Cognitive decline is associated even with mild Covid – Can we now talk about indoor air hygiene?

Cognitive impairment occurs even after asymptomatic or mild Covid disease, as a study conducted at Imperial College London shows. Cognitive decline affects memory, verbal intelligence and spatial thinking in particular. Long Covid sufferers experience the greatest losses.

Whether it is demands for a review of the containment measures adopted during the pandemic or shouts of “Covid is over!” to people who still have to or choose to wear masks: It is hard to avoid the impression that Germany has now definitely entered a post-pandemic reality. Consequently, air hygiene measures are now also being scaled back. HEPA filters in educational establishments – if ever they did exist – are now mostly meeting one of the following fates:

  • Scrapping
  • Storage “until the next pandemic”
  • Lack of maintenance
  • Ban on use.

Where none of these apply, they are often simply being ignored. After all, children, teenagers and young adults are supposed to forget the difficult times they experienced during the pandemic. Air hygiene would only get in the way. Or would it?

Study reveals cognitive decline after Covid-19

SARS-CoV-2, the virus responsible for the pandemic, seems unconcerned: Wastewater monitoring by the Robert Koch Institute, Germany’s federal agency for disease control and prevention, showed an unprecedented peak in wastewater contamination with SARS-CoV-2 in December 2023. This went largely unnoticed, while Germans never ceased to be puzzled over the numbers of staff on sick leave. However, the virus still demands to be noticed, as even mild Covid leads to a decline in cognitive abilities, as shown in a study by a research team led by Adam Hampshire of the Medical Faculty at Imperial College London. Their article was published in the New England Journal of Medicine.

People with long-term impairments following a Covid infection in particular report concentration problems, word-finding difficulties and similar symptoms which suggest a decline in cognitive abilities, i.e. abilities related to a person’s intelligence. In their study, the researchers therefore collected data using the Cognitron battery, an intelligence test that measures verbal and non-verbal reasoning and planning as well as working memory. More than 140,000 people took part in the study and almost 113,000 completed all tasks.

Since the researchers were primarily interested in finding out whether cognitive impairments depend on the duration and severity of a Covid illness, the participants were assigned to the following groups:

  • Not infected with Covid
  • Infected with Covid, but asymptomatic
  • Infected, with symptoms that resolved after less than four weeks
  • Infected, with symptoms that resolved after four to twelve weeks
  • Infected, with symptoms that resolved after more than twelve weeks
  • Infected, with symptoms that remained unresolved after more than twelve weeks and were still present at the time of data collection (“Long Covid”).

Earlier variants caused more damage than later ones

The fact that the virus kept changing during data collection was taken into account by sorting participants into groups according to the variant prevailing at the time of infection:

  • Original virus: up to 30/11/2020
  • Alpha (B.1.1.7.): 01/12/2020 to 30/04/2021
  • Delta (B.1.617.2): 01/05/2021 to 15/12/2021
  • Omicron (B.1.1.529): 16/12/2021 to 31/03/2022, which marked the end of data collection.

Good news first: With the transition to the Delta and, later, the Omicron variant, cognitive decline decreased. While it was still 0.32 standard deviations (SD) for the original virus, it is only 0.16 SD with Omicron, i.e. half as much.

Results may be represented in points on the IQ scale

Standard deviations are, for instance, used in intelligence tests to represent results. Thus, the IQ scale is defined by a  mean of 100 and a standard deviation of 15. Hence, if you move 15 points to the left and 15 points to the right from the mean value, this defines the range of intelligence that is commonly referred to as “average”. Average intelligence therefore means an IQ of 85 and 115. Slightly more than 68% of the population attain an IQ between 85 and 115. Values below 85 count as below average, values above 115 as above average.

A loss of 0.32 SD with the original virus therefore means a decline in cognitive abilities corresponding to 4.8 IQ points. Accordingly, a decline of 2.4 IQ points is to be expected with Omicron.

