🎓 World University Rankings — Power BI Report

A star-schema data model analyzing 592 universities across 73 countries with 29 DAX measures, tracking year-over-year ranking performance from 2025 to 2026.

Project Overview

Higher education institutions, ranking agencies, and policy analysts face a fragmented landscape when trying to understand university performance dynamics at a global scale. Year-over-year ranking shifts, regional disparities, and the interplay between research output, internationalization, and sustainability demand a structured, decision-ready analytical environment — not a spreadsheet.

This Power BI report delivers an enterprise-grade analytical model for the QS World University Rankings, covering 592 universities across 73 countries and 5 global regions for the years 2025 and 2026. The model is built on a clean star schema with a central fact table and 6 dimension tables, enabling fast, reliable analysis across 11 academic performance dimensions without ambiguous filter propagation or calculation errors common in flat-file Power BI designs.

Stakeholders can use this report to benchmark institutional performance against global and regional averages, identify ranking movers and consistent performers, analyze the relationship between research intensity and internationalization, and evaluate sustainability commitments across institution types — all through a 5-page interactive report with cross-filtered slicers, drill-through hierarchies, and calculated classification columns that pre-segment the data for immediate insight.

Key Insights

These findings emerge from the 2025→2026 dataset and reflect patterns that are analytically meaningful beyond surface-level ranking tables.

#	Finding	Analytical Commentary
📊	49.7% of universities improved their ranking (294 of 592)	The near-equal split between improvers (294) and decliners (295) suggests systemic ranking methodology adjustments rather than isolated institutional improvements — a single-year snapshot would miss this nuance entirely
🌍	73 countries represented across 5 regions	Europe and Asia dominate in volume, but regional average rank reveals that representation ≠ performance — key for policy benchmarking
📈	Average ranking volatility of 23.4 positions	High volatility signals that QS rankings are sensitive to short-term metric shifts, making multi-year tracking far more meaningful than any single edition
🔬	International Index (60.1) outpaces Research Excellence (49.8)	Universities are more internationally connected than they are research-productive — a strategic gap that elite institutions actively close
🏆	Only 16.7% of universities are ranked in the Top 100 (99 of 592)	The Top 100 acts as a natural benchmark denominator; country/region comparisons against this threshold reveal true dominance vs. participation
🌱	170 universities are Sustainability Leaders (score ≥ 80)	Nearly 29% of all ranked institutions hold sustainability leadership status — a meaningful structural shift in higher education priorities post-2023
🔒	Only 26% are Consistent Performers (≤5 position change)	The majority of institutions are in active ranking flux, making this dataset particularly valuable for longitudinal tracking
⚡	Average Overall Score: 50.3 with StdDev of 18.4	High standard deviation confirms a wide performance spread — average-based benchmarks alone are misleading without percentile context

Data Model Architecture

Schema Design

The model follows a star schema — the gold standard for analytical Power BI models. A single central fact table connects to 6 dimension tables via Many-to-One relationships. All cross-filter directions are set to single direction (dimension → fact) to prevent circular dependency risks and ensure predictable filter context in DAX calculations.

                    DimCountry  ──────┐
                    DimRegion   ──────┤
                    DimStatus   ──────┼──→  FactUniversity  ←──  DAX Measures
                    DimSize     ──────┤          (592 rows)       (0 data rows)
                    DimResearch ──────┤
                    DimFocus    ──────┘

Design decision: A dedicated DAX Measures table (with no data rows) centralizes all 29 measures into a single, organized repository. This prevents measure sprawl across tables and makes the model substantially easier to maintain and hand off.

Tables

Table	Type	Rows	Purpose
`FactUniversity`	Fact	592	Central table — rankings, scores, foreign keys, calculated columns
`DimCountry`	Dimension	73	Geographic country classification
`DimRegion`	Dimension	5	Continental region grouping (Americas, Europe, Asia, Oceania, Africa)
`DimStatus`	Dimension	4	Institutional public/private status
`DimSize`	Dimension	4	Student population size bands with min/max thresholds
`DimResearch`	Dimension	4	Research output intensity classification with level indicator
`DimFocus`	Dimension	4	Institutional specialization type with focus level
`DAX Measures`	Measures	0	Centralized measure storage — no data rows

Relationships

All 6 relationships are active, Many-to-One, with single-direction cross-filtering.

