Expert-annotated Hungarian reflective essay dataset

Product or Programme · All products and programmes
Evidence · 3 claims (3 for) · 1 study (1 design), q2 · 0 of 1 report an effect size · 3 claims rest on one study

Description

A corpus of approximately 1,954 Hungarian reflective essays from Early Childhood Education students, expert-annotated on a four-level reflection scale by Zsolt Csibi and colleagues.

Components

• Expert-annotated Hungarian reflective essay dataset (1,954 essays, four-level reflection scale): A large expert-annotated corpus of Hungarian reflective essays collected over 4 consecutive years from the Early Childhood Education programme at a large public university in Central Europe. "The dataset contains almost 1954 annotated essays from roughly 450 students." Essays were written after pedagogical practice placements in nurseries and rated by experts on a 0–3 reflection scale (0 no reflection to 3 high reflection). The distribution is heavily imbalanced: "1.8% of the essays are rated as level 0, 2.8% as level 1, 27.4% as level 2 and 68% as level 3." The dataset is not publicly available due to privacy and ethical considerations. (Zsolt Csibi et al. (2026))

Claims

• In the shallow-model ablation, class balancing lowered every metric, oversampling did not improve overall performance, and semantic embeddings raised ROC-AUC but lowered accuracy and F1 [+W]
• Hungarian transformer models achieve slightly lower overall score (about 68%) but better minority-class sensitivity than shallow models [+W]
• Shallow machine learning with SMOTE oversampling and Qwen3 embeddings achieves the best overall reflection-classification performance (0.7176) on Hungarian essays [+W]

Related Products and Programmes

• Edustories dataset

Key Sources

• Zsolt Csibi, Kristian Fenech, Mónika Sándor, Mónika Serfőző, Kinga Gyöngy. (2026). Automatic Reflection Level Classification in Hungarian Student Essays. https://arxiv.org/abs/2605.02402

Page metadata

Field	Value
type	product
id	expert-annotated-hungarian-reflective-essay-dataset
title	Expert-annotated Hungarian reflective essay dataset
description	A corpus of approximately 1,954 Hungarian reflective essays from Early Childhood Education students, expert-annotated on a four-level reflection scale by Zsolt Csibi and colleagues.
status	review
product_kind	dataset
generated	by process:settle-candidates on 2026-10-11
trust	unverified — no one has confirmed this page's content

Frontmatter is the page's machine-readable record — see the Schema & Guide.