Overview

One of the frequent complaints about ESOL assessment is the amount of time it takes for an assessor to write up the assessment and complete the relevant Assessment Record Sheets, particularly due to the requirement for actual learner examples. With this in mind, we have been looking at ways to reduce the assessment and record-keeping burden while keeping the integrity of our assessments and the judgement of assessors.

Phase one of this project would allow the use of transcription devices for all centres to help assessors pinpoint examples of where learners have met the standards and provide easy to use evidence and examples. The potential second phase would be a pilot project (on Entry Level Functional Skills) run by our Senior Standards Verifier to explore how an assessor can interface with AI to make AI-assisted judgements on Speaking and Listening assessments. This pilot, planned for the 2025/26 academic year at two centers, aims to evaluate the approach's effectiveness and potential broader application across other centres.

Phase 1

The creation of Assessment Record Sheets for ESOL is time-consuming for assessors. Using transcription devices would reduce the amount of time needed to complete these forms.

As part of the assessment process, centres are encouraged to record their learners, either via video, audio or online team meetings. Uploading the video and audio recordings to OneDrive for storage, as is common, would give providers access to transcription tools. This can be done at the time of recording, with online team meetings using tools such as Copilot.

This transcription can then be used in conjunction with the recording to write up the assessments and make the evidencing of standards and the use of exemplar quotes much easier. Pearson will be creating guidance for centres to ensure that they can check the accuracy of the transcripts as they write up their Assessment Record Sheets.

As well as saving assessors time, there is also the bonus that using transcription will help improve the quality of the Assessment Record Sheets and the judgements overall. Assessors will need to watch the assessment recordings back alongside the transcripts to ensure that they are accurate, and this will help to improve the overall quality of judgements.

Phase 2

Phase 2 is more complex than Phase 1 as it will require the assessors to check the accuracy of the judgements being made by AI. The plan is to run a pilot of this not only to ensure that AI can make these judgements accurately, but also to improve the inputs from the assessors to ensure that the judgements being made are merited. This will be a closed pilot, and we will give feedback on our findings in the summer of 2026.

What Do Assessors Think of This?

Speaking to Richard Wensley, our Senior Standards Verifier and an assessor, who came up with the initial idea and will lead the potential phase 2 pilot, he said this “sounds exciting. I know this will be greatly welcomed by centres.” He went on to say "This is just an option for centres to use. It is not mandatory. Although not all assessments are to be recorded, it certainly helps assessors with the write-ups, whilst allowing Internal Verifiers to check consistency and quality. Also, the videos themselves could be used in-house to standardise teams of assessors.”

Frequently Asked Questions

Is it mandatory?

No.

Which Speaking and Listening activities will suit this best?

We believe that this will be really useful for the discussions at all levels and for the presentations at higher levels.

What if I do not record my Speaking and Listening?

While recording Speaking and Listening is mandatory for ESOL, some learners, if they have an exempt do not have to be. If an assessment is not recorded, then unfortunately, you will not be able to make a transcript to ease the writing burden on your assessors.

Could there be any issues with the use of recordings and transcription?

Assessors need to be careful with the accuracy of what has been said. Inaccuracy is usually due to the candidate mumbling or having pronunciation issues, or even background noises/poor quality recording. An accent could also play a part. I know from my experience in creating captions for my webinars, the transcription process has difficulty picking out my name correctly.