DIGITALPLAUSIBLE VALUES IN THE PISA REPORT AND THEIR TREATMENT IN EFFICIENCY ANALYSIS USING EDUCATIONAL DATA
1 Miguel Hernandez University of Elche (UMH), Centre of Operational Research (CIO) (SPAIN)
2 University of Extremadura, Department of Economics (SPAIN)
3 Ministry of Education and Culture, Autonomous Community of the Region of Murcia (SPAIN)
2 University of Extremadura, Department of Economics (SPAIN)
3 Ministry of Education and Culture, Autonomous Community of the Region of Murcia (SPAIN)
About this paper:
Appears in: INTED2020 Proceedings
Publication year: 2020
Page: 765 (abstract only)
ISBN: 978-84-09-17939-8
ISSN: 2340-1079
doi: 10.21125/inted.2020.0287
Publication year: 2020
Page: 765 (abstract only)
ISBN: 978-84-09-17939-8
ISSN: 2340-1079
doi: 10.21125/inted.2020.0287
Conference name: 14th International Technology, Education and Development Conference
Dates: 2-4 March, 2020
Location: Valencia, Spain
Dates: 2-4 March, 2020
Location: Valencia, Spain
Abstract:
In recent decades, there has been widespread literature dedicated to evaluating efficiency in the educational sector, both at the micro level, analyzing the performance of students or schools, and at the macro level, exploring the behavior of regions and even countries. These types of studies have been driven by the proliferation of international databases such as PISA, TIMSS or PIRLS, which constitute a very useful analysis tool for researchers, as they provide information on many factors of the educational process, in addition to several measures of the results in different competencies, usually considered as the outputs of the educational production function.These output measures are usually represented by different values drawn randomly from the distribution of results, the so-called plausible values, understood as a representation of the range of skills each student has (Mislevy et al., 1992, Wu, 2005). The usual practice in the efficiency analyses that use this source of information consists in the arbitrary use of one of these values or an average of all of them, although, supposedly, the most advisable thing would be to use all the available information of each student, that is, all plausible values.
In this paper, we intend to analyze to what extent the results of an efficiency analysis may be affected depending on whether or not all the information provided by the set of plausible values available in international databases is incorporated. To do this, we will carry out an analysis at the school level adopting a very simple production function based on the information contained in the PISA 2015 database, in which there are a total of 10 plausible values for each competency evaluated (reading, mathematics and sciences). Looking for robustness in our results, we will resort to several types of analyses to assess the suitability of the use of a single plausible value or the average of the ten plausible values. Specifically, the efficiency analysis will be carried out using the following methodologies: frontier techniques (parametric and non-parametric) and non-frontier techniques (using regression).