[Sigia-l] software for persona design

[Andrew Boyd]

Eric Scheid wrote:

>what sort of research do you do, and what sort of analysis processes do you
>use?
>
>  
>
Hi Eric,

without getting drawn into process evangelism, I worked with a group who 
developed their own process based on Hackos and Redish (User and Task 
Analysis for Interface Design), Alistair Cockburn's Use Case book, with 
some of the language changed to reflect the RUP-using engineers that we 
worked with. I've used the same process ever since, which is, simply put -
1. observe the users in the course of their work
2. question them based on all available preconceptions from duty 
statements, current version design material, and anecdotal material.
3. draft plain language use cases that describe each task/function they 
perform
4. feed these back to them as user/task personas
5. map relationships to ensure that there are no unrecorded information 
flows
6. when this is available, then start to assume that we know something 
about the true issue set, not before. This is where the traditional 
analysis processes start to become useful.

The "research" was in internal application development inside large 
government organisations. That team continues the work at that sort of 
level. I'm in a different field now, but still do things in much the 
same way when I have time.

>really -- how many personas do you have per project?
>
>  
>
I have worked on projects where we had several dozen - less than ten 
major user areas, but with functions thrown in as described above, they 
quickly add up. I remember the first time this was used at the above 
depth - user and developer alike said that it was the first time they 
believed that they understood what was happening. If there were room in 
the world for one more design book, I would throw my notes together and 
POD self-publish :)

Cheers, Andrew