After testing the prototype with several teachers and showing to several schools personnel it was time to extract our learnings and make designs for a first version of the service. We adjusted, changed and extended our designs in Figma to the learnings from the prototype and to make it more ready, to be able to start development. 

Goals

A main goal with the service is to allow teachers to offer learners a Duolingo-like experience with bite-sized learning content to help learners find time to do their homework and gamification and nudging to help motivate learners to continue practizing. We have earlier written about learners' challenges with homework and it is a goal to address these.

The second main goal with the service is to help teachers. We want to gather essential tools in one service and add real benefits. We want to be a tool to create and distribute exercises in bite-sizes, but we also want to help create or get those exercises by use of automation, AI and sharing with each other. We don't want to be another flash card service, where the teacher always have to add all content.

A supporting goal is to make the system understandable and easy to use, and remove the complexity of traditional LMSes. In addition, the service must work well on mobile. In our interviews and explorations, we see that teachers mostly use laptops to create and distribute material with their students and to do other communication, while the learners often use their mobile devices - unless the LMS doesn't work or is too difficult to use on mobile.

One app

We considered whether to create a single app or to create different teacher and learner apps. We chose to pursue a single app, which may introduce some user experience challenges. However, we believe that the teacher in many cases shall be able to see the materials similar to how the learner experiences it, so most of the learner functionality shall be present in the teacher app. We also want the teacher to be able to use their phone, so it would not be restricted to say web. Although teachers may prefer to use a laptop and have more space when working, we still wanted it to be possible to quickly make changes and additions without having to bring out the laptop.

For the development we chose Flutter, a cross-platform framework that allows us to create iPhone, Android and web applications from the same code. This may pose some challenges as there may different design patterns and ways of doing things in mobile and web apps. However again, as most of the functionality will be the same, we prefer to face these challenges and try to find solutions.

Navigation

With the prototype we considered that a learner will typically have a single course and we chose similar to Duolingo that the navigation should be centered around a single course and then having a selector to switch between courses if the learner actually has more. Not explicitly exposing courses or more general choices in the "top" navigation helps on mobile with restricted space. It may also help the learner by reducing distractions and choices.

With the prototype we also considered that the teacher need to be able to operate at a higher level than the individual course. The teacher may want to create new courses, archive all courses at the end of the semester, or copy a course to adjust it and run it again next semester. Such operations seem difficult from within a single course. It may be doable, but it feels like if I would always be inside a file on my computer and there was one button to select a new document, but no file explorer to work with files without being inside.

We created the previous prototype in a desktop format and thinking it would be possible to adjust easily to mobile with the choices and perspectives above. However, when starting development we ran into a few challenges. On mobile it is a common pattern to have a bottom navigation bar, which we had also used in our designs. However, it is an anti-pattern to change the bottom navigation for different views in the app. The outer navigation - or file-explorer like views for working with courses - do not fit inside a specific course and when not part of the bottom navigation, the navigation items shall go somewhere else and these views will have to be "modals" where the bottom navigation is not present and the user has the option to go back or close the modal to get back to inside the course. 

We chose to put the outer navigation items in a "drawer" that can be drawn out by clicking on a profile icon and to have the views be modals where the bottom navigation is not shown and the user needs to navigate back to get to the other navigation options. 

The three screenshots above show three mobile views with a focus on navigation. The first shows the regular view when "inside" a course with the navigation bottom bar being oriented around the course and with a course selector at the top to change to other courses. The remaining navigation options can be seen in the drawer in the second screenshot. The last screen shot shows when one of the outer navigation options have been chosen; the bottom navigation has disappeared and the only navigation option is to navigate back.

On larger display sizes we chose to put the navigation in a left side menu, similar to the drawer, but also showing a course selector and the navigation options for a current course. In this way the outer views become visible for teachers and at the same level as the individual courses. As such there is a difference in the navigation on different display sizes, but it seems to make sense as the teacher will mostly be using a larger display while the learner will mostly use the smaller display size. 

Learning paths

For teachers we want to create a "home" screen that gives access to the most relevant information like upcoming courses, the most recent courses worked on, communication, and in addition introduces help to get started with Edumo as well as various features. However, for the beta we decided on just placing the teacher on Drafts, which shows the things the teacher is working on but hasn't shared with anybody. In addition, we introduced two other outer views; Active, for courses that have been shared with learners and Archived for archiving course and material when it is not being used anymore, but the teacher still wants to just store it for later reuse.

We want teachers to be able to create single exercises to full Duolingo like courses. This allows teachers to start small and then learn and use more over time. For teachers having individual learners rather than classes, we hope they will create a "course" per learner and add material, so the learner still gets the Duolingo experience, where they can just open the app and get started on homework and can see previous and maybe future homework. However, we also expect some teachers will prefer to just create individual materials and distribute them individually, i.e. not as part of a larger course.

