skills.rest
AKCqhzdy
AKCqhzdyCOMMUNITY
Explore all skills by this author →
GitHub

hkdse-math-evaluation

Compute Spearman ρ and Kendall τ for HKDSE Mathematics AI grading against ground-truth levels.

Updated Aug 27, 2026
Search Tags:#metrics#evaluation#grading#ground-truth#hkdse#spearman#kendalltau
One-click install
npx skills add https://github.com/AKCqhzdy/dse-subject-grading --skill hkdse-math-evaluation
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: hkdse-math-evaluation
Source: https://github.com/AKCqhzdy/dse-subject-grading/tree/main/skills/hkdse-math-evaluation
Command: npx skills add https://github.com/AKCqhzdy/dse-subject-grading --skill hkdse-math-evaluation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Evaluates HKDSE Mathematics (Compulsory Part, Paper 1) AI grading results for year 2024 against ground truth levels (1–5). Computes Spearman ρ, Kendall τ, exact/adjacent match rates, and generates evaluation figures and DOCX reports for 10 students.

Core Features & Use Cases

  • Unified evaluation workflow that compares AI grading outputs against ground-truth levels to quantify rank-based agreement (ρ, τ) and match accuracy.
  • Produces per-year metrics, level division usage when available, and ground-truth mappings to support robust validation.
  • Generates evaluation figures (scatter plots, box plots, confusion matrices) and per-student DOCX reports to facilitate grading quality reviews.

Quick Start

Execute the HKDSE Math evaluation workflow to compare 2024 grading outputs with ground-truth levels and produce evaluation metrics and reports for 10 students.

Frequently Asked Questions about hkdse-math-evaluation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I evaluate AI grading results for HKDSE Mathematics against ground truth levels?▼

Evaluating AI grading results for HKDSE Mathematics involves comparing per-year grading outputs against ground-truth levels (1–5) by computing rank-based metrics like Spearman ρ and Kendall τ. This workflow generates evaluation figures and per-student DOCX reports for 10 students.

What metrics are used to validate HKDSE Math AI grading accuracy?▼

HKDSE Math AI grading accuracy is validated using Spearman ρ and Kendall τ rank-based correlation metrics, alongside exact and adjacent match rates. These metrics quantify the agreement between AI-generated scores and ground-truth level mappings.

How do I generate evaluation reports for HKDSE Mathematics grading outputs?▼

Generating evaluation reports for HKDSE Mathematics grading outputs uses per-year grading data and ground-truth mappings to produce scatter plots, box plots, confusion matrices, and per-student DOCX reports. This facilitates comprehensive grading quality reviews for the evaluated students.

Can I use this evaluation workflow for more than 10 students?▼

The evaluation workflow is specified with an evaluation scope of 10 students using per-year grading outputs. It computes rank-based metrics and generates DOCX reports specifically tailored to this defined student scope for HKDSE Mathematics grading validation.

What inputs are required to compute Spearman and Kendall tau metrics for HKDSE Math grading?▼

Computing Spearman and Kendall tau metrics requires per-year grading outputs, optional level divisions, and ground-truth mappings located under specified paths. These inputs enable the system to compare AI grading results against levels 1–5.

Related Skills

Explore More Education & Research Skills→
Ikram-AlamIkram-Alam

evaluation

Evaluate agent outputs against rubrics and ground truth with multi-dimensional scoring.

Community
Basic
paleoterrapaleoterra
2

code-metrics-analyzer

Compute code quality metrics and generate HTML reports.

Community
Intermediate
flashinfer-aiflashinfer-ai
264

add-reference-tests

Generate and validate pytest tests for flashinfer_trace reference implementations against FlashInfer or SGLang ground truth.

Official
Advanced
YoungLeadersDotTechYoungLeadersDotTech
10

ground-truth-template

Generate canonical ground-truth documentation templates with structured sections and versioning.

Community
Basic