Can computers really mark exams? Benefits of ELT automated assessments

app Languages
Hands typing at a laptop with symbols

Automated assessment, including the use of Artificial Intelligence (AI), is one of the latest education tech solutions. It speeds up exam marking times, removes human biases, and is as accurate and at least as reliable as human examiners. As innovations go, this one is a real game-changer for teachers and students. 

However, it has understandably been met with many questions and sometimes skepticism in the ELT community – can computers really mark speaking and writing exams accurately? 

The answer is a resounding yes. Students from all parts of the world already take AI-graded tests.  aԻ Versanttests – for example – provide unbiased, fair and fast automated scoring for speaking and writing exams – irrespective of where the test takers live, or what their accent or gender is. 

This article will explain the main processes involved in AI automated scoring and make the point that AI technologies are built on the foundations of consistent expert human judgments. So, let’s clear up the confusion around automated scoring and AI and look into how it can help teachers and students alike. 

AI versus traditional automated scoring

First of all, let’s distinguish between traditional automated scoring and AI. When we talk about automated scoring, generally, we mean scoring items that are either multiple-choice or cloze items. You may have to reorder sentences, choose from a drop-down list, insert a missing word- that sort of thing. These question types are designed to test particular skills and automated scoring ensures that they can be marked quickly and accurately every time.

While automatically scored items like these can be used to assess receptive skills such as listening and reading comprehension, they cannot mark the productive skills of writing and speaking. Every student's response in writing and speaking items will be different, so how can computers mark them?

This is where AI comes in. 

We hear a lot about how AI is increasingly being used in areas where there is a need to deal with large amounts of unstructured data, effectively and 100% accurately – like in medical diagnostics, for example. In language testing, AI uses specialized computer software to grade written and oral tests. 

How AI is used to score speaking exams

The first step is to build an acoustic model for each language that can recognize speech and convert it into waveforms and text. While this technology used to be very unusual, most of our smartphones can do this now. 

These acoustic models are then trained to score every single prompt or item on a test. We do this by using human expert raters to score the items first, using double marking. They score hundreds of oral responses for each item, and these ‘Standards’ are then used to train the engine. 

Next, we validate the trained engine by feeding in many more human-marked items, and check that the machine scores are very highly correlated to the human scores. If this doesn’t happen for any item, we remove it, as it must match the standard set by human markers. We expect a correlation of between .95-.99. That means that tests will be marked between 95-99% exactly the same as human-marked samples. 

This is incredibly high compared to the reliability of human-marked speaking tests. In essence, we use a group of highly expert human raters to train the AI engine, and then their standard is replicated time after time.  

How AI is used to score writing exams

Our AI writing scoring uses a technology called . LSA is a natural language processing technique that can analyze and score writing, based on the meaning behind words – and not just their superficial characteristics. 

Similarly to our speech recognition acoustic models, we first establish a language-specific text recognition model. We feed a large amount of text into the system, and LSA uses artificial intelligence to learn the patterns of how words relate to each other and are used in, for example, the English language. 

Once the language model has been established, we train the engine to score every written item on a test. As in speaking items, we do this by using human expert raters to score the items first, using double marking. They score many hundreds of written responses for each item, and these ‘Standards’ are then used to train the engine. We then validate the trained engine by feeding in many more human-marked items, and check that the machine scores are very highly correlated to the human scores. 

The benchmark is always the expert human scores. If our AI system doesn’t closely match the scores given by human markers, we remove the item, as it is essential to match the standard set by human markers.

AI’s ability to mark multiple traits 

One of the challenges human markers face in scoring speaking and written items is assessing many traits on a single item. For example, when assessing and scoring speaking, they may need to give separate scores for content, fluency and pronunciation. 

In written responses, markers may need to score a piece of writing for vocabulary, style and grammar. Effectively, they may need to mark every single item at least three times, maybe more. However, once we have trained the AI systems on every trait score in speaking and writing, they can then mark items on any number of traits instantaneously – and without error. 

AI’s lack of bias

A fundamental premise for any test is that no advantage or disadvantage should be given to any candidate. In other words, there should be no positive or negative bias. This can be very difficult to achieve in human-marked speaking and written assessments. In fact, candidates often feel they may have received a different score if someone else had heard them or read their work.

Our AI systems eradicate the issue of bias. This is done by ensuring our speaking and writing AI systems are trained on an extensive range of human accents and writing types. 

