Nic Spaull & Adaiah Lilenstein Stellenbosch University

Slides:



Advertisements
Similar presentations
« P ROGRAMME DA NALYSE D ES S YSTÈMES E DUCATIFS DE LA CONFEMEN » (PASEC) INITIAL TECHNICAL MEETING OECD P ARIS (F RANCE ) - D U 27 AU 28 JUIN 2013.
Advertisements

Measuring and Monitoring the Quality of Education Christopher Colclough University of Cambridge.
A.Review of PISA constructs and indices (and variables) to identify those that are likely to be operable and not operable in developing country contexts,
2.E Developing an Assessment Framework. Starting from Strand A: Modules Cognitive instruments developed under Strand A for 15-year-old students attending.
Identifying and profiling out of school populations – lessons from the UNICEF/UIS Out of School Children Initiative PISA for Development, Paris,
From Access to Learning: System effects that continue to delay children’s ability to learn. Ania Chaluda Education Policy and Data Center, FHI 360
Measuring Gender and Education Quality - The Need for Social Outcomes Huma Zia, Sahar Saeed and Saba Saeed Comparative and International Education Society.
Figurative Language Development Research and Popular Children’s Literature: Why We Should Know, “Where the Wild Things Are” Kathleen Ahrens.
Literacy Assessment and Monitoring Programme (LAMP) UNESCO Institute for Statistics.
NEVER TOO LATE TO COMPLETE SCHOOL Results of the UIS survey on adult education and literacy programmes in Latin America and the Caribbean UIS INFORMATION.
Multiple Indicator Cluster Surveys Data dissemination and further analysis workshop Literacy Education MICS4 Data Dissemination and Further Analysis Workshop.
EDUCATION FOR ALL : Achievements and Challenges BELLA Nicole Launch Paris, 9 April 2015.
Literacy IGCSE Global Perspectives. Literacy According to UNESCO’s 1958 definition, the term refers to the ability of an individual to read and write.
Overview of U.S. Results: Digital Problem Solving PIAAC results tell a story about the systemic nature of the skills deficit among U.S. adults.
Multiple Indicator Cluster Surveys Data Interpretation, Further Analysis and Dissemination Workshop Literacy and Education.
« P ROGRAMME D’A NALYSE D ES S YSTÈMES E DUCATIFS DE LA CONFEMEN » (PASEC) W ASHINTON DC 8 – 11 A PRIL.
Overview of U.S. Results: Focus on Literacy PIAAC results tell a story about the systemic nature of the skills deficit among U.S. adults.
M&E progress in EFA Goals Prepared by Nyi Nyi THAUNG, UIS (Bangkok) Capacity Building Workshop on Monitoring and Evaluating Progress in Education in the.
PIAAC: SOME THINGS IT TELLS US ABOUT HIGHER EDUCATION New England Association of Colleges Conference 2014 William Thorn Senior Analyst, OECD.
NAEP 2011 Mathematics and Reading Results NAEP State Coordinator Mark DeCandia.
UNESCO Institute for Statistics 1 Education related MDG indicators: methodology and issues Ioulia Sementchouk UNESCO Institute for Statistics November.
Reproducing Inequality: Family Background and Schooling in Peru Santiago Cueto, Alejandra Miranda, Juan León, and María Cristina Vásquez GRADE - Young.
Education and Assessment in France Bruno Trosseille DEPP - Assessment, Forecasting and Performance Directorate Ministry of Education, France International.
Free Education and Student Test Scores in Chad Gbetonmasse B. Somasse Worcester Polytechnic Institute (WPI) International Conference on Sustainable Development.
Provincial Assessment Results Anglophone West School District November 26, 2015.
Do girls and boys learn differently? 17 th July 2014 Dr. Tia Linda Zuze HSRC.
Growing up in Poverty: Findings from Young Lives in India Round 4 Launch 22 September 2014 CESS, Hyderabad.
11 PIRLS The Trinidad and Tobago Experience Regional Policy Dialogue on Education 2-3 December 2008 Harrilal Seecharan Ministry of Education Trinidad.
Patterns in educational outcomes for boys and girls in South Africa and the surrounding region Ntsizwa Vilikazi, Stephen Taylor & Nompumelelo Mohohlwane.
Quality of education in Cambodia, Lao PDR, and Viet Nam: Policy messages from PASEC study 2011/2012 CIES, Vancouver, March 9th 2016.
International treaties with relevance to education Universal Declaration of Human Rights Free elementary education International Covenant on Economic,
PASEC2014 EDUCATION SYSTEM PERFORMANCE IN FRANCOPHONE SUB-SAHARAN AFRICA Competencies and Learning Factors in Primary Education 7 December 2015 Almadies.
1 Perspectives on the Achievements of Irish 15-Year-Olds in the OECD PISA Assessment
1 Main achievement outcomes continued.... Performance on mathematics and reading (minor domains) in PISA 2006, including performance by gender Performance.
TCAI: Lessons from first Endline TCAI Development Partners Feb 27, 2013.
1 PISA 2006 Main achievement outcomes and factors associated with performance on science Eemer Eivers, Gerry Shiel & Rachel Cunningham Educational Research.
Monitoring SDG4 (with focus on Target 4.2): national, regional and global data Silvia Montoya Director, UNESCO Institute for Statistics.
Looking for statistical twins
The education indicators concerning completion
Getting Reading Right Nic Spaull 24 May 2016.
The Long-Term Effects of Universal Primary Education:
What is PIAAC?.
Monitoring of EFA (Literacy)
PROMISING PRACTICES OF MULTI-GRADE TEACHING IN LAO PDR
ASER PAKISTAN ASER - The Annual Status of Education Report (ASER) is a citizen led large scale national household survey about the quality of.
Australian Council for Educational Research in the International Context Marion Meiers.
Literacy Assessment and Monitoring Programme (LAMP)
Context : the experience in 2010 GMR estimated annual gap for achieving universal primary and lower secondary education in at US$25.
Monitoring education inequality at the global level
國立臺灣師範大學英語系陳秋蘭 PISA 與英語閱讀素養 國立臺灣師範大學英語系陳秋蘭
The progress we’ve made… and the road to travel
NEVER TOO LATE TO COMPLETE SCHOOL
Types of Large Scale Assessment
Presenter: DR. Michaela Gulemetova Elnaz Safarha Laurence Dessein
Sheldon Shaeffer Chair, Board of Directors of the Asia-Pacific
Learning Assessments Regional and Global Initiatives
Gender and Development: Issues in Education
Kuwait National Curriculum
New Internationalist Ready Lesson: Intermediate
Schooling and Adolescent Reproductive Behavior in Developing Countries
Using PAL Network data to identify children who are being left behind
VIETNAM SPONS ECD MIDLINE RESULTS
Module 7: Monitoring Data: Performance Indicators
Program for International Student Assessment (PISA) 2009 Results
2015 International Development and Early Learning Assessment (IDELA)Baseline Results: ELM project Afar and South Omo, Ethiopia.
Data Analysis Workshop on Education Indicators
Young Lives, University of Oxford
Chart 1.1 Educational pathways for adolescents, Page 14
Getting Reading Right Nic Spaull 24 May 2016.
ARE COUNTRIES ON TRACK TO ACHIEVE SDG 4?
Presentation transcript:

