Usability Testing Flashcards
Useful
utility + usability = useful
To create “useful” products and services that are designed with objectivity and empathy for others in mind.
In order to be useful, our designs need to provide the appropriate utility, or features, as well as be usable (i.e., these features must be sufficiently simple and enjoyable to use).
Utility
The features and functionality needed to complete important tasks
Usability
How easy your project features are to use, as well as whether they bring satisfaction
Validate the usefulness of your design
validate the usefulness of your design or prototype is to test it with real people using evaluative research methods such as usability testing.
Peter Morville’s UX Honeycomb model
In addition to designing Useful and Usable products and services, they must also be Desirable, Accessible, Credible, and Findable to deliver real value to real people
Design research
Design research falls in the same general area as research for social sciences—fields like psychology, sociology, and political science.
Tangible vs Intangible
Despite its focus on intangibles (human psychology, for instance), research in social sciences isn’t all guesswork. Researchers in these areas draw upon the same general experimental structure you see in natural sciences (physics, chemistry, and medicine).
Aggregated Data
As long as data is aggregated and not focused on an individual’s personal information, companies such as Facebook or Google can analyze mountains of data and conduct social experiments in the absence of formal regulation and ethics reviews.
Informed consent
Informed consent is the process of helping research participants make an educated decision as to whether they want to take part in your research. Informed consent describes the nature, intent, and details of your study to participants, allowing them to make a conscious decision on whether or not they want to participate.
Facebook User Manipulation 2014
You may recall the uproar over a 2012 Facebook study that manipulated users’ feeds to see what impact doing so would have on users’ emotions. Some reactions to the study were released in 2014 in “Facebook emotion study breached ethical guidelines,”
Sampling bias
sampling bias can mean your results aren’t generalizable to a larger audience or customer segment.
Whilst it’s most important to test with your primary persona, ensuring that you run some usability tests with those dissimilar to you or your persona is a good way to ensure your designs are useful to as broad an audience as possible. This will also help you avoid general research issues such as sampling bias, which causes errors in results due to a non-random selection of test participants.
“Guerrilla Testing”
Hallway Testing or “Guerrilla Testing”: This type of recruitment simply involves approaching friends, coworkers, or strangers in their natural environment.
Amazon’s Mechanical Turk
Amazon’s Mechanical Turk is useful if you need to recruit a high volume of participants at a low cost. It allows you to set up a test and pay anonymous people a small amount to complete quick tasks.
Panel Agencies
Panel Agencies such as ResearchNow have large databases of potential participants available for unmoderated tests. Costs range between $15 and $55 per response.
Usability.Gov.
The circumstances of the project may vary, but in general, there are a few types of common templates/messages researchers use to contact potential participants. Usability.gov is a great place to look for recruitment, confirmation, and reminder email templates.