We don’t want perfect native-speaking accents or writing styles to train our engines. We use representative non-native samples from across the world. When we initially set up our AI systems for speaking and writing scoring, we trialed our items and trained our engines using millions of student responses. We continue to do this now as new items are developed.

The benefits of AI automated assessment

There is nothing wrong with hand-marking homework tests and exams. In fact, it is essential for teachers to get to know their students and provide personal feedback and advice. However, manually correcting hundreds of tests, daily or weekly, can be repetitive, time-consuming, not always reliable and takes time away from working alongside students in the classroom. The use of AI in formative and summative assessments can increase assessed practice time for students and reduce the marking load for teachers.

Language learning takes time, lots of time to progress to high levels of proficiency. The blended use of AI can:

  • address the increasing importance of formative assessmentto drive personalized learning and diagnostic assessment feedback 

  • allow students to practice and get instant feedback inside and outside of allocated teaching time

  • address the issue of teacher workload

  • create a virtuous combination between humans and machines, taking advantage of what humans do best and what machines do best. 

  • provide fair, fast and unbiased summative assessment scores in high-stakes testing.

We hope this article has answered a few burning questions about how AI is used to assess speaking and writing in our language tests. An interesting quote from Fei-Fei Li, Chief scientist at Google and Stanford Professor describes AI like this:

“I often tell my students not to be misled by the name ‘artificial intelligence’ — there is nothing artificial about it; A.I. is made by humans, intended to behave [like] humans and, ultimately, to impact human lives and human society.”

AI in formative and summative assessments will never replace the role of teachers. AI will support teachers, provide endless opportunities for students to improve, and provide a solution to slow, unreliable and often unfair high-stakes assessments.

Examples of AI assessments in ELT

At app, we have developed a range of assessments using AI technology.

Versant

The Versant tests are a great tool to help establish language proficiency benchmarks in any school, organization or business. They are specifically designed for placement tests to determine the appropriate level for the learner.

PTE Academic

The  is aimed at those who need to prove their level of English for a university place, a job or a visa. It uses AI to score tests and results are available within five days. 

app English International Certificate (PEIC)

app English International Certificate (PEIC) also uses automated assessment technology. With a two-hour test available on-demand to take at home or at school (or at a secure test center). Using a combination of advanced speech recognition and exam grading technology and the expertise of professional ELT exam markers worldwide, our patented software can measure English language ability.

Read more about the use of AI in our learning and testing here, or if you're wondering which English test is right for your students make sure to check out our post 'Which exam is right for my students?'.

Explore our courseware and exams

More blogs from app

  • A girl sat at a desk looking at an exam paper, there are people behind her sat at desks in a row doing similar

    Reducing exam day stress

    By Amy Malloy

    What are the origins of exam-day stress?

    There’s no doubt about it – exams are scary. But why exactly is this? What is it about an exam scenario that stresses us out and how can we make it feel okay?

    To answer these questions, we’re going to have to take a trip back in time and look at how our ancestors evolved their abstract thinking skills. At the same time, we’ll look at how mindfulness can actively help shift the way we think about exams.

    About 70,000 years ago, or thereabouts, a critical development happened in how the human brain processed experiences.

    Over a relatively short time (in evolutionary terms), scientists believe we stopped simply experiencing primal urges (safety/hunger/tiredness etc.) and responding to them.

    Instead, we started to be able to imagine, analyze and believe in things that didn’t exist. This meant not only could our brains respond to the threat of a real predator in front of us, but also to the perceived or imagined threat of a predator. This shift had really interesting consequences for our future relationship to threat.

    Developing imaginations also meant we started to believe in the same things as our peer groups. As a result, if a member of the tribe stopped behaving in a way that supported the tribe’s survival, other members might start to doubt their usefulness as a member of the community.

    At this time, social isolation meant no share of the food and no protection against predators. Inclusion was directly linked to survival. Essentially the physiological consequences of not meeting societal expectations were the same as coming face-to-face with a lion: fight or flight.

    Why is fight or flight mode a problem for exam performance?

    Fast forward to the present day and the context of English language exams. Exams are a type of societal expectation: a standard of language proficiency which a student is required to meet to prove their usefulness in a community who speaks that language.

    Thinking about it this way, it’s no wonder we can experience stress and panic at the thought of an exam or in the exam room itself. In evolutionary terms, an exam = a lion or the possibility of social exclusion!