Nic Spaull & Adaiah Lilenstein Stellenbosch University ASSESSING EARLY LITERACY OUTCOMES IN BURKINA FASO AND SENEGAL  Using DHS and PASEC to Combine Access and Quality  Nic Spaull & Adaiah Lilenstein Stellenbosch University CIES 2019 San Francisco 15 April 2019

The need for this book?

Additional focus on West Africa (IBE) “…22% of primary school pupils were in schools that had no toilet facilities whatsoever (CONFEMEN, 2015, p. 104). In Togo, Chad, and Côte d’Ivoire, the figure is 40%. In Niger, the average Grade 2 class had 48 pupils in it, with much higher class sizes in Senegal (52 pupils per class), Benin (57 pupils per class), and a bewildering 80 Grade 2 pupils per class in Burkina Faso (CONFEMEN, 2015, p. 99)” Additional focus on West Africa (IBE)

Arabic + Hindi + Bengali + Swahili = more than 1-billion people 69-70% of all article in the 3 journals are on Eng/French 6% of total (rounding). Journal of Research in Reading, Scientific Studies of Reading, Reading Research Quarterly

5 T’s of early literacy

Pretorius 2019

Menon et al 2019

Menon et al 2019

Geneaology of the work… SACMEQ (Southern/East Africa)  PISA (Turkey)  now PASEC (Francophone W-Africa

Burkina Faso and Senegal

In this presentation we report on forthcoming research using more recent data (PASEC 2014) than that used in the IBE chapter. We also report on more countries than just Burkina Faso and Senegal.

