En Ru
+375 (29) 692-02-78 +375 (17) 219-48-27
Contact sales

Usability- testing

16 March 2011 | admin | Theory

The essence of usability testing lies in watching users while their interaction with the system during performance of real tasks. Usability testing allows to find out if the system really works, i.e. if it solves the assigned design tasks, whether it’s productive enough and, simply, if users can perform their tasks at all.

Advantages of testing on real users.
Usability testing results in alteration of architecture and functionality of a program or a web site. Such changes can positively influence your business purposes: web site traffic, users’ loyalty and conversion level. Usability testing for B2C-systems (for instance, online stores) has the following advantages:

1. Navigation improvement (it becomes clear to the user)
2. Increase of conversion level
3. Improvement of brand perception

Usability-testing makes software design cheaper due to the ability to identify design problems at the initial stage of the project.

Types of testing
Not all usability tests are the same. The recommended approach to real users testing depends on the purpose. The purpose may be to collect quantitative data (error level, task performance test), to analyze competitors, to identify reasons for low ROI (e.g. low level of sales in spite of large traffic). Plan and structure of user sessions should be designed in the way which best facilitates reaching the programming purposes. Testing also depends on the object of assessment: whether it’s a paper prototype, interactive prototype or a working scheme.
For some types of products, like mobile phones for example, longitude research can be most suitable for lasting data collection in the natural sphere of phone usage. In other cases, single sessions can be quite enough (for instance to answer the question, if the user is able to find address book in the menu and input a name and contact number at first dash). Some types of testing should be carried out in special labs, where the specialist has total control over testing environment, while others require more realistic environment (for instance, at home, on user’s computer, at user’s work place). Details on how to choose the appropriate usability-testing method can be found on UasbiliyNet

Why use professionals?
Any kind of usability-testing is better than no testing at all or shot in the dark, trying to guess how a user will work with the site. However there are very real types of distortions and deviations, which can happen during testing session. A specialist is aware of possible types of problems and is able to interpret test results properly. A user can spend 10 minutes trying to solve a simple task, then get confused, lost and embarrassed, never solve it and, after being asked what changes he would like to make, he would answer that everything is fine the way it is. Cultural aspects should also be considered. For instance, in Korea users are unwilling to publicly criticize the work of others, which can seriously distort test results. Professional arsenal includes practically all methods for solution of problems connected with interpretation and analyses of testing data. That’s why planning and control as well as result analyses are better to be trusted with professionals.

Test results
Contents and types of documents after usability-testing depend on project purposes and on how the client is going to use the information. Universally accepted standard is the following: a complete list of detected problems, problem criticality, suggested solutions, productivity metrics (for example, according to ISO 9241: error level) and video material (display image and/or user’s face).

Standards and recommendations
ISO 9241 contains Ergonomic requirements for office work. Article 11 of this standard contains usability materials. It particularly has metrics, which can be used for quantitative user productivity assessment while task-performance during usability-testing sessions
ISO 20282 covers everyday usability. First part of the standard specifies the way to define properties of the context the product will be used in. The second part defines methods for usability measurement (i.e. efficiency, productiveness and satisfaction measurement) for everyday things.
ANSI/NCITS 354-2001: Common Industry Format for Usability Test Reports. This document specifies the common format for usability-testing results. It costs 30$ but you can check it out here for free.

Subsidiary software for usability-testing.
Western specialists have developed a number of programs to automate usability-testing. Market of subsidiary software offers the following programs:
• For qualitative and quantitative testing of web sites and software in lab conditions with moderator

TechSmith Morae is an example of such software. This program can:

  1.  log low-level OS events (keyboard click, mouse click), high-level actions (clicking the links in a browser)
  2.  record user actions by logging everything that is going on on the screen or in the window
  3.  Record user’s emotions with a digital or web cam
  4.  Follow user’s actions from anywhere in the world where high speed internet connection with respondent’s computer is available
  5.  Add its comments on user’s actions or emotions live
  6.  Record different metrics, for instance task performance time
  7.  Create presentations in Windows Media Format displaying user’s face, recorded display, keyboard, mouse and browser actions as well as observers’ comments

For qualitative and quantitative distant testing of web sites and software without registering users’ emotions
Example of such software – Ethnio. This program allows doing everything Morae does, except for recording with video or web cam. Moreover, such software allows hiring respondents “here and now”: as soon as a user logs in to the site, he receives an invitation to participate in the usability testing. If he is willing to participate, he can start a test session in a few moments: the user receives an e-mail, containing a link to the program, which should be installed on his computer for creating the log of events and recording the screen. After the end of session this program is uninstalled from the computer, and returns it to its initial state. A big advantage of such kind of programs over software of the first type is their low requirements to net carrying capacity, as only recorded screen or window is transmitted, not the video traffic from video or web cam.

• For automated testing
Such types of programs are independent as they allow collecting various useful information without user’s or usabilitist’s participation. An example of such software is ClickDensity. This web application records all user’s clicks on a web site page. As a result a click map is created, which helps to judge about the popularity of certain functions or navigation problems. Moreover, there’s a whole class of programs which collect various statistics of site traffic. Only surface conclusions can be made on the basis of such information, for instance that register form does not work (as the statistics showed that 80% of users quit registration without completing it). A vivid example of such programs is Google Analytics. In order to identify real registration problems a full scale usability-testing should be carried out and real users should take part in it.

There are various external means except software, which help to
• Record eye movement on the screen and on the interface of tested software or web site
• Register various biometric data, such as: pulse, pupil extension, skin capacity, muscle tension

Related Articles
Usability standards and recommendations

Liked this post?