ALEXA-Grade These Tests

Artificial intelligence in the educational world

Introduction

Big data is an up and coming term used interchangeably among many industries. However, within the realm of education, big data refers to “a large volume of data produced through online courses, teaching and learning activities. (Oi et al., 2017, as cited in Baig et al., 2020). In other words, lots of information about students that is supposed to help drive instruction. A common problem amongst educators and their ability to analyze big data appropriately all points back to time, resources, and training. Most educators simply don’t have enough time to offer quick feedback on assessments. Many districts lack the resources to properly comb through data to help drive instruction and training and professional development opportunities are always an issue. Can big data transform teaching and learning? Yes. Can it create other issues that invade student privacy, profile learners, and create test-driven teaching? Also, yes. (Cope & Kalantzis, 2016).

The CBTA, The CAT, and The NLP

Enter computer-based text analysis (CBTA). The future success of big data lies in the hands of technology. With the proper use of the proper tools, educators can begin assessing students in a variety of ways that cater to the variety of skill levels within their classroom. The list of acronyms for this type of testing is exhaustive. One definitely needs an organized table to understand the many types of computer-based testing available. However, to keep it simple. Here’s the skinny on just a few of the most dominantly used programs.

Computer Adaptive Testing

Computer Adaptive Testing (CAT) is a program that generates questions according to what the student knows about a topic. “Computer adaptive tests serve students progressively harder or easier questions depending on whether they answer correctly” (Cope & Kalantzis, 2016). This type of assessment produces more valuable feedback to educators. Another proactive matter of this assessment is the ability to lessen cheating because very rarely do two students end up taking the same test.

Natural Language Processing

Natural Language Processing (NLP) is a form of artificial intelligence that creates connections between humans and computers in a variety of ways (What is Natural Language Processing? Definition and Examples, n.d.). For assessment purposes, most CATs can use NLP to grade written response questions using two main methods: statistical corpus comparison and analytical text parsing (Cope & Kalantzis, 2016). Statistical corpus comparison the computer system is trained through the connection of human-graded texts entered into the machine. The machine then compares the human-graded texts with the student responses to produce a grade. Analytical text parsing is when machines are specifically programmed to look for certain features within the writing that show competency of the subject.

What do they all have in common?

While the progression of such systems is still developing, one thing remains common-artificial intelligence (AI) is here to stay. Jimenez & Boser (2020) produced a report entailing all established uses of AI in our world today and predicted ways it can accelerate education forward in positive ways. AI machines can be used to monitor individual student activity, create learning paths tailored to the unique learner, and accurately predict student outcomes on targeted standards-just to name a few. Using tools like vision-based AI, teachers can use an optical system to take a picture of a student’s work, i.e. a math problem, and then the program will grade it. Voice-recognition systems, those used to power Alexa and Siri, have researchers “exploring ways to use (it) to diagnose reading and other academic issues” (Jimenez & Boser, 2020).

The report goes on to explain that while AI is already widely used across educational settings, many experts believe the possibilities are endless. Whether you are already on the AI bandwagon or refusing to even pack your bag, I think we can all agree that this type of technology can benefit students and teachers alike. As an educator, the predictions made about AI eventually creating tailored assessments that target student interests, using early warning systems to catch students in danger of failing, or grading papers just by taking a picture sounds incredibly promising. Faster grading leads to faster feedback and faster feedback leads better students.

Integrity of It All

As far as the academic integrity of each type of computer based assessment, it all lies within the learner. I believe there are ways to prevent the sharing of tests and cheating like creating different forms of the assessment or even utilizing the CAT that evolves with the learner and gives a variety of questions depending on skill. Setting time parameters on tests or requiring they be taken in person are also ways to cut down on cheating. At the end of the day, the learner themself is responsible for their actions and how they choose to show their integrity. However, when learners are well prepared for assessment, there is less temptation to find ways to cheat. Good teaching equates to good learning.

References

Baig, M. I., Shuib, L., & Yadegaridehkordi, E. (2020). Big data in education: a state of the art, limitations, and future research directions. International Journal of Educational Technology in Higher Education, 17(1). https://doi.org/10.1186/s41239-020-00223-0

Cope, B., & Kalantzis, M. (2016). Big Data Comes to School. AERA Open, 2(2), 233285841664190. https://doi.org/10.1177/2332858416641907

Jimenez, L., & Boser, U. (2020). Future of Testing in Education: Artificial Intelligence (pp. 1–8) [Review of Future of Testing in Education: Artificial Intelligence]. Center for American Progress. https://www.americanprogress.org/wp-content/uploads/sites/2/2021/09/Future-Of-Testing-In-Education-Artificial-Intelligence1.pdf

What is Natural Language Processing? Definition and Examples. (n.d.). Coursera. https://www.coursera.org/articles/natural-language-processing