Explaining computerized English testing in plain English

app Languages
a pair of hands typing at a laptop

Research has shown that automated scoring can give more reliable and objective results than human examiners when evaluating a person’s mastery of English. This is because an automated scoring system is impartial, unlike humans, who can be influenced by irrelevant factors such as a test taker’s appearance or body language. Additionally, automated scoring treats regional accents equally, unlike human examiners who may favor accents they are more familiar with. Automated scoring also allows individual features of a spoken or written test question response to be analyzed independent of one another, so that a weakness in one area of language does not affect the scoring of other areas.

was created in response to the demand for a more accurate, objective, secure and relevant test of English. Our automated scoring system is a central feature of the test, and vital to ensuring the delivery of accurate, objective and relevant results – no matter who the test-taker is or where the test is taken.

Development and validation of the scoring system to ensure accuracy

PTE Academic’s automated scoring system was developed after extensive research and field testing. A prototype test was developed and administered to a sample of more than 10,000 test takers from 158 different countries, speaking 126 different native languages. This data was collected and used to train the automated scoring engines for both the written and spoken PTE Academic items.

To do this, multiple trained human markers assess each answer. Those results are used as the training material for machine learning algorithms, similar to those used by systems like Google Search or Apple’s Siri. The model makes initial guesses as to the scores each response should get, then consults the actual scores to see well how it did, adjusts itself in a few directions, then goes through the training set over and over again, adjusting and improving until it arrives at a maximally correct solution – a solution that ideally gets very close to predicting the set of human ratings.

Once trained up and performing at a high level, this model is used as a marking algorithm, able to score new responses just like human markers would. Correlations between scores given by this system and trained human markers are quite high. The standard error of measurement between app’s system and a human rater is less than that between one human rater and another – in other words, the machine scores are more accurate than those given by a pair of human raters, because much of the bias and unreliability has been squeezed out of them. In general, you can think of a machine scoring system as one that takes the best stuff out of human ratings, then acts like an idealized human marker.

app conducts scoring validation studies to ensure that the machine scores are consistently comparable to ratings given by skilled human raters. Here, a new set of test-taker responses (never seen by the machine) are scored by both human raters and by the automated scoring system. Research has demonstrated that the automated scoring technology underlying PTE Academic produces scores comparable to those obtained from careful human experts. This means that the automated system “acts” like a human rater when assessing test takers’ language skills, but does so with a machine's precision, consistency and objectivity.

Scoring speaking responses with app’s Ordinate technology

The spoken portion of PTE Academic is automatically scored using app’s Ordinate technology. Ordinate technology results from years of research in speech recognition, statistical modeling, linguistics and testing theory. The technology uses a proprietary speech processing system that is specifically designed to analyze and automatically score speech from fluent and second-language English speakers. The Ordinate scoring system collects hundreds of pieces of information from the test takers’ spoken responses in addition to just the words, such as pace, timing and rhythm, as well as the power of their voice, emphasis, intonation and accuracy of pronunciation. It is trained to recognize even somewhat mispronounced words, and quickly evaluates the content, relevance and coherence of the response. In particular, the meaning of the spoken response is evaluated, making it possible for these models to assess whether or not what was said deserves a high score.

Scoring writing responses with Intelligent Essay Assessor™ (IEA)

The written portion of PTE Academic is scored using the Intelligent Essay Assessor™ (IEA), an automated scoring tool powered by app’s state-of-the-art Knowledge Analysis Technologies™ (KAT) engine. Based on more than 20 years of research and development, the KAT engine automatically evaluates the meaning of text, such as an essay written by a student in response to a particular prompt. The KAT engine evaluates writing as accurately as skilled human raters using a proprietary application of the mathematical approach known as Latent Semantic Analysis (LSA). LSA evaluates the meaning of language by analyzing large bodies of relevant text and their meanings. Therefore, using LSA, the KAT engine can understand the meaning of text much like a human.

What aspects of English does PTE Academic assess?

Written scoring

Spoken scoring
  • Word choice
  • Grammar and mechanics
  • Progression of ideas
  • Organization
  • Style, tone
  • Paragraph structure
  • Development, coherence
  • Point of view
  • Task completion
  • Sentence mastery
  • Content
  • Vocabulary
  • Accuracy
  • Pronunciation
  • Intonation
  • Fluency
  • Expressiveness
  • Pragmatics

More blogs from app

  • Hands typing at a laptop with symbols

    Can computers really mark exams? Benefits of ELT automated assessments

    By app Languages

    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.  and 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, including , 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.