How Trajectories of Disadvantage Help Explain School Attainment

(2019) - Stephen Gorard, Nadia Siddiqui

Journal: SAGE Open
Link:: http://journals.sagepub.com/doi/10.1177/2158244018825171
DOI:: 10.1177/2158244018825171
Links::
Tags:: #paper #Transition #school-to-work #SocialClass #Education
Cite Key:: [@gorardHowTrajectoriesDisadvantage2019]

Abstract

This article illustrates the links between different ways of assessing disadvantage at school and subsequent qualification outcomes at age 16 in England. Our previous work has compared variables that represent current or recent snapshots of disadvantage (such as eligibility for free school meals [FSM]) with long-term summary variables and found the latter to improve measures of both social segregation between schools and explanations of raw-score differences in attainment. This new work takes an even more detailed longitudinal approach, modeling the course of one age cohort of 550,000 pupils from the National Pupil Database through their entire schooling to the age of 16 in 29 distinct analytical steps, using “effect” sizes, correlations, and a regression model. The steps represent stages such as what is known about each pupil when they were born, who they attended school with at age 10, and where they lived at age 14. The model also includes variables representing where data are missing for any pupil in any year. Using capped Key Stage 4 points as an outcome measure, these stages can predict the outcome with R = .90. This is considerably higher than for models using either snapshots or summaries of disadvantage. Key predictors are poverty and special educational needs at age 5, and throughout schooling, coupled with prior attainment at ages 6, 10, and 13. With predictors fed into the model in life order, there is little evidence of differential progress for different language and ethnic minority groups and no evidence of regional differences or a type of school effect. The article concludes with the implications of these results for assessing disadvantage when considering school contexts and for policy makers. Given the small but apparently consistent negative school composition “effects” in every year, one clear implication is that school intakes should be as mixed as possible both socially and academically.

Notes

“Schools can make these inequalities worse through their method of allocating places to pupils. A key issue, therefore, is the clustering of poverty within particular schools—the extent to which poor pupils go to schools with others like them” (Gorard and Siddiqui, 2019, p. 2)

“Evidence from around the world shows that such segregation is unnecessary and harmful to students (Gorard, 2018). It is associated with greater unfairness in practice, worse opportunities for the most disadvantaged, lowered aspirations, and lower participation rates in later education (Schmidt, Burroughs, Zoido, & Houang, 2015).” (Gorard and Siddiqui, 2019, p. 2)

“This composition effect might suggest that pupils in very disadvantaged schools do much worse than expected. If accepted, this has implications for policies on allocating school places. Composition effects appear stronger where the sorting of pupils into different tracks by ability is stronger (Danhier, 2018).” (Gorard and Siddiqui, 2019, p. 3)

“There is a reported danger of under- or overestimating school compositional effects due to measurement error, making them appear as phantom results (Televantou et al., 2015). Pupil-level measurement error can produce spurious schoollevel compositional effects.” (Gorard and Siddiqui, 2019, p. 3)

“Our prior work has confirmed that considering long-term indicators of disadvantage, rather than simply flags for current or recent status, leads to better understanding of both disadvantage and its impact on outcomes.” (Gorard and Siddiqui, 2019, p. 11)

“In the full model, the coefficients for ethnic group (major) are small, and removing these flag variables makes very little difference to R in each year. The values for each ethnic minority are always small and almost always positive.” (Gorard and Siddiqui, 2019, p. 11)

“Having a first language other than English is similarly no long-term barrier to progress” (Gorard and Siddiqui, 2019, p. 11)

“There is no substantial school type effect here. Schools are largely defined by who attends them. Once that is accounted for, there is no great difference between the outcomes of any of them (coefficients of .002-.008).” (Gorard and Siddiqui, 2019, p. 12)