By Dr Claire Tyler, Prof Lindsey Macmillan & Padmini Iyer

 

In our latest blog, we’re delighted to be joined by Dr Padmini Iyer, Head of Research and Advocacy at social mobility charity Making The Leap to explore the similarities in the findings from our recent research on where inequalities arise in UK graduate recruitment processes.

Our recent UCL research about inequalities in access to professional jobs and Making The Leap’s recent research about what UK employers are doing to advance social mobility paint a consistent picture of employers prioritising outreach work over removing barriers for underrepresented groups during recruitment processes. Here we discuss our key findings and share some practical recommendations to support employers on their social mobility journey.

Still in first gear?

Access to good-quality employment is essential to achieving the financial security and job stability that supports social mobility. Making The Leap’s recent report ‘Still in First Gear?’ shows there is a good level of commitment to advancing social mobility among employers, in spite of challenges to diversity, equity and inclusion (DEI) in the UK in 2025. By investing in data, strategy and leadership, employers are increasingly building strong foundations for their social mobility work. However, Making The Leap’s findings indicate that employers’ social mobility work may still be in ‘first gear’, with most efforts focused on outreach and much less action to recruit individuals from less advantaged socio-economic backgrounds (LSEBs).

The research shows that 94% of employers’ social mobility work focuses on outreach, compared to only 53% on recruitment. What does this mean in practice? Employers’ outreach work aims to inform and inspire people from LSEBs to consider pathways they have not previously considered. Activities therefore include targeted work experience programmes, events to raise awareness about career pathways, and initiatives to develop young people’s employability skills. These are all valuable interventions – but they do not directly lead to young people from LSEBs getting jobs.

This where social mobility-focused recruitment activities come in: fair and inclusive recruitment processes, high-quality apprenticeships linked to permanent employment, and targeted employment programmes for LSEB individuals. However, with only 53% of employers taking action in these areas, there is clearly a disconnect between informing LSEB individuals about career opportunities, and actually hiring them. Moreover, as discussed below, many employers’ recruitment processes are more likely to reinforce socio-economic inequalities than to resolve them.

Adverse impact during recruitment process is a key barrier to equalising career opportunities

UCL’s recent research about ‘Inequalities in Access to Professional Occupations’ report, funded by the Nuffield Foundation, shows the potential consequence of a greater focus on outreach and a lesser focus on adverse impact during the recruitment process. The data shows that students from LSEBs and ethnic minorities are well represented in application pools for graduate roles, but are significantly less likely to receive job offers. For example, working class applicants are 32% less likely to receive a job offer than applicants from professional backgrounds. The penalty for Black applicants is 45% and for Asian applicants is 30%. Even when adjusting the findings to compare applicants who ‘look the same on paper’, including having similar educational credentials, large inequalities remain. This suggests that even if greater investment in outreach is effective for diversifying applicant pools, it is unlikely to solve persistent inequalities in who receives job offers. Efforts to ensure both a diverse applicant pool and hiring pool are required.

UCL findings (Figures 1 and 2) further show that while overall working-class inequalities occur relatively evenly across screening/testing and face to face stages of recruitment, the inequalities by ethnicity are much more pronounced at the screening/test stage, especially for applicants from Black, mixed and other ethnic backgrounds. For example, Black applicants face a 46% penalty at screening/testing and an 8% penalty at face to face stages compared to white applicants. A small proportion of these disadvantages can be explained by other demographic, educational and application characteristics. However, even when we compare candidates who ‘look the same on paper’, Black applicants still face an unexplained 37% penalty at screening/testing and a 5% penalty at face to face stages compared to otherwise similar white applicants. In comparison, working class applicants face an 18% penalty at screening/testing and a 14% penalty at face to face stages compared to applicants from a professional background – again, some of this can be explained by other demographic, educational and application characteristics, however an unexplained 9% penalty remains at each stage relative to otherwise similar applicants from a professional background.

