Back to search

University of Strathclyde

Effective task design to improve worker performance and engagement in crowdsourcing image-labelling tasks

Abstract

dc:description.abstract

Nowadays, crowdsourcing has become a baseline approach for large-scale data annotation, yet maintaining workers’ engagement, and outcome quality in repetitive tasks remains a persistent challenge. Current systems often rely on individual workers, offer minimal social interaction and apply gamification in a fragmented manner without a systematic mechanism to fulfil motivational needs. Consequently, there remains limited understanding of how collaboration combined with gamification could improve worker performance, and engagement in crowdsourcing environments. This thesis argues that effective and high-quality crowdsourcing requires a shift from an isolated, incentive-driven task execution paradigm to a collaborative socially embedded, and motivation-aligned system design. It investigates how collaborative gamification (e.g., teamwork, real-time interactions and motivational design components) can enhance performance and engagement in image-labelling tasks. The work is grounded in Self Determination Theory (SDT) and Flow Theory, which provide a theoretical foundation for understanding how the system design influences worker motivation and behaviour. To address this, the thesis adopts a multi-phase explanatory sequential framework research programme that incorporates three interrelated studies. First, an exploratory study empirically investigates how workers’ engagement and performance differ according to the crowdsourcing task types. Second, a systematic scoping review examines how gamification features have been applied across crowdsourcing archetypes, identifying the most popular design strategies, the underpinning theory, results and gaps in the literature. Third, a real-time collaborative gamified image-labelling platform is designed, developed, and evaluated through four controlled between-subjects conditions to assess how collaboration and gamification affect worker performance and engagement, and their perceived workload. The thesis makes four primary contributions: 1) it establishes empirical baselines of worker behaviour, engagement, and performance across different crowdsourcing task types, 2) it provides a systematic scoping review that maps and summarises current studies gamification features and the motivational theories across crowdsourcing archetypes, 3) it presents a custom-built platform for real-time collaborative gamified crowdsourcing with a robust evaluation pipeline, and 4) it provides large-scale experimental evidence that collaborative gamification improves worker performance (operationalised as labelling accuracy and workers’ agreement strength), and engagement while reducing perceived workload in image-labelling task compared to traditional individual approaches. Collectively, these contributions advance our understanding of motivational design in crowdsourcing and offer practical direction for the future of human-centred crowdsourcing, and human–AI applications.

Degree

thesis:*
Name dc:type.qualificationname
phd
Level dc:type.qualificationlevel
doctoral-pg
Grantor dc:publisher.institution
University of Strathclyde
Year dc:date.issued
2026

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Banuqitah, Huda Umar
Advisors dc:contributor.advisor
  • Dunlop, Mark
  • Abulkhair, Maysoon F.
  • Terzis, Sotirios, 1973-
Author Identifier dc:creator.authoridentifier
  • 202084148

Rights

Identifiers

dc:identifier.*
Identifier
T18172

Chain of custody

source
Harvested from
University of Strathclyde
Base URL
stax.strath.ac.uk/catalog/oai
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Banuqitah, Huda Umar. Effective task design to improve worker performance and engagement in crowdsourcing image-labelling tasks. doctoral-pg thesis, University of Strathclyde, 2026. https://stax.strath.ac.uk/concern/theses/c534fp751