Organizations in the education sector such as the Institute of Education Sciences have been using traditional statistical analysis to drive policy and administrative decisions. More recently, innovative data science techniques are being applied to improve content delivery, student learning and affordability. This blog post highlights three companies at the forefront of these innovations: Coursera, KnowRe, and Knewton.
Coursera is an educational platform that provides massive open online classes (MOOCs) to users worldwide for free. The company partnered with top universities, such as Stanford, Yale, and Princeton, and other global organizations to fulfill the goal of making higher education a “basic human right” (Bast, 2012). The four key ideas that assist Coursera in its vision include online learning effectiveness, peer assessment, mastery learning, and blended learning (Coursera, n.d.). These components form the backbone of the Coursera experience, allowing the platform to offer over 500 courses with more than 22 million enrollments from students in 190 countries (Coursera, 2014).
Coursera primarily utilizes data analytics through personalizing the learning experience to facilitate mastery learning. Mastery learning involves ensuring a student understands a topic before moving on to the next section. The immediate feedback of the online platform as well as randomized versions of assignments makes mastery learning a reality. In addition, professors are able to pinpoint the most common misconceptions and their root causes, allowing them to send targeted error messages to students. In addition, if students possess limited experience with regards to a certain topic, additional content are made available to them, allowing students of varying skill levels to take the same course (Bast, 2012).
Recently, Coursera has also been experimenting with keystroke biometrics in their Signature Track verification system. Through this identification verification system, Coursera offers their students the opportunity to receive a verified certificate for a given course to show their employers and help them further their careers (Coursera, n.d.). The system analyzes the user’s keystroke characteristics, such as rate and press duration, to verify user identity to ensure that the individual receiving the certificate is the one completing the assignments and quizzes. In addition to keystroke biometrics, the user’s facial image via a webcam is also matched to his/her photo ID for further validation (Young, 2013).
KnowRe is an educational platform that provides online adaptive learning for students in mathematics. Their main focus is to provide a personalized learning experience for secondary student. KnowRe’s approach is to first assess an individual’s strength and weaknesses and then personalize a curriculum for each student depending on their own personal needs. KnowRe’s uniqueness comes from their interaction to students through Gamified Learning. This type of learning attracts students’ attention through interactive game-like fun and beautifully enhanced graphics (KnowRe, n.d.).
KnowRe also includes a teacher dashboard in which teachers can easily see the progress of each students, hand out assignments, and discover what areas their students are struggling in. KnowRe also provides automated grading and very accurate assessments of each student so it makes real-time data actionable for teacher (KnowRe, n.d.).
KnowRe’s mission is to be able to provide students with the most personal and engaging learning experience to every student worldwide. They do this by trying to figure out the strength and weakness of each student and then personalizing a curriculum that is uniquely designed to help them be their best. They also want to provide the most interactive learning to students such that the students will continue to actively pursue learning in the future (KnowRe, n.d.).
Knewton adopts a service oriented model that provides tools to enable adaptive learning. Under the hood, Knewton is similar to Coursera and KnowRe in that it uses data gleaned from student performance to answer questions about ongoing student proficiency, content efficacy, and suitable next steps (Knewton). These underlying data collection and inference engines are made available to partner institutions such as Pearson Publishing (Pearson Ed, 2012) and Arizona State University (Knewton, 2011). The result is that traditional educational institutions can merge their content into an online adaptive learning experience for the student, educator and content creator.
In its simplest form, Knewton’s software evaluates student proficiency at the question level (rather than the aggregate subject level which is common in schools and universities). This allows the software to study individual performance for each question and hone in on specific concepts and educational material which are correlated with good versus bad student performance at a more aggregate level (Knewton Developer Blog, 2012). Patterns and information gleaned from these connections are then served to the educator, student and content creator in the form of recommendations that will improve student performance (Knewton).
