The case for A National Flashcard Platform

Here is a Feature Post from Mathias Maurer about his idea for a National Flashcard Platform.

Feature Post

A National Flashcard Platform 

Hwb offers Welsh pupils access to unique digital learning opportunities, but its true potential for supporting the learning of Welsh as a second language is yet to be unlocked. 

I would like to suggest the creation of a national digital flashcard platform, accessible through Hwb and tailored to the needs of English medium pupils in Wales. 

Context: One million speakers

In the Cymraeg 2050 strategy paper, the Welsh government sets out the vison of one million Welsh speakers by 2050. The strategy consists of three interdependent themes: 

1. Increasing the number of Welsh speakers
2. Increasing the use of Welsh
3. Creating favourable conditions – Infrastructure and context

As a primary teacher in the English medium sector, my professional interest lies in particular with the expectation on English medium schools to contribute towards increasing the number of Welsh speakers. The official target for English medium education is that, by 2050, 50% of all pupils finishing English medium education should report themselves as Welsh speakers. This sounds significant, but the number is difficult to put into context, as I have so far been unable to find current corresponding figures that could act as a base line. (Maybe the uptake of A-levels in Welsh as a second language can act as an indicator for the popularity of the subject: in 2015/16, 1.3% of pupils in English medium education entered A-levels for Welsh as a second language, the target for 2021 is 3%. We might be some way off.)

It is clear that, in order to have any chance of achieving such an ambitious target, we must draw on the very most effective methodology and resources for teaching Welsh as a second language. In the following, I am going to outline how Hwb could facilitate the use of smart technology for supporting the teaching of Welsh as a second language.

Flash cards: using technology the smart way

I suggest a bespoke flashcard platform that combines the best features of commercially available retrieval platforms with up to date language acquisition theory and methodology, creating an effective and child friendly learning tool. High-quality pupil management and data analysis tools will allow teachers to adapt children’s individual accounts to their needs, and to monitor progress and time spent practising.

From my childhood days learning first French then English, to learning Welsh with Welsh for Adults, Say Something in Welsh and Cardiff University as part of the Welsh in a Year sabbatical scheme, I have always used flashcards. In fact, most successful language learners of my personal acquaintance are using flashcards following the Leitner system in one way or another. While a few still sort actual cards into real boxes, just as I did many years ago as a school child, many, including myself, have moved on and now embrace the opportunities technology offers in the form of digital flashcard platforms. The advantages are manifold, from portability to the use of highly sophisticated algorithms and the gleaning of formative data, to mention just a few. 

Daisy Christodoulou, author of Teachers vs Tech, has written about the potential for spaced retrieval platforms in teaching, and about her personal use of the flashcard platform Anki (my personal choice as well). Her posts and the many positive conversations I have had with colleagues and fellow learners about the topic, I have been encouraged to write this blog.

Here is what a purpose built flashcard platform could look like:

One platform for all

A national flashcard platform for all pupils learning Welsh as a second language. All pupils have a personal account which they can access through Hwb. This account stays with them for the whole duration of their school career. The platform follows the nationally agreed language continuum envisaged in Cymraeg 2050 and supports all learners throughout their school career, from KS2 onwards. One platform for all. 

A pure flashcard platform

Daisy Christodoulou identifies adaptive learning platforms that use a spaced repetition algorithm as one of the most effective ways to study. Most retrieval platforms like Quizlet and Memrise use multiple choice options or give various other clues that reduce the retrieval effort and make it easier to find the answer. 

Pure flashcard platforms on the other hand, like Anki, require the learner to remember or predict the back of the card without providing any clues. This is the option I consider most effective.

Hear the Welsh – say the Welsh

Both sides of the cards contain print and a voice recording, the front English, the back Welsh. The learners read and listen to the English side before working out the Welsh translation and trying to say it aloud. Only then do they turn the card over to check if they were right. For beginners in particular, it is important at this point not to just check the print, but to also listen to the Welsh voice recording, as it provides them with repeated exposure to correctly pronounced Welsh. This is essential for shaping their own pronunciation and internalising high frequency language patterns. 

Adaptive learning

After answering a flashcard, the learner decides where to sort it back into the queue. The platform will strike a balance between using a highly sophisticated algorithm, and keeping the sorting procedures simple and child-friendly. The basic principle is this: a card the learner needs to learn or re-learn goes back to the front of the queue and returns after only a short interval. Cards they are more confident with go to the middle or the end of the queue, and return after medium or long-term intervals respectively. Retrieval intervals grow with increasing competence. 

Interleaving and cumulative learning

As the learner progresses, old and familiar patterns and vocabulary are revisited and interleaved with new material, guaranteeing spaced revision and cumulative learning. In this aspect, the principle is similar to the successful Say Something in Welsh method.

