Koutsolampros, P; Sailer, K; Haslem, R; Austwick, M; Varoudis, T; (2017) Big Data and Workplace Micro-Behaviours: A closer inspection of the social behaviour of eating and interacting. In: Heitor, Teresa and Serra, Miguel and Silva, João Pinelo and Bacharel, Maria and Silva, Luisa Cannas da, (eds.) Proceedings of the 11th International Space Syntax Symposium. (pp. 149.1-149.16). Instituto Superior Tecnico, Departamentode Engenharia Civil, Arquitetura e Georrecursos, Portugal: Lisbon, Portugal.

Preview

Text Koutsolampros_Sailer_etal2017_BigDataWorkplaceMicrobehaviours_SSS11.pdf - Published Version Download (1MB) | Preview

Abstract

Evidence-based design aims to understand human behaviour so that strategic decisions are well-informed when creating a new space. Workplace research to date has provided interesting insights, but has mostly done so on a case-by-case basis. This approach does not yield generalisable patterns, making results problematic to use in an evidence-based design context. This paper builds upon previous large-scale analysis done by the authors and focuses on two aspects of workplace behaviour – eating and interacting. We aim to understand the nuances of these behaviours, thus we explore them as independent phenomena, separate them into subcategories and set out to understand the reasons behind these observations. The examined dataset includes 23 organisations in the UK, with a wide variety of sizes, numbers of floors and buildings. It consists of human activity data collected through direct observation, Visibility Graph Analysis and organisational parameters such as industry and flexibility of desk occupancy. The first behaviour we focus on – interaction – has already been explored in previous research and has been found to happen primarily in workspace and meeting rooms. In this instance we initially classify interactions according to the activity of the members and the type of space they occur in. The analysis of the second behaviour – eating – revolves around the activities and locations of people at lunchtime. We aim to discover where people choose to eat and how this is affected by the characteristics and availability of eating spaces. For the two behaviours studied, we examine how each activity relates to the space it is happening in, taking into account a set of spatial and organisational factors. In the first case we test each interaction against proximity to circulation and local visibility of the space, while in the second we examine the popularity of different types of spaces, for example canteens and breakout spaces, against their proximity to workspace and what possibilities of inter-visibility they offer. This paper provides detailed insights into the phenomena of interacting and eating, and reflects on limitations of traditional statistical analysis. It will also highlight further opportunities for handling these types of big datasets using different techniques such as Principal Component Analysis and machine learning.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item