Figure 1: Relative likelihood of passing (1) screening/testing stage and (2) face to face stages for the graduate programmes, conditional on observable differences across applicants, by parental occupation

Figure 2: Relative likelihood of passing (1) screening/testing stage and (2) face to face stages for the graduate programmes, conditional on observable differences across applicants, by ethnicity

 

How can employers move to ‘second gear’?

Making The Leap and UCL’s research identifies several areas of best practice to improve recruitment processes which will help employers move into second gear (and beyond!), including:

1. Collect applicants’ demographic data during recruitment, including socio-economic background data. This will allow you to identify trends and understand whether specific groups are being adversely impacted at each of your assessment and recruitment stages, and in turn, to target your efforts to ensure fairer recruitment processes.
2. Broaden the focus of diversity monitoring to include the screening and online testing stages of recruitment, in addition to any existing focus on face-to-face assessments. If testing is outsourced, employers can challenge external providers of recruitment assessments to provide data on success rates of candidates from under-represented groups for each type of test conducted.
3. Consider introducing contextualised recruitment. This means that those making hiring decisions can evaluate a candidate’s potential and achievements by considering the personal, social, and educational circumstances in which they were gained. This data might include demographic data alongside other information such as type of school and university attended.
4. Offer pre-assessment coaching and support. Anonymised, skills- or task-based assessments may be preferable to using minimum grade/qualification requirements during recruitment. However, as UCL’s research highlights, assessments can still disadvantage ethnic minority and LSEB candidates, so it is important to pilot and evaluate the diversity impact of any new assessments you introduce. You can then offer targeted, pre-assessment support to candidates from less advantaged backgrounds, as they may be less familiar and less confident with assessment tasks compared to their more advantaged peers. You can also work with universities’ and schools’ careers teams to share insights about which under-represented groups may need additional information or guidance to be prepared for online assessments.

What next, and how to get involved?

At Making The Leap, we are looking forward to two forthcoming publications. In January 2026, we will publish The Social Mobility List – a comprehensive overview of individuals, charities, and companies making significant strides to advance social mobility in the UK. And in spring 2026, we will publish a set of in-depth case studies exploring social mobility best practice among UK employers and educators. We are always open to partnerships to support employers’ social mobility journeys, from direct delivery with young people to research and advocacy work. Please do contact us (research@mtl.org.uk) for more details.

At UCL, we are launching a new project funded by the Nuffield Foundation aiming to shine a light on adverse impact during different stages of recruitment processes. We have built an ‘early careers data hub’ containing detailed anonymised data on over 2.5 million applicants (and growing!) to entry-level roles in over 20 large UK employers over several years – the largest research resource of its kind. Using this previously ‘untapped’ data, we will continue our investigation of the role of employers in the career outcomes of young people, specifically now focussing on the barriers faced by underrepresented groups in each type of assessment and selection method. This will provide an evidence base for employers, careers services and universities to drive change and target interventions and support more effectively. Please get in touch (c.tyler.14@ucl.ac.uk) if you would like to hear more about how to participate in our research collaborations. Projects are anonymous, free of charge and full of insights for participating employers.

 

By Claire Crawford

Last week the Department for Education published the first comprehensive statistics on the take-up of the new early education entitlements for 0-2-year-olds in working families. As of January 2025, around 42% of children in this age group were using at least some hours of government-funded early education, with take-up lowest amongst 9-11-month-olds (20%) and highest amongst 2-year-olds (55%). This may be good for the lives of these children and their families, but the working families entitlement is arguably not reaching those most in need of support. For the government to achieve its Plan for Change target of 75% of children being ‘school ready’ by age 5 by 2028, it may need to provide more support to children in non-working or very low income working families who are not eligible for the working families entitlement.

One programme that does benefit these relatively more disadvantaged children is an entitlement to 15 hours of government-funded early education per week for disadvantaged 2-year-olds – now rebranded the ‘Families Receiving Additional Support’ (FRAS) entitlement. But one of the concerns raised by us and other experts was that children eligible for the FRAS entitlement might miss out as a result of the expansion entitlements for working families. This could happen via a number of routes. Families could be confused by the strong messaging around the working families entitlement and mistakenly believe that they are not eligible for any government-funded early education. Higher demand for childcare places from families eligible for the working families entitlement could also have crowded out these children, whose families may be less likely to pay for additional hours.

