Editors: Lars Müller, Veronica Rivera-Pelayo, Andreas Schmidt, FZI.

Link to the work package (WP3): WP 3 Capturing learning experiences

Delivered: M12 (June 2011)

Link to the deliverable (pdf): D 3.1 User studies, requirements, and design studies for capturing learning experiences

Executive summary:

This document describes the research process of work package 3 (WP3) in the first year of the MIRROR project. Work package 3 aims at “Capturing learning experiences” and thus support reflective learning. A learning experience is understood as an experience relevant to the reflective learning process. In the context of MIRROR, we focus on learning experiences that relate to work practices. Therefore, an experience is relevant if it has the potential to lead to a work-related outcome of a reflection session.

 
The collection of data is complicated by the unpredictability of the (captured) data relevance and the subjectivity of experiences, which demands interpretation of the captured data by the employee. WP3 plans to capture affective aspects and consciously guide the user’s motivation to facilitate the selection and interpretation of the available data.

In the first year, WP3 took an exploratory approach to identify sensing technologies that are realistically usable and that contribute to capture meaningful data for a “learning experience”. This compromises technologies to capture the task context (objective 3.1) and the user’s physical and emotional stress as context information (objective 3.2). First steps towards a context management and representation (objective 3.3) have been taken. This means especially creating a common understanding of captured context and its usage across all work packages.

Five studies were conducted to reach these goals:

Research on available technologies in the testbeds and possible new sensors has been carried out in the pre-study “Capturing technologies”.

• The design study on using Mood Maps has confirmed that this simple conceptual approach to capture the affective dimension is effective and accepted by users.
• A domain model was created in discussion with the other work packages. The domain model has achieved a shared understanding of the various work packages of the captured data.
• Psychophysiological sensor data was collected in a study at NBN and annotated with notes from a parallel ethnographic study. This data is an important basis for developing and testing new algorithms and applications in the upcoming months. Furthermore, motivational barriers and drivers have been identified.
• An offsite user study has been carried out by KMRC, whereby they have asked employees from all testbeds about their current usage of technology and their acceptance of additional sensors.

The insights of these five studies have been amalgamated into a first draft of a conceptual model that provides a set of criteria to analyze learning experiences.
In summary, we can conclude that there is no single solution for all testbeds when it comes to capturing task context. Capturing task context needs approaches that take into account specific characteristics of the different testbeds. There is not a single solution as each of the approaches has its advantages and disadvantages, such as privacy, and usability in work practices of sensors, the applicability of mobile applications, or the potential of connecting to existing information systems. Careful analysis of workplaces is required to choose the appropriate capturing tools.

Our study at NBN has shown the big inter-individual differences in requirements between users within a testbed. Applications that capture and present physiological data for reflection may not be universally useful, but each of them has to target a specific type of users. We expect similar issues with other reflection apps, which will be investigated in the coming months. To facilitate that, we have collaboratively developed empirically grounded personas to feed it into the requirements process.

Creating transparency about one’s own experiences can interfere with existing coping strategies that have been developed by individuals. This can lead to resistance to this transparency due to anxiety of its impact. This needs further investigation of the impact of interventions and how to address anxiety as part of scaffolding learning processes. This requires further investigation of the impact of MIRROR interventions into existing coping strategies.

• In year 2, WP3 will build on the above described results and will shift its main focus towards the context infrastructure:
• The above mentioned results and the conceptual model will be integrated to define a base for context augmentation and filtering that can be used by all capturing apps.
• The in-depth analysis of the physiological sensor data (annotated with real-world work practices) that was collected at NBN and the positive acceptance of sensors will support the design and development of capturing applications that allow the capturing of emotion and work related aspects like interest or stress.