Funky Data: working with unconventional data in surveys and research

Technological developments as diverse as the rise of social media, the smartphone, voice recognition and now even facial recognition, combined with an explosion in the kinds of data being retained by organisations or harvested from the Internet means the landscape of what survey researchers or market researchers now have to work with as raw data is changing fast.

Some have characterised this as the challenge of dealing with “big data”, but size is not the only issue: the real challenge is the their unconventionality or even eccentricity; in other words, how “funky” they are. These are kinds of data that are new to researchers and who often lack the tools and skills to process and analyse them. They can be intrinsically difficult to merge in with existing, valued and validated methods, yet if that can be achieved, they offer the promise of enriching existing data and even bringing down the costs of collecting data conventionally.

The aim of this conference is to bring together practitioners across the related fields of market and social research, and survey statistics, who are working with unconventional data, in order to share knowledge and experience of the technologies and techniques they have developed, or discuss the challenges and present potential solutions.

The focus of ASC is the role that software and technology plays in advancing the practice of research.

  • Methods to code or extract meaning from images or video
  • Analysing/combining multiple sources of data with survey data, particularly where unconventional data are involved such as smartphone data, geolocational data, SMS text message data, pictures or other attachments contributed by survey participants
  • Innovations in bringing together qualitative and quantitative data
  • Innovations and experience in applying text analytical tools to survey research data
  • Innovations in analytical methods for social media research
  • The challenges in adapting tools that deal with unconventional data (e.g. unstructured text data) and which were developed for non-research application; how sentiment analysis can provide value to the researcher.
  • Research applications or experiences with automated voice to text recognition
  • Applying automation in handling mixed or multimedia data – how best to deal with these richer data types where scale and volume make conventional methods ineffective
  • Is there a role for Amazon’s Mechanical Turk or other crowdsourcing activities to analyse text or perform other micro-analytical tasks?
  • Turning logfiles or other incidental data to insight, either from websites or mobile apps
  • Analysing data from mobile behavioural apps
  • Neuroscience, biometrics, eye-tracking, facial expressions – experiences in making these relevant and viable as data sources for research and insight generation
  • Or other new developments or experiences in related topics

(Website in development – papers & presentations to be uploaded soon!)