The crucial difference is that the platform is adaptive and reacts to the learners’ performance.  

High frequency chunks

Whilst having single words on each side might be an effective way to learn individual words, recalling single words and grammatical conjugations in isolation does not develop speaking fluency. New words should be introduced and practised in a meaningful context, as part of a sentence or a key communicative pattern (chunk). 

Most of the cards on the platform will use full sentences and chunks combined with interchanging relevant vocabulary. 

Allowing for young learners’ needs

Making the increased cognitive effort of retrieving the back of the card without clues leads to better long-term memory and stronger retrieval paths, but it also comes with the price of higher effort and frustration levels. This can be a challenge for primary school age children in particular.

I therefore suggest using the flashcard platform only from Key Stage 2 onwards, and only if the continuous investment is made to train and monitor the children in the effective use of  it. Most children do not have the same self-regulatory proficiency as successful adult language learners, and for many, the temptation to race through the cards by being disingenuous about their performance will be too great a temptation to resist without support. 

Effective pupil management

Effective pupil management tools are essential. For each pupil, teachers will be able to monitor time spent practising and individual progress. Teachers will have the essential option of setting additional cards, or reducing the amount of cards according to pupils’ prior attainment and progress through the continuum. This procedure will be quick, straightforward and easy to use for all teachers. 

Achievement awards: a different approach

Most educational retrieval apps offer achievement awards based on the learner’s progression through the learning material. In many cases, learners can speed up their progress towards the awards by choosing to undertake additional practice. Often, this comes at the price of shortening the time space between retrieval attempts, and forfeiting the benefits of spaced retrieval. This is not desirable.

In my opinion, achievement awards can play an important role in enticing and motivating young learners, but only if they reward for time spent practising, following fixed retrieval intervals, and if there is no option of shortening the time space between retrieval attempts. 

The platform will provide child-friendly rewards for time spent practising, rather than for achieving progression targets. 

Pre-empting frustration: study breaks

A word of warning: Anki can be a bully. Whenever I have not been practising for a few days, I am greeted in the morning with a long list of cards to work through. This is not practical for young learners. 

The flashcard platform will offer a pause feature or a way of limiting practice, so that learners who do not have the means or motivation to work on weekends or through school holidays can pick up where they left once back in school, without having to deal with a pile up of cards. I believe this to be a reasonable compromise. 

Putting in additional work at home will be a voluntary option, not an expectation.

What’s on the market?

In order to make this idea work, I consider all above specifications as essential. I have spent considerable time on market research, but I am yet to find a commercial platform that offers everything in one package. Anki is a strong contender, but it falls short on child-friendliness and pupil management options. 

The strongest tool comes from Germany: Phase 6 (https://www.phase-6.de) allows pupils to choose the language textbook their school is following, and then provides them with matching vocabulary flashcards, including audio, using an effective spaced repetition algorithm. Unfortunately, there is no class management tool, which renders it unsuitable for our purpose. 

A case for a prototype

This blog contains a lot of technical information. I believe that a fully functional prototype would be a better way of demonstrating the strengths of the flashcard platform I propose.  

I therefore am considering building such a prototype in collaboration with my brother, Andreas Maurer, who is a German IT specialist with considerable programming expertise (currently with some spare time on his hands due to the Covid19 situation.)

Our aim is clear: one platform for all, following the nationwide agreed language progression envisaged in Cymru 2050. Pupils access their account through Hwb and keep it for the duration of their school career. 

We are still planning. If you yourself have specialist knowledge in any of the domains this blog draws on, and would be able to offer support in the form of constructive feedback, suggestions or collaboration, then please get in touch via direct message @MaurerMathias.

Thank you!

Diolch o galon

Through exploring several pathways whilst learning to speak Welsh myself, I have encountered a rich variety of methodologies and resources. These experiences have fed back into my own teaching of Welsh as a second language to primary children. 

Although the skills and experiences acquired along the way give me a broad perspective and good understanding of the different domains contributing to an effective language platform, I am equally aware of the limitations of my own knowledge. 

I would therefore like to take this opportunity to thank the following fabulous people for all I have learned through them: Barri Mock @BarriMoc,  Dr Gianfranco Conti @gianfrancocont9, Daisy Christodoulou @daisychristo, Damian Benney @Benneypenyrheol, Lowri Davies @davieslw_lowri and her team at @ysgolygymraeg , Aran Jones @aranjones from @DailyWelshWords, Emma Dermody @EmmaDermody1 from @CSC_Cymraeg, Andreas Maurer (serious programmer – and my big brother!) and Sonia Maurer  @SoniaMaurer (Learning Technologist, Flashcard pro and wife to the author). 

All of you have far deeper specialist understanding in your domains than I have, and will no doubt have identified aspects in need of correcting or challenging.  Please do so! We welcome your feedback and support.  

Mathias Maurer