Methods and Tools

Develops a fine-grained annotation framework grounded in psychological theory that distinguishes six empathy-seeking and seven empathy-giving practices in online discourse, moving beyond binary or unidimensional empathy labels. Fine-tunes language models to detect these practices reliably, then applies them at scale across six communities, demonstrating a replicable pipeline from qualitative codebook development to large-scale computational analysis.

Also in: platforms and power

Reviews fifty years of network analysis since Wigand's 1977 article, tracing the evolution from hand-drawn sociograms to computational layout algorithms like Fruchterman-Reingold. Argues that advances in network visualisation and modelling have made networks not just analytical tools but performative systems—platforms like Facebook and LinkedIn are themselves network formalisms that shape the communication they purport to represent.

Also in: networks as epistemology, platforms and power

Combines social network analysis with non-participant observation and in-depth interviews in a realist evaluation of neonatal care units in Kenya. The mixed-methods design uses SNA sociograms and network metrics (in-degree centrality, clustering coefficient) to triangulate qualitative findings, demonstrating how network methods can reveal the relational structures that underpin workplace culture and quality improvement in resource-constrained healthcare settings.

A meta-ethnographic synthesis of qualitative studies on link workers' experiences, developing a novel conceptual framework from across the existing literature. Demonstrates how systematic qualitative synthesis methods can produce theoretical understanding of a relational role that quantitative evaluation alone cannot capture.

Also in: networks as epistemology

A Python textbook organized around the practical skills social scientists need: working with DataFrames, accessing web data and APIs, merging and cleaning datasets, and interpreting results. Bridges the gap between social science research questions and the computational tools required to answer them at scale.

Also in: teaching and translation

Introduces Network Canvas to the epidemiological community as an end-to-end open-source workflow for designing and conducting network interviews. Provides an accessible entry point for health researchers who need social network data but lack the technical infrastructure to build custom data collection tools.

Candidly documents the data collection failures of a social network study of university-industry collaboration in controlled radical polymerization, extracting transferable lessons about flexibility and adaptability in network data collection. A rare honest account of methodological failure that serves the field better than another success story.

Draws on qualitative interviews from the fourth British National Survey of Sexual Attitudes and Lifestyles (Natsal-4) to examine how adults use digital technologies for partner seeking. Applies a social practice framework to a sensitive research domain, demonstrating how qualitative methods can complement large-scale survey data.

Documents the key design decisions behind the Network Canvas software suite: Architect (protocol design), Interviewer (touch-optimized data collection), and Server (deployment management). Prioritizes usability and accessibility across platforms, addressing the chronic problem that self-reported social network studies are complex and burdensome for both researchers and participants.

Conference abstract presenting preliminary findings on online partner seeking from the Natsal-4 study, later developed into the full 2022 Journal of Sex Research paper. Documents the methodological approach to studying digital sexual health practices within a national survey framework.

Develops a comprehensive multimodal data repository integrating epidemiological, governmental, mobility, and weather data across countries at national and regional levels. A large-scale data infrastructure project demonstrating how interdisciplinary collaboration can produce research tools that serve the broader scientific community during a crisis.

Tests whether apparent instability in egocentric networks across survey waves reflects genuine network churn or interview artifacts like anchoring and motivated underreporting. Uses Network Canvas to demonstrate that structural embeddedness predicts alter retention, providing empirical grounding for interpreting longitudinal ego network data.

Also in: networks as epistemology

A mass collaboration of 160 teams using machine learning on rich longitudinal data (Fragile Families) found that life outcomes remain stubbornly unpredictable, with the best models barely exceeding simple benchmarks. Demonstrates the value of the "common task method" for social science while revealing practical limits to prediction that should concern policymakers using such models.

Also in: semantic structures

Uses mixed-method social network analysis across 85 interviews in six UK communities to understand how conversations influence energy practices. Identifies social factors that "open up" or "close down" energy discussions, including perceived stigma, thereby showing how network methods can reveal the social dynamics underlying behavioral change.

Also in: networks as epistemology

Updated methodological guide for collecting and analyzing online social network data, reflecting the significant changes in platform architectures since the first edition. Addresses the growing tension between researchers' analytical needs and platforms' progressive restriction of data access.

Also in: teaching and translation

Evaluates the digital translation of paper-based participant-aided sociograms to a touchscreen interface for use with high-risk populations (young MSM in Chicago). Finds that the screen-based tool produces comparable network data while reducing interviewer burden, validating the move from paper to the Network Canvas platform without sacrificing data quality.