The challenge Knewton tries to solve, similar to companies like Coursera and KnowRe, is to perform this task at scale across disciplines and across students. Thus, the models driving the learning platform need to maintain generalizability without sacrificing the ability to provide personalized recommendations to assist students who do not conform to group trends (Knewton Developer Blog, 2012). In addition, from a technical perspective, there is a computational challenge associated with constantly updating data models to account for new student activity while delivering real-time recommendations based on the data models (Knewton | FAQ, n.d.).
Traditional Institutions
In addition to private companies, formal educational institutions have also taken an interest in leveraging data science techniques to improve educational content delivery. For example, Carnegie Mellon’s Open Learning Initiative (OLI) “brings together evidence-based research in learning, science, and technology to create web-based learning environments” (Soares, 2011). OLI’s homepage lists a number of published studies and reviews of adaptive learning platforms to facilitate ‘life-long learning’ objectives for students and educators alike. The challenges faced by companies like Coursera, KnowRe and Knewton are just as applicable for OLI. In addition, traditional institutions have to convert the insights from these systems into actions on the part of users. This last piece can be challenging in an educational environment where the primary content delivery mechanism, the textbook, has essentially remained unchanged since the 19th century (Morris, n.d.).
References
Bast, M. (2012, June). What we’re learning from online education. TED. Retrieved April 2, 2014, from http://www.ted.com/talks/daphne_koller_what_we_re_learning_from_online_education/transcript.
Coursera. (2014, January 7). Community | Coursera. Retrieved April 2, 2014, from https://www.coursera.org/about/community
Coursera. (n.d.). About | Coursera. Retrieved April 2, 2014, from https://www.coursera.org/about/
Coursera. (n.d.). Signature Track Guidebook. Retrieved April 2, 2014, from https://www.coursera.org/signature/guidebook
Knewton | FAQ. (n.d.). Frequently Asked Questions. Retrieved April 2, 2014, from “http://www.knewton.com/platform/faq/
Knewton. (2011, January 6). Arizona State University Partners with Knewton. Retrieved April 1, 2014, from http://www.knewton.com/about/press/arizona-state-university-partnership/
Knewton. (n.d.). About Knewton. Retrieved March 30, 2014, from http://www.knewton.com/about/
Knewton. (n.d.). Adaptive Learning Whitepaper. Retrieved March 20, 2014, from Knewton: http://www.knewton.com/adaptive-learning-intro/
Knewton Developer Blog. (2012, June 7). The Mismeasure of Students: Using Item Response Theory Instead of Traditional Grading to Assess Student Proficiency. Retrieved April 1, 2014, from http://www.knewton.com/tech/blog/2012/06/understanding-student-performance-with-item-response-theory/
KnowRe. (n.d.). The KnowRe Story. Retrieved April 3, 2014, from http://about.knowre.com/about/who-we-are/
KnowRe. (n.d.). What Is KnowRe? Retrieved April 3, 2014, from http://about.knowre.com/
Morris, S. (n.d.). Sanity Check: Does Technology Have a Role in Our Classrooms? Retrieved April 2, 2014, from Lenovo Blog: http://blog.lenovo.com/education/sanity-check-does-technology-have-a-role-in-our-classrooms
Pearson Ed. (2012, January 11). Pearson and Knewton to Bring Adaptive Learning Technology to Thousands of Underprepared College Students. Retrieved April 1, 2014, from http://www.pearsoned.com/pearson-and-knewton-to-bring-adaptive-learning-technology-to-thousands-of-underprepared-college-students/#.Uz2loKIvmGQ
Soares, L. (2011). The ‘Personalization’ of Higher Education. Center for American Progress.
Young, J. R. (2013, January 9). Coursera Announces Details for Selling Certificates and Verifying Identities. The Chronical of Higher Education. Retrieved April 2, 2014, from http://chronicle.com/blogs/wiredcampus/coursera-announces-details-for-selling-certificates-and-verifying-identities/41519