Reading seminar 1 notes.
There are several different ways to gather data for
instance: questionnaires interview, natural observations and etc. Data
gathering is mainly done to capture users’ reactions and performances with a
system or prototype so that a set of stable requirements can be established for
said system. When gathering data, triangulation, which means to gather several
kinds of data from different perspectives, is usually a good way to go about
it. This is something written in the text and which I also agree on. Only going
about data gathering in one way can be a bad idea as you only are gathering the
views from one perspectives. Triangulation is something I feel can be
implemented in the project in this course, as a way to improve our results.
I also agree on combining qualitative analysis and
quantitative analysis. As written in the text, “Karapanos et al (2009)
go further and suggest that averaging treats diversity among participants as
error and proposes the use of a multidimensional scaling approach instead”, I
agree and feel that the “multidimensional approach” could be in the form of a
combination between qualitative and quantitative analysis. In data gathering
and data analysis, I believe that a holistic view of the system and/or prototype
being examined is important, this equates to examining as many aspects as
possible from as many different viewpoints as possible.
The next step is to use all the data that has been gathered
and analysed in order to create stable requirements. The literature mentions
“personas” (e.g. novice or professional) and how requirements need to be
adjusted to different kinds of personas. This made me wonder how large systems,
that try to create requirements, deal with having to adjust to a widespread distribution
of various personas. Are there any instances, when having too many personas (/
users with different experiences), in a system can create problems? Are some of
the different experience levels of users ignored when building system?
Inga kommentarer:
Skicka en kommentar