Countries No research on combined measures of access and quality: Although some estimates of enrolment statistics, and learning outcomes from PASEC reports, no research exists which looks at access and quality simultaneously

ACCESS - DHS QUALITY - PASEC

Quality Burkina Faso > Chad? 58% 15% Use available data on access and learning Combine Access-to-learning

Access Chad > Burkina Faso? 36% 40% Use available data on access and learning Combine Access-to-learning

Access and Quality Chad? Burkina Faso? Use available data on access and learning Combine Access-to-learning

Access to Learning 23% 7% Burkina Faso > Chad

Motivation this issue of expanding access and attainment has been one of the defining features of education systems in developing countries for the last two decades. Access

Anglophone Sub-Saharan Africa Spaull & Taylor, 2015

Data DHS PASEC Comparable Reliable Generally available “Highest year of education attained” Can be dated (2010-2014) PASEC 2014 Well developed assessment Released items are available… in French Add notes on IRT

DHS Data Country DHS year Years dated by Benin 2011 3 Burkina Faso 2010 4 Chad 2014 Ivory Coast Niger 2012 2 Senegal Togo 2013 1

PASEC vs SACMEQ PASEC Pupils are able to combine two pieces of explicit information from a document or can carry out simple inferences in a narrative or informative text. They can extract implicit information from written material while giving meaning to implicit connectors, anaphora or referents. Pupils locate explicit information in long texts and discontinuous documents. SACMEQ Interprets meaning (by matching words and phrases, completing a sentence, or matching adjacent words) in a short and simple text by reading on or reading back

Method Multiply completion rate by literacy/numeracy rate Simple Multiplication Overestimate completion for poorest 80% of sample – up to 10 p.p. Underestimate completion for richest 20% of samples – up to 10 p.p.

Method Simple Multiplication Example 50% Completed Gr6 Multiply completion rate by literacy/numeracy rate Example 50% Completed Gr6 60% Literate in school = 0.5 x 0.6 = 30% access to literacy Overestimate completion for poorest 80% of sample – up to 10 p.p. Underestimate completion for richest 20% of samples – up to 10 p.p.

Method Simple Multiplication Adjust for older DHS datasets Use a younger cohort & apply estimated dropout rates Overestimate completion for poorest 80% of sample – up to 10 p.p. Underestimate completion for richest 20% of samples – up to 10 p.p.

Method Simple Multiplication Adjust for older DHS datasets   PASEC DHS Benin 2005 2011/12 Burkina Faso 2006 2010 DRC 2013/14 Ivory Coast 2009 Senegal 2012/13 Togo Method Use a younger cohort & apply estimated dropout rates Simple Multiplication Adjust for older DHS datasets Overestimate completion for poorest 80% of sample – up to 10 p.p. Underestimate completion for richest 20% of samples – up to 10 p.p.

Method Simple Multiplication Adjust for older DHS datasets Adjust for late completion Overestimate completion for poorest 80% of sample – up to 10 p.p. Underestimate completion for richest 20% of samples – up to 10 p.p. Apply estimated dropout rates to those still enrolled in Gr.3-5

Method Simple Multiplication Adjust for older DHS datasets Adjust for late completion Adjust for underrepresentation of poor kids in schools Overestimate completion for poorest 80% of sample – up to 10 p.p. Underestimate completion for richest 20% of samples – up to 10 p.p. Apply estimated completion rates from DHS data to PASEC sample

Results (1): Access 72% 34%

Results (2): Quality 61% 8%

Results (3): National

Results (3): National

Results (3): National

Results (3): National

Results (3): National <1/3 children reading in all countries <1/10 children reading in Chad & Niger

Results (4): SES

Results (4): SES In Benin 2/5 literate children are wealthy From the wealthiest 20% of the population 1/10 literate children are poor From the poorest 40% of he population Rich children are 4x more likely to be literate than the poorest Results (4): SES Generally statistically significant Except Chad, Niger IvC poor/mid 9% 41% 40%

Results (4): SES In Niger 7/10 literate children are weathy From the richest 20% of the population 1% of the poorest 80% of children are literate Rich children are 20x more likely to be literate than the poorest children Results (4): SES 1% <1% 9% Generally statistically significant Except Chad, Niger IvC poor/mid

Results (6): SES & Gender (1) <20% poor learn to read: Less than 20% of the poorest boys and girls will learn to read or do math at a basic level (2) Poor boys > Poor girls: The poorest boys always have better outcomes than girls, except in Senegal, although differences are not statistically significant (3) Total illiteracy for poorest girls: In 5 of 7 countries, standard errors mean that it’s possible that zero females are learning to read at a basic level

