john Gill technology header image

Accessible Universal Design of Interactive Digital Television


Dr John Gill
Dr. Sylive Perera, formerly of the Scientific Research Unit, RNIB
March 2003

 

Abstract

When universal design is not applied to a developing technology such as interactive digital television, it is plausible that significant numbers of people will be excluded from accessing or easily using it. Foresight and research into users' needs, if applied, throughout the process can inform a design and help create a system that supports people. Possible solutions involve inclusive (multimodal) design, customisation and standardisation. The RNIB is researching visually impaired users' needs in terms of functions, the design of information, and how they understand and navigate an interactive digital television system.


Keywords

Inclusive design, usability, users' requirements, customisation


Introduction

There are various degrees of interactivity of digital television from a simple remote control button press to sending information back and forth to service providers by means of a return path. So interactive digital television [iTV] can be defined in a multitude of ways. Within this paper it will referred to under two guises [Commission of the European Communities, 2002], "Enhanced Broadcasting" which is when applications are mounted on a virtual carousel and transmitted in a continuous loop, with the viewer selecting applications via their remote control. Data and / or additional multimedia information can be incorporated into the video stream and made available either upon viewer's selection in real time, or stored on the hard disk, thus allowing for 'local interactivity' and permitting viewers to break free from the constraints of linear broadcasting. Whereas, "True Interactivity" means the user transmits requests through a return channel, and the provider supplies individually requested data and services separately from the main video program.

Interactive digital television is a potential revolution in home entertainment, enabling the convergence of many types of media. The scope of application is as great as one's imagination. Beyond entertainment, iTV could become a lifeline for some people by maintaining their independence and subsequently empowering them. When people perceive themselves as active and responsible members of society, they are more likely to maintain their capabilities. Used as a communication device or information point nestled in the heart of the home, for many elderly and visually impaired people iTV could improve their quality of life, if it is accessible.


People and the Technology

The proportion of elderly people is increasing with time and will most likely continue to do so. By 2015, it is estimated that 50% of the British population will be over 50 [Freeman, 2001]. They are the fastest growing proportion of society and are getting wealthier, possessing 80% of the wealth and disposable income. Their expectations will also be higher than those of previous elderly generations. Myths seem to have arisen surrounding elderly people and cutting edge technology. They are thought to be unwilling to learn and interact with such devices and anxious of the effort required. This is documented [Quigley and Tweed, 1999] as being unfounded with most apprehension being based on a lack of familiarity or even contact rather than inability. Recent research by the RNIB Scientific Research Unit involved visually impaired people with a mean age range of 40 - 50 years old. Eight out of ten of the subjects had used a computer previously averaging 15.4 hours per week. One of the subjects currently had no television but two people had analogue and seven had some form of digital television. Those with digital were attracted by the number of channels, better subtitles, possibility of email or because their children had insisted upon it. This sample of visually impaired people indicate that on the whole they are not afraid of technology and they want features that would make it easier for them to use.


Usability

Higham [2001] thought there was a need to persuade designers to think harder about usability i.e. to identify and understand users' expectations, characteristics, limitations, needs, as well as the task functionality and environment. Usability is also important because the public will not invest time to learn the system, they do not want training and rarely read manuals. Counterpoint [2001] found that subjects had different levels of training on delivery e.g. some people were just shown the electronic programme guide (EPG) whereas others were introduced to all the relevant services. Some people only sought the range of channels available but did not investigate other services. Sometimes new services were introduced by another person. When functions were found by accident, the mistake was learnt so it could be redone when needed. Many people learnt to use the service by trial and error [Counterpoint, 2001], finding services by accident or remaining unaware that they had interactive options available. These reactions illustrate unintuitive design.

Inclusive Design

It is very difficult to determine the numbers of people that are excluded from accessing digital television as data is sparse. A subset of those who can not access iTV, includes people who have digital television, but are restricted in their access because the usability of the system does not correspond to their needs so only some of the features are accessible. For example a totally blind person whose children coerced them to purchase an interactive digital television but they themselves cannot read the EPG.

The potential of the television system is unlimited but what if people can not access this developing form of technology. They will be denied a fundamental element of our culture. Radios and televisions differ sufficiently for some visually impaired people to prefer the company of their television to a purely auditory output. Although it may be surprising that 94% of visually impaired people watch television, it is not so implausible when presented with the fact that only 7% of these people are totally blind with no light perception [Bruce, McKennel, & Walker, 1991]. From this we can infer a high proportion of visually impaired people have some useful vision although this may vary in its manifestation e.g. some people may have only peripheral vision whereas others have patchy vision covering their entire visual field. "Already some people, most notably the partially-sighted, have heard of interactive television's convenience, sought to use it - but found services inaccessible" [Kay, 2000].

