MOBILITY IN PUPILS' COGNITIVE ATTAINMENT DURING SCHOOL LIFE

(2022) - Leon Feinstein

Journal: OXFORD REVIEW OF ECONOMIC POLICY
Link::
DOI::
Links::
Tags:: #paper #NCDS #Mobility #CognitiveAbility
Cite Key:: [@Feinstein2022]

Abstract

This paper considers the extent of mobility in pupils' attainment relative to peers as they move through school. Considerable shifts in position are demonstrated using data from the 1958 and 1970 UK birth cohorts and from the National Pupil Database for 2002. These shifts in attainment in primary and second ary school are shown to relate strongly to social class, demonstrating that the large social-class attainment gap in the UK is not a one-off effect prior to school entry but a compounding effect throughout school life. These changes in relative attainment during school are also shown to relate strongly to adult economic outcomes. These results suggest that although there are good arguments in support of an increase in pre school expenditures, equality of opportunity also requires enhanced investments for the worst-off through out schoo

Notes

“the extent of mobility in pupils' attainment relative to peers as they move through school. Considerable shifts in position are demonstrated using data from the 1958 and 1970 UK birth cohorts and from the National Pupil Database” (Feinstein, 2022, p. 213)

“the significance of preschool experience is undeniable, it is during the school years that the initial foundations of educational achievement are real ized, or fail to be realized, in concrete educational performance, and when the gap between disadvan taged children's family life and the educational process emerge and crystallize.” (Feinstein, 2022, p. 213)

“What Kagan (1998) describes as 'the lure of infant determinism' thus tends to mask the continuing importance of the interactions between children and their social and physical environments that extend throughout childhood and beyond (Schaffer, 2002). The early years matter tremendously, but the critical question remains for education policy: in what form and in what quantity should resources be allocated across childhood” (Feinstein, 2022, p. 214)

“For the 1958 cohort, tests of reading and maths were sat at ages 7,11, and 16. The tests are treated here as indicators of general cognitive development and the reading and maths scores are combined. To create a score at each age, both scores at each age are first standardized to create a common variance. The mean of the two age-specific tests is then taken and the final variable constructed is the rank of the child in the mean of the two tests. This provides an effective age-standardization and the method ena bles consideration of changes in each child's posi tion in the distribution of combined maths and read ing scores relative to that of other children over the same period (Feinstein, 20036; Feinstein andBynner, forthcoming).” (Feinstein, 2022, p. 215) important level of operationalisation

“For the 1970 cohort, the tests were undertaken at ages 5 and 10, which means that the specific transitions made will be slightly different to those between ages 7 and 11. In addition, although the age 10 score is the sum of reading and maths scores, the cognitive development score at age 5 is taken not from reading and maths scores but from copying tests and vocabulary scores that are more appropri ate for age 5 children (see Feinstein (20036) for more detail on the scores and validity testing). These differences reduce comparability slightly. Sadly, it is not possible to assess the transitions between 10 and 16 for the 1970 cohort because of a teacher's strike during the crucial phase of data collection” (Feinstein, 2022, p. 215) same as above

“strength of the 1958 cohort data is that informa tion is available on relative shifts in both primary and secondary phases of education. The relative impor tance of shifts over these periods is an important research question, and some comparisons between implications of shifts between 7 and 11 and between 11 and 16 are drawn, but the main aim is to assess the extent to which these shifts alter the prediction that an observer would make on the basis of the age 7 position. This can be compared to the results from the 1970 cohort, which indicate the extent to which predictions are altered, given information between the ages of 5 and 10.” (Feinstein, 2022, p. 217)

“The first conclusion I draw from these results is that there is considerable mobility in apparent academic abili” (Feinstein, 2022, p. 226)

“For good reason, most econometric or statistical analyses of cognitive attainment as children mature focus on correlations or partial correlations in re gression analyses in which prior attainment is en tered as a covariate (see, for example, Haveman and Wolfe, 1995; Hanushek et” (Feinstein, 2022, p. 226)

“Thus, for example, for the 1958 cohort we find that cognitive capabilities are far from being set in stone at age” (Feinstein, 2022, p. 226)

“The second finding is that there is a strong relation ship of social class and the large relative changes as children mature” (Feinstein, 2022, p. 227)