Can ChatGPT answer multiple-choice questions?

Looking at the chatbot's answers

can-ChatGPT-answer-multiple-choice-questions-ai-icon-openai-logo

PC Guide is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Read More

Last Updated on

In the past few years, a multitude of generative AI solutions have emerged. They have transformed the way we interact with technology and information online. While many generative AI solutions are available now, OpenAI’s ChatGPT stands tall as a frontrunner. Its prowess in understanding and generating human-like text seems far ahead of several other solutions. But among all its capabilities, one intriguing question is whether ChatGPT can answer multiple-choice questions. Multiple-choice questions are a standard assessment format that many users need help with. Let’s find out if ChatGPT can answer multiple-choice questions.

Can ChatGPT answer true or false questions?

The answer is a resounding yes. ChatGPT is adept at comprehending the context and nuances of students’ learning styles and MCQs. As a result, it can provide accurate and contextually relevant right answers. By leveraging its extensive training on diverse textual data, ChatGPT can analyze the given options and select the most suitable response based on the information provided. This capability, thanks to the abilities of large language models like GPT-4, makes ChatGPT a versatile tool for educational purposes, test preparation, and information retrieval. It performs especially well on true/false questions.

In an article published in JAMA Internal Medicine, a team of researchers found that the model on average scored more than four points higher than students.

“We were very surprised at how well ChatGPT did on these kinds of free-response medical reasoning questions by exceeding the scores of the human test-takers,” says co-author Eric Strong, a hospitalist and clinical associate professor at Stanford School of Medicine.

Co-author Alicia DiGiammarino expanded that “with these kinds of results, we’re seeing the nature of teaching and testing medical reasoning through written text being upended by new tools.” DiGiammarino, the Practice of Medicine Year 2 Education manager at the School of Medicine, went on to say that “ChatGPT and other programs like it are changing how we teach and ultimately practice medicine.”

Can ChatGPT answer multiple-choice questions?

However, ChatGPT has some limitations you need to know before using it for answering multiple-choice questions. ChatGPT is trained in a diverse range of internet text up until September 2021. This includes websites, books, articles, and other text sources to develop a broad understanding of human language.

This means that while ChatGPT can give possible answers to multiple-choice quiz questions, you cannot rely on its accuracy, especially when it comes to questions on data past September 2021. So if you ask ChatGPT to answer a multiple choice question on who is the current Prime Minister of France or similar questions, it will likely give an inaccurate answer. However, for many other questions, ChatGPT generally gives correct answers. Access to the internet effectively makes all exams an open-book test for the chatbot – without any need for prior study.

Nonetheless, there may still be tell-tale signs an examination has been completed by AI. Although with multiple-choice answers it is hard to distinguish chatbots from humans, questions that require more creativity and greater analytical insight are usually better answered by humans, as studies by James Fern at the University of Bath show. In these, chatbots can make nonsensical errors and often struggle with getting a real-looking reference in paragraph-long answers or mathematical processes. ChatGPT’s score is systematically lower than humans in this respect.

Can AI teach accounting or medicine?

Speaking on the ability of AI to teach accounting, lead study author David Wood, a BYU professor of accounting called this technology a game-changer. “When this technology first came out, everyone was worried that students could now use it to cheat,” recalls Wood, “but opportunities to cheat have always existed. So for us, we’re trying to focus on what we can do with this technology now that we couldn’t do before to improve the teaching process for faculty and the learning process for students. Testing it out was eye-opening.”

Financial accounting and managerial accounting are, of course, data-based skills. However, practitioners cannot afford a wrong multiple-choice answer, as the undergrad BYU students mentioned previously (at least, while practicing on a test dummy).

“It’s not perfect. You’re not going to be using it for everything,” admits Jessica Wood, BYU freshman. “Trying to learn solely by using ChatGPT is a fool’s errand.”

Beyond Answering: ChatGPT as a Multiple Choice Question Creator

The capabilities of ChatGPT extend beyond mere question-answering. It can also contribute to creating multiple-choice questions, demonstrating its potential of artificial intelligence to assist educators, content creators, and researchers. Simply input text instructions 

By generating well-structured multiple-choice questions, ChatGPT becomes a valuable asset in curriculum development, assessment design, and the creation of engaging learning materials. The AI chatbot’s ability to mimic human-like complex questions, while considering various levels of complexity, adds an extra dimension to its utility. Multiple question types can be created. This provides another means of testing real-life students’ abilities through creation of multiple-choice tests, without laborious work from teachers and professors. This is simple thanks to its fast access and processing of an entire corpus of internet content, and could revolutionize the teaching process.

Essential AI Tools

Editor’s pick

7-in-1 AI Content Checker – One-click, Seven Checks

7 Market leading AI Content Checkers in ONE click. The only 7-in-1 AI content detector platform in the world. We integrate with leading AI content detectors to give unparalleled confidence that your content appear to be written by a human.
Only $0.00015 per word!

Winston AI detector

Winston AI: The most trusted AI detector. Winston AI is the industry leading AI content detection tool to help check AI content generated with ChatGPT, GPT-4, Bard, Bing Chat, Claude, and many more LLMs.
Only $0.01 per 100 words

Originality AI detector

Originality.AI Is The Most Accurate AI Detection.Across a testing data set of 1200 data samples it achieved an accuracy of 96% while its closest competitor achieved only 35%. Useful Chrome extension. Detects across emails, Google Docs, and websites.
EXCLUSIVE DEAL 10,000 free bonus credits

Jasper AI

On-brand AI content wherever you create. 100,000+ customers creating real content with Jasper. One AI tool, all the best models.
TRY FOR FREE

WordAI

10x Your Content Output With AI. Key features – No duplicate content, full control, in built AI content checker. Free trial available.

Final thoughts

Chat GPT’s capabilities regarding multiple-choice test questions has come into the spotlight due to how recent studies revealed the AI bot outperformed medical students from the United States in a test on clinical reasoning skills – the ACG assessment, which gauges how you would do on the American Board of Internal Medicine exam. This posits questions on medical education and the future of students’ learning. The chatbot performed better in the case-report portion of the exam, although in other sections it showed false memories, with the adding-in of false details.

The kinds of results provided by an AI system can be systematically different from those of a human. Challenging clinical care exam questions, and other medical reasoning questions, are still best served by trained human doctors. The details of a patient case may even exceed the token length of context that the AI system works with.

A school of medicine faculty has access to patient medical records that should not be used in text prompts. This is true of any personal data. Such medical records could clue medical professionals into unrelated chronic medical conditions due to myriad extraneous details to which artificial intelligence is not privy. The nutrition status of a patient should also not be prompted by an AI chatbot.

Online chatbots are the fascination of the entire internet. ChatGPT’s influence cannot be overstated, but it won’t constitute the wholesale replacing of doctors. They’re great for open-ended questions (free-response question) with original answers – things that calculators and Google can’t solve. However, case-based questions involving a particular case study or clinical reasoning cases are best left to medical professionals. In that case, For now, it’s fun to see how ChatGPT’s clinical practice answers compare to those of students, but it also poses worries about the future test-taking integrity of tomorrow’s doctors. 

Can ChatGPT answer multiple-choice questions? FAQs

Can I use ChatGPT for exams?

You can use ChatGPT for exams but cannot rely on its answers. Many experts have used ChatGPT for this and have concluded that the answers can be false and inaccurate. Moreover, universities can detect ChatGPT. Hence, it is best to avoid using it for exams.

Why does ChatGPT give different answers?

ChatGPT uses natural language processing for understanding and generating human-like text. Hence, even slight changes in the questions can make ChatGPT produce different answers.