Session Submission Summary
Share...

Direct link:

How to use longitudinal education assessment data for secondary analysis

Sun, March 25, 3:00 to 6:00pm, Hilton Reforma, 4th Floor, Don Emiliano

Group Submission Type: Pre-conference Workshop

Description of Session

The workshop will provide a basic training in techniques for longitudinal data analysis using a mixture of theoretical and practical sessions to illustrate concepts. We will work with data from large-scale school surveys, drawing on Young Lives’ unique study of childhood poverty, conducted in Ethiopia, India, Peru and Vietnam since 2002.

We will start by reviewing the advantages of collecting and analysing longitudinal data and introduce survey designs and research methods with examples from the 2016-17 round of school surveys. The implications of complex data structures will be addressed – including the repeated measurement of student achievement and the linking of students to teachers and schools.

In the second part, participants will be led by trainers in the analysis of an example research question before developing their own enquiries and conducting analysis using STATA or SPSS. Results will be presented and the workshop will close with ideas for further investigation.

Proposal

Workshop leads: Caine Rolleston, Rhiannon Moore, Padmini Iyer, Bridget Azubuike, Jack Rossiter

Workshop Rationale:

As the world gears up to meet the targets set by SDG 4, much work is under way to understand how to measure the extent to which children are really ‘learning’ in schools across the world. In line with this, the 2018 World Development Report encourages stakeholders to ‘measure and track learning better’ and to ‘use the results to guide action,’ as one of three key policy actions. In order to track learning progress globally, improved collection and analysis of longitudinal education data is required. Our CIES 2018 workshop is designed to guide practitioners through the basics of such analysis, drawing on Young Lives’ data.

In its latest round of school surveys, with a sample of over 31,000 children studying in three low- and middle-income countries (Ethiopia, India [Andhra Pradesh and Telangana] and Vietnam), Young Lives has used a novel approach to measure student achievement and link it to student backgrounds and to the characteristics of schools and school systems. Data from these surveys can provide valuable insights into the learning trajectories of students and student groups in three quite different education systems.

This workshop aims to (i) provide participants with an understanding of the basics of longitudinal education data analysis, (ii) provide interactive sessions that reveal the potential of longitudinal education data and stimulate discussion for further analysis; and (iii) encourage later use and uptake of Young Lives’ rich, relevant and publicly available data.

Young Lives is a unique longitudinal study of childhood poverty which has been conducted in Ethiopia, India, (Andhra Pradesh and Telangana), Peru and Vietnam since 2002. The study traces the lives of 3,000 children in two age cohorts in 20 sentinel sites located in each country and data are collected on a variety of human development topics, including nutrition, education and livelihoods. Young Lives’ survey data are all publicly archived with the UK Data Service (http://ukdataservice.ac.uk).

In 2010, a school component was introduced to explore Young Lives children’s experiences of schooling and education in depth. Primary school surveys were conducted between 2010-2013 in all four countries, and in 2016-17 an ambitious cross-country school survey was conducted at the upper primary level (in Ethiopia) and secondary level (in India and Vietnam).

Data from these surveys allow researchers to understand, describe and explain school, teacher and system effectiveness – including how it varies between groups and changes over time. Data can be applied to research questions such as: how much progress are students making in one academic year and what are the drivers of learning trajectories over time, including how these relate to equity?

This workshop will be the first opportunity to work hands-on with Young Lives data from the 2016-17 round of school surveys – one example of large-scale longitudinal school survey data. It will provide participants with a full understanding of the survey design, instrument development, approaches to measuring ‘meaningful learning’ and data structure that links students to teachers, classrooms and schools. This will serve as the foundation for the first interactive session with the data, which will explore an example research question targeted at revealing the variation in learning levels within and between countries and how this relates to specific groups of interest (e.g. according to gender).

Participants will then be guided to develop their own example questions and practice analysis using STATA or SPSS with the support of the workshop trainers before sharing results with the group. Finally, the workshop will lead to discussion of options for further research with a broad range of stakeholders including academics, assessment developers, international development agencies, and practitioners.

The skills obtained from this training will be directly applicable to processing and analysing Young Lives’ publicly available education data. Many of the issues discussed and practised will be relevant to other international large-scale assessments that seek to compare student achievement across countries.

The workshop will consist of brief presentations and contextual information about the data sets but will focus primarily on providing participants with opportunities for hands-on experience analysing these data in response to proposed research and analytical questions. This course requires that participants attend with their own computer with Stata or SPSS installed; everything else will be provided by the workshop hosts.

Duration and size: 3hrs. 20-30 participants.

Special requests: LCD projector in room, please

Sub Unit

Workshop Organizers

©2018 All Academic, Inc.   |   Privacy Policy