Cognitive decline occurs with all degrees of severity

A decline in cognitive abilities can even be observed with asymptomatic infections (see Figure 1). A loss corresponding to around 2.6 IQ points can be expected here. With symptomatic infections, the decline is similar regardless of how long it takes for the symptoms to resolve: A decline of between 3.5 and 3.9 IQ points was observed in those groups. People affected by Long Covid face the greatest problem: The frequently reported issues with reasoning, word finding and memory correspond to a loss of around 6.3 IQ points.

Chart shows cognitive decline according to duration and severity of Covid-19
Figure 1: Cognitive decline according to duration and severity of Covid-19

Memory is one of the most severely affected cognitive abilities (see Figure 2). Although all people who have been infected with Covid are affected, Long Covid sufferers have to cope with the greatest losses regarding immediate and delayed memory.

Chart shows cognitive decline regarding immediate and delayed memory after Covid-19
Figure 2: Cognitive decline after Covid-19 – Immediate and delayed memory

The area of verbal intelligence is also affected (see Figure 3). While there are similar losses with asymptomatic infection and symptoms which have not persisted, Long Covid sufferers again show the greatest decline for both word definitions and verbal reasoning based on analogies.

Chart shows cognitive decline regarding verbal intelligence after Covid-19
Figure 3: Cognitive decline after Covid-19 – Verbal analogical reasoning and word definitions

Spatial planning is also impaired in all groups (see Figure 4). Again, Long Covid sufferers are those who have to live with the steepest declines, although all groups who have been infected with Covid are concerned.

Chart shows cognitive decline regarding spatial planning after Covid-19
Figure 4: Cognitive decline after Covid-19 – Spatial planning

Educational institutions must protect learners’ cognitive skills

What does this mean for education? Schools take pride in providing learners with the best possible support to match their cognitive abilities. This should include ensuring that those cognitive abilities are not knowingly jeopardised. It is therefore still important to avoid Covid infection wherever possible.

The city of Bradford has shown how this can be done: The use of HEPA filters reduced the number of Covid infections amongst pupils by more than 20 %. The number of lessons cancelled due to teachers on sick leave can be reduced in the much the same way. Air filters will also reduce the number of Long Covid cases amongst teaching staff. Thus, the quality of teaching can be maintained, as cognitive impairments will lead to changes in lesson organisation and classroom management, and will also increase stress in teachers.

Indoor air hygiene must become a normal fact of life if we wish to maintain the cognitive performance of the current and future workforce.

Hampshire, A., A. Azor, C. Atchison, W. Trender, P.J. Hellyer, V. Giunchiglia, M. Husain, G.S. Cooke, E. Cooper, A. Lound, C.A. Donnelly, M. Chadeau‑Hyam, H. Ward & P. Elliott (2024). Cognition and memory after Covid-19 in a large community sample. The New England Journal of Medicine, 390(9).

Teacher burnout is related to physical illness: “We rely on teachers to not give a *** about themselves.”

For years, the US has been talking about The Great Teacher Resignation. “The education crisis was very much here way before Covid,” says Anna Sutter, former teacher and director of school counselling in an opinion video by Agnes Walton and Nic Pollock, published by the New York Times. “I was spinning 100 plates, and I only had the capacity to spin 25 plates.” OECD data shows that even in 2016 – years before the Covid pandemic hit – sizable proportions of teaching staff were quitting their jobs. Germany, on the other hand, still considers itself shielded from the phenomenon. German teachers are well paid, right? Most of them are in tenured positions, civil servants even. What’s not to love?

Yet, just like other countries the world over, Germany is facing shortages of teaching staff. Some federal states have come up with a solution: Part-time options – frequently chosen by teachers to reduce the mental load and physical burden that comes with the profession – have been restricted. This will cost us dearly, because teacher burnout is related to physical illness, as a metastudy by Daniel J. Madigan, Lisa E. Kim, Hanna L. Glandorf and Owen Kavanagh of York St John University and the University of York, UK, shows.

