Posted in Website Usability Testing	January 7, 2008

How to prepare a usability lab for several different types of studies

Usability testing is a controlled experiment that tests the architecture and user-friendliness of your website. Reasons why businesses conduct usability tests are often to provide feedback to information architects and designers.

The ultimate goal of a usability test is task completion, however, there are many different parts to this:

• How long did it take the user to complete a task?
• Did the user know they completed the task when they did?
• How did the user feel while trying to complete the task?
• Did the user give up while trying to complete a task?

Only 47 seats left at the Mequoda Summit Napa Valley 2008 where you can learn more about conducting usability labs, and how to optimize your website architecture. Register now.

It has been observed that, as a group, users used to be a fairly low-esteem bunch. In the past, when users ran into problems on the Web, they were quick to blame themselves. However, nowadays when consumers encounter problems purchasing online, they are blaming the website and not the medium for failures.

We suggest that you do usability testing in between all four phases of your website development: prototyping, coding, quality assurance and production as well as whenever you plan to make changes to the interface or a critical process.

Over the next couple of days, we’re going to provide you with the different ways that you can conduct usability labs, and today we’re starting with lab preparation.

Only 47 seats left at the Mequoda Summit Napa Valley 2008 where you can learn more about conducting usability labs, and how to optimize your website architecture. Register now.

Lab Preparation

What are the associated costs?

• Participants cost approx $200 per participant
• $100 for incentive, $100 for recruiting. Unless you are recruiting on your own, which may be less, depending on your budget.
  
• Cost of the Morae software (this is a one-time purchase):
• Morae Bundle: $1599 (Includes 1 Morae Manager, 1 Recorder, 1 Observer)

• Each additional remote observer - $195.

• You need a lab environment. If you don’t have a computer that meets the needs of the software, you may need to buy one. You also need a quiet office, free of any type of distractions.

What does the ideal lab environment look like?

• A room free of distractions
• Moderator
• User
• Webcam
• Computer with Morae (or other) usability software

How many participants should I have for each study?

• 5-8 participants detect 80% of usability problems
• Competitive tests require more participants

How do I recruit participants?

• Develop a screener using a user profile
• Facilitate recruiting internally

• Existing customers
• Craigslist.org
• Newspaper ads
• College campus bulletin boards

• Recruiting agency
  • Approximately $100 per participant

Only 47 seats left at the Mequoda Summit Napa Valley 2008 where you can learn more about conducting usability labs, and how to optimize your website architecture. Register now.

Determine Task Analysis

• Focus on key strategic tasks
• Frequent
• Critical
• How many tasks?
• Task Wording
• Determine what is successful completion?
• How long should each task take a user?
• How many errors are OK?
• Is use of online help OK?
• Are assists OK?

Gather Test Materials

• Testing script for moderator
• Consent / Non-disclosure Form
• Task list
• Incentives
• Stopwatch to time participants
• Post Questionnaire

Perform a Pilot Test

• Dry run of lab using one or two participants
• Can use readily available person or a real user
• Identify major issues with process or technology
• Create any user profiles ahead of time
• Finalize participant schedule
• Allows test team to practice
• Allow at least 1 day to fix problems

What is the Goal of a Usability Lab?

• Increased customer satisfaction
• Positive word of mouth
• Decreased support costs
• Decreasing development

On Wednesday, we’ll run down the ideal lab environment as well as the different types of usability labs that you can run including expert review, heuristic evaluation, and the many different types of usability studies with real users including cart sorting, cognitive walkthroughs, questionnaires, paper prototypes, and high fidelity prototypes.