    Fight or flight mode is designed to help us overcome or escape danger as quickly as possible.

    When fight or flight is activated, our brain shuts down the bodily functions it doesn’t need to survive. This includes digestion, language creation, creative and strategic thinking, writing, and deeper breathing.

    Unfortunately, some of these are key to performing well in a language exam!

    that, especially in an exam course context, “extremely high stress levels[…] can affect your cognition, negatively impacting your memory and ability to complete a task. Over time, chronically high stress levels can also impair your ability to form new memories, so high stress levels all semester long could impact your final exam performance later.”

    This means that high stress levels in the exam make it hard to perform at one’s best, and stress in anticipation of the exam can mean that the content is not learned properly in the first place.

    What can teachers do to help?

    Ultimately, we want to make the exam scenario seem safe. When our mind and body do not perceive danger, they will allow the ‘rest and digest’ functions to become active and allow our performance to flow.

    There will of course be some nerves. But the key is to help the student stay in that sweet spot of clear thinking before the stress takes over and turns into panic.

    Here are three ways we can help:

    1. Practice mindful breathing

    Practice some simple mindful breathing with your students. Mindfulness refers to the awareness that we find by focusing consciously on the present moment, objectively, with compassion and without judgment.

    It takes us back into the physical reality of the present moment and out of anxious thought cycles, which can make our stress spiral out of control. Practicing regularly ahead of the exam will help build up neural connections in the brain around the areas that help us relax and .

    The following straightforward routine can be used both regularly at the start of class and by the student in the exam itself if they feel panic rising:

    • STOP: Close your eyes. Take a moment to notice the physical sensation of the chair beneath you and the floor underneath your feet. Notice the sounds around you in the room and allow them simply to flow in and out without needing to follow them.
    • BREATHE: Focus your attention gently on your breath going in and out of your nose at the tip of the nostrils. Count ten breaths in and out, seeing if you can notice something new about each one. If you feel panicky, allow the breath out to continue a little longer than the breath in.
    • WATCH: Become aware of your thoughts and allow them to simply come and go, like clouds across the sky. Notice if you are holding on to something from the past or striving towards something in the future. Allow these to simply release and take your attention back to the breath in the present moment. Count another ten breaths. Then open your eyes and come back into the room.

    2. Familiarize your students with the exam scenario

    Make the exam scenario as familiar as possible in advance. Anything you can do to make the exam conditions appear less threatening ahead of exam day will be of great benefit.

    This may be as simple as making conditions surrounding regular progress assessments as close to the main exam as possible, so it is less strange on the day. If you have control over the exam day itself, make this as low pressure as possible.

    3. Help students develop a growth mindset

    Foster a growth mindset with your students by using the word ‘yet’ with them. Encourage them to replace ‘I can’t do it’ with ‘I can’t do it yet’. This simple word takes them out of comparison mode with where they think they should be in terms of language ability and helps them stay focused on where they are, reducing self-pressure.

    All of the above strategies are designed to take away the threat of a lion in the exam room when the day comes.

    Wishing your students all the very best for their exam preparation and good luck!

  • A teacher holding books in a classroom

    Mindfulness for teachers: managing expectations over the holidays

    By Amy Malloy

    Mindfulness and your routine

    In the run-up to the holidays, it is common to feel like your routine has completely broken down, especially when you’re not giving classes or working at school. The holidays also often bring with them lots of people, family, and excitement. That sometimes means we also experience fluctuating emotions, stress and the feeling that everything should be perfect.

    On top of this, shops and social media are filled with advertising – and there’s definitely more ‘stuff’ to buy. In addition, we can see messages telling us we need to feel ‘merry’ and ‘bright’ wherever we look. Even the popular greeting, ‘Merry Christmas’, can sometimes feel less of a greeting and more of an instruction.

    Sometimes it feels like the people around us expect us to always feel happy and joyful over the holiday season. This is fine if we do feel merry, but we will always have ups and downs. If you don’t feel happy, for whatever reason, it can feel even harder than it might at times where there is less expectation all around us.

    Overcoming the challenge

    Finding a way to introduce mindfulness into the holiday season can be a wonderful way for us to understand our emotions at this time of year. It will help you think about your expectations and let you find a moment to pause to accept whatever the reality actually is.

