ITV, user interface design, usability test, evaluation, user requirements, graphic design, remote control, direct manipulation, interaction design

As the Internet is growing rapidly experts predict that it will move into all areas of our private life [ 1] . Hence, new services in the area of infotainment, edutainment and electronic shopping have great commercial and social potential. However, the question is how private users will gain access to these services and which technology will be employed [ 2] . With the development of interactive TVs new user groups may access online services. People who are inexperienced with the Internet or the PC in general may now use the new technology to access electronic services with a familiar device – the remote control and the TV set.

At the same time experienced surfers may not want to buy an extra PC when they can use one at work – but still want to check their e-mail or the flight schedule while watching their favourite show at home. Thus, online services will be available on both platforms in parallel in order to be successful. Then, however, users will demand a consistent navigation concept and a common graphical design.

A user-centred approach was followed in the development of a cross platform graphical interface for an online service that can be accessed via PC and ITV. Partner of this study was the service provider InfoCity, which offers a pilot online service for people in the Ruhr Area, Germany.

The design of a graphical user interface (GUI) for both PC and ITV platforms and its iterative evaluation in user tests were the goal of this study. The focus of the design was ease of use. People with no or little background with PCs and the Internet. Should be able to easily navigate on services on both platforms. Thus, a common design had to be found despite different screen resolutions and hardware interfaces (mouse versus remote control).

Besides the technical requirements the following restrictions for design and evaluation of the prototypes had to be considered:

• Control of the graphical ITV interface with a remote control and "focus-jump" (key-navigation up/down and left/right with the remote control)
• Direct manipulation with the mouse on the PC platform.
• Zoning, i.e. designation of functional areas on the graphical user interface (selection area, navigation area and contents area / menu area)
• Orthogonal structure between and within areas to provide for the limited degrees of freedom of the remote control (up/down, left/right)
• Page-oriented layout (reduction of the need to scroll).
• Integration of existing corporate-design of the service provider.
• Easy-to-use navigation elements and comprehensible information structure with regard to different user profiles (newcomers, experienced users) [4].

These requirements were translated into the a graphic design concept with multiple navigation elements and implemented as software prototypes (Figure 3). The graphical interface is divided into three functional areas: Task/selection bar, navigation bar, graphic menu/content area.

The area on the left side of the screen comprises the task bar with the selection menu. It offers access to the basic functions and main services of the application.

Unlike the TV version the PC version is equipped with a shortcut bar in the selection menu/task bar. Here, the text links to the first level of the service menu are given. Thus, when the user is in the service overview, he can either select the category (travel, leisure etc.) via the shortcut bar or by selecting the service in the graphic menu area.

The navigation bar in the upper display area consists of two lines. The first line represents a history list with links to those levels the user has already visited in the hierarchical information structure of the service (e.g.: "Services > Entertainment > Movies"). Thus, the user can simply click on one of the links to go back to the level and its menu of her choice. In the second line the options of the current level are shown, for example the list of movie theatres in the movie menu. If there are more entries that can be shown in this list one can scroll the list via the "arrow buttons". Below the second line of the navigation bar the title of the active page is shown (e.g. "movies").

In the large menu / content area in the centre of the screen either the menu options of the current level are depicted as thumbnails or, in the lowest level, the selected content is shown, e.g. the new movie in your neighbourhood theatre. The PC version can provide up to nine menu entries whereas the TV version is limited to six entries due to the reduced space available.

The PC version offers direct manipulation by mouse control, however, without the functions of the browser bar. The interface of the TV version is accessed by remote control with four directional keys (left-right-up-down). Hence, there is no continuous pointer movement as with the mouse. Rather, the direction keys move the cursor in an invisible grid over the screen from cell to cell. Each link or option can be thought as one cell (or the "focus") of that grid ("focus jump"). The selection of the link or an option in the focus is done by pressing the "ok" button on the remote control. A "back" button brings the user to the last upper level.

