Attendance by Early Years Foundation Stage Attainment
1 Introduction
This report details the findings of analysis of how school attendance varies by the ‘school readiness’ scores which are reported by teachers at Early Years Foundation Stage (EYFS), and how that relationship varies with other characteristics. This is a replication of a research study (Wood et al, 2024). The study took data on 62,000 children in Bradford to establish a link between school readiness (as assessed in the EYFS scores) and persistent school absence later in life.
This article gives an overview of the Bradford study: https://www.nurseryworld.co.uk/news/article/link-between-meeting-early-learning-goals-in-eyfs-and-later-school-absence-research
And the full research paper can be found in the PDF https://royalsocietypublishing.org/doi/pdf/10.1098/rsos.240272
The Bradford study found a clear link between EYFS attainment levels and school attendance, with children who do not meet the expected level being more than twice as likely to be persistently absent.
This document shows the findings of a similar analysis using local Sheffield data. We took around 44,000 results for EYFS in Sheffield between 2013 and 2019, and looked at their school attendance later in life. COVID years are generally removed or ignored. Unlike the Bradford study, here we have repeated the analysis for a set of other factors from the attendance data - demographics, geography, special educational needs etc. And just like the Bradford study, we found that the EYFS attainment levels are highly predictive of a child’s future attendance levels, and clear intersectionalities with other factors known to be associated with lower attendance.
2 Data processing
The data is loaded from the attendance data model, which contains pre-aggregated & cleansed attendance data, as well as other descriptive characteristics. EYFS scores are loaded from the SCC OSCAR database, and the attendance data is joined on for each national curriculum year (NCY). Finally the data is aggregated, with COVID years filtered out, to calculate average attendance rates for each NCY, first by the EYFS ‘good’ indicator, and then by other characteristics of interest.
2.1 Analysis
Across the years that this data covers, attainment levels improved, levelling off with about 30% of children meeting the expected level.
Plotting the overall attendance rates for each year, by the EYFS good score shows a clear gap in attendance between children meeting expected level and those not. This gap is present in all years and grows during secondary school.
Next we’ll explore how this difference varies by different characteristics, in each case recreating the plot above.
2.1.1 Gender
2.1.2 Ethnicity
This table provides the numbers for the following plot
| y1 | y2 | y3 | y4 | y5 | y6 | y7 | y8 | y9 | y10 | y11 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Any Other Ethnic Group | |||||||||||
| 0 | 557 | 437 | 423 | 427 | 411 | 382 | 287 | 293 | 308 | 244 | 162 |
| 1 | 672 | 509 | 497 | 521 | 544 | 541 | 408 | 361 | 326 | 238 | 152 |
| Asian or Asian British | |||||||||||
| 0 | 1768 | 1475 | 1399 | 1356 | 1337 | 1250 | 865 | 921 | 1013 | 799 | 580 |
| 1 | 2666 | 2103 | 2061 | 2014 | 2031 | 2064 | 1514 | 1453 | 1338 | 951 | 598 |
| Black or Black British | |||||||||||
| 0 | 785 | 661 | 647 | 590 | 565 | 558 | 420 | 435 | 468 | 347 | 240 |
| 1 | 1305 | 1041 | 1011 | 953 | 963 | 1023 | 822 | 761 | 655 | 436 | 254 |
| Chinese | |||||||||||
| 0 | 88 | 72 | 63 | 52 | 45 | 40 | 34 | 41 | 40 | 34 | 19 |
| 1 | 175 | 132 | 131 | 128 | 121 | 118 | 78 | 74 | 67 | 51 | 31 |
| Mixed / Dual Background | |||||||||||
| 0 | 1141 | 948 | 946 | 909 | 889 | 865 | 624 | 645 | 722 | 553 | 395 |
| 1 | 1995 | 1587 | 1514 | 1513 | 1555 | 1578 | 1284 | 1220 | 1048 | 750 | 443 |
| White | |||||||||||
| 0 | 8825 | 7407 | 7144 | 6990 | 6742 | 6268 | 4534 | 4840 | 5342 | 4208 | 2986 |
| 1 | 16951 | 13724 | 13321 | 13172 | 13280 | 13472 | 10772 | 10322 | 9530 | 6809 | 4294 |
| not known | |||||||||||
| 0 | 308 | 257 | 247 | 249 | 223 | 182 | 114 | 142 | 212 | 182 | 144 |
| 1 | 465 | 338 | 324 | 366 | 332 | 314 | 250 | 214 | 262 | 207 | 141 |
2.1.3 SEN level
Children with an EHCP plan are generally unlikely to meet the expected level at EYFS, and so the numbers do not allow a useful comparison here, and they are removed from this chart.
Children with no special educational needs make up the majority, and so closely follow the overall averages (shown here as dotted lines).
Children requiring SEN support attend worse than those with no SEN regardless of their EYFS score - those who do meet the expected standard still attend worse than average throughout their school career.
2.1.4 SEN primary specific need
2.1.5 Geography and deprivation
As in other areas of analysis within the Inclusion & Attendance project, we’ll use the 2019 Indices of Multiple Deprivation (IMD) scores of the child’s ward of residence. Those scores are divided into four quartiles, with 1 represents the most affluent 25% of the city, and quartile 4 the most deprived 25%.
There is a clear relationship between deprivation score and EYFS attainment:
To look at attendance, here the middle two are removed, so the pairs of lines below represent the most deprived 25% and least deprived 25% of the city, each split by the EYFS good measure.
Two things are worth noting in this chart: 1. The gap between those meeting expected EYFS level and not is greater in poorer areas of the city, and grows faster through secondary school. 2. Children in the poorest areas who do not meet the expected level shows the steepest drop off in attendance between Y6 and Y7, as they transition to secondary school. 2. The blue line (children not meeting the expected level in the least deprived wards) is higher than the green line (children who do meet the expected level in the most deprived wards). So deprivation makes more of a difference to attendance than the EYFS measure.
Plot by ward
2.2 School
2.3 Annual cohort analysis
