The value of emerging technologies for investigating academic practice
Download as PPTX, PDF0 likes103 views
This document discusses how emerging digital technologies can provide new types of data for investigating academic practice. It notes that current research methods rely heavily on surveys and interviews that measure perceptions rather than actual behaviors. The document proposes using "reality mining" techniques that capture continuous behavioral data streams through sensors to study behaviors over long periods of time. An example uses ceiling cameras, software, and analysis tools to automatically capture and analyze video data from an academic's office over 6 months to study durations of time in the office and with visitors. A second example proposes capturing stress measurements, computer activity, and location data streams over time from doctoral students to map stress patterns and associated stressors.
The value of emerging technologies for investigating academic practice
- 1. The value of emerging technologies for investigating academic practice Russell Butson Higher Education Development Centre University of Otago HERDSA 2015
- 2. The purpose of this paper is to highlight developments in digital technologies that offer access to new types of data.
- 3. The idea: that developments in digital technologies could offer access to new types of data. The possibility of seeing something new
- 4. Do we need something new? Data collection methods in higher education research are predominantly centred on surveys and interviews. As a result, much of our current knowledge of higher education behaviour/practice is built on perception-based data.
- 5. Review of methods from empirical studies published in a HE journal (2015) 1-1 Interview Survey Focus Group Reflective Writing Grade Video Photograph 17 14 7 5 2 1 1
- 6. The Problem These methods measure perceptions not behaviour In terms of contributing to studies focused on practice or behaviour – they offer a post-event recollection …. What is recalled or perceived to have occurred rather than what actually occurred.
- 7. Solution - Reality Mining To understand behaviour or practice requires approaches that can harvest naturally occurring behavioural data – ‘mining reality’. Reality Mining is a method that employs ‘sensor- based’ systems to capture continuous naturally occurring data feeds/streams – language and/or behaviour as it occurs.
- 8. Example-1 Academic Practice: The Office Focus: Capture & Data Analysis
- 9. The idea of seeing something new One of the core claims is that these new methods allow us to explore patterns of behaviour over long periods of time. In the past this has been impossible: 1. No instrument to capture naturally occurring behavioural data over extended periods. 2. Lack of analytical tools to analyse the volumes of data that would result from extended periods of data capture.
- 10. Ceiling cameras – HD motivation detection Automated intelligent systems requiring no human interaction. Data is fed continuously to a data warehouse.
- 11. Capturing Software Automated – email/text notification of any issues
- 12. Capture – the ‘Big Data’ issues Five cameras operating over 6 months generated - 5000 hours of footage - 15TB of data (five days to copy/move) This would take 19 months to view through once. Solution: ‘new’ developments in data analysis.
- 13. The ‘New’ - Analysis Software
- 14. SQL AGENT Vi-Search Server Dell PowerEdge 12G R320 rack server Client PC Dell Precision M4700 ObserverXT Cameras AXIS P3354 6mm PoE | 12fps Data transfer via Internal Ethernet Network Data transfer via Internal Ethernet Network File Conversion Video Streams Feature Streams Queries Milestone Smart Client Milestone X-Protect X4-4TB Surface Analysis Depp Analysis Complicated Process
- 15. The benefits once in place? • The duration the academic is in the office • The number of times the academic enters/exits • The number and duration of visitors • The number | duration in zones • Any changes to the office – objects • Traces of movement over time periods • Heatmaps (duration) across zones
- 16. Example-2 Mapping Stressors in Doctoral Students Map stress patterns over extended period + Map associated stressors Focus: Data Capture x3 streams
- 17. Stream-1: Stress Measure Wristband Sensors
- 18. Results generated through live feeds
- 19. Stream-2: Behaviour | Computer Activity Computer use logged at 5min intervals ?
- 20. Stream-3: Location 09 am 10 am 11 am 12 pm 01 pm 02 pm 03 pm 04 pm 06 pm 07 pm 08 pm 09 pm 10 pm 05 pm
- 21. THANK YOU