In particular, the following questions were addressed in the user tests:

1. Design and Navigation Concept:

• Which navigation elements are used by the participants (e.g. shortcuts, navigation bar, etc.)?
• How well can users navigate in the hierarchical information space (5 levels)?
• What is the role of graphic design in navigation concepts?

1. Comparison of Interfaces:

• Is the navigation concept user-friendly and effective on both platforms?
• Which one of the two versions is preferred and why?
• How do users handle the remote control compared to the mouse?

The design and usability of the user interface was tested on both platforms (PC/TV) by 13 users (six women and seven men) aged 17 to 69 years. The users were computer literate but had no or little experience with the Internet. They had to solve seven tasks each for the PC version and ITV version. They had to search for certain information within the service categories and were also required to use some options of the task bar ("about us").

All participants were able to solve all tasks without direct help. In general, users were able to navigate successfully in an information depth of five levels, even without prior Internet experience. According to the participants, they never felt lost in hyper space. All navigation tools were used to a substantial degree, which suggests that multiple navigation elements allowed successful navigation (Figure 2).

For a quantitative comparison of both versions, a two-factor analysis of variance was computed based on the number of mistakes per user with the factors sequence of using a version (TV or PC used first or last) and type of platform/interface (TV/PC). Those operation steps were considered as mistakes, which exceeded the minimum number of selection actions needed to accomplish the task.

In the TV version, participants found the shorter navigation paths not as easily as in the PC version which differed only by the shortcut bar and the larger amount of information depicted. However, this difference is statistically not significant. There was no indication of a specific learning effect regarding the number of mistakes: It did hardly make a difference whether the users started with the TV version or the PC version. Thus, navigation performance was equivalent on both platforms in terms of effectiveness.

The items of the usability questionnaire following the tasks were adapted from ISO 9241, part 10 - 11 [3] according to their relevance for the present study. In general, users rated both interfaces as positive (Figure 3).

The PC version did slightly better than the TV version in four categories: Error tolerance, conformity with user expectations, suitability for learning, and general acceptance. However, the rating differences between the platforms are only marginal.

In a preference questionnaire there was no preference for one platform regarding the question "which version was easier to learn?" which confirmed the results of the usability questionnaire. However, the PC interface was clearly preferred regarding ease of operation, buying preference, and general preference (Figure 4).

Twelve out of thirteen users stated they preferred the PC version regarding ease of use even though both versions had done about equally well in the usability ratings and although the users did not need differ regarding the number of clicks to find information. The explanation may lie at the hardware interface: twelve participants preferred using the mouse to the remote control regarding comfort of use because the former offers control by direct manipulation: it took longer to select a link using the "focus jump".

All users who preferred the ITV version in one of the remaining categories (buying preference, general acceptance) did rarely or never use a PC. They preferred the TV version because they are used to the remote control.

Using „multiple navigation elements" appears to be an efficient "design for all" approach in the case of a cross platform interface. Not only does it provide for different navigation styles and user needs, but also do peoples’ usage of navigation elements change during the course of interaction due to the increasing proficiency with the interface and knowledge of the information space. In general, the performance and ratings of the participants who tested the interfaces confirmed the usability and potential acceptance of the design approach

Usability of graphical user interfaces and navigation concepts is certainly a prerequisite for commercial success of online services, but is not the only criterion for customer acceptance. With regards to ease of use, most users preferred a PC platform. This is due in particular to the still relatively inconvenient remote control of the ITV. Compared to the functions available on current ITV controls (e.g., mini-joystick), the simple remote control with four-arrow-keys and "focus jump" is an acceptable alternative, at least with an orthogonally structured GUI. However, the development of an adequate control device for direct manipulation of an ITV graphical interface would certainly increase user acceptance.

Whereas users without PC experience were more willing to purchase an ITV as a platform to access electronic services, experienced PC users will certainly not dismiss the extended possibilities of a PC. However, they will possibly want to use already available ITV platforms, as long as those offer a similar degree of convenience. This emphasises the importance of user-centred developments of online services as well as that of a platform overlapping design in order to reach new user groups and at the same time satisfy the needs of experienced users.