Teacher burnout affects countries across the world

Burnout is a condition which develops when a person experiences chronic work-related stress. They will feel emotionally exhausted, they might become cynical as a response, and the passion which they once felt for teaching may wane. They feel less competent than they used to do. As a result, they get the impression that they get less work done.

To investigate how burnout is related to physical illness in teachers, Madigan and his colleagues reviewed 21 studies involving a total of more than 5,000 teachers. Teachers in the studies came from Brazil, Canada, China, Germany, Luxemburg, Poland, South Africa, Slovakia, Spain, Ukraine and the USA. Almost 60 percent of them were female. On average, they were 43.3 years old and had been teaching for 14.4 years. In some of the studies, burnout was treated as a categorical variable, which means that teachers were sorted into two groups according to whether they were experiencing a high or low level of burnout symptoms. In others, burnout was represented as a continuum. To measure burnout symptoms, most of the studies used the Maslach Burnout Inventory, which covers three dimensions:

  • Emotional exhaustion (feelings of emotional overextension caused by the workplace)
  • Depersonalization (an impersonal and seemingly unfeeling response to one’s profession and students)
  • Personal accomplishment (feelings of competence and achievement).

Burnout is associated with a variety of physical health conditions

As is the case for other professions, teacher burnout is related to myriad physical health conditions. Teachers with a high level of burnout symptoms suffer from headaches in general, and migraines in particular. They experience back pain, skin rash, and disorders of the digestive system, which may be as specific as gastroenteritis, an inflammation of the stomach and/or intestine. One of the more teacher-specific physical health problems concerns their voice. Some teachers may, for instance, experience a complete loss of voice. Teachers have also been shown to suffer from temporomandibular joint dysfunction, a condition which affects the muscles that move the jaw, and which can cause difficulties in speaking and eating, pain, tinnitus and dizziness.

Cardiovascular disorders, which concern the heart and blood vessels, have also been shown to be related to teacher burnout. Amongst the heart function indices used in the studies which were reviewed, burnout was positively associated with sympathetic activity. The sympathetic nervous system is responsible for increasing heart rate when responding to stressful situations. A positive association therefore means that teachers with burnout symptoms were displaying increased heart rates. On the other hand, burnout was negatively related with parasympathetic activity, which is responsible for controlling heart rate during times of rest. A negative association means that rest and recovery do not work well in teachers with burnout symptoms.

Physical health conditions in teachers with strong burnout symptoms
Physical health conditions in teachers with strong burnout symptoms

Teacher burnout affects their cortisol levels. Cortisol is a hormone which, in healthy people, regulates a person’s response to stress, their blood pressure, blood sugar and metabolism, which is the process that converts food into energy. Chronic stress has a two-fold impact on cortisol levels. Initially, cortisol levels will be raised as the body is struggling to regulate its responses. The longer this is required though, the less cortisol the body will produce, and cortisol levels will fall below what is the norm in healthy people. It is therefore unsurprising that teacher burnout was found to be associated with both above- and below-the-norm cortisol levels.

Burnout also affects cytokines. Cytokines are proteins which may be viewed as messengers between cells of the immune system. They are responsible for the body’s response to infection, inflammation and cancer. Teachers with a high level of burnout symptoms displayed lower anti-inflammatory responses and higher pro-inflammatory responses. In other words, their body fights disease less well and even contributes to the development of new inflammatory diseases.

Pathways to physical illness: Immune system, stress response, unhealthy behaviour

The researchers suggest three different pathways which might lead to the association between teacher burnout and physical health problems. As has been shown above, stress can lead to a person’s immune system being compromised. In addition, chronic stress may cause the body to become less effective in responding to stress in adequate ways. Last but not least, chronic stress may lead a person to engage in unhealthy behaviour more that they would normally do. For instance, they may smoke more and drink more alcohol, which, in turn, causes health problems.

