Training Resources
General Resources
- Bit by Bit: Social Research in the Digital Age — Matthew J. Salganik, 2018. A comprehensive, open-access introduction to computational social science covering surveys, experiments, mass collaboration, and ethics. Book
- Computational Social Science — David Lazer et al., 2009. The foundational paper arguing for a field that leverages large-scale digital data to understand human behaviour. Paper
- National Statement on Ethical Conduct in Human Research — NHMRC / ARC / Universities Australia, 2023 (updated 2025). Australia's core framework for research ethics, essential for anyone working with human or digital trace data. Guide
- Computational Analysis of Communication — Wouter van Atteveldt, Damian Trilling & Carlos Arcila, 2022. An open-access textbook with R and Python code covering text, network, and image analysis. It is not expected of the participants to be familiar with R or Python, but it is advisable to explore what they are and some foundational terms. Book
- R for Data Science — Hadley Wickham & Garrett Grolemund, 2nd ed. A practical, free introduction to data science with R and the tidyverse for researchers with little coding experience. Another resource that would be useful as an introduction into the key aspects of working with data using R.Tutorial
- Introduction to Cultural Analytics & Python — Melanie Walsh, 2021. A free online textbook covering Python basics, text analysis, and social media data specifically for humanities and social science scholars. Tutorial
Session-Related Resources
Day 1 — Introduction to Computational Social Science · Monday, 22 June
What is Computational Social Science and Why It Matters in Australia?
Pre-reading
- Computational Social Science — Lazer et al., 2009. The landmark essay that defined the field and its potential for studying society at scale. Paper
- Bit by Bit — Chapter 1: Introduction — Salganik, 2018. An accessible overview of how digital data is transforming social research. Book
Tools & Platforms
- SICSS Website — Hub for all Summer Institutes in Computational Social Science worldwide.
- Australian Internet Observatory (AIO) — Australia-wide research infrastructure for digital platform observability.
Further reading
- Computational Social Science: Obstacles and Opportunities — Lazer et al., 2020. A decade-on reassessment of CSS's challenges, including access to proprietary data and issues of generalisation. Paper
- Computational Analysis of Communication — van Atteveldt, Trilling & Arcila, 2022. Full open-access textbook with hands-on R/Python code. Book
Social Bias in Computational Social Science
Pre-reading
- Reflecting on Social Bias: Challenges and Opportunities for Computational Social Science — Kathirgamalingam et al., 2025. The keynote speaker's forthcoming paper on social bias in CSS. Preprint
- Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries — Olteanu et al., 2019. A comprehensive survey of biases arising from digital data sources, collection strategies, and analytical methods. Paper
- Bit by Bit — Chapter 2: Observing Behavior — Salganik, 2018. Covers limitations and biases inherent in using digital trace data for research. Book
Further reading
- Algorithmic bias: review, synthesis, and future research directions — Nima Kordzadeh and Maryam Ghasemaghaei, 2021. A discussion of alogrithmic biases and the importance of considering them in research and data-driven decision-making. Paper
Day 2 — Working with Data: Ethics and Practices · Tuesday, 23 June
Ethics in Computational Social Science
Pre-reading
- Bit by Bit — Chapter 6: Ethics — Salganik, 2018. A practical guide to ethical considerations in digital-age research, including informed consent and privacy. Book
- National Statement on Ethical Conduct in Human Research — NHMRC, 2023. Australia's authoritative framework for research involving human participants. Guide
Tools & Platforms
- FAIR Principles — Guidelines for making research data Findable, Accessible, Interoperable, and Reusable.
- CARE Principles for Indigenous Data Governance — Framework centred on Collective benefit, Authority to control, Responsibility, and Ethics.
Further reading
- Tutorial: Legality and Ethics of Web Scraping — Vlad Krotov, Leigh Johnson, & Leiser Silva, 2020. An examination of ethical and legal challenges when collecting data from the web. Paper
Data Donations and Participant-Centric Research
Pre-reading
- A framework for privacy preserving digital trace data collection through data donation — Boeschoten et al., 2022. Introduces data donation as a methodology, covering design principles, participant experience, and infrastructure. Paper
Tools & Platforms
- Data Donation (datadonation.eu) — European research hub for data donation methodology, tools, and projects.
- Australian Internet Observatory (AIO) Data Donation System — A demo version of AIO's data donation platform demonstrating current functionality.
Further reading
- OSD2F: An Open-Source Data Donation Framework — Araujo et al., 2022. Technical overview of the original data donation framework (now archived, succeeded by Port). Paper
Nectar Research Cloud
Pre-reading
- ARDC Nectar Research Cloud — ARDC Page — ARDC. Introduction to Australia's national research cloud, what it offers, and who can use it. Guide
Tools & Platforms
- Nectar Research Cloud Dashboard — The main portal for accessing virtual machines and cloud compute.
- Nectar Cloud Tutorials — Step-by-step guides for getting started with instances, storage, and networking.
Further reading
- Nectar Cloud Getting Started Guide — ehelp.edu.au. Detailed walkthrough of account setup, authentication via AAF, and launching your first project. Tutorial
Day 3 — Data Collection and Working Across Disciplines · Wednesday, 24 June
Does Computational Social Science Lack Theory?