The statistics published last week showed a substantial drop in the percentage of eligible 2-year-olds taking up the FRAS entitlement, falling from 75% in 2024 to 65% in 2025. But – and it is a big but – the Department urged caution in comparing the figures across these two years, as some children eligible for the FRAS entitlement are also eligible for the working families entitlement. Despite the statutory guidance provided to local authorities (LAs) that any children eligible for both should be recorded as being entitled for the FRAS entitlement rather than the working families entitlement, the Department had reason to believe that not all LAs had done this consistently.

What are the consequences of this uncertainty? It means we basically don’t know whether the fall in take-up means that some disadvantaged children have missed out. We don’t know how much action might be needed to correct for any fall. And if the working families entitlement was even partially responsible for this fall, then we might be especially concerned about will happen over the coming year, given that any negative effects may worsen once 0-2-year-olds in working families are entitled to double the number of hours as they are now (rising to 30 hours per week, from 15) in September.

So, with thanks to my colleague Oliver Cassagneau-Francis for assistance, we set out to see if we could work out whether there had been a genuine fall in take-up, or whether the fall was all down to misclassification. Using figures from the Institute for Fiscal Studies, around 18% of 2-year-olds eligible for the FRAS entitlement are also likely to be eligible for the working families entitlement (see below for all the gory details).

Using the estimated number of children eligible for the FRAS entitlement shared by DfE last week (just under 146,000), this would mean that just under 27,000 children were eligible for both entitlements this year. The actual fall in the number of children taking up the FRAS entitlement between 2024 and 2025 was just under 21,000 children. We can chalk up about a third of this to declining eligibility (see below). Of the remaining 14,000 children, how many are still using early education and how many are potentially missing out? Given that 27,000 children are potentially eligible for both entitlements, if just over half of these 27,000 children had been mis-recorded, that would be enough to explain the entire decline in take-up.

Is this plausible? It’s hard to know based on what the Department has said, but it’s certainly possible. So, is this the end of the story? Not quite. One of the great things DfE does when publishing these kinds of statistics is share the data underlying the headlines, including data at local authority level. We’ve used this data to good effect in previous research, funded by the Nuffield Foundation, to identify LAs who were ‘out-performing’ the characteristics of their local populations and achieving higher take-up rates than expected amongst disadvantaged 2-year-olds, to try to characterise the ingredients of their success. Here, we wanted to see how much the decline in take-up differed across LAs. Was it spread fairly evenly across the country, or was there variation? Was the extent of the decline correlated with any other characteristics of these LAs that might suggest a genuine fall rather than misclassification?

Perhaps an even more compelling part of the story is the correlation between the magnitude of these declines in take-up and LA characteristics. LAs that have experienced larger falls in take-up relative to what might have been expected tend to have higher female employment rates and lower unemployment rates, and to have a higher proportion of validated codes for the working families entitlement, perhaps indicating greater demand from children in these families and hence more competition for places. Moreover, these relationships hold over and above the size of the LA, so are not just driven by what’s happening in big cities.

While not conclusive, this suggests to us that the decline in take-up of the FRAS entitlement may not be all about misclassification. Yes, it is possible that misclassification could be higher in areas with these other characteristics, but is there a good reason to expect it to be so? A more plausible reason, based on our findings, could be that some disadvantaged children may be being squeezed out by higher demand for formal early education created by the working families entitlement. Given that take-up of the FRAS entitlement has never been higher than 75% and that there were large variations in take-up across LAs well before this latest round of data, we certainly wouldn’t want the government  to bank on misclassification as the sole reason for the drop in take-up of the FRAS entitlement, particularly as they have the Plan for Change target to think about. We would love to see action being taken now to remove barriers to take-up, including but not limited to those relating to the working families entitlement. This will help to ensure that all disadvantaged 2-year-olds have the opportunity to take advantage of government-funded early education and contribute to the success of the government’s Best Start in Life strategy.