Unsurprisingly, burnout and its associated physical health problems affect teachers’ personal lives in a variety of ways. They will influence their mood, general life satisfaction and mental health, creating long-term physical and mental health conditions. Interpersonal consequences include conflict, irritability and a reduced capacity for communication. The latter in particular will impact the quality of teaching and the wellbeing of students.

Ignoring teacher burnout will cost us dearly

The researchers emphasize that there is a need for more awareness of the consequences of teacher burnout. As burnout interventions, they cite mindfulness, cognitive-behavioural, social and emotional learning-based programmes. They do, however, add, that changes will have to take place at the organizational level as well: “[G]iven the potential cost of burnout for physical health and associated problems, it is likely intervention benefits will be significant, and possibly exceed outlay costs.”

Nevertheless, few countries seem to seriously consider this approach. “We rely on teachers to not give a *** about themselves,” says Anna Sutter. “We rely on teachers to not care about themselves. We rely on teachers to not know their worth and to not know what they deserve.” This ignorance will cost us dearly.

Madigan, D. J., L. E. Kim, H. L. Glandorf & O. Kavanagh (2023). Teacher burnout and physical health: A systematic review. International Journal of Educational Research, 119, 102173.

German teachers face more mental health problems than general public

A study conducted at the Medical Centre of the University of Freiburg found that the mental health of teachers in Baden-Württemberg is significantly worse than that of the population as a whole. The researchers’ conclusion: Better risk assessment and implementation of programmes to maintain the mental health of teachers.

They were looking for a suitable instrument to analyse the mental health of teachers. But they found much more: Teachers suffer significantly more from mental health issues than the general population in Germany. They may, for instance, feel constantly under strain, unable to enjoy day-to-day activities, or just not reasonably happy, as a study by Sarah Susanne Lütke Lanfer, Ruth Pfeifer, Claas Lahmann and Alexander Wünsch shows.

Collection of mental health data requires suitable tools

Previous studies from various countries had already shown that teachers feel weighed down by their workload, administrative tasks, inadequate breaks and negative experiences with learners, parents and colleagues. Mental health issues not only lead to absences, but also to a drop in the quality of education and relationships. The researchers have therefore developed the Freiburg Manual-Based Psychological Group Programme, which focuses on social support, rethinking one’s own coping strategies, practising relaxation techniques and learning about neuroscientific aspects. To measure success in the prevention of mental illness, the scientists examined whether the General Health Questionnaire-12 (GHQ-12), a questionnaire with twelve questions on mental health, is a suitable tool for professionals in education.

Representative samples for the general population and teachers

GHQ-12 has been translated into various languages and was tested for Germany in 2013 by Matthias Romppel, Elmar Braehler, Marcus Roth and Heide Glaesmer. More than 3,000 people between the ages of 14 and 93 took part in their study. The result was a representative sample of the German population. Around 4,000 teachers from the Freiburg Group Programme took part in the study by Lütke Lanfer’s group between 2012 and 2020, more than 80% being female. Different age groups, school types and life situations are also included in the sample. Hence, although the sample is not representative of the population as a whole, it may be viewed as being representative of teachers in Germany. GHQ-12 can therefore be used to compare the mental health of teachers with that of the population as a whole.

Increased risk of emotional exhaustion, depression and burnout

GHQ-12 collects data on mental health through self-assessment on a scale from 0 to 3. 0 stands for the lowest level of stress, 3 for the highest. When comparing the scores obtained by teachers with those of the population as a whole, it is noticeable that teachers are more stressed in almost all areas (see figure):

  • Teachers have a stronger feeling of being constantly under strain.
  • They find less enjoyment in day-to-day activities.
  • Overall, they do not feel reasonably happy with their lives to the same extent.
  • They are less able to concentrate on what they are doing.
  • They lose more sleep over worry.
  • They feel more often unhappy and depressed.
  • They are under the impression that they are less able to overcome difficulties.
  • They often feel that they are unable to play a useful part in things.
  • They find it more difficult to make decisions.

