Notes for seminar 1, Mårten Norman

Reflections on ch7 Data gathering

Dividing the work process in the steps Data, Information and Conclusion looks like a good description structure. Especially when communicating to an non-initialized audience. The chain of activities are visible signs of a planned structure and there is a promise of a result that may improve something.

Design of questionaries is a science in itself. I feel rather familiar with the topics. Still it is obvious that I should make a draft and ask for guidance to avoid mistakes.

Stopping observation and consider it ready when you start recognizing patterns appear like a good rule of thumb for me. This reminds a bit of my learning strategy where I move on when things make sense. On the other hand it's easy to get stuck when things do not make sense. Maybe this can happen with observations too; when there is no obvious pattern to recognize within a realistic timeframe.

 

Reflections on ch8, Data analysis.

Quantitative data vs qualitative data is still a bit confusing to me. I struggle to find an intuitive and still correct definition better than quanitative = "lots of numbers" and qualitative = "a few informal essays".

The outspoken acceptance of Convencience sampling is good. Not since it is good to "take those you find" rather than taking random samples, but it is good to admit reality and deal with it in a realistic and scientific way.

Reflections on ch10, Identifying needs, Establishing requirements. 

Requirements reflects a goal and are related to a persona or user characteristic or activity. Higher level requirements should be broken down into more detailed requirements. Class diagrams and sequence diagrams may be required too before actual HW and SW design may start.


Questions:

• Good intuition for better separation of quantiative vs qualitiative? They are very different, but exactly how?
• Ethics for extracting information from server logs and other semi-anonymous data that is easy to access in quantity? Are there any gray zones or even acceptable ways to use this without consent from the users?