All the gory methodological details

We started from an article from the Institute for Fiscal Studies which tried to estimate the extent of overlap in eligibility between the two entitlements. Their figures suggested that around 18% of 2-year-olds eligible for the FRAS entitlement might also be eligible for the working families entitlement.

(Their Figure 1 shows that 21% of the poorest quintile (fifth) of working families are eligible for both the working families and FRAS entitlements, in addition to 8% of the second poorest quintile and 2% of the middle-income quintile. Given that 85% of families with 2-year-olds are working, this means that just over 5% (((0.21*0.2)+(0.08*0.2)+(0.02*0.2))*0.85=0.0527) of the cohort are estimated to be eligible for both entitlements. Meanwhile, just under 11% (((0.49*0.2)+(0.11*0.2)+(0.03*0.2))*0.85=0.1071) of the cohort from working families are eligible for the FRAS entitlement only, as well as just under 13% (85% of the 15%=0.1275) of the cohort from non-working families. Altogether, this means that almost 30% of families of 2-year-olds (0.0527+0.1071+0.1275=.2873) are estimated to be eligible for the FRAS entitlement and 5% are estimated to be eligible for both the FRAS entitlement and the working families entitlement, meaning that around 18% (0.0527/0.2873 = 0.183) of those eligible for the FRAS entitlement are potentially eligible for the working families entitlement.)

Using the estimated number of children eligible for the FRAS entitlement provided by DfE last week (145,824) – and assuming that the degree of overlap, which was estimated on slightly older cohorts, is still approximately correct – this would mean that 26,686 (0.183*145,824) children were eligible for both entitlements this year.

The actual fall in the number of children taking up (registered for) the FRAS entitlement between 2024 and 2025 was 20,821 (115,852-95,031) children. Some of this decline would have happened anyway, both because the birth rate is declining, and hence the population of 2-year-olds is falling over time, and because the proportion of 2-year-olds who are eligible for the FRAS entitlement is also falling. The decline in eligibility between 2024 and 2025 – which reflects both these elements – was just under 6% ((154,957-145,824))/154,957=0.059). If take-up (registration) had fallen in line with this decline in eligibility, then we might have expected 6,835 (0.059*115,852, which is the number of 2-year-olds registered for the FRAS entitlement in 2024) fewer 2-year-olds to be using government funded early education in 2025 than 2024, even if nothing else had changed.

If we subtract 6,835 from the observed decline of 20,821 children, that leaves 13,986 children. The question is: how likely is it that these children were still using early education and had merely been incorrectly classified as taking up the working families entitlement rather than the FRAS entitlement, and how many were genuinely missing, i.e. not taking up government-funded early education at all?

By Professor Lindsey Macmillan, Dr Claire Tyler and Dr Catherine Dilnot

For the last decade we’ve been collecting and analysing recruitment data from over 20 of the UK largest graduate recruiters to highlight barriers to professional careers for underrepresented groups of young people. We have shown that young people from ethnic minority and/or working class backgrounds are well represented in the applicant pools for entry level professional careers but are less likely to be hired even when they look the same on paper. We have used our innovative ‘early careers recruitment data hub’ containing data on over 2.5million job applicants to document new evidence about why these barriers exist, highlighting the role of online testing, applying early, educational background and intersectional disadvantages.

But is this picture getting better or worse over time?

In this latest blog post we summarise new findings from our updated report, showing trends in inequalities in access to professional entry level careers over time. For our updated analysis we look over the period from 2022 to 2024 using a sample of over 350,281 applicants to graduate programmes, 73,525 applicants to internships, and 298,057 applicants to school leaver and apprenticeship programmes across a range of sectors (predominantly accountancy, law and public sector).