Some visual impairments such as presbyopia, macular degeneration and cataracts are age related. There is a high positive correlation between the prevalence of visual impairment and age, so designing for visually impaired people will benefit the increasing ageing population. Presbyopia occurs around middle age to most of the population and causes a decrease in the ability of the eye to focus at different distances. People may require longer to view information on the screen as their visual accommodation adjusts. Some people have to move closer to the screen to view textual information but they are reluctant to do this. They may also have to put their glasses on and off as they focus on the remote control and screen in quick succession as is necessary with interactive functions. Good design for people with impairments is frequently good design for everyone. An inclusive design premise aims to embrace everyone's needs and includes not designing separate devices or 'add-ons' which are not upgraded at the same rate as mainstream technology. To decrease the digital divide - between those that have access to digital technology and those that do not - the concerns of those not up-taking digital television i.e. the elderly, need to be addressed [Consumers' Association, 2001]. A by-product of decreasing the digital divide would be a lessening of the burden on society.

Conflict with Industry

Industry is aiming for a stylish, engaging, entrancing, enhanced television environment with a high return on investment. But interactive digital television is a nascent technology still trying to find it's feet, lacking compatibility, interoperability and solid guidelines. There are major differences in the technological capabilities between computers and televisions so how transferable are guidelines for computers or the web? For example the resolution of a television screen is much lower than that of a computer monitor. The television screen's 'safe area' is smaller than a computer monitor's. Additionally people tend to sit at greater distance from their televisions - 7H (where H = the height of the screen) [Silver, Gill, Sharville, Slater & Martin, 1998] - compared to their monitors - 50cm - 75cm optimally [Pheasant, 1991]. These factors require information to be presented in a larger format than is appropriate on computers. Also unlike on computers, it is not yet standard to be able to alter the text size. Other constraints also cause a difference from computers e.g. a limited means of input and output control. Users, their goals and attitudes of interaction differ from those of other screen based systems. Guidelines need to be determined based on scientific principles and specifically for television that will enable designers to create designs according to the needs of the people that will utilise the systems.

Many elderly people and visually impaired people want the minimum amount of confusion within a practical system. This appears to create a conflict of needs between industry and some viewers. Traditionally television is a social relaxing diversion so even the general public do not want to work at it. This disparity is not insurmountable. By adopting a user-centred approach and finding out what people want from this media, industry's mission can be complemented. For example, making an iTV application accessible would enable an old lady (without a computer) to buy her groceries though her television and she may then purchase things she doesn't usually buy e.g. because they are too heavy to carry. Instead of a trip to the video shop, 'Video-on-demand' can supply films at any time, with trailers and digests readily available especially to people who find shops inaccessible. An untapped potential market is lying dormant. Being able to shop from home will not only aid mobility impaired people but those who may have other impairments whether temporary or permanent e.g. people who have had an accident.


Customisation

Consumer surveys indicate that people with 50 channels usually only use about 7 of them [Sweeney, 1995]. Counterpoint [2001] found that although the choice of channels was an influential purchasing factor, real choice was limited for some people to 10-20 channels. Mental customisation was used by some people to exclude some channels and make scanning easier. This indicates that people are attracted to the high functionality but despite this they still require ease of use.

Customisation is the ability for the user to specify their own configuration. An option to display a system specifically to suit any user's needs could be useful to everyone e.g. the interaction of a novice and expert will differ due to the influence of prior use. Elderly people or those with impairments would also benefit.

Universal design is imperative to ensure that everyone can use iTV. Due to the capabilities of digital television, clearer typefaces than the analogue mosaic typefaces can be displayed. Involving specialised user groups, Tiresias Screenfont [Silver, Gill, Sharville, Slater & Martin, 1998] was successfully developed for use on digital television and adopted by the Digital Television Group as the standard for subtitling and text services. For seamless functionality, usability based on a thorough understanding of users' needs should be a primary concern. As people age they can suffer from a decrease in their sensory, mental and physical abilities. Additionally, multiple impairments are more devastating than the sum of the separate impairments. The wide range of television users' needs and requirements may differ considerably so it is not surprising that they may conflict. For example if a visually impaired person wants audio description - when the gaps in the dialogue are used to aurally describe visual information - this could prove annoying to other people who while simultaneously watching do not want this feature. Headphones or another solution would have to be adopted.

iTV multimedia must also be multimodal [e.g. with audio and visual cues] with the option to dis/enable specific features. Options would provide a means of access to iTV to those with special needs but could also be attractive to other sectors of the population e.g. young peoples' infatuation with mobile ringtones and covers. Over specified functionality with numerous options decreases the ease of use of a system, customisation could alleviate this. This would allow the configuration of the presentation - layout, contrast etc. - and functionality of the system - simplifying any interaction.