Pre-reading
- Preface: Big Data Is Not About the Data! — Gary King, 2016. A provocative argument that CSS suffers from an overemphasis on methodology at the expense of substantive theory. Paper
- Integrating Explanation and Prediction in Computational Social Science — Hofman et al., 2021. Argues for balancing predictive accuracy with explanatory power in CSS research. Paper
Further reading
- Text as Data: A New Framework for Machine Learning and the Social Sciences — Grimmer, Roberts & Stewart, 2022. An authoritative guide linking computational text methods to social science theory and inference. Book
The AIReD Platform for Australia-wide Social Media Discovery and Usage
Pre-reading
- Australian Internet Observatory — About — Overview of the AIO's mission and the platforms it covers. Guide
Tools & Platforms
- AIReD — Australian Internet Research Dashboard — Platform with 500M+ social media posts from BlueSky, Mastodon, YouTube, GDELT, and historic X/Twitter data. Setup (requires AAF institutional login)
Collecting and Analysing Data Download Packages
Pre-reading
- A framework for privacy preserving digital trace data collection through data donation — Boeschoten et al., 2022. Covers design principles and infrastructure for data download package studies. Paper
Tools & Platforms
- Port — Open-source framework for locally processing donated data download packages.
- Data Donation (datadonation.eu) — Resources on best practices for requesting and using DDPs from major platforms.
Further reading
- A framework for privacy preserving digital trace data collection through data donation — Boeschoten et al., 2022. A detailed methods paper on designing and implementing DDP-based research. Paper
Working with Text Using Computational Techniques
Pre-reading
- Text Mining with R: A Tidy Approach — Silge & Robinson, 2017. Free online book covering tokenisation, sentiment analysis, and topic modelling in R. Book
- Computational Analysis of Communication — Ch. 10-12 — van Atteveldt et al., 2022. Covers text as data from preprocessing to supervised and unsupervised methods. Book
Tools & Platforms
- tidytext (R) — R package for tidy text mining workflows.
- quanteda (R) — Comprehensive R package for quantitative text analysis.
- spaCy (Python) — Industrial-strength NLP library for tokenisation, NER, and text processing.
Further reading
- Text as Data: A New Framework for Machine Learning and the Social Sciences — Grimmer, Roberts & Stewart, 2022. The definitive guide to computational text analysis in social science. Book
- LADAL — Language Technology and Data Analysis Laboratory — Australian-hosted tutorials for NLP in R and Python, designed for linguists and social scientists. Tutorial
Day 4 — Tools and Approaches to Data Analysis · Thursday, 25 June
Screen Capture for Data Collection
Pre-reading
- MOAT — Description of the Australian Internet Observatory's Mobile Ad Observatory Toolkit (MOAT) and relevant work conducted.Guide
Tools & Platforms
- AIO Mobile Screen Capture Tools — Research-grade mobile and browser extension tools for capturing personalised platform content (ads, feeds, recommendations).
Further reading
- The Australian Ad Observatory: Background Paper — Jean Burgess, Mark Andrejevic, Daniel Angus & Abdul Karim Obeid, 2022. Outlines the rationale, citizen science approach, and methodology of the project. Paper
Using LLMs to Create Data Analysis Pipelines for Text-as-Data Research
Pre-reading
- quallmer: AI-Assisted Qualitative Data Analysis — Maerz & Benoit, 2025. CRAN page for the R package that enables codebook-based, LLM-assisted text coding with built-in replication and validation. Tool
- Can ChatGPT Replace Manual Annotation? — Gilardi et al., 2023. Assesses GPT-4's accuracy as a text annotator compared to crowd workers across multiple tasks. Paper
Tools & Platforms
- quallmer (GitHub) — R package for structured LLM-assisted coding, validation, and audit trails in text-as-data research. Setup
- quallmer.app — Interactive Shiny companion app for manual coding, reviewing LLM annotations, and computing agreement metrics.
- ellmer (R) — Backend package for connecting to multiple LLM providers (OpenAI, Anthropic, Google, Ollama).
Further reading
- Large Language Models Outperform Expert Coders and Supervised Classifiers at Annotating Political Social Media Messages — Törnberg, 2024. Evaluates LLM annotation quality across multiple political science classification tasks. Paper
RAG 101
Pre-reading
- Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks — Lewis et al., 2020. The seminal paper introducing the RAG framework that combines retrieval with generation. Paper
Tools & Platforms
- LangChain — Popular Python framework for building RAG pipelines, with pre-built modules for document loading, embedding, and retrieval. Setup
- LlamaIndex — Framework for connecting LLMs to external data sources, purpose-built for RAG applications.
Further reading
- Retrieval-Augmented Generation for Large Language Models: A Survey — Gao et al., 2024. Comprehensive survey categorising the evolution of RAG into Naive, Advanced, and Modular paradigms. Paper
Image Analysis for Qualitative and Quantitative Research
Pre-reading
- Computational Analysis of Communication — Ch. 14: Multimedia Analysis — van Atteveldt et al., 2022. Introduces computational image analysis concepts and methods for communication researchers. Book
Tools & Platforms
- Image Machine — Open-source tool for clustering visually similar images using machine vision embeddings and identifying visual patterns in large datasets.
