Accelerated Learning using Gradiance

Two important elements of the learning experience are assignments and quizzes. They allow the instructor to control the progress of individual students as well as allowing the students to asses  their level of understanding.

Still, traditional assignments and quizzes suffer from several drawbacks:

  1. they might introduce unnecessary stress for students
  2. they require instructor’s time and effort for evaluating the results
  3. they are hard to protect against plagiarism
  4. multiple submissions with detailed feedback are hard to support

At the same time, once done and evaluated, they offer little help for the students to improve their knowledge.

The Gradiance accelerated learning tool was developed in Stanford in order to solve the problems raised above:

  1. it uses innovative technology to automatically test answers to long and complex questions
  2. it gives automatic and informative feedback to students the moment they have finished an assignment or lab
  3. it provides, in addition to assignments, also lab environments which allow students to test skills like programming
  4. they support multiple submissions by changing the answer sets for the same questions
  5. they prevent plagiarism by creating individual answer sets for each student
  6. they support various text books in computer science and mathematics with predefined questions and lab assignments.
  7. new courses can be defined by uploading questions in a certain format
  8. feedback includes references to the course textbook

The normal workflow with Gradiance is for the students to complete an assignment on material seen in class. On submission, the students receive immediate feedback which includes hints of how to correct their answers and references to sections in the textbook (or slides) which elaborate the material encountered. At this point the student improve his understanding of the errors and is allowed to retake the assignment. This time, the assignment will contain different selections of multiple choice answers and the students must, in order to improve their grade, fully understand the problems at end. The goal for each student is to achieve maximal score.

Similarly, Gradiance also contains lab settings for evaluating students practical work like programming and SQL code. The same procedure of immediate feedback and instant evaluation is present here as well.

Information about the ideas behind Gradiance can be found on their website.

Please follow this tour in order to see a demo of the functionality described above.

I have used this tool as an aid for teaching a Database course and the feedback from some students was good. Students have taken the exercises multiple times and some achieved maximal score after several iterations. In addition, the system freed me from the need to constantly check, grade and give feedback to multiple submissions for many students. It also allowed the students to solve the exercises in their own time and decreased any stress which might have accompanied some exercises. Lastly, the system gave the students an opportunity to practice with writing SQL queries and to get an immediate feedback for their work. A thing which is hard to achieve under normal circumstances.

I would like to conclude this post with a class token for a  programming with Java course which I have created as a demo to accompany this post. You will have to register here and after logging in, enter the token C3B505BB .

This demo includes one assignment and one lab about loops and iterations. The class follows the textbook Introduction to Java Programming by Y. Daniel Liang. Below are screenshots of the various assignments, labs and handouts available for this book in Gradiance. Supported textbooks are:

  • Databases (ElMasri-Navathe or Garcia-Widom-Ullman)
  • Compilers
  • Automata
  • Discrete Math
  • Java Programming
  • Data structures
  • Data mining (Tan-Kumar-Steinbach or Leskovec-Rajaraman-Ullman)

Additional textbooks can be supported by uploading new questions to the system.