Project CEMAPRE internal
| Title | Statistical modelling in Education Research: Multilevel models applied to large and complex data - year four |
| Participants | Maria Teresa G Alves, Amélia Bastos, Alcino Couto, Maria Eugénia Ferrão (Principal Investigator), Paula Prata |
| Summary | The fourth year of the Statistical Modeling in Education Research project is focused on two main topics: 1) Educational inequalities; 2) Patterns of missing data in educational research. Most of the research conducted on the topic of educational inequalities has been based on cross-sectional data. Regarding the first topic, the ongoing studies provide innovative evidence based on longitudinal data and on panel cross-sectional data. Moreover, several scholars have warned on the use of big data for research purposes, which may be simultaneously an opportunity or a threat. If the research purposes include inferences from big data statistical modelling and the big data does not cover the entire target population or there is a selective mechanism that produces missing data that are not completely at random, the research itself may be compromised. In addition, the use of quantitative methods in the area of education has been strongly conditioned either by the availability of data or by their quality. Thus, the second topic concerns what should the education researcher do with incomplete data, how to identify the pattern of missingness, and the implications for educational inequity research of naïve assumptions on such patterns. Different statistical methods are used. Identifiable big data sets refer to Brazil (Prova Brasil and Census of Education) and Portugal. |