Crowdsourcing safety perceptions of people: Opportunities and limitations

Martin Traunmueller*, Paul Marshall, Licia Capra

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference Contribution (Conference Proceeding)

8 Citations (Scopus)

Abstract

Online crowdsourcing has successfully been used as a paradigm to collect large amount of perceptions about our cities quickly and cheaply, enabling social scientists to quantitatively test urban theories at scale. While doing so, researchers have not focussed on getting answers from specific demographics, relying on a self-selected crowd instead. However, existing theories suggest that knowing who the respondents are is crucial for understanding safety perceptions about people (instead of, for example, about the built environment). In this case to quantitatively validate theories, it is not just the amount of data that matters, but also what demographics participate (or not). In this paper we investigate to what extent online crowdsourcing can be used for the specific case of safety perceptions about people.We built an image-based online crowdsourcing platform, collected safety perception ratings and background information from more than 700 people and used them to quantitatively evaluate established theories based on qualitative research. On one hand, we show in this paper that online, image-based crowdsourcing can be used to gather perceptions about people too, not just architecture, confirming established theories based on qualitative work. Furthermore, we are able to uncover detailed interactions that would be challenging to grasp using qualitative methods. On the other hand, we show limitations of using crowdsourcing as a method. By not controlling who makes up the crowd, we were not able to investigate all theories as we did not reach all user groups that have been discussed in qualitatitve research.

Original languageEnglish
Title of host publicationSocial Informatics - 7th International Conference, SocInfo 2015, Proceedings
PublisherSpringer-Verlag Berlin
Pages120-135
Number of pages16
Volume9471
ISBN (Print)9783319274324
DOIs
Publication statusPublished - 1 Jan 2015
Event7th International Conference on Social Informatics, SocInfo 2015 - Beijing, China
Duration: 9 Dec 201512 Dec 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9471
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference7th International Conference on Social Informatics, SocInfo 2015
CountryChina
CityBeijing
Period9/12/1512/12/15

Keywords

  • Crime studies
  • Crowdsourcing
  • Perception
  • Social studies
  • Theory validation

Fingerprint

Dive into the research topics of 'Crowdsourcing safety perceptions of people: Opportunities and limitations'. Together they form a unique fingerprint.

Cite this