the process:

background research
software style guidelines
ti connect 2.0
audience identification
initial user testing
prototyping of ti connect
user testing guidelines with usability professionals

background research
We began by doing research into how technology is currently being used in classrooms, how students learn and how teachers teach. Some of the resources we consulted include:

  • Oversold and Underused: Larry Cuban
  • How People Learn: John Bransford
  • Improving Learning with Information Technology Workshop Report: the National Research Council
  • Stanford Research Institute: www.sri.org
  • Nonprofit educational research: www.concord.org
  • Center for Learning Technology: www.cilt.org
  • University of Michigan, PDAs in Education: www.hice.org

software style guidelines
The software style guidelines were created to help Texas Instruments develop consistent, usable applications. Developing consistent software under a common set of guidelines will help users increase recognition and transfer learning across programs.

The guidelines are intended to serve as a reference point when specifying, developing and testing new TI software. Individual guidelines may address specific design decisions but the document should always be considered as a whole when designing and developing applications. Refer to the checklist [to be included] near the completion of the specification and development phase to ensure all areas of the guidelines have been considered. If a new case appears that is beyond the scope of the current guidelines, note the cause and solution and update the document accordingly to ensure future applications address the issue in a consistent manner

While the guidelines provide an extensive checklist to consider while developing TI applications, it cannot cover all situations. It is important to get a global perspective of the guidelines and always develop software with the intention of keeping the program consistent and familiar to users.

ti connect 2.0
Though our initial research, we discovered that many teachers are unaware of existence of TI computer software, are unable to differentiate between the applications, and have difficulty understanding what they can do with TI software. Our vision was to create a unifying application that could provide easy access to all TI computer software from a single place, inform users about the software’s capabilities, and serve as a central reference point for other TI programs & services.

audience identification
Based on Texas Instrument’s recommendations and Carnegie Mellon’s user research high school teachers between 9th and 12th grade were considered target users for TI applications. Since TI applications are directed towards high school level math classes and teachers are the primary decision makers on how lessons are conducted within the classroom, the software must meet immediately meet their needs to have the largest impact. Students were considered secondary stakeholders since they will use a limited set of features within the software and will respond to the requirements of the teacher. The reference to TI software “users” throughout the software style guidelines refers primarily to teachers, with secondary consideration given to students.

initial user testing
TI applications were evaluated to identify inconsistencies and design opportunities within the interface. Guideline recommendations were also tested with these programs to ensure compliance with existing software features and functionality.

To begin the guideline development process, TI software was tested with teachers at T^3 conferences, the Carnegie Mellon user studies lab and teacher’s classrooms. Novice and experienced users were tested. Studies were recorded through notes, video tapes and screen capture utilities. For a full list of tasks, user profiles and test recordings refer to Appendix [B] of the online version of the guidelines.

To identify problem areas within the program users were asked to complete representative tasks while a test monitor observed. As users completed the tasks they were asked to think aloud and describe their expectations at each step. This testing technique provides insight to the assumptions and inferences that users make while using the program and exposes areas of the program that do not behave as expected. If any area of the software was confusing or negatively affected the completion of the task it was recorded as a “breakdown”.

prototyping of TI Connect
The prototype of this redesign was built and tested using Think Aloud protocols with 4 users.

user testing guidelines with usability professionals
Once the guidelines were in their final draft stage, we hired five usability professionals to spend an extended period of time trying to either redesign an existing TI software, or design a new one from only a spec. What they produced allowed us a glimpse of how these guidelines would actually be used. We incorporated their sugesstions, confusions and successes into our final guidelines document.

July, 2003