From (Fact)	To (Dimension)	Join Column	Cardinality
`FactUniversity[CountryKey]`	`DimCountry[CountryKey]`	CountryKey	Many → One
`FactUniversity[RegionKey]`	`DimRegion[RegionKey]`	RegionKey	Many → One
`FactUniversity[StatusKey]`	`DimStatus[StatusKey]`	StatusKey	Many → One
`FactUniversity[SizeKey]`	`DimSize[SizeKey]`	SizeKey	Many → One
`FactUniversity[ResearchKey]`	`DimResearch[ResearchKey]`	ResearchKey	Many → One
`FactUniversity[FocusKey]`	`DimFocus[FocusKey]`	FocusKey	Many → One

✅ 100% referential integrity — zero orphaned foreign keys, zero nulls, all 592 fact rows have valid dimension matches.

Calculated Columns (7)

All calculated columns are stored on FactUniversity and enable slicer-based segmentation without runtime DAX overhead.

Column	Type	Categories	Analytical Use
`Ranking Tier`	String	Top 10 / Top 50 / Top 100 / Top 200 / Top 500 / Beyond 500	Ranking band grouping for tier-based analysis
`Rank Direction`	String	Improved / Declined / Stable	Row-level YoY movement direction
`Performance Category`	String	High Performer / Mid Performer / Emerging	Overall score segmentation (≥75 / 50–74 / <50)
`Rank Change Magnitude`	String	No Change / Small / Medium / Large / Very Large	Magnitude of position shift (0 / 1-10 / 11-50 / 51-100 / 100+)
`Research Intensity`	String	Research Intensive / Teaching Focused	Dual-threshold classification (Academic Rep + Citations both ≥80)
`International Level`	String	Highly International / Moderately International / Locally Focused	Composite of 3 international scores (≥70 / 40–69 / <40)
`Sustainability Level`	String	Leader / Good / Fair / Emerging	Sustainability score banding (≥80 / 60–79 / 40–59 / <40)

Hierarchies (4)

Hierarchy	Levels	Primary Use Case
Performance Profile	Performance Category → Research Intensity → International Level	Segmentation analysis — understand what drives high performance
Ranking Movement	Rank Direction → Rank Change Magnitude	Volatility analysis — characterize movers vs. stayers
University Excellence	Ranking Tier → Performance Category	Identify underperformers within top-tier bands
Sustainability Focus	Sustainability Level → Performance Category	Correlate ESG commitment with academic achievement

DAX Measures Showcase

29 measures organized across 6 analytical categories. Collapsible sections below contain the full inventory; 4 spotlight measures are featured with annotated DAX to demonstrate advanced calculation techniques.

Categories at a Glance

Category	Count	Primary Use
Core Metrics	9	Fundamental KPIs: counts, averages, top-N counts
Benchmarking	8	Comparative performance vs. global and regional averages
Composite	2	Custom composite indexes from multiple raw dimensions
Statistical	3	Distribution analysis, outlier detection, volatility
Movement & Trend	5	YoY ranking dynamics, consistency, improvement rates
Specialty	2	Sustainability leadership and research citation analysis

Spotlight Measures

1. `Universities Above Avg` — Filter Context with Variable Isolation

Universities Above Avg =
VAR GlobalAvg = CALCULATE(AVERAGE(FactUniversity[Overall]), ALL(FactUniversity))
RETURN
CALCULATE([Total Univerisities], FactUniversity[Overall] > GlobalAvg)

Why this matters: A VAR captures the global average before the CALCULATE filter is applied. Without the VAR, using a measure reference inside a filter argument would cause the inner measure to evaluate under the same filter context as the outer CALCULATE — producing a circular or incorrect result. The variable locks the global average as a scalar before any row-by-row filtering occurs.

2. `International Index` — AVERAGEX Over a Composite Expression

International Index =
AVERAGEX(
    FactUniversity,
    (FactUniversity[International Faculty] +
     FactUniversity[International Students] +
     FactUniversity[International Research Network]) / 3
)

Why this matters: AVERAGE can only operate on a single column. Using AVERAGEX iterates over each row, computes the three-column composite expression per row, and then averages the results — producing a true row-level composite average rather than an average of column averages, which would yield different (and incorrect) results when filters are applied.