This puts teachers at a higher risk of emotional exhaustion, mental illness and burnout, with older teachers being more stressed than younger ones. On the other hand, teachers are better able to face up to their problems and are less likely to feel worthless. Lack of self-confidence is about as common amongst teachers as in the population as a whole.

Bar chart displaying differences between teachers and the general population regarding aspects of mental health
Mental health in German teachers and the general public: Comparison of data from Romppel et al. (2013) and Lütke Lanfer et al. (2022); differences were not tested for statistical significance

Researchers recommend systemic prevention

On the basis of their data, the researchers make the following recommendations:

  • The mental health of teachers should be closely monitored in order to recognize which times are the most stressful ones.
  • Teachers or groups of teachers with particularly high levels of stress should be offered and enabled to participate in prevention programmes.
  • Structurally, ways must be found to promote the mental health of teachers and reduce stress.

Pitfalls in everyday implementation

But how can the researchers’ suggestions be implemented in everyday life? There are a few aspects to consider here:

  • Unfortunately, it is often not in the teachers’ best interests to inform their superiors that they are facing mental health issues. It must therefore be ensured that the surveys are completely anonymized and that support programmes can be attended without the knowledge of superiors.
  • If teachers take up offers of help on the basis of mental health monitoring, this may, again, lead to stigmatisation as anonymisation would be removed. A more reasonable approach might be an implementation of prevention programmes for all teachers, and within their working hours. Supervision, intervision and prevention programmes must become part of the everyday life of teachers – regardless of the amount of acute psychological stress.
  • Required structural changes affect many areas that pose a risk to mental health: Number of working hours, work intensification due to more and more non-teaching activities, number of contacts due to increasing class sizes, lack of self-determination.

Above all, however, it must be recognized that teachers face more mental health challenges than the general population. Thus, any teacher experiencing stress realizes that they are no exception. Ideally, an understanding of the situation would lead to more open discourse on mental health amongst teaching staff.

Lütke Lanfer, S. S., R. Pfeifer, C. Lahmann & A. Wünsch (2022). How to Measure the Mental Health of Teachers? Psychometric Properties of the GHQ-12 in a Large Sample of German Teachers. International Journal of Environmental Research and Public Health, 19, 9708.

Romppel, M., E. Braehler, M. Roth & H. Glaesmer (2013). What is the General Health Questionnaire-12 assessing? Dimensionality and psychometric properties of the General Health Questionnaire-12 in a large scale German population sample. Comprehensive Psychiatry, 54(4), 406-413.

“The perils of returning to school”: More suicide prevention required at schools after holidays

The following text summarizes a study on youth suicide.
If you or someone you know struggles with suicidal thoughts, reach out to a trained professional for help.
UK: National Suicide Prevention Helpline UK: 0800 689 5652
USA: National Suicide Prevention Lifeline: 1-800-273-8255
Other countries

A study on youth suicides in Germany reveals that there are fewer suicides during school holidays than during term time. Conversely, more school-age students take their own lives when school resumes, not only after the summer holidays, but after every school holiday throughout the year. Boys and young men are affected most.

Posting details of mental health emergency services before Christmas has become a common feature in social media. 2023 has been no exception. But while we talk about mental health more than we used to do, suicide prevention in particular does not usually receive a lot of attention: For most, it is an uncomfortable topic to talk about. This might be a reason why, as a society, we mostly overlook the fact that suicide rates actually drop during the Christmas holidays. But this comes at a cost, as there are more attempted and completed suicides during the New Year. Yet, mental health warnings about that day – or other holidays for that matter – seem to be rare. A Canadian-German team of health economists has investigated the distribution of suicides across the year in children and adolescents, with a clear result.

