Recommendation Letter for Krystian Samp’s Thesis: The Design and Evaluation of Graphical Radial Menus
I am writing this letter to enthusiastically recommend that you consider Krystian Samp’s thesis, “The Design and Evaluation of Graphical Radial Menus”, for the ACM Doctoral Dissertation Award.
Don Hopkins, October 31, 2012
Dear Sir or Madam:
I am writing this letter to enthusiastically recommend that you consider Krystian Samp’s thesis, “The Design and Evaluation of Graphical Radial Menus”, for the ACM Doctoral Dissertation Award.
I believe his dissertation is a significant contribution to basic computing research that focuses on user interface design, because it clearly presents a carefully planned, controlled study of a thoughtfully justified and practically executed design, which valuably advances the state of the art. Recognition for Samp’s work from ACM would dramatically raise visibility of research excellence that contributes to theory and practice.
Samp first contacted me six years ago, after reading my articles on pie menus, while he was developing a prototype of his radial menu design, researching related work, and asking people for feedback, suggestions and references to prior art.
Jack Callahan, Ben Shneiderman, the late Mark Weiser and I published a paper in CHI’88 about “An Empirical Comparison of Pie vs. Linear Menus”, which inspired Samp to build and evaluate his own original designs for radial menus.
During my career, I have continued to build on that work, and encouraged other people to explore the possibilities of user interface design. So I was delighted to hear from Samp and help him out, because he is dedicated to learning and building on other people’s prior art, developing his own original ideas, applying sound research, practical design, empirical testing, and contributing to computer science literature and user interface design through peer reviewed academic publications. I’m proud that he has persevered for so many years, written such a high quality dissertation, and successfully completed his PhD.
He enthusiastically described his design and his prototype, and asked many pointed questions. I encouraged him to develop his prototype, try out his ideas first hand, and measure their effectiveness. We discussed use cases like mobile touch screen interfaces, desktop interfaces, eLearning environments, user defined and dynamically generated menus, Semantic Web visualization, taxonomy browsing, fixed design interfaces for games like SimCity, dynamic interfaces for games like The Sims, customizable interfaces for games like World of Warcraft, and implementation technologies like web browsers, Python, Lua, Qt, GTK, and the One Laptop Per Child “Sugar” user interface.
He sent me a link to a video demo of his prototype, and asked for advice about scientifically evaluating his work, and writing a good paper. He understood that it’s not about winning a horse race or beauty contest, but about objectively measuring performance using scientific methodology.
So I brought Ben Shneiderman into the conversation, who was impressed by the design and video demo. Ben offered some constructive criticism, and encouraged Samp to build more prototypes, perform controlled studies to measure performance on key tasks, build a model on which to base an evaluation, get many people to use the interface, give talks about it, and he referred Samp to papers to read and people to contact.
Ben Shneiderman has reviewed this letter and adds his support for recognition of Samp’s contributions.
Here is a link to a web page with references to some articles and videos that I have shared and discussed with Samp over the years:
Samp’s designs are well informed by previous work, including successful models like Fitts’ Law and GOMS/KLM. His goals are practically focused on shortening selection time, decreasing navigational difficulty, lowering error rates, supporting both novice and experts, and smoothly accelerating the transition from novice to expert.
One exceptional feature of his thesis is the set of well-defined goals that he has articulated, and how he has used them to evaluate each design decision, by weighing each alternative against specific goals, and making conscientious trade-offs. His goals include increasing selection speed, decreasing error rates, improving ease of use, guiding exploration, enabling effective visual search, supporting navigation, accelerated novice-to-expert transition, handling large quantities of items, and reducing screen space consumption.
His focus on supporting both novices and experts with one consistent design, and easing the continuous transition from expert to novice, is sharpened by scoring each design decision against those goals.
What disappoints me about the current state of the art in commercial user interface design is that it’s so compartmentalized into “user experience design”, “graphical design”, “programming”, “software architecture”, etc. Few people have the perspective to take those and other aspects of design into consideration. The web, desktop and mobile devices are awash with user interfaces designed by people who have mediocre expertise in one or at best a few of those areas. They produce “cargo cult” imitations of the surface characteristics of traditional designs, and don’t bother to evaluate their efficacy. I’ve seen many disappointing implementations of gimmicky “ersatz pie menus” that at first glance look cool and shiny, but totally fail to reap the benefits of Fitts’ law, and are hard to use, slow, error prone, and inflexible.
That is why Samp’s work is so important, stands apart in refreshing contrast from the rest, and serves to inspire both commercial and academic user interface designers and researchers: he creates empirically tested designs based on sound principles, and articulates practical goals against which he and other designers can evaluate new ideas. More good work like this needs to be done to explore and evaluate the design space and study the characteristics of radial menus.
Although I chose a more commercial, less academic path for my own career, I continue to reformulate and use pie menus to create commercial products, tools and games, exploratory research prototypes, and reusable open source components. So it is beneficial for me to read papers like Samp’s, because they provide insightful ideas and useful tools like his designs, models and goals, which inspire and help me to re-evaluate my own ideas. I am sure other people will appreciate and benefit from his research in the same way that I have.
Samp’s work deserves recognition for carefully and constructively building on the work of others, then giving back innovative designs and refined theories that advance the discipline and practice of computing.
Sincerely,
Don Hopkins