Usability testing is all about making sure the user is getting what they
asked for by continually having the user test the design as much as is
possible. The tests should be easy and and totally un-stressful for the
User-centred design is based on iterative design. The process evolves
and shapes the final product. User input is always required and can often
lead too critical changes too the system.
Three principles of user-centred design
- The primary focus is on the users and the tasks. The users need too
be known and the tasks they will be performing
- Obtain empirical measurements of product usage:
learnability, performance effectiveness, flexibility, error
tolerance and system integrity, user satisfaction
- Test early in the design and modifying where needed.
Prototypes are samples or incomplete designs that are quick and cheap
too develop. They give a user a hands on ability too use the system.
They come in 2 varieties
- Lo-Fi: Often Card based. Depicts the concepts, design alternatives
and screen layouts. Users navigate the cards like they would on a
- Hi-Fi: prototypes are computer based. Much the same as Lo-Fi but
constructed in an easier too navigate system.
Methods for usability testing
Think aloud protocol - A single user interacts with the system and is asked too
vocalize there thoughts and opinions.
Co-discovery learning - Same as above but with 2 people.
Question-asking protocol - Users do a given task and are then questioned about
any troubles they had understanding or learning the system.
Interviews - Users again use the product and are then questioned about the
system in an interview
Focus groups - a number of users (6-9) discuss pre-determined aspects and issues
about a given system. It is moderated by a "human factors engineer".
Performance measurement - Quantitative data is collected while examining a user
using the system.
Logging actual use - Users use the system and there use is recorded by the
computer automatically. Allows alot of use from many users.
Cognitive walk-through - Users follow a given scenario and are then examined and
problems within the system such as learnability are discovered.
Pluralistic walk-through - users, developers, and usability specialists do an
entire system/site walkthrough examining every aspect of it.
Heuristic evaluation - Users are examined on a set of relatively simple and
Consistency inspection - Usability specialists test the system too see if it
performs the same way as their own design projects.
Feature inspection - Analyse the features of the design from a common scenario.
Most effective midway through the development.
Developing a test plan
Guidelines for conducting usability tests
- Goals of testing
- Which usability technique(s) too use
- The users that will be involved
- The set of tasks
- The test environment
- Performance measures
- Session should be neutral
- All participants are separate
- Helping the participants should be last case scenario
- Relaxed atmosphere
- Examiners need too remain neutral
- Mistakes are not the end, don't panic