User Experience News, Blogs, and Videos
In the youthful field of usability, we may be surprised to hear about "venerable" measures of usability. However, that's just what we have with the DEC "System Usability Scale" (SUS) copyrighted in 1986 and publicly discussed in 1996 by John Brooke.
Say you are tackling a vexing task for redesign. Would you like to get a baseline questionnaire measure of subjective "ease of use" on the original design? And then would you like to compare your redesign with
a follow-up measure (to see if the design works better)?
Do you want a questionnaire to be fast, fast, and fast (otherwise known as "Quick and Dirty")? Of course.
And that's what motivated John Brooke and colleagues to invent a ten-question assessment that takes about 90 seconds to fill out. In fact, it's so easy your test participants could fill it out several times during a longer usability test session."Quick and (not so) dirty" means you can get data that measures user-friendliness by task! Plus you can compare your design progress over time!
Wouldn't it be nice to see whether our design is just "OK" versus "good," "excellent," or "best imaginable"?
Put another way, wouldn't it be nice to know...
I've read a lot of research. Because of their visionary efforts across ten years, my hat goes off to the team of three researchers who systematically collected almost 3,500 results of SUS surveys over those ten years. The team consists of two gentlemen from AT&T Labs, Aaron Bangor and James Miller, and a professor at Rice University, Philip Kortun.
They used the SUS survey shown here across six different interface design contexts: Web (41% of the 3,500 responses), Cell phone (17%), IVR (Interactive Voice Response) (17%), GUI (Graphical User Interface) (7%), Hardware (7%), and Television (5%) interface products. (This graphic comes from their published study.)
See Brooke, J. (1996) for scoring methodology and the original text. (Our Bangor et al researchers found that the word "awkward" in statement 8 worked better than the original word "cumbersome". They also use the word "product" instead of "system".)
Mirror, mirror on the wall, who's the fairest of us all?
Our authors asked an important question, which affects us all.
What does a specific SUS score mean in describing the usability of a product?
Do you need a score of 50 (out of 100) to say that a product is usable, or do you need a score of 75 or 100?
The first part of their answer was to look at the distribution of scores across the 3,463 questionnaire results contained within their 273 studies.
So now we have a sense of the middle score (70) and the average score for the top 25% of studies (scores ~78 and above).
Does the idea of your design falling in the top 25% of studies give you a feeling for "good"?
Well, we hope that the top 25% must have some value. But we need more evidence.
We could look at other studies. For example, Tullis and Albert (2008) show that an SUS score of 81.2 puts you in the top 10% of their particular sampl.... So, we have a second snapshot of quality – the top 10% of another group of SUS surveys.
Pop the question to seal the deal
But even better, our three authors analyzed a final, single question added to 959 of their recent SUS questionnaires. Participants picked one of these adjectives after answering the SUS questions.
Here's how the SUS scores matched the adjectives (all graphics come from the study publication):
So now you have some adjectives to include in your report. You can give the score, and this chart shows you which adjective matches the score.
Do it yourself, too; and make the grade
Better yet, include the adjectives with your own SUS questionnaire. Let your participants give you an overall evaluation directly. See how closely your average SUS results match the chart given above.
Which leads us to how you can give a "grade" to your design.
Our three authors make a speculation. It goes like this.
The traditional school grading scale uses numeric scores that represent the percentage a student has earned out of the total possible score. Remember?
A score of 70% to 79% on a test got you a "C". A score of 80 % to 89% got you a "B" and a score of 90% or more got you an "A". (You never got a "D" or an "F", right?)
See what our authors propose in the context of matching their adjective results with a proposed "grading vocabulary". Does it make sense to you? It sure makes sense to me.
Recall our second goal for usability metrics: to determine "how well our web design encourages visitors to come back to our site". This refers to site "loyalty".
For this we turn to another active web researcher, Jeff Sauro, magi of www.measuringusability.com. Sauro (2009) recently published in his newsletter the results of his study somewhat similar to the adjective study above.
Sauro examined SUS data from 146 participants tested in a dozen venues such as rental car companies, financial applications, and websites like Amazon.com.
In addition to the usual SUS queries, test participants got one extra question: "How likely are you to recommend
this product to a friend or colleague?"
This extra question, turns out, has quite a pedigree. It gives a "Net Promoter Score" (NPS). Some authors report this one question offers good prediction of long term growth for a company.
Upon correlating scores from the NPS with overall results of the SUS, Sauro found that the SUS explains about 36% of the variability in the Net Promoter Scores.
He points out that people identified as a "Promoter" have an average SUS score of 82 (plus or minus about five points). So, if you want your web site to serve as a beacon for loyal customers ("promoters"), strive for an SUS score above 80.
Recall that earlier in this article, an SUS score of 80 gets a "B" reflecting roughly the midpoint between the adjective phrases "Good" and "Excellent" (see Figure 3 above).
So making the grade of "B" also gets you some customer loyalty. Not bad.
SUS-tainability: the gist of it all
Here are a few points to help you attain SUS at-one-ment.
Check out the references below for details. I've used the SUS many times. It helps you communicate to any client the overall subjective response of usability test participants.
If user experience is important to you, then the SUS gives voice to the experience of your users.
Bangor, Aaron, Kortun, Philip, and Miller, James, 2009. Determining What Individual SUS Scores Mean: Adding an Adjective Ra... Journal of Usability Studies, 4 (3) May 2009, 114-123.
Brooke, John, 1996. SUS: A "quick and dirty" usability scale. In P. W. Jordan, B. Thomas, B. A. Weerdmeester, & I. L. McClelland (Eds.), Usability evaluation in industry (pp. 189–194).
London: Taylor & Francis.
Sauro, Jeff, 2009. Does better usability increase customer loyalty? The Net Promoter S... 7 Jan 2010.
Sorflaten, John, 2009. Wherefore Art Thou O Usability? – Cognitive lock-in to the rescue. HFI Newsletter, Nov, 2009.
Tullis, T. S. & Stetson, J. N. (2004, June 7-11). A Comparison of Questionnaires for Assessing Website Usability, Usability Professionals Association (UPA) 2004 Conference, Minneapolis, USA.
Tullis, T. S. & Albert, B. (2008, May 28). Tips and Tricks for Measuring the User Experience. Usability and User Experience 2008 UPA-Boston's Seventh Annual Mini UPA Conference, Boston, USA.