A comparisson of teacher and test based assessment for Spanish primary and secondary education Óscar D. Marcenaro-Gutiérrez* Anna Vignoles** *Departamento de Estadística y Econometría (UMA) y Fundación Pública Centro de Estudios Andaluces (FCEA) ** University of Cambridge XI Conference FES Madrid, 10-12 th July 2013 G13 1
A COMPARISSON OF TEACHER AND TEST BASED ASSESSMENT FOR SPANISH PRIMARY AND SECONDARY EDUCATION póster de Óscar D. Marcenaro-Gutiérrez y Anna Vignoles Congreso FES grupo Sociología de la educación Madrid 2013
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A comparisson of teacher and test based assessment for
Spanish primary and secondary education
Óscar D. Marcenaro-Gutiérrez*Anna Vignoles**
*Departamento de Estadística y Econometría (UMA) y Fundación Pública Centro de Estudios Andaluces (FCEA)
** University of Cambridge
XI Conference FESMadrid, 10-12th July 2013
G13
1
Outline:
Motivation
Aim
Data
Methodology
FES Madrid
Conclusions
3
FES Madrid
Motivation:
1. Needless to say: importance of education as an engine of socio-economic change.
4
The profitability of investment in education is higher
that investment in capital goods (Psacharopoulos, 1985)
FES Madrid
Importance of training as an engine of socio- economic change: Budgetary effort.
Motivation:
FES Madrid
Motivation:
Central to Economics of Education: Do we have precise measures of students’ outcomes? To what extent does the nature of the assessment criteria produce social class inequalities in academic outcomes?
FES Madrid
Motivation:
2. Teacher assessment matters, because:- Affects how students perceive their own academic ability (self-esteem, etc.).- It is used to determine whether a child can proceed to the next grade.
1. Virtual lack of datasets allowing the comparison “Teachers’ actual assesment-external test assesment”; novelty of the paper.
7
It is crucial to determine whether such assessment is a good guide to pupils’
actual level of achievement:PISA? Actual exams?....
FES Madrid
Aim:
Are assesment tests (test-based assessment –TestA-) and official marks (teachers-based assessment –TeachA-) telling different things? We explore this empirically.
Do these differences vary systematically with pupil characteristics?
8
gender
ethnicity
socio-economic background
FES Madrid
Aim:
Are assesment tests (test-based assessment –TestA-) and official marks (teachers-based assessment –TeachA-) telling different things? We explore this empirically.
Do these differences vary systematically with pupil characteristics?
Differences in achievement levels of pupils in public (100%) schools and semi-private schools.
Do this potential gap discourage further study, in general or in any particular subject area (such as science, arts, etc.), after completing compulsory education? (i.e. future success)
9
FES Madrid
10
Literature:
Gender or Ethnicity of the teacher might interact with that of the student to produce systematic differences in pupil achievement (Dee, 2005 & 2007): teachers over grade students of the same gender to themselves.Comparing teacher assessments of pupil achievement and test scores: Reeves et al (2001) , Gibbons and Chevalier (2008) –UK-, TestA-TeachA consistent.
Gutierrez and Adserá (2012) –as opposed to Calero and Waisgrais (2008)- public school students higher grades: might be explained by differences in grading practices between public and private school teachers.
FES Madrid
Data:
PISA: 15-years-old students -4th year ESO-(OECD countries).
Dependent variables: - Math Test scores (normalised).- Reading Test scores (normalised).
Independent variables:- Student characteristics- Family background 26 variables- School characteristics
PISA-2009: Spain: 25.887 students within 889 schools.Andalusia: 1700 students within 51 schools
11
Data (1st stage):
FES Madrid
Linked databases (novelty):
- Andalusian Social Survey: Education and Homes in Andalusia (Encuesta social: Educación y Hogares en Andalucía, ESOC10); includes diagnostic assesment tests.
- Administrative marking records for those in the survey sample.
Data (2nd stage):
FES Madrid
ASS10:- Students born in 1994 + students born in
1998.
15 years 11 years
- Pseudo-panel: past and future outcomes.
- Very important: Includes interviews aimed at parents and children.
13
Crossing information
(consistency)
Data (2nd stage):
FES Madrid
ASS10:- ADDITIONAL added value: link ASS10 with
information from an administrative information system.
14
Information entered by the educators from public and semi-
private schools
• Repeaters.• Disabled (or with special educational requirements).• Private schools (only 2.5%) –data non avaialble-.
Data (2nd stage):
Excluding
FES Madrid
Andalusia 2009, huge differences with
the rest of Spain?
16
• Differences between natives and immigrants men more substantial in Andalusia as compared to the rest of Spain
• Public school Vs Private school: same differences as for the rest of Spain.
• Mother educational level: variables even more relevant.
• Once again…. IMPORTANCE OF HUMAN CAPITAL ACCUMULATION OF THE PROGENITOR AS A WAY OF ACADEMIC SUCCESS FOR CHILDREN, PARTICULARLY FOR WOMEN.
A) School failure:
B) Performance determinants:
FES Madrid
17
¿Consistency? (PISA-ASS)0
.1.2
.3.4
Den
sity
-4 -2 0 2 4Normalized scores
PISA reading TestA
ASS10 reading TestA
Reading
0.1
.2.3
.4D
ensi
ty
-4 -2 0 2 4Normalized scores
PISA math TestA
ASS10 math TestA
MathSubstantial overlap
FES Madrid
High grade of consistency between PISA-2009 y ASS10-SEN conditional effects:
Semi-private school (=1) 0.41 0.19 0.50*** 0.41**Constant -0.71*** -1.31*** 0.97*** 0.90***Number of observations 1164
Χ2 90.39***
“pupils who are under estimated by teachers in terms of their
achievement are more likely to drop out”
Choice at 16:
FES Madrid
Conclusions
High relevance of parents human capital (particularly mothers); this is deeper in Andalusia, relevant to economic growth and labour market opportunities of women!!! (higher performance and lower failure)
29
PISA09 and ASS10 similar results, TestA consistent.
FES Madrid
Conclusions
Systematic differences between teacher assessments and actual test scores.
Girls’ achievement is over estimated by teacher assessments relative to test scores (preconceived ideas?????).
Teachers under estimating achievement of students at Semi-private schools (measuing relative acievements –peer effect-?????).
Pupils who are under estimated by teachers in terms of their achievement are more likely to drop out→ TestA-TeachA has a longer term impact on pupil outcomes.
30
FES Madrid
Thank you!
31
FES Madrid
A comparisson of teacher and test based assessment for
Spanish primary and secondary education
Óscar D. Marcenaro-Gutiérrez*Anna Vignoles**
*Departamento de Estadística y Econometría (UMA) y Fundación Pública Centro de Estudios Andaluces (FCEA)