Take home points Overall SES Gender (1) <1/3 literate or numerate in Benin, Burkina Faso, Ivory Coast, Senegal, Togo Overall (2) <1/10 literate or numerate in Chad, Niger (3) Vast socioeconomic inequalities In all countries SES Gender (4) Total illiteracy for poorest girls In 5 of 7 countries at lower confidence intervals

Thank you

Numeracy (national)

Numeracy by SES

Problem Statement Problem 1: Generally misleading results Using small non- representative sub-samples as a proxy for a cohort of 15 year olds or an entire schooling system can lead to misleading results. Problem 2: Inaccurate cross-country comparisons when using only PISA data, especially when countries have different levels of incomplete access or different proportions of delayed (& ineligible) students. Problem 3: Underestimating progress of the real educational improvements over time for countries that have improved access/attainment. Problem 4: Underestimating inequality When there is a sample selection process involved such that poorer students are more likely to be excluded from the sample, PISA will underestimate socioeconomic inequalities

Creating “access to numeracy” Get estimates of: Percentage of 15-16 year-olds that are eligible for the PISA sample (enrolled and in Gr7+) (from DHS) (I choose to use 2 age years to increase the sample size and decrease the standard errors) Percentage of PISA sample that achieved Level 2+ in Maths (from PISA) Multiply the two together, assuming that ineligible students would not have reached Level 2… eg 80% of 15-16 year-old girls reach Gr7+ 50% of girls in PISA reach Level 2+ Access-to-numeracy = 0.8*0.5 = 40% Correct for differential access-to-PISA-sample by SES…

Adjusting PISA SES 40-40-20 categories to correspond with DHS categories PISA Data

^ on quality AND access

Sensitivity Analysis & Limitations Ineligible = illiterate? Large literature attesting to the fact that delayed students and those who drop out are more likely to come from poorer families, rural areas, be female and be low- performers prior to dropout (Lambin, 1995; Filmer & Pritchett, 1999; UNESCO, 2005; Lewin, 2007; Lewis & Lockheed, 2006) Similar correlates in Turkey as shown in Köseleci’s (2015) review of the Turkish literature. In PISA 2012, those delayed 2-3 grades 362 (Reading) In PISA 2012, those ‘on-track’/modal-grade 500 (Reading) (i.e. 1.6 SD’s lower)

Conceptual overview Access & Quality – Post-SDG’s it is now a truism that we need to focus on both access and quality. But when Filmer, Hasan & Pritchett (2006) were arguing for it it wasn’t common cause. Coming from Sub-Saharan Africa this is a MAJOR issue. Comparing SACMEQ scores across countries for example (Spaull & Taylor, 2015) Centre & periphery – While PISA was designed with OECD countries in mind, there has been a rapid expansion into ‘partner’ (typically developing) countries. PISA assumptions built on the ‘centre’ not the periphery “PISA provides an assessment of the cumulative yield of education and learning at a point at which most young adults are still enrolled in initial education” - OECD & UNESCO-UIS (2003, p. 249). Of the 88 economies/regions that have participated in PISA only 36 are OECD (https://www.oecd.org/pisa/aboutpisa/pisa- participants.htm). Overlooking huge contextual differences - Problems that were initially overlooked as being overly technical or unimportant (sample selection & coverage) need to be revisited in light of changing PISA demographic. Much more nuanced analyses and interpretations are necessary for developing countries.

New PASEC Evaluation Tools: 2nd and 6th grade tests: They are built upon : Expected basic skills from students in 6th grade of primary school, essential for schooling continuation and recognized by the international community of researchers in pedagogy and education psychology, Official curricula implemented in the countries.

Tests objectives at 6th grade are to : Identify the proficiency level in writing comprehension and in the domain of mathematics (numeracy, measures, geometry) expected skills : decoding, extract information, deduce, analysing and giving opinion / to know, to implement, to reason in mathematics). Spot education systems efficiency factors in primary school. Enable students assets’ international comparison between all PASEC countries on one hand , but also with SACMEQ countries. (2 hours test in maths and writing comprehension, 4 booklets -60 items- in the two disciplins with a block of common items)

Tests objectives at 2nd grade: Detect students gaps , at the very beginning of the schooling, in order to adapt and re-orient teaching, curricula, pedagogy and practices, and more globally sector policy (at 2nd grade the comparison is possible only for countries sharing the same teaching language ). Assess directly the students level upon unique competences and not upon a set of competences (one item = one competence assessed) in oral comprehension, reading, writing comprehension, numeracy Individual test administration, orally to detect the real level in very targeted skills.