Digital Methods
- Introduced by
- Richard Rogers, Amsterdam Digital Methods Initiative
- Year
- 2013
- Type
- Empirical research approach for online data
- Family
- Web epistemology
What does it do?
Digital methods is an empirical research approach, developed by Richard Rogers and the Amsterdam Digital Methods Initiative, that takes the methods of online platforms and devices—the link, the like, the hashtag, the search engine ranking—and repurposes them for social and cultural research.
Its guiding maxim is to 'follow the medium': rather than importing offline methods like the survey onto the web, the analyst learns what the medium already counts, ranks, and recommends, and turns those native operations into research instruments.
The intuition behind the method
Imagine you want to study how a public debate is unfolding. The old reflex is to run a survey or a questionnaire as if the web were just another field site. Digital methods reverses this. It notices that search engines already rank pages by relevance, that platforms already cluster users by their likes, and that hyperlinks already encode who endorses whom. Instead of fighting these built-in operations, the researcher repurposes them—asking what Google's top results reveal about source hierarchies, or what a hashtag's co-occurrence pattern says about an issue. The medium is treated as a recording device for social life, and learning to read its native objects becomes the method.
History
Digital methods crystallised at the University of Amsterdam, where Richard Rogers founded the Digital Methods Initiative and the associated govcom.org work in the early 2000s. Rogers's 2013 book Digital Methods consolidated the programme, contrasting it with both 'virtual methods' (which migrate offline instruments online) and digitisation, and proposing instead that researchers 'follow the medium'. The approach grew alongside science and technology studies and the broader 'computational turn' in the social sciences, and Marres and Gerlitz's notion of 'interface methods' (2016) sharpened the reflexive negotiation between social research and the live methods of digital devices.
When to use it
Use digital methods when your research object lives substantially online and when the platforms involved produce traces—links, posts, rankings, engagement metrics—that can be queried, scraped, or accessed via API. The approach suits the study of issues, publics, controversies, and source hierarchies as they form on the web, and it rewards researchers willing to learn each platform's native logics. It assumes that the medium is a legitimate site and instrument of social research rather than mere noise to be filtered out. It is less appropriate when the phenomenon leaves few digital traces, when platform access is closed or unstable, when offline context is decisive, or when the medium's own biases would overwhelm the social signal you seek.
Limitations and considerations
- Findings are entangled with platform logics and algorithm changes, so the medium's biases can be mistaken for social facts.
- API restrictions, rate limits, and platform 'deplatforming' of researchers make data access fragile and results hard to reproduce over time.
- Trace data captures recorded behaviour, not meaning, motivation, or offline context, limiting interpretive depth without complementary methods.
- Sampling is bounded by what the medium counts, so silent actors, private spaces, and non-users are systematically underrepresented.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Leverages data that platforms already produce at scale, granting access to publics and issues that conventional methods reach slowly or not at all.
- The 'follow the medium' principle yields medium-specific, situated findings rather than imposing ill-fitting offline instruments onto online life.
- Repurposing native operations (ranking, recommendation, co-link) gives a reflexive grip on how platforms themselves order knowledge and visibility.
- Supported by a mature, openly documented toolset from the Digital Methods Initiative, enabling reproducible and teachable research designs.
Applications
- Mapping the hierarchy of sources and the composition of publics around an issue by analysing search-engine rankings and hyperlink networks.
- Studying the dynamics of online controversies through hashtag co-occurrence, retweet networks, and platform engagement metrics.
- Investigating how platforms and recommendation systems shape visibility, reputation, and the circulation of (mis)information.
- Tracing the geography and temporality of attention to events using timestamps, location data, and trending topics.
- Auditing search engines, marketplaces, and social platforms to reveal their ordering logics as objects of STS inquiry in their own right.
How it works
How it works
Begin from a social, cultural, or political research question rather than a purely technical one. The analyst asks what can be learned about a phenomenon—an issue, a public, a controversy—through online traces, and frames the inquiry so that the medium's native operations can plausibly answer it rather than forcing an offline instrument onto digital terrain.
Identify the natively digital objects relevant to the question: hyperlinks, tags, timestamps, likes, shares, search-engine rankings, recommendations, or comment threads. These objects are distinguished from digitised objects (scanned texts, online surveys) because they are born of and counted by the medium itself, carrying the medium's own logics of relevance and association.
Repurpose the methods that platforms and devices already use to order content—crawling and ranking, recommendation, trending, co-occurrence—as research operations. As Marres and Gerlitz argue, the analyst negotiates an 'interface' with the device's methods, neither naively trusting them nor discarding them, but reflexively adapting platform metrics into analytic measures.
Collect trace data using dedicated tools that query platform APIs, scrape result pages, or crawl link networks. Sampling follows the medium's own boundaries (a query, a hashtag, a seed set of URLs), and the analyst documents the device settings, query design, and time of capture, because results are situated in the platform's current state.
Analyse and interpret the harvested data—often as networks, rankings, timelines, or frequency profiles—while reflexively accounting for how the medium shaped what was collected. Findings are read both as evidence about the social phenomenon and as evidence about the platform's own ordering work, since the two cannot be cleanly separated.
Common pitfalls
- Treating platform metrics as transparent windows onto society rather than as device-shaped artefacts that require reflexive interpretation.
- Failing to document query design, device settings, and capture date, making a study impossible to situate or repeat as platforms change.
- Confusing digitised objects (an online survey, a scanned archive) with natively digital ones, thereby losing the method's specific purchase.
- Scaling up data collection without a clear social research question, producing impressive but uninterpretable network maps.
Frequently asked questions
How do digital methods differ from 'virtual methods' or general computational social science?
Virtual methods take established offline instruments—surveys, interviews, focus groups—and run them online, treating the web as a neutral container. Digital methods instead 'follow the medium', repurposing the platform's own native operations (ranking, linking, liking) as research instruments. Compared with computational social science generally, digital methods is distinguished by its STS sensibility: it foregrounds how the medium and its devices co-produce the data, rather than treating online traces as raw behavioural records.
What is a 'natively digital' object?
A natively digital object is one that is born of the medium and counted by it—a hyperlink, a tag, a like, a timestamp, a search ranking, a recommendation. It is contrasted with a digitised object, such as a scanned book or an online questionnaire, which merely migrated to the web. The method's distinctive leverage comes from studying natively digital objects, because they already carry the platform's logics of relevance and association that the researcher can repurpose.
Do platform restrictions and the 'post-API age' make digital methods obsolete?
They make it harder but not obsolete. As major platforms have closed or throttled their APIs, the Digital Methods Initiative and its tools have adapted toward platform-compliant scraping, data donation, and the study of recommendation and interface dynamics. The reflexive core—negotiating an interface with the device's live methods—remains, but researchers must now document access conditions even more carefully and treat data fragility as part of the object of study.
Related methods
Sources
Rogers, R. (2013). Digital Methods. MIT Press.
ISBN: 9780262018838Copied.Marres, N., & Gerlitz, C. (2016). Interface methods: renegotiating relations between digital social research, STS and sociology. The Sociological Review, 64(1), 21-46.
Open DOI referenceCopied.