3. `Overall vs Average` — Removing Filter Context for True Global Benchmarking

Overall vs Average =
AVERAGE(FactUniversity[Overall]) -
CALCULATE(AVERAGE(FactUniversity[Overall]), ALL(FactUniversity))

Why this matters: ALL(FactUniversity) strips all active filters from the table — including user slicer selections — to return the true global average regardless of what a user has selected. This ensures the benchmark is always anchored to the complete dataset, making this measure reliable as a KPI comparison card in any filter context.

4. `Ranking Volatility` — Statistical Measure with AVERAGEX + ABS

Ranking Volatility =
AVERAGEX(
    FactUniversity,
    ABS(FactUniversity[RankingY2026] - FactUniversity[RankingY2025])
)

Why this matters: ABS ensures that both improvements and declines contribute positively to volatility — a university moving from rank 50 to 100 is just as volatile as one moving from 100 to 50. Using AVERAGEX with row-level ABS produces a true mean absolute deviation, which is the correct statistical measure for ranking stability analysis.

Core Metrics — 9 measures

These measures power the fundamental KPI cards and summary statistics across all 5 report pages.

Measure	Returns	Current Value	Technique
`Total Univerisities`	Count of universities in context	592	`COUNTROWS`
`Total Countries`	Distinct countries in context	73	`DISTINCTCOUNT`
`Avg Overall Score`	Average overall performance score	50.3	`AVERAGE`
`Top 100 Univerisity`	Count of universities ranked ≤100	99	`CALCULATE` + filter
`Rank Change`	Sum of YoY position change (positive = improved)	—	`SUM` arithmetic
`Rank Change Category`	Text: Improved / Declined / Stable	—	`SWITCH(TRUE())`
`Universities Improved`	Count with better 2026 rank	294	`CALCULATE` + row comparison
`Universities Declined`	Count with worse 2026 rank	295	`CALCULATE` + row comparison
`Universities Stable`	Count with identical rank	3	`CALCULATE` + row comparison

Benchmarking — 8 measures

These measures support comparative analysis, allowing users to evaluate selections against global or regional reference points.

Measure	Returns	Current Value	Technique
`Median Overall Score`	Median overall score	45.9	`MEDIAN`
`Overall vs Average`	Delta from global avg	—	`ALL()` for context removal
`Regional Avg Rank`	Average 2026 rank in context	299.6	`AVERAGE`
`Top 100 Percentage`	% of universities in Top 100	16.7%	`DIVIDE`
`Top 500 Universities`	Count ranked ≤500	499	`CALCULATE` + filter
`Universities Above Avg`	Count above global avg	249	`VAR` + `CALCULATE`
`Best Rank in Context`	Minimum (best) rank in selection	—	`MIN`
`Above Avg Percentage`	% above global avg	42.1%	`DIVIDE`

Composite Metrics — 2 measures

Custom composite indexes built from multiple raw dimension columns, enabling cross-dimensional performance comparison.

Measure	Returns	Current Value	Technique
`International Index`	Avg of 3 international dimension scores	60.1	`AVERAGEX` over composite expression
`Research Excellence`	Avg of Academic Reputation + Citations per Faculty	49.8	`AVERAGEX` over composite expression

Statistical — 3 measures

Statistical measures for distribution analysis and outlier detection.

Measure	Returns	Current Value	Technique
`StdDev Overall`	Population standard deviation of overall scores	18.4	`STDEV.P`
`Top Quartile Threshold`	75th percentile of overall scores	62.3	`PERCENTILE.INC`
`Ranking Volatility`	Mean absolute YoY ranking change	23.4	`AVERAGEX` + `ABS`

Movement & Trend — 5 measures

These measures quantify ranking dynamics and are the backbone of Page 3 (Ranking Movement & Trends).

Measure	Returns	Current Value	Technique
`Consistent Performers`	Count with ≤5 position change	154	`CALCULATE` + `ABS` filter
`Avg Rank Improvement`	Mean YoY position gain	-0.4	`AVERAGEX`
`Improvement Rate`	% of universities that improved	49.7%	`DIVIDE`
`Consistency Rate`	% of consistent performers	26.0%	`DIVIDE`
`Avg Employment Score`	Average employment outcomes score	48.4	`AVERAGE`

Specialty — 2 measures

Targeted measures for sustainability and research citation analysis, powering Page 5.

