The full-day workshop on Assessing Multimodal Interaction will be part of the 7th Nordic Conference on Human-Computer Interaction (NordiCHI) held in Copenhagen, Denmark from 14-17 October 2012.


The main challenge in assessing multimodal interaction is that present usability and user experience metrics fail to account for the specifics of different modalities. How do the features of input and output methods affect their evaluation and the questions we as researchers should be asking? For example, recognition-based input (speech and gestures) is likely to cause different kinds of problem situations compared to input that requires no subsequent processing (e.g. mouse and keyboard).

At the moment, we do not know how the different modalities affect usability/user experience ratings provided for the multimodal system. Regardless, current research often makes the implicit assumption that methods for evaluation of unimodal interaction can be applied to other modalities or multimodal assessment.

Workshop theme and goals

The workshop is intended to offer a platform for practitioners from academia and industry to share their experience with collecting and analyzing data on multimodal interaction. The goal of the workshop is to get an extensive overview and summary of useful metrics and data integration techniques for assessing multimodal interaction. For that purpose, we kindly ask participants to submit a short paper (up to 4 pages in the NordiCHI paper format) that describes how they deal with metrics for multimodal interaction. It can be based on previous research, but should give a more detailed description on how the data were collected, processed, which methods were used for analysis, and what issues were encountered during the evaluation. This can also include feedback from reviewers.

The focus should be on empirical studies that cover the evaluation of interactions with different input and output modalities, such as speech, touch and vision, and different interaction techniques such as gestures, eye tracking, physical touch, haptic feedback, speech recognition and synthesis, non-speech audio, etc. The metrics can cover different data types, qualitative as well as quantitative, and both subjective and objective measurements; for example interviews, questionnaires, observations, log files. The requirement is that the analysis led to an overall judgment of the multimodal usability/user experience of the device or system, beyond an assessment of a single modality.

We want to bring together experienced people from the HCI community dealing with issues of multimodality and provide room for the exchange of ideas and opinions. Particularly the workshop provides an opportunity to share knowledge, to get feedback for own work and to initiate corporations. The workshop will be held interactively; specific subtopics will be discussed together on the basis of evaluation scenarios provided by the organizers and selected presentations of the participants.

Download the call for papers here.