Worryingly, we find that the inequalities in access to professional careers that we highlighted in our original report have increased between 2023 and 2024, by socio-economic background (SEB), ethnicity and their intersections, across all entry routes.

Larger barriers for lower SEB applicants

For the graduate entry route, independent school applicants were 20% more likely to receive an offer relative to similar state school applicants in 2024, up from 7% in 2023 (Chart 1 – ‘full models’). By ‘similar’ we mean the applicants we compare are of the same gender and ethnicity, from the same UK region, attended the same type of university, studied a similar undergraduate subject, reported the same personal/professional networks and applied for the same role with the same employer in the same location.

This trend of increased inequalities is also consistent for working class applicants relative to professional background applicants. Comparing similar applicants, those from working class backgrounds were 21% less likely to get offers than similar professional background applicants in 2024, up 5 percentage points from the 16% penalty in 2023.

 

 

Larger barriers for ethnic minorities

We also found an increase in inequalities in offer rates by ethnicity, with Asian and Mixed and other ethnic groups being increasingly less likely to receive graduate offers in 2024 compared to White applicants, relative to 2023 (Chart 2 – ‘full models’).

By contrast, inequalities in graduate offer rates for Black applicants relative to White applicants reduced slightly from 38% in 2022 to 31% in 2024, leading to an equalisation of inequalities across minority ethnic groups. This suggests any gains from employers hiring more Black applicants has come at the expense of lower hiring rates for other ethnic minority groups rather than White applicants.

Chart 2: Offer rates to graduate programmes, conditional on observable differences across applicants over time, by ethnicity

 

Intersectionality increasingly matters

Our analysis highlighted a double disadvantage for state educated ethnic minority applicants to graduate entry level roles which has increased over time. We found a widening of the gap in offer rates for this group relative to White state educated applicants from 16% to 32% and a corresponding increase in the gap in offer rates for White independent school applicants from 9% to 21%. State educated ethnic minority applicants are therefore losing out on graduate offers relative to White independent school educated applicants with similar characteristics. (Chart 3 – ‘full models’).

Chart 3: Offer rates to graduate programmes, conditional on observable differences across applicants, by ethnicity and school type

 

School leavers, apprentices and interns

Inequalities in offer rates to internships and school leaver and apprenticeship programmes show a similar increase in inequalities by school type from 2023 to 2024. While state and independent school applicants had a similar chance of achieving an offer to an internship and school leaver or apprenticeship programme in 2022 and 2023, independent school applicants were 15% more likely to receive an offer to an internship programme in 2024 relative to state school applicants who were otherwise similar. They were also 10% more likely to receive an offer to the school leaver and apprenticeship programmes.

And while Black applicants were more likely to get an offer to an internship programme in 2023 relative to similar White applicants, this trend reversed in 2024 where they were 23% less likely to receive an offer relative to otherwise similar Black applicants. This coincided with some employers removing internship schemes with a specific focus on diversity.

Recommendations

Our findings suggest an important role for both employers and universities in tackling inequalities in access to professional careers (see the full list of ten recommendations in our report.)

This worsening picture for inequalities in access to professional careers we have outlined here has led to us include these specific recommendations for employers:

• Inequalities appear to have worsened over time as competition for places has increased to record high levels. It is therefore important to be particularly vigilant about diversity across the recruitment process when competition for places is higher. Employers should be more proactive in managing the adverse impact for underrepresented groups who are more likely to lose out as competition increase.

• Consider keeping existing / introducing new schemes that have specific focus on hiring diverse candidates. The existence of these schemes for internship positions saw increased diversity while the removal of them coincided with a significant decline in diversity.
• Consider how to contribute to building a robust evidence base of ‘what works’ to support your colleagues, your sector and policy makers to refocus diversity initiatives and policies more efficiently.

Notes:

‘Raw’ models in this blog contain employer fixed effects, gender, SEB, and ethnicity. ‘Full’ models adds controls for region of origin, visa status, university category, subject studied, networks, region of office, and job role applied to. Points represent % estimates while lines represent 95% confidence intervals.

