Research Paper Rubric for Doctoral Computer Science

Key Indicators

• •Synthesizes recent State-of-the-Art (SOTA) literature to isolate specific, unsolved research gaps.
• •Articulates a clear, falsifiable research hypothesis or problem statement derived directly from the literature review.
• •Differentiates the proposed contribution from existing methods through direct theoretical or empirical comparison.
• •Validates the solution using appropriate benchmarks, datasets, or proofs that reflect current community standards.
• •Contextualizes the implications of findings, distinguishing between incremental engineering tweaks and fundamental theoretical advancements.

Grading Guidance

The transition from Level 1 to Level 2 hinges on the move from isolation to awareness. Level 1 work typically treats the problem in a vacuum, ignoring existing solutions or citing irrelevant sources, whereas Level 2 work acknowledges the broader field by summarizing related papers. However, Level 2 remains fragmentary because it lists previous work without synthesis, failing to accurately pinpoint where the student's specific contribution fits or why it is necessary. Moving from Level 2 to Level 3 represents the threshold of competence, defined by the alignment of the gap and the solution. Level 3 work shifts from merely summarizing literature to constructing an argument that identifies a valid gap in the SOTA. At this stage, the student proposes a logical solution to that gap, although the comparison against baselines may be standard rather than rigorous. The key distinction is that Level 3 work presents a cohesive narrative where the proposed solution is a justifiable response to the literature, whereas Level 2 often suffers from a disconnect between the background research and the actual implementation. To advance from Level 3 to Level 4, the work must demonstrate rigorous validation and precise positioning. While Level 3 claims novelty, Level 4 proves it by benchmarking against the most competitive current methods (SOTA) rather than weak straw-man arguments. The leap to Level 5 requires an epistemic shift; distinguished work does not merely report better metrics but explains the underlying mechanisms of *why* the improvement occurred, offering generalizable insights or reframing the problem space itself. Level 5 work contributes a durable change to the body of knowledge, whereas Level 4 is a solid, well-executed contribution to a specific application.

Proficiency Levels

Key Indicators

• •Structures the argument to progress logically from problem formulation to technical solution
• •Synthesizes evidence to support the central thesis without logical gaps
• •Constructs clear transitions that link distinct sections and experimental results
• •Justifies technical design choices within the narrative context of the research gap
• •Employs signposting to guide the reader through complex algorithms or derivations

Grading Guidance

Moving from Level 1 to Level 2 requires organizing scattered technical details into a recognizable research paper structure (e.g., Introduction, Methodology, Evaluation) rather than a disjointed collection of notes or code descriptions. To cross the threshold into Level 3 (Competence), the student must establish basic logical connectivity where the problem statement directly motivates the proposed solution, and the evaluation measures the specific claims made, ensuring the reader does not encounter significant non sequiturs or missing links in the reasoning. The transition to Level 4 involves shifting from a functional structure to a persuasive narrative; the student must use effective transitions and signposting to explain 'why' specific design decisions were made, rather than merely listing 'what' was done, creating a cohesive thread that connects the hypothesis to the conclusion. Finally, achieving Level 5 requires constructing an 'inevitable' logic where the solution feels like the natural mathematical or engineering consequence of the problem framing; the narrative architecture is seamless, anticipating reader skepticism and synthesizing complex technical details into a lucid, compelling argumentative arc.

Proficiency Levels

Key Indicators

• •Formats algorithmic pseudocode and mathematical notation according to standard LaTeX or domain conventions.
• •Generates vector-quality data visualizations with precise labeling, legends, and scale representation.
• •Adheres strictly to designated citation styles (e.g., ACM, IEEE) with complete bibliographic integrity.
• •Maintains professional academic syntax free of mechanical errors, colloquialisms, or ambiguous terminology.
• •Integrates technical definitions and variable bindings consistently throughout the manuscript.

Grading Guidance

Moving from Level 1 to Level 2 requires shifting from chaotic or informal presentation to recognizable academic formatting; the work may still contain broken references, pixelated bitmaps, or inconsistent notation, but it attempts to follow a standard template. To cross into Level 3 (Competence), the student must eliminate distracting mechanical errors; mathematical notation becomes syntactically valid, code blocks are readable, and citations are consistently formatted, ensuring the reader focuses on content rather than decoding the layout. Transitioning from Level 3 to Level 4 involves a leap in visual and technical refinement; data visualizations shift from default spreadsheet outputs to high-resolution, professionally captioned figures, and prose becomes concise and strictly formal. Finally, achieving Level 5 (Excellence) requires publication-ready precision where every variable is defined before use, vector graphics are used exclusively for scalability, and the document adheres flawlessly to specific venue constraints (e.g., font embedding, spacing) without sacrificing readability.

Proficiency Levels