Measure	Returns	Current Value	Technique
`Sustainability Leaders`	Count with sustainability score ≥80	170	`CALCULATE` + filter
`High Citation Count`	Count in top 10% for citations	61	`CALCULATE` + `PERCENTILE.INC`

Report Pages

The report is structured as a 5-page analytical narrative, each page answering a distinct business question for a different audience need.

Page 1 — Executive Overview

Business question: What is the state of global university rankings in 2026, and how has the performance landscape shifted year-over-year?

Designed for leadership and non-technical stakeholders who need the big picture at a glance.

6 KPI Cards: Total Universities · Top 100 % · Avg Overall Score · Universities Improved · International Index · Research Excellence
Donut Chart: Ranking movement split (Improved / Declined / Stable)
Clustered Bar Chart: Top 10 countries by university count
Gauge Chart: Improvement Rate vs. 50% target
Sparkline Card: Ranking Volatility indicator
Slicers: Region, Performance Category

Page 2 — Geographic Analysis

Business question: Which regions and countries dominate the rankings, and where is academic performance concentrated versus distributed?

Enables geographic benchmarking for institutional strategy and regional policy analysis.

Map Visual: Countries sized by university count, colored by average overall score
Matrix: Region → Country drill-down with rank, score, and movement metrics
Clustered Column Chart: Regional performance comparison across key dimensions
Scatter Chart: Country size (university count) vs. average quality score
Slicers: Region, Ranking Tier

Page 3 — Ranking Movement & Trends

Business question: Which universities experienced the most significant ranking shifts, and what patterns distinguish movers from consistent performers?

The volatility and momentum analysis page — critical for year-over-year narrative reporting.

Waterfall Chart: Ranking movement breakdown by direction
100% Stacked Bar: Rank change distribution by region
Clustered Bar: Rank change magnitude by performance category
Line Chart: Volatility by country (Top 20)
KPI Cards: Improved / Declined / Stable counts with rates
Table: Biggest movers (top gainers and fallers)
Slicers: Rank Direction, Region

Page 4 — Performance Deep Dive

Business question: What drives high-performer status — research excellence, internationalization, sustainability, or all three?

The analytical core of the report — multi-dimensional performance decomposition using all 11 QS scoring dimensions.

Hierarchy Slicer: Performance Profile drill-down (Performance → Research → International)
Radar Chart: All 11 performance dimensions — shape reveals institutional profile
Scatter Plot: Research Excellence vs. International Index — identify quadrant positioning
Funnel Chart: Performance category distribution (High → Mid → Emerging)
Ribbon Chart: Top performers by key dimension, ranked
KPI Cards: StdDev Overall · Top Quartile Threshold · Median Overall Score
Slicers: Research Intensity, International Level, Sustainability Level

Page 5 — Sustainability & Specialization

Business question: How do institutional focus and sustainability commitment correlate with academic ranking performance?

Examines whether sustainability leadership translates to ranking advantage, and how institutional characteristics (size, research type, focus) segment performance.

Tree Map: Universities grouped by Sustainability Level
Stacked Area: Sustainability score distribution across the full dataset
Clustered Column: Institutional size distribution by performance category
Matrix: Performance breakdown by Research Type × Institutional Focus
Donut Charts: Research Type · International Level · Focus Type splits
KPI Cards: Sustainability Leaders · High Citation Count · Avg Employment Score
Slicers: Status, Size Category, Sustainability Level

Tools & Skills Demonstrated

Tool / Skill	Application in This Project
Power BI Desktop	End-to-end report authoring — data model, DAX, visualization, layout
DAX	29 measures across 6 categories; advanced use of CALCULATE, AVERAGEX, VAR, SWITCH, ALL, PERCENTILE.INC, STDEV.P
Star Schema Design	1 fact table + 6 dimensions + measures table; deliberate single-direction cross-filtering
Power Query (M)	Data ingestion from CSV and XLSX, type enforcement, column normalization
Dimensional Modeling	4 drill-through hierarchies, 7 row-level calculated columns for analytical segmentation
Data Modeling	Cardinality management, referential integrity validation, foreign key mapping
Statistical Analysis	Standard deviation, percentile thresholds, mean absolute deviation (volatility)
Microsoft Excel	Dimension workbook authoring — Status, Size, Research, Focus lookup tables
Analytical Storytelling	5-page report structured as a progressive analytical narrative by audience type

