducted a survey mainly focusing on the importance and availability of certain information to answer those questions. Some ‘what to do’ are also presented as ways for software teams to get the information in need more effectively and efficiently.
The observation focuses on the following three questions:
- What information do software developers seek?
- Where do developers find this information?
- What inhibits the acquisition of such information?
What I really learnt from this paper:
- You can conduct a survey based on a previous observation and it looks so much better;
- How you can structure and qualify an observation (say one consists of notes only);
- It is possible to frame the same (objective) thing in different ways. (The central topic is: the asking behavior of software developers. And ways to research this include: what do they ask? when do they ask? to whom do they ask? how do they ask? etc.)
Some things I am not so sure about:
- How do they address the disadvantages of using observation?
- About external validity: how do they address the difference in project nature, personal working styles, cooperate characteristics, etc.?
- The relation of the observation and survey and the final results are not very clear (which study contributes to which results?)