research-design-positioning

Original：🇺🇸 English

Translated

Use when developing academic research ideas, framing research questions, defining scope, novelty, contribution claims, hypotheses, operational definitions, evidence plans, or publishable positioning before SOTA, experiments, or writing.

3installs

Sourcevincenzoimp/academic-research-skills

Added on2026-05-20

NPX Install

npx skill4agent add vincenzoimp/academic-research-skills research-design-positioning

SKILL.md Content

View Translation Comparison →

Research Design Positioning

Turn rough ideas, literature gaps, and hunches into researchable questions and defensible contribution claims. Use this before experiments, SOTA synthesis, or paper drafting when the project is not yet falsifiable and scoped.

Read First

references/research-positioning-policy.md

```
references/repository-contract.md
```
```
references/output-contracts.md
```

Workflow

State the problem, target community, and why the field should care now.
Extract candidate gaps from
```
sota/gaps.md
```
,
```
sota/synthesis.md
```
, source pages, and venue expectations.
Draft candidate research questions and reject questions that cannot be answered with available sources, data, experiments, or defensible analysis.
Convert viable questions into contribution claims and identify the closest prior work for each claim.
Define constructs, variables, units of analysis, operational definitions, and scope boundaries.
State hypotheses or expected patterns when appropriate.
Define the minimum evidence needed: SOTA, dataset, experiment, qualitative coding, reproduction, benchmark, expert judgment, or artifact evaluation.
Mark claims as hypothesis, partial, supported, rejected, or out of scope.
Save the stable version in
```
docs/methodology/research-design.md
```
.
Add open issues to
```
wiki/open_questions.md
```
or
```
wiki/questions/
```
.

Contribution Types

empirical finding
method, algorithm, model, or system
dataset, benchmark, taxonomy, or measurement instrument
reproduction, replication, or negative result
theory, conceptual reframing, or design framework
survey, mapping study, or structured literature synthesis

Quality Gate

A research design is not ready for experiments or writing until it has:

problem, audience, and venue fit
research questions with scope and evidence path
closest prior work and specific delta
operational definitions for key constructs
success criteria and failure condition
minimum evidence needed for each claim
threat that could invalidate the contribution

Output

Prefer concise tables in

docs/methodology/research-design.md

:

rq_id

,

question

,

scope

,

evidence_needed

,

method

,

risk

,

status

claim_id

,

claim

,

closest_prior_work

,

delta

,

evidence

,

failure_condition

,

status

construct

,

definition

,

observable_proxy

,

limitations

Do Not

Treat "no one has done X" as novelty without searching.
Claim generality beyond the intended population, benchmark, domain, or source base.
Start drafting the paper before identifying the strongest adjacent work.
Let an interesting question survive if it has no evidence path.