jams-theory-development
GitHub为JAMS论文构建概念框架与机制，推导假设、指定中介与边界条件。适用于理论薄弱或缺乏营销理论支撑时，避免套用通用理论，强调逻辑推导与管理启示。
Trigger Scenarios
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill jams-theory-development -g -y
SKILL.md
Frontmatter
{
"name": "jams-theory-development",
"description": "Use when building the conceptual framework and mechanism for a Journal of the Academy of Marketing Science (JAMS) manuscript — deriving hypotheses from marketing theory, specifying mediators and boundary conditions, and carrying the logic through to a managerial implication. Builds the argument; it does not select the question (jams-topic-selection) or run the analysis (jams-data-analysis)."
}
Theory Development (jams-theory-development)
When to trigger
- The question is set but the conceptual framework that generates the hypotheses is thin
- Reviewers will ask "what is the mechanism?" and "why these moderators?"
- You have a strong empirical pattern but no organizing marketing theory
- You are tempted to bolt a generic psychological or economic theory onto a marketing finding
JAMS's theory bar: a framework that does work
JAMS papers are expected to develop theory, not merely apply it. The house archetype is the conceptual framework with a nomological network — antecedents → focal construct(s) → consequences, with mediators that name the process and moderators that name the contingencies. Marketing reviewers want to see the framework drawn (a conceptual model figure) and defended construct by construct, each path justified by a stated theoretical logic, not by a citation alone.
JAMS draws on the field's standard theoretical lenses — resource-based view and dynamic capabilities (strategy), relationship-marketing / commitment–trust and transaction-cost / governance (B2B and channels), service-dominant logic and SERVQUAL-tradition gaps models (services), signaling and brand-equity theory (branding), diffusion and the marketing–finance interface (innovation, firm value), and consumer information-processing and self-construal theories (behavior). Pick the lens the phenomenon actually demands and use it to derive predictions, not to decorate them.
Build the conceptual logic
- Define every focal construct conceptually before it is operationalized — JAMS reviewers police construct definitions and conceptual/operational alignment hard.
- Draw the nomological net: antecedents, focal construct(s), consequences. Each arrow is a hypothesis with a theoretical reason, not just a sign.
- Specify the mechanism (mediation): name the process variable that carries the effect, so the story is "X → M → Y," not "X → Y, somehow."
- Specify boundary conditions (moderation): when does the effect strengthen, weaken, or reverse? Contingency theory is JAMS's bread and butter; a contingency makes the framework generalizable and the managerial advice conditional.
- Carry the logic to the manager: if the mechanism holds, what should a decision maker do, and when? The managerial implication must be derivable from the theory, not appended.
Two legitimate routes
- Theory-first: lens → framework → hypotheses → test. The standard JAMS structure.
- Empirics-first / abductive: a robust, surprising marketing pattern leads, and a conceptual account is built to explain it. Acceptable at JAMS if declared honestly — do not present a post-hoc account as a priori theory (HARKing).
Hypothesis craft
JAMS reviewers read hypotheses closely, so write them to do real theoretical work:
- One claim per hypothesis. A hypothesis bundling a main effect and an interaction is two hypotheses; split them so each can be supported or rejected cleanly.
- State the direction and the reason. "H2: Because [mechanism], the effect of X on Y is stronger when [moderator] is high." The clause after "because" is what separates a JAMS hypothesis from a hunch.
- Mediation as a stated path. Hypothesize the indirect effect (X → M → Y) explicitly so the analysis can test it directly, rather than inferring process from a significant total effect.
- Boundary conditions that are theoretically motivated, not data-mined — name the moderator before you see the interaction, and tie it to the lens.
- Keep the wording identical to the conceptual model figure and the results section; drift between them is a classic reviewer catch.
Conceptual papers (the framework genre)
JAMS also publishes purely conceptual / framework articles, and the bar is higher than a literature review:
- Deliver a genuinely new organizing structure — a typology, a process model, a set of propositions, or an integrative framework that reconciles competing streams — not a summary of what is known.
- State propositions (not testable hypotheses) with the logic that generates each, and specify the conditions under which the framework applies.
- Make the research agenda concrete: what the framework lets future empirical work test that it could not test before.
- Carry the framework to a managerial reading — even a conceptual JAMS paper should tell a decision maker how the framework changes how they think about a marketing problem.
Avoid the construct-validity traps at the theory stage
Many measurement problems are really theory problems, so catch them now: a construct that cannot be defined distinctly from a neighbor will fail discriminant validity later; a construct defined so broadly that it absorbs its own antecedents creates tautology; and a mediator that is conceptually the same as the outcome cannot carry a mechanism. Before handing off to jams-methods, confirm each construct has a single, bounded definition, that no two constructs overlap conceptually, and that mediator, predictor, and outcome are genuinely distinct ideas — not three labels for the same thing.
Checklist
- Every focal construct conceptually defined and distinct from its neighbors
- Conceptual model figure drawn; each path is a reasoned hypothesis
- Mechanism named via at least one mediator (process, not black box)
- Boundary conditions / moderators specified, with the theoretical reason they matter
- Theoretical lens fits the marketing phenomenon (not a generic borrowed theory)
- Managerial implication is derived from the mechanism, not bolted on
- Route (theory-first vs. empirics-first) declared honestly
- Constructs are conceptually distinct (no tautology, no predictor/outcome overlap)
Anti-patterns
- Hypothesis-by-citation: "Prior work finds X→Y (Cite); thus H1" with no mechanism
- Theory theater: an elaborate framework that adds no new substantive insight
- Main-effect-only logic: no mediator, no moderator, no contingency
- Borrowed-theory veneer: a psychology/economics theory pasted on without marketing fit
- Construct sprawl: many overlapping constructs with no discriminant logic
- A mechanism with no managerial consequence anyone could act on
Anti-patterns (continued)
- Borrowing a lens you do not use: naming a theory in the front end that never constrains a hypothesis or interpretation.
- A model figure that overpromises: boxes and arrows the empirics will not actually test.
- Propositions with no logic in a conceptual paper — assertions are not theory.
Output format
【Theoretical lens】RBV / commitment-trust / S-D logic / signaling / diffusion / consumer IP / ...
【Focal constructs (defined)】[...]
【Conceptual model】antecedents → focal → consequences; figure drawn? yes/no
【Mechanism (mediator)】X → M → Y: [...]
【Boundary conditions (moderators)】[...] + why they matter theoretically
【Route】theory-first / empirics-first (declared)
【Managerial implication derived】if mechanism holds → manager should [...]
【Next skill】jams-literature-positioning
Version History
- 1839142 Current 2026-07-05 13:59
