Accessible Digital TV: a research agenda Giorgio Brajnik Dip. di Matematica e Informatica Universit` a di Udine Udine, Italy giorgio@dimi.uniud.it EXTENDED ABSTRACT The potential of Interactive Digital TV (IDTV) is enormous, given its reachability, the potential ability to provide highly personalized services (when users identify themselves us- ing smartcards, for example to get access to individual e- government services, to health information and services, to social networks), and to act as yet another outlet for using the web. However these advantages can be hindered by difficulties of using it, including accessibility barriers. Similarly to the web, however, developers of IDTV applications will have no control on the actual devices the user will be using, and will have to address the needs of a very heterogeneous audience. Input devices are likely to be a major hurdle for achieving accessibility. Remote controllers are already very complex devices based on modal interaction and overloading of con- trols. Furthermore, they support a limited input channel, that poses significant constraints on the kind of applications that can and should be developed. Another hurdle to overcome is the possible resistance of users to actually use and engage in applications delivered through the IDTV. While over time we should expect that the experience of users in managing digital technologies will in- crease, during the first years after the switchover, the IDTV will be used also by people that are laggards in terms of tech- nology adoption. Most of them will probably be older adults, with their own accessibility needs. Only if these needs will be met, then IDTV will be able to reach its full potential. I think that a research agenda on accessible IDTV should include methodological aspects. Accessibility is not only conformance to a set of guidelines 1 . While adoption and de- velopment of accessibility guidelines is important, it should not be the only aim of the research community. In [3] I dis- cuss some of the shortcomings of an accessibility-as-conformance viewpoint, which we might be able to avoid for IDTV. I believe the following issues are central to the topic of accessibile IDTV. 1 In [2] I listed 9 different definitions. The one I usually prefer in the context of quality assurance is: accessibility means that disabled people can use the artifact with the same effectiveness as non-disabled people; this definition, for the web, was first given by [6, 5] Accessibility Model: Developing a model of accessibility that helps understanding, characterizing and operational- izing the concept. In particular the model should ad- dress the following questions: 1. what accessibility is and which properties should be central to it (eg. effec- tiveness, productivity, usability, satifaction), 2. which factors influence accessibility and how can they be controlled (eg. experience level with IDTV, attitude towards digital devices, physical conditions under which the IDTV is used, etc.) 3. how accessibility is go- ing to be evaluated and measured (through analyti- cal methods like conformance review, barrier walk- through; through empirical methods like user testing, subjective assessments, etc.). Accessibility Evaluation Methods: Several AEMs can be used to assess the level of accessibility of an appli- cation, but little is known about them. In particular, properties like validity, reliability, usability of Acces- sibility Evaluation Methods are little studied at the moment. Yet this is crucial to build confidence on methods and obtain a reasonable level of consensus. For the web this did not happen (yet), but more and more researchers are focusing on methods. For the IDTV we will need a set of agile and valid evaluation methods, i.e. 1. such that they can be deployed with limited resources (effort, knowledge, time, infrastruc- ture), 2. that can be scaled up depending on the desired level of quality of their results, 3. that can be used to predict the true problems that will show up when the application will be used. Database of typical barriers: For practictioners assessing accessibility of IDTV, and for developers as well, it would be useful to be able to draw onto scientific lit- erature presenting evidence of certain kinds of acces- sibility barriers, and their consequences on users. For the web, specific interpretations of existing (generic and technology-agnostic) guidelines exist, and can be used as a (limited) reference so far. A particular im- portant point here will be the focus on older adults [1, 4]. Sustainable Accessibility: We should be able to provide ways to estimate the return on investment for accessibil-