Presentations and media from the third seminar, “What Data and How?“
A summary of our discussions from this seminar can be found here: Summary of Seminar 3
Productive Subjects: Workplace wellness programmes, big data and affective control
This paper will draw on qualitative research into the implementation of corporate wellness (CW) programmes using digital self-tracking (DST) to consider the use of big data practices as a means of enabling a critical, reflexive feedback mechanism aimed at increasing productivity and improving affective relationships at work. When seen as part of broader trends of self-tracking and Quantified Self the large-scale observation and analysis of human behaviour it can be suggested that exercise activity is in the process of being conceptually reconfigured as work. Exercise activities are increasingly being experienced and understood through quantified measures which enable accumulation and comparison with others and generate valuable data. Personal and corporate health are being conflated and private companies increasingly see it as part of their ethical responsibility to intervene in the everyday (non-work) life of individuals. It will be suggested that a significant part of the rationale for the implementation of CW DST is the promotion of an affective and aesthetic relationship between employer and employee. The health of the individual and the health of the economy/organization are increasingly intertwined and the definition of health (through a focus on ‘wellness’) is being aligned with productive capacity to form a new corporate health ethic. It will be proposed that CW DST have commercial, economic interests in the incitement to health (the generation of data or a more productive workforce) which are merged with an ethical concern for the health of the population and a drive to ‘do good’ as part of social responsibility. While such initiatives are notionally open to all they subtly (and probably unintentionally) target particular groups who increasingly demand “meaningful” from an employer who is driven by values. Consequently, groups who are most in need of intervention are tacitly excluded.
Big data: Big opportunities and big challenges
We often hear the term ‘Big Data’ being used in discussions of data science and analytics. But haven’t many organisations been dealing with large amounts of data for decades? Is there something new and different about Big Data or is this just another example of ‘techno-hype’? This presentation will consider the concept of ‘Big Data’, its definitions (such as the 4 V’s) and what this means in practice. Examples of areas in business that are utilising Big Data will provide an opportunity to discuss the big opportunities that Big Data affords along with the big problems orchallenges that handling and using BigData also entail. The content of this talk will help to provide some preliminary context to Big Data that will then be picked up in later presentations.
‘Using big data in practice: Computational infrastructures’
Big Data is challenging to process due to its volume and velocity dimensions. It is also difficult to make sense from it because of its variety and veracity challenges. In this talk we will discuss modern computational infrastructures to scale-out the processing of large datasets by means of distributed computing. We will introduce the concept of Map/Reduce and present Big Data systems like Apache Spark which is currently the market leader in Big Data processing solutions. We will explain by means of examples how large datasets can be processed in batch or as a stream.
‘Big data in the context of research on wages and related gender gaps’
For capturing dynamics in work stress and well being, conducting multiple measurements across time is necessary, but it is not sufficient. Any measurement reflects a snapshot at a discrete time point only. Important questions are (a) how discrete measurements can be used to analyze continuous (i.e., dynamic) relations among work stress and well being, and (b) how to optimally design the spacing (lags) of measurements and the length of studies. I will introduce some recent developments, which have demonstrated how discrete measurements can be related to their underlying continuous time (CT) processes and how this can be implemented in structural equation models (SEM). I will also show how the optimal spacing of measurements can be computed (using continuous or discrete time analyses), and discuss implications for the design of studies.