A level England - Analysis of 2025 DfE national dataset - February 2026

A level England - Analysis of 2025 DfE national dataset - February 2026

2025 A Level Results in England 


This article presents an overview of the findings from the Alps review of the DfE national dataset from 2025.

It provides a comparison with
  1. the 2019 DfE national dataset, currently on toggle 1 of the benchmark selector 
  2. the Alps customer benchmarks for England produced in August 2025, currently on toggle 2 of the benchmark selector
It is widely expected that the outcomes from the 2025 examinations for KS5 in England will now be the 'baseline' for future grading standards. Therefore the hope is that the 2025 DfE national value-added picture represents an opportunity to move to a new standard benchmark, moving away from the 2019 DfE national benchmark that we have been using in Connect and Summit since before the pandemic. 

The article concludes with an overview of the recommendations for benchmark selection for retrospective analysis of your 2025 results for A level, and for monitoring your current Year 13 and 12 cohorts. 

Watch video summary of article:

 

National Attainment for A level in 2025

A level Attainment Profile - England
Results for attainment showing gaps between 2019 and 2025, and between 2024 and 2025. 


  1. The key points: 
    Compared with 2019,
    1. the percentage A* is 1.7% higher
    2. the percentage A*-A is 3.0% higher
    3. the percentage A*-B is 3.9% higher
    Compared with 2024,
    1. the percentage A* is 0.1% higher
    2. the percentage A*-A is 0.6% higher
    3. the percentage A*-B is 1.4% higher

Value-added context for A level 2025

The 2025 results see the return towards a pre-pandemic average prior attainment baseline for Key Stage 5 students, although it remains slightly higher than in 2019.
Using the data submitted to Alps since results day, we have carried out a thorough analysis of the dataset and present our findings below. 

Prior attainment baseline


The table above shows an average prior attainment GCSE score for Alps customers for this summer's cohort of 6.17 (national 6.18), which is lower than the 2024 cohort, but still higher than the 2019 value of 6.11. This means that if we were to represent the value-added profile diagrammatically over the past few years, we would see the following pattern.




The starting point has fluctuated over the past few years, but the outcome end point has also fluctuated. The arrow shows the relative size of the value-added progress journey.

At the beginning of the 2024/25 academic year, we advised that the 2019 national dataset represented the most beneficial benchmark on which to base your target setting and tracking priorities. The value-added progress journey in 2025 was likely to be more similar to that of the value-added picture in 2019 than that of 2024.

On Results Day, 14 August 2025, the default benchmark for KS5 was therefore the DfE 2019 dataset. At the end of August we published the Alps Customer benchmark for the 2025 outcomes. In 2024 this dataset had been very representative of the 2024 DfE national and we expected this to continue into the 2025 results. Therefore we recommended:

Analysis of Results 2025
  1. The 2019 DfE national dataset does appear to be slightly more exacting than the 2025 Alps Customer benchmark. Using the DfE 2019 to measure analysis would potentially produce a more cautious overall grading than might ultimately be the case when the 2025 DfE national dataset becomes available.
  2. Therefore, at this point in time, we recommend using the 2025 Alps Customer benchmark set for your 2025 A level examination analysis. 
Target setting and tracking of Year 12 and 13 in 2025/26
  1. As GCSE results were brought back close to 2019 standards in 2024 & 2025 and will be in 2026 & 2027, value-added based on 2019 national benchmarks will give a fair assessment of progress for both Y12 and Y13 towards results in 2026 & 2027.
  2. Our Alps Customer benchmark should also give you a fair assessment of progress for both Y12 and Y13 towards results in 2026 & 2027, as our Customer benchmark produced similar outcomes in 2024 as the national.
Note that we previously recommended that the 2024 DfE national dataset was not representative of the value-added picture for 2025 outcomes and beyond. 

2025 DfE National dataset: comparison with the 2019 DfE national and 2025 Alps Customer benchmarks

MEPs and MEGs

The diagram below shows the Alps prior attainment bands for A level from 1 to 11 and the proportion of students in each band. It compares the 2025 DfE national dataset with the 2019 DfE national dataset and the Alps customer dataset from 2025.





There remains a slightly larger proportion of students in the top few bands for the 2025 datasets than that of 2019. This is in line with the elevated prior attainment addressed above. 
There is a high degree of correlation between the 2025 DfE and Alps Customer datasets in terms of the proportions of students in the prior attainment bands. 

In terms of Minimum Expected Points, diagram below compares the target points required at the 75th percentile across Alps bands. 





In most prior attainment bands, the MEPs generated from the 2025 Alps Customer benchmarks are very much in line with the DfE 2025 national data. 2019 MEPs are slightly higher, presenting a more exacting dataset for 2019 than in 2025. 
 

Translating those MEPs into grades, we arrive at the Minimum Expected Grades or MEGs. We can see in the table below that the MEGs needed across all three datasets are generally in line, but that the MEPs for 2019 are higher than the 2025 as demonstrated in the graph above. 




Implications on QI (Quality Indicator) grade

In this section we will consider the impact of the MEP changes on the QI scores and grades for A level. This means that we look at the QI scores and grades for all Alps customers, and analyse the shift patterns when we switch from one benchmark to another. We will continue to compare the 2019 DfE national with 2025 Alps Customer and the new 2025 DfE national.

