Student perceptions of the autograder were mixed, with the LaTeX workflow rated more positively than the grader itself¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? The LaTeX workflow item averaged M = 3.64 (SD = 1.13; 59.7% positive), above feedback quality (M = 3.26) and overall sentiment toward LaTA (M = 3.16), which sat only slightly above neutral. → Jesse A. Rodríguez 2026

Evidence¶

Jesse A. Rodríguez 2026¶

Jesse A. Rodríguez. (2026). LaTA: A Drop-In, FERPA-Compliant Local-LLM Autograder for Upper-Division STEM Coursework. arXiv preprint. https://arxiv.org/abs/2605.05410 doi:10.48550/arxiv.2605.05410

q2 · i? · design · r2

Three 5-point Likert perception items from the anonymous post-term survey (N=159), reported as stacked percentages in Figure 4. The quote carries the printed means, SDs, and positive-response percentages.

"The LaTeX submission workflow itself was the most positively received component, with a mean of M= 3.64 (SD = 1.13) and 59.7% of respondents selecting 4 or 5; the feedback content (M= 3.26 , SD = 0.98; 42.1% positive) and overall sentiment toward LaTA (M= 3.16 , SD = 1.09; 40.3% positive) both sat slightly above the neutral midpoint but below the LaTeX-workflow item."

Discussion¶

Related Claims¶

• An on-premises LLM autograder graded a 200-student course at 1–3 minutes wall-clock time per submission across a full term — related
• Student-reported LaTeX time overhead fell about 46% across the term, from 87 to 47 minutes per assignment — related
• Students in the generative AI-supported condition reported favorable perceptions of the instruction, with feedback for correcting mistakes rated highest — related
• Students reported first-pass false-positive deductions at a mean of 2.94 per term, normalising to about 0.6% of rubric items — related