Licentiate seminar on Information Systems with Eric Andersson

Title: "Reexamining Data and Information Quality in Information Systems  Using Semiotic Theory"

Date: September 30, 2026 at 10:00 a.m. CET

Room: Mid Sweden University, Campus Sundsvall, room C312  

Main supervisor: Aron Larsson

Assistant supervisor: Felix Dubslaw

Opponent/External reviewer: Associate Professor Mikael Berndtsson, University of Skövde 

Abstract

While data and information quality is widely recognized as important for operational success in public-sector data use, present IS research insufficiently accounts for how human factors shape perceptions and evaluations of quality. Human agents operate under cognitive limitations, bounded knowledge, and potentially conflicting motivations, introducing complexity into quality evaluation and decision-making.

This thesis investigates two research questions: how semiotic theory may improve the theoretical understanding of quality, and how individuals perceive and respond to quality concerns in practice. The thesis synthesizes findings from three articles.

Article I employs a systematic literature review to align existing data and information quality concepts with semiotic theory. The findings demonstrate the extent to which quality depends on user interpretation and knowledge in both conceptualization and measurement.

Article II uses semi-structured interviews to examine how agents interpret information, reason about quality concerns, and justify decisions under uncertainty. The resulting process model highlights how uncertainty, emotional discomfort, domain knowledge, source trustworthiness, time pressure, and perceived consequences shape informal quality assessment and organizational decision-making.

Article III addresses the limited body of causal evidence on how quality affects decisions. Through a scoping review, the article critiques common methodological approaches and shows that core dimensions such as accuracy, completeness, consistency, and timeliness remain insufficiently understood. The study proposes revised approaches and demonstrates, using open data, how incompleteness increases uncertainty and alters perceived utility. Jointly, the findings demonstrate the extent to which agents function as critical components of quality control within socio-technical systems. In particular, domain knowledge and trust strongly influence whether agents can effectively evaluate quality and whether organizations achieve substantive quality or merely the appearance thereof.

The thesis argues that IS quality research must place significantly greater emphasis on agent dependency and subjective interpretation. These concerns are especially important as public-sector organizations increasingly pursue data-driven governance, large-scale data integration, and inter-organizational data-sharing. As decision-making becomes more dependent on interconnected information infrastructures, unresolved quality concerns may propagate across organizational and institutional boundaries, raising broader questions of how sustainable and durable forms of quality control can be achieved in socio-technical systems.