As background it is worth noting that when we compared the QI impact between the DfE 2024 and the 2024 Alps Customer benchmarks this time last year, we found that 94% of providers saw shifts of 0.01 or less in the QI score. This gave us the confidence in the representative nature of our customer benchmarks to make the recommendation to use the 2025 Alps Customer version for analysis of 2025 outcomes.

QI score correlation between 2019 DfE and 2025 DfE
The graph below shows the close correlation in QI score for all Alps customers for A level between the 2019 DfE benchmark (NBM 19) and the new 2025 DfE benchmark (NBM 25). 

QI score correlation between 2025 Alps Customer and 2025 DfE
This next graph shows the close correlation in QI score for all Alps customers for A level between the 2025 Alps Customer benchmark (CBM 25) and the new 2025 DfE benchmark (NBM 25). 


In both cases, the correlation indicates that the value-added trend from the 2025 DfE dataset closely matches both the 2019 DfE and the 2025 Alps Customer datasets. The resulting impact on QI grade changes across Alps customers is as follows: 

 

 QI Grade changes

DfE 2019 to DfE 2025

2025 Alps Customer to 2025 DfE

1

60%

33%

0

40%

65%

-1

0%

2%


 The resulting impact on QI distributions across Alps customers is as follows: 

QI Grade

2019 DfE

2025 Alps Customer

2025 DfE

Red

25%

33%

34%

Black

49%

45%

53%

Blue

27%

22%

13%


Summarising the above: 
  1. If you have been using the 2019 DfE benchmark for 2025 results analysis and monitoring, generally speaking you are likely to see an uplift in value-added when you are switched to the 2025 DfE benchmark on toggle 1. The evidence for this is given in the fact that no Alps customers dropped a grade in the first table, and that there are 87% of Alps customers in the black/red QI zone when looking at the 2025 DfE benchmark compared to 74% black/red when matched against the 2019 DfE benchmark.  
  2. If you used the 2025 Alps Customer benchmark to analyse results in 2025 and to track current students, the data is much closer to the 2025 DfE national picture. You are most likely to remain on the same QI grade. Against the 2025 Alps Customer benchmark the percentages in each red/black/blue category are similar with perhaps a slight movement of blue to black, Alps grades 7 to 6 for example.
  3. The 2025 Alps Customer benchmark has been  found to be equally as representative of the 2025 DfE national figures. 

Results in 2026 and 2027 are predicted to be in line with the 2025 outcomes, and with GCSE results relatively stable from 2023 -2025, the prior attainment baseline has been reset.
Therefore the value-added picture in 2026 and 2027 is likely to be broadly in line with that of this 2025 DfE benchmark for A level.

It is therefore our recommendation that you use this new 2025 DfE dataset from which to base your interventions and priorities for the remainder of the time you have left with your current Year 13 students, and for tracking Year 12 as they progress through their A level courses. 

The default benchmark on Results Day 2026 will be the 2025 DfE national dataset.

What happens now with benchmarks in Connect and Summit?


We will release the 2025 DfE benchmark for A level into Connect and Summit at 8.30am on 19 February. Note that L3 Core Maths will be released alongside the A level benchmark. 

The benchmark will be applied to the following gradepoints
  1. Historical gradepoints, namely the gradepoints for KS5 A level from September 2024 up to and including the examination gradepoint for 2025.
  2. All monitoring gradepoints from September 2025 onwards to the examination gradepoint for 2026, i.e. all gradepoints affecting current Year 13 and 12 students
  3. The 2025 DfE benchmark will be automatically applied to Toggle 1: National 
  4. We will retain the 2025 Alps Customer benchmark on Toggle 2: Customer 

Summary of the benchmarks in Connect and Summit 


The tables below show which datasets are applied to which benchmark for which Academic Year. 
     
    1. Default benchmark/national toggle. This is the DfE national benchmark. From week beginning 16 February 2026 it will become the 2025 DfE national dataset. 
      1. Applied retrospectively to all gradepoints across 2024/25 including examination results 2025 
      2. Applied to all gradepoints across 2025/26 - the current academic year

    2. When using the Alps customer benchmark:
      1. All gradepoints up to and including examinations 2020/21 are analysed against the national 2019 datatset.  
      2. All gradepoints across 2021/22 will be analysed against the 2022 Alps customer benchmark. 
      3. All gradepoints across 2022/23 will be analysed against the 2023 Alps customer benchmark. 
      4. All gradepoints across 2023/24 will be analysed against the 2024 DfE national dataset as this was the first national dataset since the pandemic. 
      5. All gradepoints across 2024/25 and monitoring across 2025/26 will be analysed against the 2025 Alps customer benchmark. 

    IMPORTANT: please note that in the interim period between the release of the A level benchmarks in the week beginning 16 February 2026 and the release of the KS5 vocational benchmarks in the week beginning 16 March 2026, there is a hybrid nature to both sets of gradepoints across 2024/25 and 2025/26 as per the grid below. This will be rectified on the release of the vocation benchmarks when we expect to update the qualification set to 2025 DfE national on both sides of the toggle.