Addresses a surprising gap: prior work correlated Facebook use with social capital, but social capital is inherently structural, and little work related it to actual network structure. Uses Facebook ego network data to demonstrate how structural features of online networks, including density, clustering, and bridging, map onto different dimensions of social capital.

Also in: networks as epistemology

Proposes decoupling temporal information about forum users into event sequences and inter-event times, creating a novel feature space for analyzing communication patterns. Applied to 30,000+ users across four internet forums, the method reveals communication dynamics invisible to traditional approaches.

Also in: semantic structures

Applies social network analysis to map information and financial support exchanges between low-carbon community groups and other actors in Oxfordshire, UK. Demonstrates how network methods can illuminate the meso-level governance structures that shape local energy transitions, making visible the relational infrastructure that policy analysis typically ignores.

Also in: networks as epistemology

Develops an interactive mapping tool that places every geolocated Wikipedia article on a world map, enabling visual analysis of geographic patterns in knowledge production. A demonstration of how visualization methods can transform large-scale platform data into legible spatial patterns.

Also in: platforms and power

Develops a conditional logistic framework for modeling tie creation in continuously-observed networks, enabling simultaneous estimation of multiple growth mechanisms (homophily, reciprocity, triadic closure, popularity). Applied to a Facebook-like community, the method advances beyond panel-data approaches to capture the fine-grained temporal dynamics of network evolution.

Also in: networks as epistemology

Develops an innovative audit methodology using 5,620 fictitious email inquiries sent to Craigslist landlords, each landlord receiving inquiries from five racialized name categories. The method enables large-scale discrimination testing at low cost by leveraging the affordances of online platforms for systematic field experiments.

Also in: platforms and power

Provides practical guidance for visualizing and interpreting Facebook network data using the NodeXL platform, translating social network analysis techniques for a broader audience of researchers and practitioners. Demonstrates how network visualization tools can make structural patterns in social media data legible and analytically useful.

Also in: networks as epistemology

Early methodological work using the Facebook API to extract and compare egocentric network structures, demonstrating that online friendship networks, while larger, exhibit structural properties (clustering, density) that meaningfully correspond to offline social organization. Establishes the viability of platform APIs as tools for social network research.

Also in: networks as epistemology

Develops methods for modeling interactions among face-to-face and electronic communication modes, demonstrating how ICT use patterns can be captured alongside travel behavior data. The data collection framework treats communication mode choice as a methodological variable, enabling researchers to study how different media serve different social network functions.

Also in: networks as epistemology

Provides a comprehensive methodological framework for analyzing social networks using internet-based data, covering ego networks, blogs, email, and social networking sites. Establishes best practices for online social network analysis at a moment when these methods were emerging but not yet standardized.

Also in: teaching and translation

Develops a data collection instrument that explicitly links social networks to travel behavior through an egocentric approach, testing the hypothesis that individuals' travel is conditional on their social network structure. Demonstrates how survey and interview methods can be designed to capture the social dimension that transportation research had largely ignored.

Also in: networks as epistemology

Bridges data engineering and social science by demonstrating how information networks, including the patterns in email, messaging, and online interaction, can be analyzed to understand social behavior. Provides a technical audience with a framework for seeing social structure in the digital traces they already collect.

Also in: networks as epistemology

Explores how social network metadata extracted from email headers can be used to help users prioritize their inbox, leveraging the social structure embedded in communication patterns. A practical demonstration of how network analysis methods can improve information management tools.

Introduces participant-aided sociograms as an interview-based data collection method that extends traditional name generators by having respondents place alters on a visual diagram. Demonstrates both practical and methodological advantages of keeping "high technology in the lab and low technology in the field," since paper-based sociograms reduce respondent burden while capturing richer structural data than standard survey instruments.

Also in: networks as epistemology

Documents the methodology and early findings of the Connected Lives study in Toronto, which combined surveys, interviews, and communication logs to study the interplay between ICTs, community, and domestic relationships. A methodological blueprint for studying networked individualism through mixed-methods approaches that capture both network structure and lived experience.

Also in: networks as epistemology

Develops and validates an eight-item survey scale for measuring email overload, showing it links well to actual user behavior across 292 subjects. A pragmatic methodological contribution that provides a standardized instrument for evaluating email management systems and studying workplace communication patterns.