CALRG Conf: xDelia

Gill Clough, Gráinne Conole, Eileen Scanlon on xDelia’s Design and Evaluation framework.

Not going to talk too much about the framework, more about the link to games.

xDelia is a pan-European project, €3.2m, looking at effects of emotional bias on financial decision-making of: traders; investors; individuals. Three year project, using bio-feedback (sensors) and serious games. Active workpackages on Traders and investors (OU – OUBS), Financial capability (Bristol), Games development (Sweden), Sensor development (Germany), Evaluation (OU – IET).

Design and Evaluation Framework.

Rationale: embedded from the beginning, engaged throughout. Lens for reflection, capture project experiences at key moments, ensures common understanding (include terminology – emotions conceptualised differently in different workpackages), guides development of research questions and methods.  Diagram and description in Clough, Conole and Scanlon (2010 – Networked Learning) and (2009 – ASCILITE).

This is macro-level support. But also doing more micro-level support, as local evaluaitons. So building evaluation toolkits – Cloudworks Evaluation toolbox.

Example micro-level exaluation: Saxo Trader pilot – data collection, looking for difference in emotional regulation between novice and experienced traders, emotional arousal, feeding in to design of learning interventions. Also Micro Games – individual capability, games development.

Saxo Trader Pilot

Most traders male, but had two female traders. Novice/experience trader pairs. 2h data collection around planned news event – bio-sensors to collect ECG heartrate data (N=5), video (N=2), think-aloud for 10 min at start, 20 min around event, 10 min at end (N=3). Also overnight resting heart rate (N=3). Have to programme sensor with personal data.

Heartrate Variability – reflects ability of heart to adapt. Stress leads to lower HRV, calm to higher HRV. This changes with age quite profoundly. Poincare plot of HRV – taken during annual review of project – was very stressful. Beforehand, was not very variable, was all over the place during the stress.

Going to do main study with far more, which will inform the learning interventions. Want toolkit for traders – mentor-supported toolbox, serious games to address emotional regulation. Help them learn via bio-feedback. Also standalone learning for investors.

Financial capability – Micro games

Hard to get hold of audience, wants young people. Using Dutch Household Panel – runs monthly survey, have asked questions about financial decision-making. Will ID inividuals with particular financial traits, do rapid prototyping of games design.

Games to target specific emotions, address issues e.g. impulsivity and gullibility.

Local evaluations

Learning design conceptual framework, informing design and interventions – both sets of games.

Conceptual Learning Design tool – analyses intervention on a series of axes.

Start with learning intervention overview –sets out Guidance & Support, Content & Activities, Reflection & Demonstration, Communication & Collaboration. Then Pedagogy Profile – spreadsheet view of different categories of learning activities (assimilative, info handling, etc). Dimensions view – breaks down course against 4 higher-level categories. Also Task swimline – Compendium LD type view of the activity unfolding over time. Then finally Learning outcomes view – map intent and challenges, against student activity, and learning output.


Richard: Are the games for the traders, or young people, or both?

Gill: Different games for different cases. Micro games for the young people, not the traders. Will do games for the traders later. Parallel with eating interventions.

Anne: Connects with ethical/privacy issues. Consider in evaluation.

Martin LeVoi: Do you get a moment-by-moment measure of HRV?

Gill: You specify what period you want – e.g. 10 minute period, or 20 minute. Lowest time is tricky.

Eileen: Issues using the data, about how long to wait, issues with putting sensors in the wrong place, traders gesticulating – very hard to capture this data. On top of issues of access.

Gill: Many technical details.

Anne: What you attribute to the variation is very complex – biosensors only show ‘arousal’, not positive versus negative.

Eileen: Can detect events but not tell whether positive or negative. Need discussion or other data source.

Gill: Can run stream of video, audio, heart-rate and trading data, so correlate – software arriving soon.

This work by Doug Clow is copyright but licenced under a Creative Commons BY Licence.
No further permission needed to reuse or remix (with attribution), but it’s nice to be notified if you do use it.

Author: dougclow

Experienced project leader, data scientist, researcher, analyst, teacher, developer, educational technologist and manager. I particularly enjoy rapidly appraising new-to-me contexts, and mediating between highly technical specialisms and others, from ordinary users to senior management. After 20 years at the OU as an academic, I am now a self-employed consultant, building on my skills and experience in working with people, technology, data science, and artificial intelligence, in the education field and beyond.