DAX Functions Used: CALCULATE · AVERAGEX · COUNTROWS · DISTINCTCOUNT · DIVIDE · SWITCH · IF · VAR / RETURN · ALL · ALLSELECTED · MIN · AVERAGE · MEDIAN · STDEV.P · PERCENTILE.INC · SUM · ABS

Data Sources

File	Format	Records	Role in Model
`2026-world-university-rankings.csv`	CSV	592 rows	Primary fact data — university names, rankings for 2025 and 2026, all 11 performance dimension scores
`Dimension Workbook.xlsx`	XLSX	Multiple sheets	Pre-built dimension tables for Status, Size, Research, and Focus classifications with surrogate keys

Data origin: QS World University Rankings dataset covering ranking years 2025 and 2026. All data reflects publicly available institutional ranking information. No personal or sensitive data is contained in this dataset.

Note: If you plan to refresh this report, ensure both source files are placed in the same directory as the .pbip file and update the data source paths via File → Options → Data Source Settings in Power BI Desktop.

Project Structure

10-World-University-Ranking-Project/
│
├── word-university-ranking-report-2026.pbip                    # Power BI Project entry point
│
├── word-university-ranking-report-2026.SemanticModel/          # Data model (fully version-controlled)
│   ├── definition/
│   │   ├── tables/                                             # One .tmdl file per table
│   │   │   ├── DAX Measures.tmdl                              # All 29 DAX measures
│   │   │   ├── FactUniversity.tmdl                            # Fact table + 7 calculated columns
│   │   │   ├── DimCountry.tmdl
│   │   │   ├── DimRegion.tmdl
│   │   │   ├── DimStatus.tmdl
│   │   │   ├── DimSize.tmdl
│   │   │   ├── DimResearch.tmdl
│   │   │   └── DimFocus.tmdl
│   │   ├── relationships.tmdl                                  # All 6 Many-to-One relationships
│   │   ├── model.tmdl                                          # Model-level settings
│   │   ├── expressions.tmdl                                    # Power Query (M) data source expressions
│   │   └── cultures/en-US.tmdl                                # Locale / format strings
│   └── diagramLayout.json                                      # Model view diagram positions
│
├── word-university-ranking-report-2026.Report/                 # Report layer
│   └── definition/
│       ├── pages/                                              # One folder per report page
│       └── report.json                                         # Report-level settings and theme
│
├── data/                                                        # All raw data sources
│   ├── 2026-world-university-rankings.csv                      # Primary fact data (592 universities)
│   ├── 22025_QS_World_University_Rankings.csv                  # QS 2025 source dataset
│   └── Dimension Workbook.xlsx                                  # Dimension lookup tables
│
├── docs/                                                        # Documentation
│   └── world-university-ranking-model-documentation.md         # Full technical model documentation
│
├── README.md                                                    # This file
│
└── assets/                                                      # Screenshots for README preview
    ├── model-diagram.png                                        # Power BI model view screenshot
    ├── page-1-executive-overview.png
    ├── page-2-geographic-analysis.png
    ├── page-3-ranking-movement.png
    ├── page-4-performance-deepdive.png
    └── page-5-sustainability.png

Note: This project uses the PBIP (Power BI Project) format, which stores the semantic model and report definition as human-readable text files (TMDL/JSON). This enables full git version control of every DAX measure, relationship, and report page. The assets/ folder requires screenshots taken directly from Power BI Desktop for embedding in this README.

