Usability evaluation for interactive television prototypes on a budget

By Konstantinos Chorianopoulos

A common thread in previous interactive television research is the use of PC and Web technology for the development and the usability evaluation of prototypes, instead of commercial digital television platforms. The majority of those implementations are presented within the PC or Web modalities. It usually follows that input devices are the desktop mouse and the keyboard for interacting with the ITV applications. This is an approach that is quite expected, since there is a scarcity of ITV authoring and simulation tools, and most of the time ITV platforms come at a high price, because they are targeted to broadcasters and follow different pricing schemes compared to office software and hardware packages. Yet, academic researchers and budget constrained practitioners should exploit the familiar PC platform, in order to mock-up an interactive television system. Next, I will specify an affordable configuration for seamless user testing at domestic settings, using off-the-shelf components.

The use of inappropriate equipment is detrimental to the validity of user evaluations for ITV applications, especially for those evaluations that employ PC-like and Web-like user interfaces, as it is revealed by the screenshots provided in some of the interactive television publications. For example, ITV applications are presented inside browser windows --complete with navigation buttons and menus-- or inside operating system windows --complete with minimize, maximize, close buttons-- and in all these cases the evaluations are taking place using high-quality --compared with TVs-- PC monitors.

Developers should choose to go beyond the PC and Web experience that intimidates most of the previous ITV applications’ evaluations from consumers. The central element for the experimental set-up could be a laptop computer. The laptop’s display software is configured using the extended desktop setting and the application is set to display at the TV screen. Thus, the laptop’s screen is available to monitor the running ITV application. Then, the laptop’s TV-out and audio-out is connected to the SCART input of a TV screen. The ITV application is designed to run in full-screen and in window-less mode. After running the ITV application, we can close the laptop’s lid and place it away from the TV.

For achieving a seamless television experience it is important to set the application border style to ‘none’, so that there is no visible window around the video. For doing experiments with consumers, it is also important to run the prototype at full-screen mode, so that there is no visible portion of the Windows desktop. Video resizing may be detrimental to the quality, so it is suggested that the video source resolution is similar to the screen resolution used for testing. For example, given a consumer TV, setting the screen resolution at 640x480 (the different aspect ratio is still a problem, but a minor one) and displaying similarly sized video files produces a result that is identical to broadcast television quality, at least in the eyes of the consumers.

For supporting relaxed control with a normal TV remote control, the laptop’s serial port is connected to an infrared sensor that receives the signals from the remote control. The Irman infra-red sensor is strong enough to collect the signals, so it could also be put at the side of the TV. The sensor’s software driver and the supportive applications are used to map the remote control’s buttons to specific keyboard buttons. Then, application developers are programming the actions of an ITV application’s user interface by assigning them to the PC keyboard buttons.

The mapping of the remote control buttons to the keyboard buttons acts as an abstraction layer between the ITV application and the different infrared remote controls that can be used to control it. The whole set-up is unobtrusive and seamless to the television viewer. It also allows the experimenter to take the ITV prototype ‘on the road’ and perform evaluations at consumers’ homes.

In the context of a full scale usability evaluation room, we can also keep the laptop lid open or connect the laptop’s VGA-out to a PC monitor, in order to observe in real time the user’s activity. Furthermore, using a video splitter, the video signal may be also sent to another big screen TV or video projector in the usability lab observation room and super-imposed with the video feed of the user, so that it resembles a full usability evaluation set-up. Further development should also integrate an external TV tuner box that supports real-time capturing and MPEG storage. Adding a TV tuner would extend the ITV prototype functionality to include synchronization between broadcast and local storage and to perform more realistic TV experiments at consumers’ homes, over longer periods of time.

The cost and effort required for obtaining and configuring a high-end laptop computer for play-out on a TV screen using a remote control for input, is very low compared with the benefit of the natural interaction between users and television. Given the current status of usability testing of ITV applications, a more naturalistic set-up would help improving ITV applications’ user interface design. I anticipate that this type of set-up would push forward the developers to think about the end user and the experience of their application on a TV screen, controlled with a simple remote control device. I have already implemented this specification for testing an interactive television prototype.

About UITV.INFO | Editorial team | Questions and Answers | Contact | Privacy Policy | Help
Copyright © 2002-2008 UITV.INFO. All rights are reserved by UITV.INFO or the respective Publishers and Authors.
Reproduction of material from UITV.INFO without permission is strictly prohibited.