Customisation would be suitable for this medium once it is set up and functioning but the initial configuration could be tiresome and frustrating if poorly designed. Another consideration is that the system may have to be reconfigured if people with conflicting needs are watching television at the same time, by means of domination or a compromise (similar to deciding which channel to watch).


Smart Cards

Smart cards could provide a viable method of customising iTV and other systems. The user's preferences are stored on the smart card so when it is inserted into the system, it will reconfigure the display appropriately according to the coding on the card. Smart cards have advantages over other methods of adapting systems because they can be transferable between systems and requires less time and input. In the future one card may suffice for use on many devices: a multi service pre-payment card [Gill, 1994].

On interactive digital television, smart cards could be used to control: text size; content layout; speech output; colour combinations; subtitles; audio description; signing; timeouts; reminders and alerts; mode changing capabilities according to level of expertise etc. [Gill, 2002].

The facility to store a user's preferred interface on a smart card was studied by the EU Saturn project [Gill, 1996]. This resulted in the European standard EN 1332-4 which is now being incorporated in a range of systems. Current work is proceeding to extend the standard to cover facilities needed by disabled users in a range of new systems and services. This has required the identification of features on user interfaces for services which have not yet been developed.

A few options may be necessary to ensure ease of smart card set-up: a form that is filled in when applying for the card, a person to fill in the form and issue the card or a machine that guides the process and the issuing mechanism.


Current Research

Different groups of disabled consumers have various problems understanding the potential services that could be provided and the possibilities for adapting the user interface to meet their needs. Therefore specific research protocols have had to be developed to study these user requirements.


Usability Evaluation

The initial work was undertaken with visually impaired viewers to determine how they interact with an interactive digital television menu system. There were ten subjects who had a range of visual impairments including: retinitis pigmentosa; cataracts; an artificial eye and cataracts in the other eye; glaucoma; nystagmus and astigmatism. They were set three tasks to prompt them to comment on the design and interactions: to change the contrast; to change the text size; and to change the colour combination. The experimenter endeavoured to allow them to muddle through the tasks but when the subjects appeared stuck or frustrated, the experimenter posed questions to keep them going.

The interactive digital television display was simulated on a computer terminal in HTML. The results showed a wide range of different requirements for user interfaces which the user would like to select for themselves. It also indicated the majority of visually impaired people tended to use numerical keys to navigate through the menu system and in terms of preference a significant difference was found for numerical buttons over arrow and OK buttons, X2(2) = 6, p<0.05. This implies that it is important for menu options to be numbered so for some people especially the totally blind, number sequences can be used to navigate to an option. Numerical buttons also required less button presses. Symbolic buttons (with direct mapping to the symbol on screen) had shorter response times than buttons with text on, as shown by the table below. A related T-test showed a significant difference between the two types of buttons, t (9) = 2.50, p<0.05.

BUTTONS Mean Reaction Time (seconds)
Symbolic
2.7
Textual
4.9

This table indicates that subjects respond more quickly to symbolic buttons as opposed to textual ones.

Information that was inappropriately positioned or too small was often missed. For example a ½ sign as used in the 'Teletext' model for page numbering was positioned in the bottom right hand portion of the screen. This was not noticed by nine out of the ten subjects and they had to be encouraged to look again so they could realise they were only on the first page and the information they required was on the second page. People tend to scan from the top left of the screen to the bottom right hand corner. The reading stereotype - from left to right - is very strong in the British culture. If cues were clear and legible the visually impaired subjects had less difficulty navigating through the system.

Physiologically, light text on a dark background is easier to read. The ocular receptors are then triggered by the characters rather than the negative space around them which is what occurs with dark text on a light background. To highlight options, a surrounding box or other method should be used which does not decrease the legibility of the text.

When asked which colour combinations they would like to see on their television screen, two subjects replied 'white on black', four replied 'yellow on black' and four replied contrast was the most important factor. This is problematic on a television screen because pure black and white do not display well so contrast is decreased. Based on what is currently available, industry are highly unlikely to produce screens in such monotonous combinations. Although the prototype was not snazzy, boldly coloured, etc. it was not thought to be dull by the subjects. If people are able to alter the presentation of the information, it will render it more accessible. Visual impairment does not lead to a homogeneous population where one solution will benefit all, this is evident even with such a small subject sample. Customisation of settings could enable increased accessibility and usability.

As the colour combinations were scrolled through, their representation was displayed e.g. when the option yellow on black was highlighted, all the text on-screen changed to that combination and so on through the options. This was commended because the interpretation of yellow was evident as was it's contrast to the other colour.