- UMAP — Uniform Manifold Approximation and Projection for dimensionality reduction, used for 2D visualisation of image similarity.
Further reading
- Automating Vision: The Social Impact of the New Camera Consciousness — Anthony McCosker & Rowan Wilken, 2020. Discussion of how machine vision techniques can be applied to analyse visual culture at scale. Book
Day 5 — Disciplines, Careers, and Industry · Friday, 26 June
Cross-Disciplinary Collaboration: Bringing Social Science and Computational Analysis Together
Pre-reading
- Interdisciplinary research has consistently lower funding success — Bromham et al., 2016. Empirical analysis of why interdisciplinary work is both valuable and difficult, with practical strategies for success. Paper
Further reading
- Computational Social Science: Obstacles and Opportunities — Lazer et al., 2020. Discusses how cross-disciplinary teams can overcome institutional and methodological barriers. Paper
Week 2 — Collaborative Research Projects · Wed 1 July
Validation in Computational Social Science
Pre-reading
- Text as Data: The Promise and Pitfalls of Automatic Content Analysis Methods for Political Texts — Grimmer & Stewart, 2013. Foundational discussion of validation strategies including human evaluation, replication, and inter-coder reliability. Paper
Tools & Platforms
- irr (R) — R package for computing inter-rater reliability statistics including Cohen's kappa and Krippendorff's alpha.
Further reading
- Integrating Explanation and Prediction in Computational Social Science — Hofman et al., 2021. Practical guidance on balancing predictive accuracy with explanatory validity. Paper
Thematic Index
Foundations & Theory
This theme covers the intellectual history, definitions, and core theoretical debates in computational social science. Participants will explore the origins of CSS, discuss the relationship between prediction and explanation, and examine critiques regarding whether CSS lacks theory. Additionally, resources under this theme address key issues of validation and research reproducibility in computational studies.
- What is CSS and Why It Matters in Australia — Day 1
- Does CSS Lack Theory? — Day 3
- Validation in CSS — Week 2
Key resources: Lazer et al. (2009), Bit by Bit, Hofman et al. (2021)
Ethics & Research Design
Computational social science introduces unique ethical challenges and data quality concerns that go beyond traditional social research. This theme focuses on identifying and mitigating algorithmic and data biases, navigating ethical review processes under Australian and international guidelines, and implementing participant-centric methodologies such as privacy-preserving data donations. Students will learn how to align their research designs with the FAIR and CARE principles.
- Bias in CSS — Day 1
- Ethics in CSS — Day 2
- Data Donations and Participant-Centric Research — Day 2
Key resources: Olteanu et al. (2019), NHMRC National Statement, FAIR Principles
Data Collection Methods
This theme introduces practical techniques and infrastructure for gathering digital trace data. Resources cover the use of API-based dashboards (specifically the AIReD platform), methodologies for collecting and processing user-donated data download packages, and deployed screen capture solutions for observing personalized algorithmic feeds. It also covers the setup and use of the Nectar Research Cloud for scaling data collection.
- The AIReD Platform — Day 3
- Collecting and Analysing Data Download Packages — Day 3
- Screen Capture for Data Collection — Day 4
- Nectar Research Cloud — Day 2
Key resources: AIReD, Port, Nectar Cloud
Computational Analysis (Text, Images, LLMs)
Once data is collected, computational techniques are required to process and analyze it at scale. This theme spans quantitative and qualitative text analysis (using tidy workflows and R/Python NLP libraries), the deployment of large language models for text-as-data annotation, retrieval-augmented generation (RAG) pipelines, and computer vision methodologies for pattern discovery in large image datasets.
- Working with Text Using Computational Techniques — Day 3
- Using LLMs for Text-as-Data Research (Quallmer) — Day 4
- RAG 101 — Day 4
- Image Analysis for Qualitative and Quantitative Research — Day 4
Key resources: Text Mining with R, quallmer, Lewis et al. (2020) RAG paper, Image Machine (QUT DMRC)
Careers, Collaboration & Publishing
This theme focuses on the professional development of computational social scientists. It provides guides and advice on building successful interdisciplinary collaborations that bridge computer science and social research, working with industry and public policy partners, writing grants for CSS projects, and demystifying the publishing process in cross-disciplinary journals.
- Cross-Disciplinary Collaboration — Day 5
- Working With and In the Industry — Day 5
- Career Success — Day 5
- Grant Writing in CSS — Day 5
- Demystifying Publishing in CSS — Day 2
Key resources: Bromham et al. (2016), Lazer et al. (2020)
The Australian Internet Observatory (https://doi.org/10.25956/twvn-ca19) is a co-investment partnership with RMIT University, QUT, University of Queensland, University of Melbourne, Swinburne University, Deakin University and the Australian Research Data Commons (ARDC) through the HASS and Indigenous Research Data Commons (DOI:10.3565/hjrp-b141). The ARDC is enabled by the Australian Government’s National Collaborative Research Infrastructure Strategy (NCRIS).