Visicog

The results from usertesting.com were pretty impressive. The tester was particularly articulate, eagerly expressing his expectations and reactions to the site. It was pretty clear, also, that this was definitely an “expert” user. I usually don’t see people editing the URL string during a usability test (how about, never?). I received the video and a written report less than an hour after I submitted my task to the service.

One of the other positives was that I was able to encourage the user as a final task to explore the site to get his impressions of the general functionality. In traditional usability tests, you usually don’t have this flexibility in time to allow such free-flow exploration– every minute is precious because of the effort involved in getting the user  to begin with.  This service makes access to participants so much easier that it makes participants into a commodity (in a non-objectified way, of course).

Negatives? Well, there is the obvious that you can’t follow up with any questions or direct the test based on their actions. You are also limited to 15 minutes, which could limit some types of real transaction-based scenarios. Perhaps the biggest drawback is the limited factors by which you can recruit users (age, gender, income, computer expertise, country). Considering that I personally do a lot of work around business processes, I’d like to be able to recruit users based on their professional background - for quality participants, I’d even be able to pay a premium for it. However, I’d recommend the service if you have a general-purpose site or relatively simple workflow.

Last semester, I wanted to test the level of usefullness for one of the textbooks I assign. Research-Based Web Design & Usability Guidelines is a solid book that I have kept as part of my class for the last couple of years, but I haven’t been completely sure as to how helpful it was to my students because of the broad range of experience levels in the classroom. So in a typical usability-teacher fashion, I ran a test in the guise of a homework assignment to assess the book’s true usefullness.

The Assignment

The assignment was for students to hand in a list of the five most “surprising” guidelines from the book. The aspect of “surprising” was open ended- it could mean that a student learned something new and helpful, that they found something that they vehemently disagreed with, or something else all together.

I took all of the answers, and put them in a spreadsheet to see what patterns could be observed in the answers. Ultimately I found that students gravitated towards guidelines that had practical implications to their current positions. I also found that there was very little overlap between what guidelines students found useful. There were two huge spikes in guideline citation, however…

The Hardest Working Guidelines in Class

These two guidelines were each cited by more than a third of the classroom in the assignment. No other guidelines came close to their popularity.

Guideline 6:9 — Avoid Scroll Stoppers
“Ensure that the location of headings and other page elements does not create the illusion that users have reached the top or bottom of a page when they have not.”

Guideline 1:10 — Use Parallel Design
“Have several developers independently propose designs and use the best elements from each design.”

Many of the students who listed the Parallel Design guideline expressed that they didn’t know how to practically implement it due to resource constraints. So, here are a couple of ways that we do this at my company that are both time and resource efficient:

1. With my designer hat on, I’ll sit down with another team member (project manager, programmer, etc) at a table for an hour. We then independently sketch our own solutions to the problem, while looking over each others shoulders for inspiration.

2. We will sometimes break out of a product meeting and hand everyone a whiteboard marker. Everyone then sketches out their own solution to the problem at the same time for 10 minutes. We’ll then review everyone’s solution, and then draft a composite of the best elements from everyone’s solutions. The designer can then flesh out the details from there.

At the end of Monday’s class, we talked about trust in the online world, and how it acts as the bedrock for transactions such as creating accounts, purchases, and entering personal information. After all, we can’t have much of a usable site if people can’t trust the site enough to actually use it.

Here is the list of qualities we created that help us to identify trustworthy sites and transactions:

• There is discussion about the site in the blogosphere
• We can find other people discussing the site with Google
• The site provides exceeding value (example: mint.com)
• Spelling and grammar is correct

• Content is relevant and meets our expectations
• The mechanics of the interface work as expected (e.g. drop downs work properly, works on all browsers, etc.)
• We can find articles about the site from reputable media outlets
• The site represents a known brand
• There is design consistency throughout the site (fonts, colors, language)
• The site has transparency in its operations and content (examples: getsatisfaction.com provides transparency in customer relations, and visicog.com provides transparency into what happens in our class)
• Privacy statements are clear and prominent (example: “we won’t use your email address for marketing purposes or share it with anyone else”)
• The site displays trust symbols such as VeriSign’s logo or industry badges such as the Better Business Bureau
• Someone we know has recommended the site
• Prominent contact information, including phone number is displayed
• The domain name appears valid (no misspellings or odd characters)