By Professor John Jerrim

All children have the right to education. They also have the right to feel safe at school.

Unfortunately, this is no longer the case in England – particularly amongst teenage girls.

My new academic working paper out today uses data from the TIMSS study to track the emotional engagement of young people before and after the pandemic. It also investigates changes in their engagement between Year 5 and Year 9.

The results are stark.

Figure 1 illustrates how the percentage of 14-year-olds (Year 9s) in England strongly agreeing that they feel safe at school changed between 2011 and 2023. Before the pandemic – between 2011 and 2019 – around 40% of teenagers strongly agreed that they felt safe at school. But after the pandemic – in 2023 – this has declined dramatically to only 26%.

 

Figure 1. The percentage of Year 9 pupils in England strongly agreeing that they feel very safe at school

 

One of the best things about international comparative data such as TIMSS is that we are able to put this decline in England into international context. These results are presented in Figure 2. The horizontal axis presents the percent of teenagers strongly agreeing they felt safe at school in 2019, while the analogous figures for 2023 are presented on the vertical axis. The diagonal 45-degree line is where these two values are equal.

Interestingly, most countries fall below the diagonal line. This indicates that – like in England – there has been a fall in teenagers feeling safe at school since the pandemic across countries. Yet the fall in England is somewhat more pronounced than in most other countries with data available.

Figure 2. The decline in 14-year-olds feelings of safety at school between 2019 and 2023. England compared to other countries.

 

Critically, in England, we find this fall in feelings of safety at school to be most pronounced amongst teenage girls. For instance, between 2019 and 2023 the fall in the percentage of 14-year-old girls strongly agreeing that they felt safe in school in England fell by 22 percentage points. This is a much larger decline than the 13-percentage point decline observed in the average country, as well as the 10-percentage point decline observed for English boys.

Within our analysis, we also explore changes in children’s feelings of safety at school between primary (Year 5) and mid-secondary (Year 9) school. Table 1 illustrates the difference in the percent of pupils that strongly agree.

The results show how – in England – there is a 31-percentage point drop in pupil’s strongly agreeing that they feel safe at school between Year 5 and Year 9. This is amongst the biggest decline out of any country with data available. This includes the average across OECD countries (23 percentage point decline) and across all countries with data available (20 percentage point decline).

Table 1. Percentage point decline in children strongly agreeing they feel safe at school between Year 5 and Year 9

Once again, this seems to be a particular problem affecting English girls. For instance, amongst Year 5s, girls in England report feeling slightly safer at school than boys, with results very similar to the average across countries (girls being three percentage points more likely to feel safe at school than boys). But – by the time pupil’s reach Year 9 – boys in England are 10 percentage points more likely to feel very safe at school than girls (compared to a four-percentage point difference on average across countries).

Together, these results paint a stark picture. England now has a real problem with the number of young people that don’t feel safe at school. This has clearly worsened since the pandemic and is now a major challenge likely impacting the wellbeing and educational outcomes of teenage girls.

This is something that requires immediate attention. A minimum expectation most parents have is for their child to feel safe while they are in the school grounds. Unfortunately, for many, this is no longer the case.

By Oliver Cassagneau-Francis

This blog was originally published on ADR UK (Administrative Data Research UK)’s website [link to original post].

In this blog post, CEPEO research fellow Oliver Cassagneau-Francis describes how he and the project team (CEPEO director Lindsey Macmillan, deputy director Gill Wyness and affiliate Richard Murphy) will use the Grading and Admissions Data for England dataset to study differences in predicted grades and compare the resulting outcomes for different groups of students. This project is funded through an ADR UK Fellowship.

Students from more advantaged backgrounds are three times more likely to go to university than their peers from less advantaged backgrounds, and they are also more likely to go to highly selective courses. These courses often lead to better careers­­. Recent work has shown that for students with the same level of academic attainment, the quality of the course they enrol into varies across socio-economic groups. In particular, students from more advantaged backgrounds enrol into more selective university courses than students from less advantaged backgrounds who achieve the same grades at school. This is true across the spectrum of student attainment.

