brycewang-stanford Theory Building (newms-theory-building)
NM&S publishes empirical, critical, and computational work, but it consistently rewards papers that deliver a concept or mechanism that travels beyond the studied platform. "Theory" here is broad: it can be a new construct, a re-specification of a travelling concept, a critical/normative argument, or a mechanism that explains how a digital-media phenomenon works socially. It is rarely a formal model.
When to trigger
- Your draft has a finding but no portable idea ("X is associated with Y" and stop)
- A reviewer called the work "atheoretical," "descriptive," or "a-just-so story"
- You need to state what concept a reader in another field could carry elsewhere
What counts as theory at NM&S
| Mode | What it delivers | Example shape |
|---|---|---|
| New construct | a named thing the field lacked | "context collapse," "networked privacy" |
| Re-specified concept | sharper definition + scope | redefining "affordance" relationally |
| Mechanism | how/why a process works | datafication → self-discipline |
| Critical / normative | what is at stake and for whom | platform power as a justice problem |
| Typology | meaningful kinds, not a scale | varieties of digital disconnection |
The portability test (NM&S-specific)
Ask: "Could a researcher studying a different platform or context import this idea and learn something?" If the answer is "only if they study the same app," you have a finding, not a concept. Make the idea portable by (a) naming it, (b) defining it independent of the platform, and (c) stating its scope conditions — where it holds and where it breaks.
Build sequence
- Name the concept or mechanism. A handle others can cite.
- Define it independent of your data. It must make sense without your platform.
- Specify the mechanism. What produces what, through what process? Not just correlation.
- State scope conditions. When does it operate, and when not? Boundary conditions earn trust.
- Derive observable implications. What would we see if the argument is right vs. wrong? (This links
to
newms-research-design's adjudication move.) - Connect to theory. Anchor in media/communication and social theory — affordances, mediatization, publics, surveillance, political economy of platforms — so the contribution is legible and located.
For critical / theoretical papers (no original data)
The argument is the contribution. Make the claim explicit, walk the strongest counter-argument, and show what the framing makes visible that prior framings hid. Avoid free-floating critique: tie it to concrete digital-media practice and name what is at stake and for whom.
Worked micro-example (illustrative)
Finding: gig couriers reorganize their day around an opaque ranking score.
Concept named: "datafied control" — control exercised through continuous datafication and ranking.
Defined platform-independently: any setting where opaque metrics govern access to work/reward.
Mechanism: visibility of the score → anticipatory self-discipline → internalized managerial logic.
Scope conditions: holds where metrics are opaque and tied to material stakes; weak where workers can
see/contest the rule. Observable implication: workers comply *before* sanction, not after.
Calibration anchors
- Name it or lose it. An unnamed mechanism cannot be cited or carried — naming is the contribution.
- Define without the platform. If the concept needs the app to make sense, it isn't portable.
- Scope conditions are strength, not hedging. Saying where the idea breaks earns a referee's trust.
Anti-patterns
- A finding with no mechanism — "X is associated with Y" and stop
- HARKing: presenting post-hoc patterns as if hypothesized in advance
- Borrowing a travelling concept (affordance, datafication) without sharpening or sourcing it
- Critique with no grounding in concrete digital-media practice
- A "theory" that only describes one platform and travels nowhere
Output format
【Concept / mechanism】the named, portable idea
【Definition】stated independent of the platform
【Mechanism】what produces what, and how
【Scope conditions】where it holds / breaks
【Observable implications】what we'd see if right vs. wrong
【Next】newms-research-design
Supplementary resources
../../resources/worked-examples/01-introduction.md— a portable concept front-loaded in an intro../../resources/exemplars/library.md— NM&S papers whose concepts travel