    Here are some quick and easy ways to find some ‘you’ time and keep checking in with how you’re feeling. These are also tips you can try with children in the classroom and for yourself at home to keep yourself on track.

    3-minute body scan

    Find a quiet moment. This may be in the few minutes after you wake up or go to bed, during break time, or even at the start of a lesson.

    • Notice the contact of your feet with the floor. Notice the sounds around you in the room.
    • Take three deep breaths and notice how they feel.
    • Scan down the body in your mind from the top of your head all the way down to your toes. Observe what you notice about your body with an air of curiosity – look for any tension, discomfort or comfort. Also, notice if there are any expectations you have of that day or moment. See if you can simply notice them and set them aside. This curiosity helps us stay detached from what we notice so we can just observe.
    • Take three more deep breaths, and carry on.

    2-minute notebook

    Writing something down can be a wonderfully mindful exercise. Have a stack of post-its or a little notebook on your desk or bedside table. You could encourage your students to do the same.

    1. Pick a point in your day. It could be at the start of each day, the start of each lesson, or just before bed. Each day, at that time, take a moment to write down three:

    • good things that have happened in your day
    • things that felt challenging
    • things you feel grateful and thankful for.

    2. Review your notes every now and again during the holiday period. This will give you a sense of your shifts of mood and energy that might have occurred.

    Noticing something you feel grateful for has been shown to .

    1-minute cupboard pause

    When things feel over-stimulating, find a quiet space just for a minute. Even if it’s in a cupboard!

    STOP: notice the contact of your feet with the floor.
    BREATHE: take ten deep breaths, breathing in for a count of four and out for a count of six.
    WATCH: watch each breath coming and going from the nose or chest or belly. Observe what your thoughts and feelings are doing. Allow them to sit without needing to respond.

    Then head back into the area you were in.

    I hope these tips help you to navigate the festive season without expectation and with curiosity for what each moment holds along the way.

    Remember that the holiday days you celebrate are really just normal days. It’s simply that expectations have changed, and what’s more, everyone’s expectations will be different.

    Simply taking time to notice this can make a massive difference to the pressure we put on ourselves. Releasing this pressure can even lead to more enjoyment overall – so why don’t you try it and see?

  • A businessman sat cross legged is meditating on a desk, around him are other business people sat at the same desk busy working

    Understanding and managing technostress

    By Amy Malloy

    What is technostress?

    Technostress affects people in different ways. I resonate with Chiapetta’s (2017) definition:

    “Technostress is a syndrome that occurs when the person, subjected to information overload and continuous contact with most digital devices, develops a state of stress.”

    This explanation isn’t limited to any particular symptoms. However, most of us can relate to being over-exposed to technological devices – especially in the past year. Consequently, I imagine many people have experienced some symptoms of technostress as a result.

    What are the symptoms of technostress?

    While the symptoms of technostress differ, there are a number of common experiences. Perhaps you obsessively check social media, or struggle to focus on everyday tasks. Or maybe you feel demotivated and low in mood. You might find yourself overly focused on getting the latest technology. Or you might be actively avoiding or feeling anxious about using technology.

    Why should we be concerned about technostress?

    People are adaptable, it’s part of our biology, but if change happens too quickly, we experience symptoms of stress. And, in our modern, continuously busy culture, it can be easy to assume something is wrong with us if we feel we can’t cope with technology. It can feel like we’re failing because we’re not focusing or achieving enough. We might also feel stupid for not understanding how a piece of software works.

    In response, we need to find ways to relieve that stress and return to a more restful state.

    Tips to prevent and manage technostress

    1. Switch off – This may be hard to do with the demands of life but sometimes, just turning your devices off and stepping away from technology is the best approach. Do something to pull you away from the information overload, like a walk in nature or reading a fictional book. Something that pulls you away from life’s goings-on.
    2. Limit your exposure – If you can’t switch off completely for whatever reason, then maybe rationing your exposure to devices is key. Time and limit your time on devices, taking regular breaks away. If you struggle to pull yourself away, there are apps and device functions that can restrict access for you.
    3. Only use with a purpose – Before you start using technology, ask yourself: “What am I using this for?”. Study? Research? Or is it just out of boredom? If you haven’t got a set aim or goal, then you are likely to mindlessly scroll and get pulled into the vast information overload of social media and the web. Try to avoid this by having a set aim/goal when using technology.