This report was originally published 26 September 2014.

Summary

Overview

The Family Occupation and Education Index (FOEI) is a school-level index of educational disadvantage related to socio-economic background. It has been selected as the basis of the equity loading for socio-economic background in the Department’s new Resource Allocation Model.

FOEI is constructed from parental education and occupation information collected from student enrolment forms and recorded in the Enrolment Registration Number system. FOEI uses a statistical regression model to produce a weighted combination of school-level parental education and occupation variables based on the extent to which each variable uniquely predicts average school performance.

This technical report details the construction of FOEI in 2013, and discusses the statistical techniques used to deal with two particular methodological issues – missing parental data and regression outliers – as recommended by statistical experts at the National Institute for Applied Statistics Research Australia through a commissioned review of the FOEI methodology.

Firstly, a multiple imputation methodology was employed to reduce bias due to missing parental data. Multiple imputation produces plausible values for the missing data based on the relationships between existing parental data and a series of other related variables. External validity evidence and a simulation study indicate that multiple imputation has reduced the bias due to missing data to an extent.

Secondly, robust regression analysis was used to reduce the influence of outliers on the regression model. Robust regression is an efficient technique for dealing with outliers by reducing their weight in the regression analysis, leading to more accurate and reliable regression parameters.

Since the introduction of FOEI as the basis of funding to schools, there has been a sizeable reduction in the level of missing data. Further work with schools is planned to continue to reduce the amount of missing data and improve the quality of parental data.

An ongoing program of work is also planned to review and further validate the FOEI methodology to ensure FOEI is the fairest and most robust measure possible of the relative socio-economic disadvantage of NSW government schools.