|
March / April 2004 Feature Article
|
|
Let's Go Dutch What does it take to present an eight hour tutorial on Usability to an audience at the EuroStar Testing Conference in Amsterdam? I thought knowledge of the subject and lots of preparation would be enough. Little did I know that a lot of faith, nerves of steel and a tolerance for sleep deprivation were also requirements. The Friday night flight to Amsterdam was delayed by 24 hours. Luckily the Saturday night flight was only delayed by an hour. After the ten and a half hour flight we arrived in a very cold and wet Amsterdam on Sunday morning. What does a person need after two nights of almost no sleep? A missing suitcase! And as dear Murphy would have it, it was the suitcase containing all my files and training aids - the stuff I needed for Monday's tutorial. At least the suitcase was delivered to the hotel at 21h00 on Sunday evening. All the trouble was made worthwhile when we won the EuroStar award for the best paper submitted. The tutorial I presented at the Amsterdam RAI on Monday, 1 December 2003, was all about usability testing techniques. I've presented similar material here in South Africa in the past, and comparing these experiences was quite an eye opener. If you are involved in web development, you might find this article especially valuable, as your interface designs are for international users. General Impressions It was quite clear that Usability is also very new in Europe and few companies know about and implement it into their development lifecycle. The people who are passionate about usability also feel they are in the minority and sometimes even feel alone in their quest to promote usability. The delegates attending my tutorial were very excited to find others they could share their ideas with. To almost all Europeans, English is a second language that is not used a lot (in some cases not even on a daily basis). Nearly all TV programmes are translated into Dutch, French, German, etc. These people are not as exposed to English as we are in South Africa. This is important to know, as it will most certainly affect the way Europeans understand and interpret terminology on a website or application. As in South Africa, the training session after lunchtime was just like a graveyard shift. Luckily there were enough practical exercises to keep everyone awake. During the course of the tutorial, I explained certain usability testing techniques and the delegates had to perform an exercise for each technique, to practice the skills they had learnt. I specifically used the same exercises I normally use in my training sessions here in South Africa, to enable me to draw a comparison between the outcomes from the different groups. Let's look at the four exercises the delegates had to perform and how their performance compares to the way South African delegates do these exercises.
Heuristic testing is a technique used by a tester where they can evaluate a system interface according
to usability guidelines. The interface is studied in depth and the testers look for properties they
know will lead to usability problems. The ten Heuristics are:
The delegates are given a list of the ten Heuristics, as well as printed screens from a "fake"
website. They need to use the Heuristics to evaluate the website and report back on which Heuristics
were broken on each screen.
I was quite surprised to see that the European audience found at least twice as many broken Heuristics
than their South African counterparts. They did this in the same amount of time I normally allow the
South African audiences. Please note that this is my subjective opinion, but the European audience seemed
to be much more analytical and gave a lot more attention to detail.
KLM is a way of predicting the time an expert user would take to complete a task on a user interface.
According to this model, certain common physical-motor operators are involved in performing most tasks.
In order to predict the execution time of an expert user to complete a particular task, we add together
the time taken for each of the physical-motor operators used. These times have been derived from Card et
al from numerous user tests performed with expert users. Examples of some of the 6 physical-motor operators
and their times are:
Time to press a key on the keyboard 0.12 sec Time to change between input devices
0.40 sec
Delegates are given a fictitious scenario where they must withdraw money from their PC. They need to enter
a PIN number via the keyboard and select $20 with the mouse. They have to calculate how long it will take
a user to perform this task, by using the Keystroke Level Model.
|
The results of the calculation compared very well to the same exercise done by South African audiences.
Everyone tends to make the same mistakes, for example where a "Mental Operator" of 1.35 sec needs to be
included in the task, most people omit it or put it in the wrong place. What was different though, is
that the European audience felt the correct answer was highly negotiable and tried to convince me that
their answers were better than mine. I enjoyed the heated debate that followed, because it showed me
that they take learning very seriously (as well as their own opinions).
In a Focus Group Session, six to nine users are brought together to discuss issues relating to the system
or product. The aim of the session is to capture spontaneous user reactions and ideas that evolve in the
dynamic group process. A facilitator prepares the list of issues to be discussed and moderates the discussion.
One delegate was "voluntarily" appointed to be the facilitator of the Focus Group. He received
instructions on his role. The objective of the exercise was to learn how a facilitator should conduct a
Focus Group session. The facilitator needs to make sure everyone contributes and that the discussion
stays focused. The group were given a product to discuss. Just to thicken the plot, I gave each delegate
a role card, which he had to act out. Examples of some of these role cards are: "You hate this
product" and "Your manager forced you to come to this session and you don't want to
participate".
The South Africans are obviously much better "drama queens" than the Europeans. I normally
have lively discussions and the participants really act out their roles very well. I found the
Europeans to be reserved, but I am convinced that it was the language barrier. Most delegates may
not have had confidence in their ability to speak English in front of a group of people. All things
considered, they did complete the objective of the exercise and the facilitator did a very good job.
Card Sorting is a technique for exploring how people group items, so you can develop your application
or website to maximise the probability of users finding items. Representative users work (usually
individually) with a pack of cards containing topics from a navigation menu of an application or website.
Users group related topics and tasks into an organisation that makes sense from their perspective. Card
Sorting is, therefore, an appropriate technique to consider when you have identified items that you need
to categorise. It is particularly useful for defining website structures. Card Sorting is beneficial in
that it helps the development team:
The delegates were split into groups of two. Each group received a pack of cards with topics for a
library website. They had to categorise the cards into groups that made sense from their perspective.
They were allowed to add new cards with topics they wanted to see on the website and leave out cards that
they either did not understand or thought did not belong on the library website.
It was quite amazing to see how people from opposite sides of the globe sort the cards exactly the same way.
What was different though, is the way they understood one or two of the card names. For example: one of the
cards says "Short Loan". This is when you take out a book for one day only, not the normal 14 days.
Most of the European delegates did not understand this, because to them, the word "loan" only means
lending money and cannot be applied to anything else.
It is safe to conclude that most people think and experience things the same way, regardless of nationality.
It is this phenomenon that enables usability testers to evaluate software interfaces according to guidelines
based on general human characteristics. It is important to note that language can be an obstacle, and therefore,
when developing websites or applications for international use, aim to make the language and terminology as
clear and as simple as possible
We, in South Africa, are in a very fortunate position to have eleven official languages. As South Africans, we need to learn how to communicate with people from a diverse range of ethnic and cultural groups. We are dependant on English as the common language most people can understand. Quite a lot of our television programmes are in English as well. We therefore hear and speak English on a daily basis, which equips us very well for international communication. Viva la rainbow nation!! Antonet Bekker |
<< January / February 2004 |
May / June 2004 >> |