Martin Frlin Novak (2019) Razvoj programskih orodij za oceno kognitivne obremenitve. EngD thesis.
Abstract
The thesis is motivated by the problem of fragmented attention. Multitasking is useful, however, uncontrolled interruptions coming from a large number of mobile devices and applications lead to poor task performance, frustration and other negative consequences. There have been attempts to control technology-induced interruptions. For example, by delaying notifications until more appropriate times. These moments are often characterized by low cognitive load of a user. However, inferring cognitive load is challenging. To date, the only options we have rely on rather intrusive techniques, such as pupil dilation measurements, requiring expensive sensing equipment, etc. The option we explore is whether cognitive load can be inferred through cheap, off-the-shelf devices, such as wearable computers (fitness wristbands, smart watches, and similar). In this thesis we present the design and implementation of an experimentation environment for cognitive load estimation. Our setup consists of a desktop application that presents the user with tasks of varying difficulty, another desktop application that runs in parallel which, through a side-task, aims to gauge the user's cognitive engagement, a wearable sensor that through a mobile phone app collects data about the user's physiological signals, and pre- and post-test questionnaires that evaluate the user's overall cognitive capacity and personality. In the thesis, we provide detailed explanations of the testing environment, especially the parts that were originally developed for the thesis -- the secondary task desktop application, the web-based pre/post questionnaires and tests, and the database orchestration -- and report on the experiences with real-world experiments conducted with 27 volunteers.
Actions (login required)
![[img]](http://eprints.fri.uni-lj.si/style/images/fileicons/application_pdf.png)
