Methodology

Attacks on Education dataset from GCPEA

The TRACE Data Portal currently uses a dataset by the Global Coalition to Protect Education from Attack (GCPEA) from its Education under Attack database which monitors attacks on schools, universities, students, and educators, by armed forces and armed groups, as well as military use of education institutions.

Data sources and data types

GCPEA collects data for the Education under Attack database through three methods: (a) a review of relevant reports and datasets, (b) media searches, and (c) outreach to staff members of international and national organizations working in relevant countries.

The first method is a desk review of relevant reports, datasets, and documents from United Nations (UN) agencies, development and humanitarian non-governmental organizations (NGOs), human rights organizations, government bodies, and think tanks. Each source is reviewed for relevant information in English, French, Spanish, and Arabic.

The second method is media searches. To identify media reports of attacks on education or military use of schools and universities, GCPEA relies heavily on the Armed Conflict Location and Event Data Project (ACLED). The project compiles media and other reports of worldwide political violence and protest events into a database which GCPEA searches for relevant incidents. ACLED has its own standards for achieving data reliability. In addition, GCPEA conducts online searches of local and international media outlets for relevant global and country-specific news articles in English, French, Spanish, and Arabic.

The third approach is gathering relevant information from local organizations and international agencies operating in countries affected by attacks on education. The aim is to (a) identify additional incidents of attacks on education and military use not yet included in the database, and (b) verify and gather additional information on attacks and military use GCPEA had already identified from other sources.

Each method yields three types of data on attacks on education and military use: individual incident reports, tallies of attacks or military use, and qualitative information. GCPEA enters these data into country-specific Microsoft Excel spreadsheets. Reports of individual incidents, tallies, and qualitative information are categorized by type of attack on education or military use, where possible, and the date and location of the attacks are recorded, alongside details on the individuals and institutions affected, disaggregated by gender, level of education, and professional role. GCPEA carefully reviews incidents, tallies, and qualitative information to prevent duplication and double counting.

Data reliability

The reliability of the data in the Education under Attack database varies. Some information comes from organizations such as UN agencies or Human Rights Watch, which maintain rigorous standards for verifying incidents before reporting them. Other incidents are compiled from media outlets that maintain reporting standards but ones that are typically lower than for the organizations mentioned. GCPEA corroborates reports whenever possible and relies on local partners to assess the reliability of sources.

More information

For more information on the sources, reliability, and limitations of the data in GCPEA’s Education under Attack database, see the methodology section of Education under Attack 2024.

GCPEA’s role in the development of the TRACE data portal

GCPEA provided technical guidance and 2020-2023 data for the development of the TRACE Data Portal. Data came from GCPEA’s Education under Attack database, which monitors attacks on schools, universities, students, and educators, as well as the military use of educational facilities. GCPEA draws on this database to publish its flagship Education under Attack report series and other thematic and country-specific reports. GCPEA is an advocacy coalition of organizations that includes: Amnesty International, the Education Above All Foundation (EAA), Human Rights Watch, Plan International, Save the Children, the United Nations Children’s Fund (UNICEF), and the United Nations Educational, Scientific and Cultural Organization (UNESCO). GCPEA is a project of the Tides Center, a non-profit 501(c)(3) organization.

Colombia dashboard from Kobo

Kobo carried out a large-scale data collection project in Colombia to document attacks on education and measures to prevent and respond to these attacks. Data from the project, which covered January 2020 though mid-2025, were then visualized on the Colombia dashboard.

Data collection involved two in-person, semi-structured surveys carried out in April, May, and June 2025. Conducted in Spanish, the surveys were administered in rural areas of Antioquia, Chocó, Nariño, and Norte de Santander – departments selected because they were heavily impacted by attacks on education and for accessibility reasons. The surveys covered 504 rural schools (with head teachers serving as respondents) and 300 educators, for a total of 804 interviews.

Verification of attacks on education

GCPEA’s definitions of attacks on education were used in the study design, to draft the data-collection tools, and during data analysis. Kobo collaborated with GCPEA throughout the study design and data visualization process. To verify attacks on education, the Kobo Research Team triangulated the incoming data - comparing attacks to recent patterns of violence in the area, reports from NGOs and the media, and recent incidents identified by GCPEA. Violent events that did not meet GCPEA's definitions or could not be triangulated were not included.

Ethics and security

Ahead of data collection, Kobo submitted the study to an independent Institutional Review Board, which evaluated the research design, sampling, and data-collection tools. Local and international education-in-emergencies and protection organizations also reviewed the study from an ethical standpoint and provided feedback, which Kobo incorporated.

Respondents provided informed consent before the interviews began. Before starting, enumerators discussed risks, made clear that participation was optional, shared topics of the survey, and informed potential respondents that their anonymized responses would be used to in advocacy efforts. No payment or other advantages were offered for participation in the survey. Only respondents over 18 years of age were included in the study.

Enumerators conducted interviews on digital devices using KoboCollect, a data collection mobile app with encryption, to protect confidential data. Data were stored on a secure cloud server.

For safety, respondents selected the interview site, enumerators were trained in confidentiality, and the optional nature of the survey was stressed from the first point of contact.

To avoid respondent retraumatization, enumerators paid attention to signs of stress, fatigue, and trauma. When appropriate, they responded by slowing down the pace, adjusting the location, and taking other steps to promote respondents’ comfort. They also offered breaks or to end the survey, and connected interviewees with psychosocial support as needed.

To learn more about the project's methodology, see the "Learning under fire" report and methodological guide.