How We Recreate the Idea of Educational Content [Part #2] | Hacker Noon

Author profile picture

The story “How We Recreate the Idea of Educational Content [Part #1]” you can read here.

Let’s continue!

Creating Digital-First Content

1. It all starts with research, quantitative and qualitative. We interview students and teachers, conduct in-house and outside polls, collect feedback after each lesson, analyze most frequent customer support requests and customer feedback on social media. Then we conduct UX tests and work with product heat maps. 

2. Then we create a roadmap for the course and two or three pilot lessons on the platform. After getting feedback, we make adjustments or reject the course. The feedback from both students and teachers is important as we view teachers as our clients too. The test period usually doesn’t take more than a week. In that week, we can test the course with thousands of students, dozens of thousands if needed. We can specifically choose new or old students or teachers, depending on our goal. Only after all those tests are completed, we start working on the main body of the course. We can make adjustments in the content and/or teaching methods based on the feedback. 

3. The next step is wave improvements (WI). We monitor these metrics weekly:

  1. likes/dislikes ratio for lessons and homework
  2. slide rating inside lessons
  3. common mistakes and their frequency
  4. % of completion for lessons and homework
  5. feedback on the content

We calculate all those metrics as we go. Slides, lessons, and exercises with low scores are excluded and sent to WI.

4. We also correct mistakes overlooked by proofreaders and layout designers (no one’s perfect) — we don’t have to wait for the next reprint.

Tuning and Setting

The right level of difficulty is key. If a task is too easy or too hard, students lose interest and motivation. We use a CMS (content management system) to adjust the difficulty.

After the first few lessons, methodologists can see how often students make mistakes in each task and adjust them accordingly. We log all students’ attempts with the exercise and analyze them both to improve our teaching methods and customize content for this very student.

Content producers collect feedback from students and adjust topics, ideas, and dynamics. In 2–3 months we can create and customize a course including all supplementary materials for teachers.

Right now, we’re working on a recommendation system that would suggest appropriately difficult content for each student and predict the score for the next task.

A Pinch of Customization

We have a pool of tasks that can be used as homework. Teachers choose tasks individually for each student. In our newer courses, students can choose how much time they want to spend on homework. That improved our homework completion metric by 5 pp which means students do more homework and are more satisfied with the results. Nothing motivates better than knowing you’ve done a great job. 

We’re also testing a script that would analyze student’s mistakes and suggest exercises to improve lacking skills. Teachers see those suggestions too and can add those exercises to the lesson right away. Another project we’re testing is a predictive model that would suggest content based on the predictions about the score for the next task. This may allow reaching the education goal 15–20 % faster.

We also employ speech recognition to analyze the mistakes students make while speaking. Now we can detect an error with the article and tense usage and some others. But there’s one problem we cannot solve yet: Speech to Text software corrects most mistakes automatically. If you know how to deal with this, we’re open to ideas.

How we choose the educational track

In a traditional school, you’re just told to learn something. You cannot ask why. “No pain no gain” and “Just push yourself’ are some common phrases in academia. School students don’t understand why they have to learn this or that. But our students always have a concrete goal — they want to travel, or go to college abroad, or pass a language exam, or they want to watch TV shows in English.

It’s our goal to choose the best course and educational track inside of it that would fit this very student. Students never need to “learn everything in English”, they want very specific results.

Let’s take an example. You need to take IELTS to go to college in Australia. In most cases, you won’t need the highest score — 6.5 might be enough. Those 6.5 points are comprised of four scores: for Reading, Listening, Writing and Speaking.

Now you have to choose your strategy:

  1. You can polish each skill to perfection. That’s what they do in traditional schools, but it’s the longest and most expensive strategy of all.
  2. You can find your weakest skills and improve them.
  3. Or you can work on skills that are in the yellow zone. Some skills will remain in the red zone but they won’t affect the score. Our colleagues from Arizona State University used this approach when developing courses for GMAT preparation.

Another example: you want to move abroad and start working there as soon as possible, You don’t need all 12 English tenses to that, 4 is enough. And 800–1000 words will cover 95% of your daily conversations. If you learn another 100–200 needed in your line of work, you’re all set. 

We prefer this approach — showing the shortest way to the result. It’s not our goal to sell to more courses. As soon as the student reaches the goal, we also do. We’re partners. We know that after achieving one goal students will have another as the language needs some upkeep. That’s lifetime learning. And if students see the result, they’ll come to us again.

We’re not afraid to give tips on how to save money. We say, “Look, to get to the next level, you need to study for 100-120 hours. You can do all of them with a teacher or you can do homework and save 40%. If you watch movies in English, you’ll save another 20%” and so on. 

Tags

The Noonification banner

Subscribe to get your daily round-up of top tech stories!

read original article here