Usability test in laboratory

When is this method appropriate?

The classical, scene-based usability test in laboratory intends a qualitative approach and evaluate the totality of experiences within the use of interactive applications by the wanted target-group. The primary target is to determine acute usability problems while using the interface. Soft factors like emotionality, joy of use, trust and optical design can be focused. Besides information and choice behaviour can be examined.

Possible questions:

Effective and efficient usage
• Are the categories, buttons and links conform to the expectations, semantically correct and comprehensibly named?
• How does the user understand the different navigation elements? Are they intuitionally operated?
• Do the terms of the navigation elements arouse expectations that cannot be fulfilled by given contents and functions?
• Does the user quickly find the wanted information?
• Is the homepage neat? Is clear which contents and services are available at a glance?
• Are the interaction sequences of central operation processes (for instance, search, contact support, order, newsletter subscription, booking, publishing of contents) “logical”?

Look & feel: Rating of the screen/content design
• What impression does the homepage leave behind? What associations and emotional effects do appear?
• Are important information placed at attention strongly spots?
• Does a visual guiding succeed (wanted eye-tracking = actual eye-tracking)?
• Are the information and contents well readable (text design)?
• How are the structures and offers of the homepage perceived?
• Is there any kind of “information overload” so that the user is less motivated to get in touch with the website?

Joy of use
• Are the contents and services of the offer conform with the requirements and expectations of the users?
• o Does the user feel “well” that means does the use of the offer bring joy? Or does the user immediately leave the offer after achieving the target?

And many more…

In what project phase applicable?

• Analysis
  Testing of “old” or existing offers can take place against the background of a complete revision to receive knowledge for the conception of the new planned website.
• Development/conception
  With help of usability tests in laboratory it is precociously possible to gain findings about usability problems as well as requirements and expectations of the target-groups from the first clickable prototypes on. Here testing of navigation and user leading as well as interaction processes are in the foreground.
• Programming/implementation
  A usability test in laboratory can be implemented as a final milestone of the project. Thereby an integrated ascertainment of positive and negative experiences referred to the interface and interaction design as well as the navigation and information architecture of a website take place. If no rough usage problems can be identified, the release to launch can be allowed. Otherwise the identified problems can be solved by the given guidance.
• Live activity/usage
  If a website is already online, a scene-based usability test in laboratory can function as succeed control. Another intention of the test is to determine the status quo of a website before the planned re-launch to learn from mistakes for the future.

What is the procedure of this method?

Test sessions with single test people are performed. That means chosen test subjects use the application (website, online-shop, etc.) either free or with concrete tasks (for example search assignments, order performance, contact support, registration for newsletter, etc.). During solving the task the test subjects are observed by experienced usability experts.

Spontaneous statements (for instance referred to usage problems), ratings and thoughts (so-called thinking-aloud protocols) are recorded. In addition, we measure – if required – the eye-tracking on the chosen website and descriptively work up these measurements (for example sequence analysis).

By observing the mouse movements und click-actions we ascertain – if required – performance values (for example needed time/clicks to solve the task) and protocol the chosen search/ surf paths of the test subjects (so-called click stream analysis).

Subsequent to the offer usage stimulus-orientated single interviews take place. Thereby test subjects rate the central websites of the offer in detail. In addition, the test leader peruses the arisen problems with the test subject and reflects these. An interview guideline serves as a base.

Finally, we perform a standardised online survey in tests with more than 20 people. The test subjects rate the offer by means of a short questionnaire in relation to the following dimensions:

• user-friendliness/usability
• content
• visual implementation
• joy of use

Due to a benchmarking data base we are able to compare the ascertained data with ratings of other offers (so-called benchmarking analysis).

Excursus:

To create highly realistic laboratory or test situations we have the following steps:

• Recruiting of test subjects and screening
  We can choose potential test people from our online-access-panel. Besides sociodemographic features we can use interests, opinions and web usage behaviour as selection criteria for creating a sample. All potential test attendees are going to be screened at first. Thereby project specific and personal features are ascertained. The eResult GmbH is very exacting about serious acquisition of test attendees. Our quality management guarantees that no duration tester falsify the results.
• Incentivation
  While choosing an incentivation concept for the test attendee we bank on producing high situative involvement. In an online-shop this works by using “own money”. That means all test attendees do not only receive a cash payment of 65 – 70 euros but also a voucher for a product from the online-shop. By this one can guarantee that the test attendees use the website under real circumstances!
• Laboratory equipment
  The eResult GmbH banks on a comfortable equipment of our rooms and a inconspicuous, non offensive, technical equipment of the usability laboratory. By doing this, we can guarantee that the test subjects feel “comfortable” and “feel at home” during the test.

• Warm up phase
  This takes place at the beginning of the user test in the lab and serves to build up a pleasant and trusted relation between the test person and the interviewer.

Strengths of this method:

• Tests in usability laboratory are a 360-degree data ascertainment. That means to observe and interview the test subject - incl. the possibility to integrate other test methods or procedures like for instance eye-tracking.
• It is possible to formulate very concrete tasks or usage scenes.
• Due to direct interactions between the test person and the test leader one can individually answer the test subject. By this misunderstandings can be clarified and if necessary particular questions can be thematised more intensively.
• The usability test in laboratory does not only identify usability problems but also approaches for the optimisation of the website and the information architecture.
• The test provides comprehensive qualitative data that show how real users act on a website.
• Video recordings of tests, or rather, test subjects that deals with the usage of the certain website are very insightfully and convincing.

What about the sample size?

Depending on the extent of the test object and the goal setting the sample size strongly varies. Generally, 8 to 10 test subjects in each target-group are advisable in particular if usability questions are in the fore. If the rating of design and layout or the preparation, amount and kind of information is important, at least 20 test people should attend the test.

Remark: Generally, one should consider whether all test subjects are supposed to be tested in the same object of examination. Or whether the website is tested with half of the sample during the first test phase to solve the identified main problems. Afterwards within a second test phase the already optimised website can be examined by the rest of the test subjects. Experience shows that at the latest after the 5th test person the greatest problems start to repeat. So why continue testing? If a test person has problems with an element or a function, other visitors of your website may have the same problems. The remaining capacities, or rather, test subjects should not see use in the confirmation of already known problems but for the identification of further problems.

How much time takes the implementation?

Depending on the sample size the study can be implemented within 3 to 4 weeks (N=10 to 15) or 4 to 5 weeks (N=20).

What input do you have to provide?

Optimally you come to our premises with concrete questions that we enquire within a test. We would like to work out with you this kind of questions in a workshop.

We need at least graphical worked out screens in a current graphics format. But to test interaction processes we need a functional prototype (click-dummy) that has enough functions to implement a task-based user test. Ideally, your website is already online or callable in the internet.

What do you receive from us?

You receive an interpretative result report with the following contents:

• Management summary
• Detailed illustration of the results based on observing and survey data.
• Interpretation of the data: identification of the reasons and causes for the ascertained data.
• Highlight video-scenes (click stream, interviews, exemplary sight videos).
• Concrete guidance to solve identified usability and design problems, verbal described or visual-functional implemented (for instance in terms of optimised wireframes or clickable prototypes).

On demand we present the results and findings of the test within a moderated workshop.

Do you have any further questions? Please do not hesitate to contact us! We would like to consult you.