To measure how effective remote teaching has been with Cerebry, it is helpful to break the results into the most important components:

Ease of conducting video-teaching:

 A quick look at the number of remote classes that teachers launched from within the dashboard and students’ recorded attendance vs. teachers’ records will be able to indicate if video conferencing integrations within Cerebry were easier for both teachers and students to use as compared to managing the classes with a more manual offline process.

For math, Cerebry allows teachers to look at question types where a significant number of students are stuck and generate problems corresponding to those types to solve in front of the class. Teachers struggle to write down the problem and solutions steps on digital whiteboards. Cerebry’s engine also generates a step-by-step solution for each problem which can save teacher the time in writing down the entire solution. The teacher can just share his screen and walk through the generated solution. If the number of questions generated by the teachers on the dashboard is high, it means they are availing of this functionality and Cerebry is helping them with remote video teaching.

Seamlessly bring math homework process online:

Number of problems solved by the students in the pilot section vs. number of problems solved by the students in other sections of the same class will help us understand how easy it was for the students to practice on Cerebry vs. other alternatives. Teachers consistently report checking homework as the biggest problem with remote teaching. Most digital solutions only allow for MCQ type or short answers, while Cerebry accepts complete algebraic expressions, equations etc. just like student write in an exam. If students in the pilot section are doing more homework problems, Cerebry is doing a great job in helping students do their homework online easily.

Setting homework and having a platform to do it on is one piece of the puzzle. The other more important pieces are students reinforcing what was taught to them when they practice in the form of completing homework and teachers checking homework to see where the students need their help. For practice to be effective, students need to work out the homework problems on their own. To stop students from copying solutions, Cerebry generates a unique question for every student. We will retrospectively calculate the number of repeated questions given to students. If this number is 0 or negligible and the number of problems solved by the students is greater than in the other sections, it means Cerebry has been effective in enabling students to do their homework seamlessly and on their own.

For teachers, a challenge in teaching remotely is having to cater to all the different areas where students are stuck. An ideal remote teaching solution helps solve most of students’ doubts and only surfaces doubts which are persistent and shared by a large number of students. Cerebry’s AI engine tries to help the student progress even when they enter a wrong answer by showing the student step-by-step solutions, getting the student to solve similar problems, breaking the question into component steps, displaying content, etc.. We will measure the number of instances where students gave a wrong answer but were able to proceed eventually (by correctly answering two or more similar but different consecutive questions). If the number of these instances is a large proportion of the cases where the student had a misconception and thus entered a wrong answer, the engine is effective in remedying most misconceptions without bothering the teacher.

Adapts to individual needs of	every student & skyrockets class Maths performance:

Students have different levels of proficiency for each topic. Students who are struggling need more practice while students who are proficient need less. We can calculate an adaptive score: number of problems solved by the slowest student to finish a sub-topic / number of problems needed by the fastest student to finish the same sub-topic. The larger this score, the more adaptive the solution is. In most programs, this number is around 1 or at most 2. For Cerebry, this number is usually > 5.

We conduct a timed mock test for each topic or group of topics. The questions in this test match the level of difficulty of the final exam and are decided in consultation with the teachers. We will measure the scores of the students in the mock tests. If students who practice on Cerebry consistently score higher, this means the program directly impacts students’ math performance in exams.

Makes Maths homework practice stimulating & fun for students:

Time spent, number of problems solved, number of distinct sessions where the student sat down to practice