This usability evaluation was conducted based on the notion that one preference card could be used on multiple systems and therefore devices e.g. at an ATM, public phone; for interactive digital television etc. The visually impaired subjects thought that it would be valuable to be able to change preferences for any of the systems through their interactive digital television in the comfort and security of their own homes.

The subjects' main concerns were with the presentation rather than the movement through the system. For those who could decipher the on-screen information, it was easy to use but impossible if this was not accessible to them. Further work was deemed necessary commencing with a user needs study. What specifically do people want and require?


Visually Impaired User's Needs

A questionnaire was sent to over four hundred visually impaired people asking them what features they would like on their digital televisions. This included: changing text size; changing the colour combinations; the use of icons; speech input and output; audio description; more time to review any information and the option to decrease the number of functions available.


Experiments

Vague guidelines for iTV within the interactive digital television community determine the areas where obvious problems arise e.g. avoid: text in graphics; multiple columns; fine lines; pure white, black and red. But where do the limits lie? How should these features be designed? If these are to be customisable options what is the best method of applying / altering options? This would provide a basis from which the standard for the coding of special needs on a smart card can be extended and a set of guidelines created.

This work is in progress but focuses on issues such as text size, flashing text, scrolling text, contrast, shapes of icons. From the previous usability evaluation, the largest text size was easy for most of the subjects to read but a few would have preferred even larger text. It is unclear what are the smallest, largest or preferred text sizes for different people. This has serious navigation implications for the system so it needs further investigation. If the text is larger, then more screens are required to display the same amount of information so the navigation becomes more complex. Information presentation is becoming increasingly important and needs to be precise and concise. Would people mind being shown less information?

Fancy text and images whilst being inaccessible themselves can actually decrease the legibility of the whole screen by confusing and negating vital information. Flashing text certainly grabs the attention but it can also be distracting, annoying and difficult to read. Is continuous flashing text acceptable for important information or should it only flash a few times and then cease? Scrolling text can also be difficult to read. This is thought to be a function of speed and size and is currently being investigated.

Contrast was highlighted as an important factor and is being explored further. What is the acceptable minimum contrast level?


Further Work

Further detailed work is thought to be necessary into navigation, the display of forms - in columns? - , colours and their combinations, the integration of the remote control and content density. At what point does the screen become too crowded? The space syntax of the screen estate requires investigation. Other issues surfaced during the research. All the subjects who had interactive digital television did not use it's text service because it was too slow, lacked progress feedback and was difficult to read and use. They continued to use the analogue version. The digital text needs to be improved before the analogue signal is terminated.

"If only those running interactive television services would … take the needs of - and advantages to - disabled and elderly people into account in their marketing and on-screen design" [Kay, 2000].


References

Bruce, I., McKennel, A. & Walker, E. (1991) Blind and partially sighted adults in Britain: the RNIB survey. London: HMSO.

Commission of the European Communities (2002) Barriers to widespread access to new services and applications of the information society through open platforms in digital television and third generation mobile communications. [accessed 29/08/07].

Consumers' Association (2001) Turn on, tune in, switched off: consumers attitudes to digital TV. [accessed 29/08/07].

Counterpoint Research (2001) Digital television - consumers' use and perceptions: a report on a research study. [accessed 29/08/07].

Freeman, J. (2001) The public's perception of the ease of use of domestic digital TV equipment. The ITC, Design Council and Consumers' Association's Easy TV Seminar, London, 28 November 2001.

Gill, J.M. (1994) Text display preferences on self-service terminals by visually disabled people. [accessed 29/08/07].

Gill, J.M. (1996) Smart cards: interfaces for people with disabilities. [accessed 29/08/07].

Gill, J.M. (2002) Tiresias - guidelines: television. [accessed 29/08/07].

Higham, N. (2001) BBC News: designing TV for Granny. [accessed 29/08/07].

Kaye, J. (2000) Why interactive will mean active. Guardian: August 16.

Pheasant, S. (1991) Ergonomics work and health. London: Macmillan Press.

Quigley, G. & Tweed, C. (1999) Added-value services from the installation of assistive technologies for the elderly. [accessed 29/08/07].

Silver, J., Gill, J., Sharville, C., Slater, J., & Martin, M. (1998) A new font for digital television subtitles. [accessed 29/08/07].

Sweeney, J. (1995) An introduction to interactive television. International Broadcasting Convention 94, Netherlands, 16-20 September 1994, 503-508.

Further information

Ofcom (2006) Summary of research on the ease of use of domestic digital television equipment. [accessed 29/08/07].

 



John Gill Technology Limited Footer
John Gill Technology Limited Footer