CJ History: FAQs
CJ History is our newest AI-enhanced assessment project. It's a Year 7 autumn term assessment on the essay question: Why did William win the Battle of Hastings?
You can view the project page and subscribe on our website here.
Below we've answered some frequently asked questions.
How do you know the AI is good enough to judge?
We have run two large scale assessments of Year 6 and Year 7 writing which have been very successful. We have also run our own trials for literature and history essays which have been very positive too. This will be our first large-scale history AI assessment.
What reports will I get?
You will get 6 PDF reports.
Our standard results report with all the data on your students and the national cohort
AI-human disagreement report
Teacher-friendly feedback
Student-friendly feedback
Individual audio comments
Whole group audio feedback reports
What is the calendar?
See here.
Your calendar doesn’t work for my school. Can we adapt it in any way?
The essays have to be uploaded to our website by 24 October. You can complete them before that if you finish the unit before then, but you will have to wait for results.
What about GDPR?
To take part in this project, you need to accept our enhanced GDPR agreement. You can read the details of the GDPR agreement here. Once you've accepted it, you can then go into your account and turn AI features on.
You can subscribe to this project before accepting the GDPR agreement. However, you will need to accept the GDPR and turn on AI features before the judging starts in the autumn term. You will not be able to take part if you do not accept the GDPR.
Do I have to enter every student?
We recommend entering your entire Year 7 cohort, but if you only want to enter some classes, you can.
How much does it cost?
The full price is £500 but we have some limited free places available.
What about the multiple choice questions?
These will be available in September on our separate Automark website.
How does the AI make its decisions? How do we know it's right? Does it have an understanding of what makes quality historical analysis?
One of the advantages of using 10% human judgements is that we can easily see if the human and AI judges are agreeing and identify any differences of opinion or biases in either set of judges. One of the strengths of human Comparative Judgement is that it allows teachers to make holistic judgements about the overall quality of an essay without being constrained by a mark scheme. Our previous human-AI writing assessments have been very successful, but there are two history-specific issues we hope to learn more about in this project: 1) what the right balance is between historical understanding and written communication 2) what the right balance is between analysis and narrative, and whether it is possible to easily split the two.
How do you know students aren’t cheating?
We ask all schools to submit handwritten responses that have been completed in class. Please do not set this task as a homework task. We will check every submission and contact schools where students have submitted very similar responses. Once you subscribe, you will be able to download a full set of instructions about the task and conditions.
How much teaching time will this take?
It's up to you, but we predict most schools will teach this as part of a half-termly unit.
You've recommended a 10% human - 90% AI judging ratio. How long will this take us?
We estimate about 15 minutes of teacher time. See here for more detail.
You can view the project page and subscribe on our website here.
Below we've answered some frequently asked questions.
How do you know the AI is good enough to judge?
We have run two large scale assessments of Year 6 and Year 7 writing which have been very successful. We have also run our own trials for literature and history essays which have been very positive too. This will be our first large-scale history AI assessment.
What reports will I get?
You will get 6 PDF reports.
Our standard results report with all the data on your students and the national cohort
AI-human disagreement report
Teacher-friendly feedback
Student-friendly feedback
Individual audio comments
Whole group audio feedback reports
What is the calendar?
See here.
Your calendar doesn’t work for my school. Can we adapt it in any way?
The essays have to be uploaded to our website by 24 October. You can complete them before that if you finish the unit before then, but you will have to wait for results.
What about GDPR?
To take part in this project, you need to accept our enhanced GDPR agreement. You can read the details of the GDPR agreement here. Once you've accepted it, you can then go into your account and turn AI features on.
You can subscribe to this project before accepting the GDPR agreement. However, you will need to accept the GDPR and turn on AI features before the judging starts in the autumn term. You will not be able to take part if you do not accept the GDPR.
Do I have to enter every student?
We recommend entering your entire Year 7 cohort, but if you only want to enter some classes, you can.
How much does it cost?
The full price is £500 but we have some limited free places available.
What about the multiple choice questions?
These will be available in September on our separate Automark website.
How does the AI make its decisions? How do we know it's right? Does it have an understanding of what makes quality historical analysis?
One of the advantages of using 10% human judgements is that we can easily see if the human and AI judges are agreeing and identify any differences of opinion or biases in either set of judges. One of the strengths of human Comparative Judgement is that it allows teachers to make holistic judgements about the overall quality of an essay without being constrained by a mark scheme. Our previous human-AI writing assessments have been very successful, but there are two history-specific issues we hope to learn more about in this project: 1) what the right balance is between historical understanding and written communication 2) what the right balance is between analysis and narrative, and whether it is possible to easily split the two.
How do you know students aren’t cheating?
We ask all schools to submit handwritten responses that have been completed in class. Please do not set this task as a homework task. We will check every submission and contact schools where students have submitted very similar responses. Once you subscribe, you will be able to download a full set of instructions about the task and conditions.
How much teaching time will this take?
It's up to you, but we predict most schools will teach this as part of a half-termly unit.
You've recommended a 10% human - 90% AI judging ratio. How long will this take us?
We estimate about 15 minutes of teacher time. See here for more detail.
Updated on: 02/06/2025
Thank you!