We initially interviewed a little more than 10 language teachers to get a better understanding of the domain and how our idea fit into their work. We developed a prototype based on the knowledge gained and inputs and then tested it with with around 10 language teachers again. Some were the same teachers we interviewed and some were new. The language teachers were mostly teaching at the Danish evening schools, but many had also taught at Danish high schools, commercial schools or privately. In the following we describe some of the learnings we got from the interviews and prototype tests, especially related to preparation.
One of our first learnings, was that the teachers interviewed were not interested in paying for the service we wanted to develop. They were all working for schools and reasonably believed that the employer should pay for such tool. Only a few indicated that they would pay themselves if the school was not paying.
Some of the teachers indicated that they had bought subscriptions to Quizlet, Wordwall or Kahoot. A single person indicated that she had been able to get some subscriptions reimbursed from some school, but not others and in general the Danish evening schools were seldom reimbursing such costs. Other teachers also indicated that they pay for their own textbooks. In public schools, teachers are full-time employed and the schools pay for learning materials, but we also talked to one public school teacher that purchased a subscription to Kahoot herself, as the influence on the learning materials the school (or municipality) chooses may be too difficult or take too long.
Another learning that we got from most interviewed evening school teachers is that they don’t believe they are paid much for preparation. Some indicated that this is the same with commercial schools and most other freelance teaching jobs in Denmark. We imagine this is probably different for full-time employed teachers like at Danish public schools or high schools, but they are also required to do more correction of student assignments and administrative work.
A few indicated that they use a little time on finding materials for their languages, but never create something themselves, as there is so much free material out there. One teacher indicated that she just use the textbook and don’t spend any time on supplemental materials and don’t even answer mails from students, as her salary doesn’t cover the time needed.
However, several of the interviewed teachers indicated that they spend significant time on creating their own content to make the teaching more relevant, up-to-date, interesting and/or personal. As a side note, when the teachers, making learning content talked about it, many mentioned that their spouse often told them to be aware that they are not paid for that much preparation. However, it seemed that besides making the classes more interesting or providing the students with more contemporary or relevant content, the teachers also liked experimenting with or working on their own content.
A challenge with some tools and services for creating learning content, is that they have great promise, but may require a lot of learning, practice or work to achieve that promise. As an example, we heard from a couple of teachers that they had tried to create escape games in Genially, but ended up giving up. In these cases the time spend ends up being wasted.
Some teachers told us that they come across texts that they think will be great for their learners, but then have to create a glossary list of new or the most difficult words, so it doesn’t get too difficult for their learners to read the text. A service can potentially keep track of vocabulary or detect a certain level of difficulty and help with a glossary or similar. A service may even use AI to suggest questions about the understanding of the text. Thus is done by existing services like Diffit and Nolej today.
We talked to some teachers that are very interested in using movies and short films in their teaching. Some of these teachers had manually transcribed videos and created learning content related to short films. A service can automate the transcription and suggest exercises for vocabulary and questions about understanding as described above. The mentioned services Diffit and Nolej offers some features to this end.
Many of the interviewed teachers write small texts, create small grammar exercises and/or create other learning content themselves to tailor the learning content to what their learners know and will benefit from practicing. They may do this by copying some existing content and adjusting it or by doing it from scratch. In either case, the teachers indicate that it can be boring, difficult to keep doing and/or just time consuming. So, there is also potential for helping with a lot of these tasks.
We learned about most of the challenges in the initial interviews and tried to address some of them in our prototype. We then tested with language teachers giving them small tasks to do in the prototype, including a task to use a small text they found and go through a flow that could create vocabulary exercises and questions about understanding.
The few teachers that did not spend time on creating learning content thought it was smart and useful, but the one that did not even answer emails from students, indicated that she would still not use the tool, as she would not be paid for the time to use it. Most of the other teachers were positive and impressed. We also discussed other types of help like for the small videos, and the teachers saw a lot of potential for saving time on these tasks.
It seems the service can help save enough time that it will be beneficial to use. Teachers may not be familiar with all AI services out there and keep updated on all of them. It may be difficult or time-consuming to stitch some services together, and in general it may also require some experimentation with prompting and continued effort to get the right results. Some of this may also be the case with a purpose build service for teachers, but it should be possible to tune it better and save some experimentation and following along and learning to use different AI services.
We hope and think it will be beneficial enough for teachers, so they will use it and also provide the benefit to their learners of bite-sized content that can be done directly on the phone as written about earlier in Challenges of Homework.