[TYPO3-hci] Usability Test

[Sebastian Erlhofer]

Hello,

----
this mail got a bit longer... in short:
I tell what usability-testing is and what methods are used.
I propose a test-design for typo3 and I suggest conducting a online-survey.
----

There is a ISO Norm, which says in core something like:
"Usability is the extend to which a product can be used with efficiency 
and satisfaction by specific users to achieve specific goals in specific 
environments."

So we have 3 maior dimensions:
- efficiency: how fast can a user do tasks (and fast means, _if_ he 
archive his goal and how long it takes)
- satisfaction: that includes how "happy" (joy of use) he can do his tasks
- Be aware of "specific goals" and "specific environments"... that is: 
there is no "world-global"-usability. It depends on the system (in our 
case TYPO3) and the users task.

Usability-Testing in the lab is conducted through miscellaneous methods. 
To enumerate some:

- The most important / often used is the Thinking-Aloud. Users are given 
a task and are watched (recorded video and voice). They are said to 
speak their thoughts out loudly, so you can afterwards see where users 
had their problems.
- For interface-design and optimization eye-tracking is often used. So 
you see whether items, menues or such things are designed suitable or 
e.g. the user just "overlook" them.
- Cardsorting: you tell the user to sort some cards in groups (for 
menu-structures for example).
- Focus-Groups: You sit on a table and have a talk (like we are doing 
here, but focus-groups are mostly done with users).
- survey: thats a method out of lab, mostly to get a wider perspective 
on some phenomena.

Enough of "what is usability and how is it measured"... In my humble 
opinion a typo3-usability test would need those steps:

1) Defining user-groups (we had this already in former mails, giving 
them names can but must not be done)
2) Defining typical tasks (coz you define the tasks for the lab from those).
3) Observing 5+ (about up to 20) users in the lab while they are solving 
given tasks.
4) Analyzing problems and producing a report.

Its mostly helpful for steps 1 and 2 to conduct a survey to get a more 
objective and wider perspective on the "real usage" of a system.

Usability-Testing in the lab is very expensive, coz you need the 
equipment (or at least rent it), you need the users, need to analyze 
etc. I'd suggest to conduct this with help of universities. Perhaps 
anybody wants to write his master thesis about this... (I did mine on 
search engine usage, which is quite similar to usability-testing).

I'd suggest that we design a online survey at first. This can be done 
without expensive effort and the analysis is easily done with SPSS 
(statistical analysis program). I will write a extra mail for this to 
get some opinions.

Greetings (& sorry for the long mail)
-sebastian






redacted user schrieb:
> Hi,
> Just to have a common idea what is usability:
> http://www.usability-knowhow.de/
> 
> unfortunately only in german.
> 
> greets, b
> 
> 
> Irene Höppner schrieb:
>> Hi,
>>
>> is someone here, who is familiar with usability tests? I'm really no 
>> expert in that, but I fear, that a lot of the discussions here are very 
>> "theoretical" without usability tests... They could maybe become part of 
>> the roadmap? Or are they already?
>>
>> greets, Irene
>>
>> --Irene Höppner
>> A.BE.ZET GmbH i.G.
>> http://www.abezet.de/
> _______________________________________________
> TYPO3-team-hci mailing list
> TYPO3-team-hci at lists.netfielders.de
> http://lists.netfielders.de/cgi-bin/mailman/listinfo/typo3-team-hci
>