Module A3: Education Indicators and Data Analysis

9. Disaggregation of education indicators

9.1 The importance of disaggregating data

‘Disaggregation’ is the process of breaking down and analysing an indicator by detailed sub-categories. disaggregation of data can reveal differences and disparities that may not be fully reflected in broad aggregate figures. disaggregation by level of education,

geographical areas and population sub-groups can allow us to understand the nature of disparities within the national education system, and where to target priority efforts to reduce disparities.

data in the school records and school census questionnaire in Modules A1 and A2 are often recorded and collected with more detailed breakdown such as enrolment by gender, age and grade; num- ber of teachers by qualification and training; school income by source and expenditure by type; etc. Analysing and comparing data and indicators according to these breakdown classifications can help us to better understand the characteristics, patterns, differences and disparities among individual schools, classes, students and teachers.

When we compare such findings at the school, district, provincial and country level (and also over time), we can obtain very useful information regarding progress, shortfalls, gaps and issues, for developing and implementing new policies and actions. it is good practice to always look at more detailed breakdowns, rather than just the aggregate totals, to see the situation changes over time.

For analytical purposes we can organize data into two types: cross-sectional data and time- series data. cross-sectional data are like snapshots that capture a situation at a particular time; for example, the number of schools, students and teachers by district at the start of the 2010 school year. Time series data consist of a series of snapshots showing the changes over a period of time; for example, the number of students who attended school each year between 1999 and 2006. data are often analysed using both cross-sectional and time-series methods.

By disaggregating data, district education officers can make comparisons between schools, students and teachers in the local area, and with schools, students and teachers in other districts as well as against the national average. Provincial education administrators can compare districts and schools in their province and with those in other provinces. The central Ministry of education can monitor and identify disparities within the education system for the country as a whole, and compare their country with other national education systems.

9.2 dimensions of disaggregation

Disaggregation of data and indicators is especially important in monitoring EFA. To ‘reach the unreached’, and to ensure equality of access and quality of education, detailed and disaggregated information by school, by location and for disadvantaged population groups are required down to the local, if not the community and household levels. Such disaggregated data can help to identify specific localities, households and individuals for priority EFA actions, which may be taken to address the kind of issues and difficulties that face them.

it is, therefore, important to produce and use education indicators that can be further analysed by relevant disaggregated classifications in order to identify specific concerns. For example, enrolment ratios can be calculated separately for boys and girls to gauge gender disparities in participation; pupil-teacher ratios calculated for different classes and grades to improve teacher assignment, and the percentage of qualified teachers by school or by district to compare education quality.

Sometimes, different types of disaggregation may be simultaneously possible for an indicator, such as enrolment ratios by gender, and also by age and grade. Selection of the right type of disaggregation will depend on what one wants to know. For example, if there are concerns about low participation in education among children of a specific age, an age-specific enrolment ratio will need to be calculated. As another example, the percentage of untrained teachers may be more relevant than the percent- age distributions of teachers by academic qualification if one is planning in-service teacher training.

A particular type of disaggregation may be very effective in highlighting differences and disparities among entities at different levels; for example comparing a pupil-teacher ratio of 1:48 in School A against 1:25 in School B, and then further comparing PTRs among classes in School A to identify even more serious cases. But beware that some disaggregations may not be as sharp and effective in com- paring different schools, districts, classes, students or teachers. The effectiveness of different indi- cators by different disaggregations can change with location and time. The choice will depend on the status or issue under review at a specific location at the time of use of the indicator.

For the purpose of monitoring disparities and inequities in EFA, and according to specific national or sub-national conditions and needs, the following types of disaggregation can be systematically included in school records management, data collection and indicators production:

  • Gender: girls and boys.
  • Geographical and administrative units: provinces, districts, localities and other
  • sub-national units, urban/rural, less developed/more developed areas.
  • Social: caste system, occupation, socio-economic status or legal status (eg. birth registration, citizenship).
  • Ethnocultural: ethnicity, religious affiliations or language minorities.
  • Vulnerable: orphans, children of poor families, migrants, working children,
  • children affected by HIV/AIDS or children affected by conflict (IDPs) and disasters.
  • Disabilities: physical or mental.
  • Education source: private/public/faith-based, formal/non-formal/community-based.

Additional types of disaggregation may be added to school records, school census questionnaire and indicators as deemed appropriate and useful to the school or district education office.

An additional advantage of disaggregated data and indicators is their use in analysing and assessing the differentiated impact of broad policies and measures on different local areas, population groups, and in addressing specific problems and issues. For example, disaggregation of the indicator ‘education expenditure per student’ by province, district and school can tell us about local cost levels and also about how the schools spend the government allocated education budget and the financial contributions from the local communities, businesses and families. Without such disaggregation, evaluation of the true impact of governmental policies cannot be made, nor can future policies be based on reliable evaluation results.

Data collection, storage format and processing need to be designed to allow for flexible compilation of data and the appropriate level of disaggregation in terms of administrative levels, geographic areas and population groups. When applying the disparity measures to indicators for the six EFA goals that have been calculated for each administrative level, area and population group, we can gauge the degree of disparities between various groups regarding the different dimensions and facets of EFA and also identify and locate the target disadvantaged population groups and areas, and to plan and deliver appropriate responses to the respective target groups addressing their specific educational needs and difficulties.

To formulate targeted action for reaching the unreached and the under-served, we need to analyse dis- parities in terms of the distribution of educational opportunities, resources, access, quality and outcomes across administrative levels, geographic areas and population groups. There are statistical methods for measuring the magnitude and frequency of various types of disparities. The measurement of disparities and inequities for the six EFA goals, and for any additional indicators we use, can be analysed using basic charts and graphs to make visual comparisons of different sub-populations. in addition, the following measurements can be used to analyse disparities between different target groups:

  • Absolute and relative percentage difference
  • Ratio (male-female, urban-rural, majority-minority)
  • Range (maximum-minimum)
  • Mean and median
  • Percentile and quartile
  • Gender Parity Index (GPI)
  • Representation Index (RI)

Recent studies on progress towards the six eFA goals use four additional measures of disparities:

  • Range ratio (maximum/minimum)
  • Coefficient of variation (deviation from the mean)
  • Gini coefficient of inequality (deviation from equal distribution)
  • McLoone Index/Adjusted McLoone Index (deviation from the median)

As specific methodological characteristics of each eFA indicator and disparity measure may depict gaps and disparities in a different manner, these should be interpreted with care (for more details, see Annex 4).

Activity 11

Review the data and indicators available to you to see how disaggregated are they, and discuss with other school managers and education officers in your district, province or c ountry about their experiences in identifying disparities and inequalities. Then, answer the following questions:

  1. How disaggregated are the data on education in your school, district, province, or country? What kind of disaggregated education indicators can be derived?
  2. What are the difficulties in producing, analysing and interpreting disaggregated education indicators?
  3. How should one go about using disaggregated education indicators in analysing disparities and reaching the unreached?

Comments are closed.