How to Open & Use

Prerequisites

Power BI Desktop — free download from Microsoft
Minimum version: Power BI Desktop 2.96+ (Compatibility Level 1600, released April 2021 or later)

Steps

Clone or download this repository to your local machine
Ensure 2026-world-university-rankings.csv and Dimension Workbook.xlsx are in the same folder as the .pbip file
Open word-university-ranking-report-2026.pbip in Power BI Desktop
If prompted about data source paths, go to File → Options and Settings → Data Source Settings and update the file paths to your local directory
Click Refresh in the Home ribbon to reload data
Navigate report pages using the tabs at the bottom of the canvas

Viewing Without Power BI Desktop

A PDF export of the report can be generated via File → Export → Export to PDF in Power BI Desktop for offline or cross-platform viewing
If this report is published to Power BI Service, a live interactive link will be added here

Compatibility Note

This report is built at Compatibility Level 1600. Opening in significantly older versions of Power BI Desktop may produce compatibility warnings. Upgrading to the latest version of Power BI Desktop is always recommended.

Version History

Version	Date	Changes
1.0	2025-12-26	Initial model creation — fact table, 6 dimensions, core relationships
1.1	2026-01-08	Core DAX measures added (9 measures)
2.0	2026-03-25	Major release — 20 new measures, 7 calculated columns, 4 hierarchies, 5-page report layout
2.1	2026-03-25	Fixed 5 DAX measures (`AVERAGE` → `AVERAGEX`); resolved filter context issue in `Universities Above Avg`
2.2	2026-03-25	Full technical documentation · README

Connect

I build data models and analytical reports that turn raw data into decisions. If you found this project useful or want to discuss Power BI, DAX, or data modeling, feel free to connect.

_{Built with Power BI Desktop · Star schema · 29 DAX measures · 5 analytical report pages}

Navigation Menu

Repository files navigation

🎓 World University Rankings — Power BI Report

Table of Contents

Project Overview

Key Insights

Data Model Architecture

Schema Design

Tables

Relationships

Calculated Columns (7)

Hierarchies (4)

DAX Measures Showcase

Categories at a Glance

Spotlight Measures

1. `Universities Above Avg` — Filter Context with Variable Isolation

2. `International Index` — AVERAGEX Over a Composite Expression

3. `Overall vs Average` — Removing Filter Context for True Global Benchmarking

4. `Ranking Volatility` — Statistical Measure with AVERAGEX + ABS

Report Pages

Page 1 — Executive Overview

Page 2 — Geographic Analysis

Page 3 — Ranking Movement & Trends

Page 4 — Performance Deep Dive

Page 5 — Sustainability & Specialization

Tools & Skills Demonstrated

Data Sources

Project Structure

How to Open & Use

Prerequisites

Steps

Viewing Without Power BI Desktop

Compatibility Note

Version History

Connect

About

Releases

Packages

Contributors

Languages

Name		Name	Last commit message	Last commit date
.vscode		.vscode
data		data
docs		docs
word-university-ranking-report-2026.Report		word-university-ranking-report-2026.Report
word-university-ranking-report-2026.SemanticModel		word-university-ranking-report-2026.SemanticModel
.gitignore		.gitignore
README.md		README.md
notebook.ipynb		notebook.ipynb
word-university-ranking-report-2026.pbip		word-university-ranking-report-2026.pbip

Latest commit

History

Folders and files

Repository files navigation

🎓 World University Rankings — Power BI Report

Table of Contents

Project Overview

Key Insights

Data Model Architecture

Schema Design

Tables

Relationships

Calculated Columns (7)

Hierarchies (4)

DAX Measures Showcase

Categories at a Glance

Spotlight Measures

1. Universities Above Avg — Filter Context with Variable Isolation

2. International Index — AVERAGEX Over a Composite Expression

3. Overall vs Average — Removing Filter Context for True Global Benchmarking

4. Ranking Volatility — Statistical Measure with AVERAGEX + ABS

Report Pages

Page 1 — Executive Overview

Page 2 — Geographic Analysis

Page 3 — Ranking Movement & Trends

Page 4 — Performance Deep Dive

Page 5 — Sustainability & Specialization

Tools & Skills Demonstrated

Data Sources

Project Structure

How to Open & Use

Prerequisites

Steps

Viewing Without Power BI Desktop

Compatibility Note

Version History

Connect

About

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

1. `Universities Above Avg` — Filter Context with Variable Isolation

2. `International Index` — AVERAGEX Over a Composite Expression

3. `Overall vs Average` — Removing Filter Context for True Global Benchmarking

4. `Ranking Volatility` — Statistical Measure with AVERAGEX + ABS