Abstract

This thesis develops a practical monitoring approach for Kosovo’s PISA outcomes by linking country-level benchmarking of PISA 2015, 2018, and 2022 with measurable system signals and a focused school-level triangulation. Quantitatively, it benchmarks Kosovo against Western Balkans Six (WB6) and selected European peers through mathematics mean scores, below-Level-2 shares, and top-performing shares. It then uses k-means clustering to define empirical peer groups and tests an early-warning screening exercise in which 2015/2018 information is used to classify held-out 2022 high-risk profiles. Qualitatively, it triangulates the statistical signal patterns with policy document analysis and a low-burden Grade 10 mathematics micro-study in Prishtina municipality (target: eight schools; realized sample: six), using anonymous teacher and student questionnaires and a rapid textbook/material audit. The findings show that Kosovo remained in a persistent lower-performance, high-risk mathematics profile across all observed cycles, with an exceptionally large low-performing group and almost no top-performing presence. The early-warning component suggests that a simple persistence rule is currently more defensible than a more complex logistic prototype under small-sample and limited-signal conditions. The field evidence does not estimate national prevalence, but it makes selected macro concerns locally plausible: pacing pressure, procedural task density, uneven explanatory clarity, limited reasoning time, and uneven support conditions were visible in participating schools. The thesis concludes that Kosovo would benefit from a compact monitoring framework that combines stable PISA-facing KPIs with repeatable local opportunity-to-learn checks. The framework is intended for prioritization and follow-up investigation, not causal diagnosis, teacher evaluation, or school ranking.

Publication Date

5-2026

Document Type

Master's Project

Student Type

Graduate

Degree Name

Professional Studies (MS)

Advisor

Ermir Rogova

Advisor/Committee Member

Leandrit Mehmeti

Comments

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Professional Studies: Data Analytics

Campus

RIT Kosovo

Download

Share

COinS