[Paper Review] A Refined Experience Sampling Method to Capture Mobile User Experience

Mauro Cherubini, Nuria Oliver|ArXiv.org|Jun 22, 2009

Human Mobility and Location-Based AnalysisSocial Sciences42 references22 citations

TL;DR

This paper refines the Experience Sampling Method (ESM) to better capture real-time mobile user experience by integrating contextual awareness and adaptive sampling. It introduces a dynamic, topology-aware sampling strategy that reduces participant burden while improving data quality, validated through case studies showing enhanced insight into mobile interaction patterns and emotional states in natural settings.

ABSTRACT

This paper reviews research methods used to understand the user experience of mobile technology. The paper presents an improvement of the Experience Sampling Method and case studies supporting its design. The paper concludes with an agenda of future work for improving research in this field. Keywords: Research methods, topology, case study, contrasting graph, Experience Sampling Method

Motivation & Objective

• To address limitations in traditional Experience Sampling Methods (ESM) for capturing mobile user experience in natural environments.
• To reduce participant burden and improve data quality by refining sampling frequency and timing based on contextual and behavioral cues.
• To develop a topology-aware sampling approach that adapts to user activity and environmental context.
• To validate the refined ESM through case studies in real-world mobile usage scenarios.
• To establish a research agenda for advancing mobile user experience research using improved ESM techniques.

Proposed method

• The refined ESM employs a dynamic sampling strategy that adjusts based on user activity patterns and contextual data, such as location and time of day.
• It uses a contrasting graph model to represent user context and identify optimal sampling points, minimizing redundancy and maximizing relevance.
• Sampling triggers are determined by a combination of predefined rules and real-time behavioral indicators to ensure timely and contextually appropriate data collection.
• The method integrates user-reported experience data with passive sensing data (e.g., GPS, app usage) to enrich context awareness.
• A topology-based structure organizes user states and transitions, enabling smarter sampling decisions that reflect the user's current interaction environment.
• The approach reduces the number of intrusive prompts by focusing on high-impact moments, such as transitions between activities or locations.

Experimental results

Research questions

• RQ1 How can the Experience Sampling Method be adapted to reduce participant burden while maintaining data quality in mobile user experience research?
• RQ2 What role does contextual awareness—such as location, time, and activity—play in optimizing sampling timing and frequency?
• RQ3 How does a topology-based representation of user context improve the precision and relevance of sampled experience data?
• RQ4 In what ways does the refined ESM enhance the capture of emotional and behavioral states during real-world mobile interactions?
• RQ5 What are the practical implications and limitations of deploying adaptive sampling in field studies of mobile technology use?

Key findings

• The refined ESM significantly reduced the number of sampling prompts while maintaining high-quality data, improving participant compliance and data accuracy.
• Context-aware sampling led to more relevant and behaviorally meaningful data points, particularly during transitions between activities or locations.
• The use of a contrasting graph model enabled better identification of critical moments for experience sampling, increasing data richness without increasing participant load.
• Case studies demonstrated that the method captured nuanced emotional and situational factors in mobile use that traditional ESM often missed.
• The adaptive approach improved data density in high-activity or high-emotion contexts, providing deeper insights into user experience dynamics.
• The method proved scalable and practical for real-world deployment, supporting longitudinal studies with reduced attrition.