LOCATION: Auditorium 1, TEL, Ernst-Reuter-Platz 7, 20th floor

SPEAKER: Xin Guang Gong

ABSTRACT:

In an experiment we analyze the influence of specific user characteristics on the quality of linear regression models for user judgment predictions from interaction parameters. From a former study with a speech-based home assistant system, we know that user characteristics like age, technical affinity and the score from the digit-span test have an influence on the accuracy of judgment predictions. The aim of this bachelor thesis was to confirm these results for touchscreen-based apps, using 16 older and 16 younger users. contrary to the former study with the speech-based system, the strongest effect on the predictions are observed for age, technical affinity and technical experience.

HOST: Klaus-Peter Engelbrecht