Similar to Duoling, we chose that a course in Edumo may have a grouping of content, similar to chapters in a book, and under each chapter there may be the 5-10 minute sessions, which holds the individual exercises or learning material. in Edumo, we have chosen the term Learning path for a course, Step for a session, and Material for the content of sessions. As mentioned above we want a teacher to be able to distribute individual materials, but also package some materials into a few steps or even into a few segments. And rather than having different terms depending on how much is distributed, we have chosen that everything is a learning path. A learning path may contain a single exercise or instructional material or group more material into steps and/or segments. 

In the outer views all the paths are shown together without distinction. We hope to learn from the first beta testers, if this is natural or it makes it difficult to find the learning content. Some ideas could be to add different icons on additional information about what a path contains in the lists of paths. As the terminology may be new and cause some confusion, we provide a start view that include a bit of information that we hope will be enough and then options to start planning the path in segments and steps or get starting just creating materials.

One difficulty we may not have solved yet, is just letting the teacher get started on creating materials without also giving a name and other information to the paths. If skipping this step the path - or materials - will just be named Untitled or similar. We are considering wether we can use generativ AI to summarize a name from the contents of the materials, but hope to get some indications of the severity of the issue and ideas for a solution.

Working with learning material

We obviously want it to be easy to work with the learning material. We find it important to be able to quickly regroup and rearrange segments, steps and materials by dragging them around, and selecting multiple items and making changes in bulk. However, we faced some challenges and prioritized getting something finished so we can get more learnings from beta testers and assess which are the most important things to fix first. It could be that it is some other limitation or something that we have been unaware of that is more important than this.

For the beta we managed to create a simple type of material (multiple-choice exercises) that the teacher can fill. We obviously want more types of content from text and video to other types of exercises. However, we will work on adding these while hopefully somebody will be beta testing and find other issues that may be more important to get fixed.

As mentioned earlier it is a main goal of Edumo to help teachers create or get content. As such we would have loved to have some first examples of this ready. We would have loved to help create suggestions for course materials or help generate exercises so the teacher sdoes not have to do this themselves, but can just specify and review. It is essential for Edumo to give such benefits to the teacher, so it is not just a another service where they can fill in their materials. However, as mentioned earlier we still think it makes sense to get some beta testing done while we continue testing. 

Doing homework

As mentioned in the beginning a main goal of Edumo is to give learners the Duolingo like experience in terms of a visible learning path with bite-sized content and gamification and nudging. For the beta version, we have developed the visible learning path and the ability to add content in small bite-sized sessions (steps).

The learner will see the learning path when opening the app and will be able to click the step that is highlighted as the next and then get started doing homework. We want this to be fast, so the learner can practice in breaks throughout the day, where the learner might otherwise read news or scroll social feeds. This is quite important, as all language course participants we have interviewed indicated that they always do all their homework in one setting as it requires too much preparation to be feasible to do just 5-10 minutes at a time. Getting textbooks out, logging into LMSes, finding mails with indications of exercises or whatever was needed to get ready to do homework was perceived as taking too long to not do all in one sitting and the reason for doing it in one go. We have talked to many teachers that actually promote practicing or immersing into the language a bit every day rather than doing a larger bit one day. 

As can be noticed, we have gobe with a design that resembles Duolingo although in our own colors. We didin't manage to get the gamification parts in place in terms of points streaks and so on. From talks with teachers, the exact same type of gamification as Duolingo may not be desirable. Another part that may differ is that classes indicated in the path as the bars (segments) will follow time wether or not the learner has done all his homework or not. Our current thoughts is to color the bar representing the next class on the day it happens, but still only color the circles (steps) if they have been solved.

We are thinking to only allow the learner to do the next homework session (step) and past homework. In essence the learner cannot skip ahead - or is at least not encouraged to. This is to not have the learner start asking questions before the class gets to the topics. If a learner has finished all the steps in one segment, but it has not become time for the next class, we also consider not allowing the learner to progress to the next segment. Instead, we will create a new repetition step, which repeats some of the previous material in a smart way. The "smart" way will ensure to include material the learner has struggled with getting right, but perhaps softened with some materials that the learner has gotten right before. The repetition may include recent material, but also have spaced repetition of material from previous segments. This help the teacher by stretching the material created to last longer, but also the learner to actually perform this kind of repetition that is important for language acquisition as well as retaining things learned in earlier lessons.

Next steps

We will continue to wrap things up and work on parts that we left out, but have some ideas on how to do. At the same time, we will also try to get foreign language teachers to use the beta version either in a test or in real classes. We want feedback and input on what is most important to implement next and ideas for how to address some of the questions we still have.

When we have developed some of the functionality that may be most important for the beta testing teachers, we will open up to more teachers. It will probably still be an early version and for teachers that are eager and willing to use it although still missing things and not being very polished. 

If that sounds like you and you want to help then contact us on one of our social media profiles below or sign up to our news mail and reply to me there.