Inside the Research Process

Begin by stating the uncertainty being investigated. Identify the terms, the relevant scale and what a useful answer would change. A model, simulation, literature review or prototype should be selected because it can address that uncertainty. This keeps the project from becoming a demonstration of a favored tool with no clear research purpose.

A prior-work map connects definitions, established results, competing approaches and known failures. It helps distinguish a new contribution from a rediscovered idea and reveals prerequisites that must be understood before a project can proceed. Sources should remain attached to the claims they support, with disagreement recorded rather than flattened into one summary.

A toy model can reveal a contradiction. A benchmark can show that a proposed method does not improve on a simpler baseline. A prototype can expose an integration dependency. The right early experiment is one that changes what the team knows. Record inputs, parameters, outputs and limitations so the lesson can survive beyond the person who ran it.

A clear explanation and a working implementation answer different questions. Review needs to inspect both when both are relevant. A technical claim should identify its supporting derivation, measurement or experiment. A strategic assessment should identify competing explanations. An invention should identify the problem, the proposed mechanism and the evidence that it works.

A contribution can be a well-defined question, a reproducible negative result, a corrected assumption, a technical note or a more mature paper. Match the label to the artifact. Preserve version history, references and corrections, then state the next unresolved step so collaborators can join the work without reconstructing its entire history.