A likely driver of these differences is the important role of teacher-predicted grades in UK university admissions. Students generally apply to university and accept their places before sitting their exams, relying on predictions of their grades made by their teachers (henceforth “predicted grades” or “predictions”). These are generally inaccurate.

 

Predicted grades became more complex during the pandemic

Unpacking UCAS predicted grades is a difficult task. Teachers are asked to be optimistic in their predictions, and so it is unclear whether achieved grades are the correct comparison for predicted grades. However, during the Covid-19 pandemic in 2020, exams were cancelled and teachers, having already given the predicted grades needed for university applications, were asked to now give their students grades that would become their actual A-level results. These were called centre-assessment grades. Note that with these grades, teachers were asked to provide a realistic judgement of the grade each student would have been most likely to get if they had taken their exam(s) in a a given subject and completed any non-exam assessment; so there is not the element of optimism that is in UCAS predictions.

Therefore, we have two groups of students with different information on each: a group who have predicted grades and actual grades (the pre 2020 cohort); and a group for whom we have predicted grades and centre-assessment grades (the 2020 cohort). By comparing UCAS predictions with centre-assessment grades and with actual grades we can learn about how teachers make predictions.

In addition, in 2020 teachers were asked to rank students within the centre-assessment grades, meaning it’s possible to see which students just achieved a given grade and which ones just missed out.

How administrative data can provide new insights

For this project, we will use this unique information to study how teacher predictions differ across different social groups (e.g. by socio-economic status, gender, or ethnicity). We will also study the impact of receiving different predictions – teacher-predicted grades for university applications, and centre assessed grades – on outcomes, such as which university and course students went on to enrol in.

To do this, we will use the Grading and Admissions Data for England (GRADE) dataset. This contains de-identified data on students from:

• the Office of Qualifications and Examinations Regulation (Ofqual)
• the Department for Education (DfE)
• the Universities and Colleges Admissions Service (UCAS).

 

The data from these different sources has been linked together, de-identified and made available to accredited researchers. The dataset is very comprehensive, covering nearly all students who were in school in England and took their GCSEs or A-levels in 2018, 2019 and the 2020 cohort whose exams were cancelled in summer 2020. This project will focus on the A-level students from these cohorts.

Measuring differences in teacher judgements across groups

In this project, we will study carefully those students placed either just above or just below a grade boundary in 2020, using centre-assessment grades and teacher rankings. If these look different – for example, if women (or students from ethnic minority or lower socioeconomic backgrounds) are more often found at the top rank of a B boundary, than at the bottom of an A boundary then this suggests bias against women (or students from ethnic minority or lower socioeconomic backgrounds) – this will suggest that teachers might be predicting more or less generous grades for students from different groups. We will expand this analysis to look at specific subjects (e.g. Maths and English) and specific grade boundaries (e.g. A* / A). We can also perform a similar exercise using exam grades and marks (pre-2020), allowing us to compare the distributions of students around grade boundaries that are determined by exam versus those due to teacher judgements. Ofqual carried out their own analysis of the centre-assessment grades, finding limited evidence that student characteristics influenced grades, and they release equalities analyses for each round of exams.

In the second part of the project, we will again look closely at students just on either side of a grade boundary and compare their university enrolments and other outcomes. These students are ranked very closely by their teachers but look quite different to universities, as they received different (centre-assessment grades. It will also be interesting to compare outcomes of students who were ranked very closely by teachers, who were given different predictions pre-Covid. By comparing their outcomes, we will be able to isolate the impact of receiving an A over a B (both at the predicted grade and at the actual grade level), for example, on students’ university pathways.

Teachers are one of the main drivers of student success at GCSEs and A levels, success which then goes on to determine future outcomes. Understanding whether there are discrepancies in teachers’ judgements in favour of certain groups over others, resulting in differences in school attainment and university choices, will help us to understand the implications for social mobility and equity.