Fewer youth suicides during school holidays, more youth suicides after school holidays

Vincent Chandler (Université du Québec en Outaouais, Canada), Dörte Heger and Christiane Wuckel (both of Leibniz Institute for Economic Research in Essen, Germany) have looked at the impact of school holidays on suicide rates in children and adolescents in Germany between 2001 and 2015. They found that fewer school-age children and teenagers died by suicide during school holidays. Conversely, more young people took their own lives as classes resumed after a holiday. Boys and young men are more at risk than girls and young women, as the researchers’ article in the Economics of Education Review shows. Rather appropriately, it is entitled “The perils of returning to school”.

Data from German federal states are particularly useful

When Chandler, Heger and Wuckel started their study, a rise in suicide rates after school holidays had already been found in other studies. However, these mostly covered summer holidays. For most countries, this means that we cannot reliably tell what exactly it is that triggers youth suicides at the end of a summer holiday. Is it really the end of the holiday and the start of a new school year? Or is it the end of summer, which means that there is less daylight and adolescents pursue fewer outdoor activities?

All German federal states have a summer holiday of about six weeks, but these weeks occur at different times in different places. In 2023, for instance, North-Rhine-Westphalia was the first German state to start its summer holiday on 22 June, returning to school on 7 August. Bavaria, on the other hand, started on 31 July, returning on 12 September. The individual states are therefore not subject to the same seasonal effects. Moreover, German schools have holidays of one to two weeks at multiple times throughout the school year: Apart from the summer holidays, most states have autumn, Christmas, winter, Easter, and Whitsun holidays. If returning to school triggers suicides, surely this effect should be visible after every holiday.

Data are derived from German official causes of death statistics

Chandler, Heger and Wuckel used data from the German official causes of death statistics, which list causes of death according to the international classification ICD-10. The age of the children and teenagers included in the study ranged from six to nineteen years. As a basic unit of measurement, the researchers opted for the probability of a youth suicide occurring on any day in any German federal state. If, for example, on 10 May 2007 one or more suicides occurred in Bavaria, that particular day was assigned the value “1” for Bavaria. If no suicide occurred, the day was coded “0”. To make sure that increases and decreases in the probability of youth suicides were indeed associated with school holidays or the return to school, the team collected data on the following aspects as well:

  • Gender
  • Type of school holiday (autumn, Christmas, winter, Easter, Whitsun, summer)
  • Transition periods (last days of holiday or school, first days of holiday or school)
  • Weather (sunshine and precipitation in the month and state in question)
  • Economic factors (unemployment rate in the month and state in question)
  • Year
  • Month
  • Day of the week.

Greatest decrease in youth suicides in Whitsun, autumn and summer holidays

Chandler, Heger and Wuckel calculate an overall probability of 3.77 % that a youth suicide will occur on any day in a given federal state. During school holidays, this probability is lower. The Whitsun holiday sees the most striking decrease, as probability goes down by 1.90 percentage points, meaning that, during the Whitsun holiday, it is no longer 3.77 %, but 1.87 %. During the autumn holiday, it drops to 2.82 %, during the summer holiday, 3.11 %. Results for the Christmas, winter and Easter holidays follow the same pattern. However, findings for those holidays are not significant. This means that, statistically, we cannot rule out that these results are due to chance.

The diagram shows that the youth suicide rate drops in all school holidays.
Figure 1: Decreases in youth suicide probability during the different types of school holidays (findings for autumn, Whitsun and summer holidays are statistically significant)

End of holidays and return to school pose the greatest danger

School holidays do not arrive unexpectedly. Children and teenagers know very well when they begin and their stress levels go down accordingly. Hence, the number of youth suicides already decreases in the last two days of school before a holiday. Conversely, in the last two days before classes resume, the probability of young people taking their lives increases by 0.46 percentage points. Instead of 3.77 % the probability on those days is therefore 4.23 %. And it gets worse: In the first two days of school, figures rise by 1.21 percentage points, to 4.98 %. Compared to the average probability of 3.77 %, this means that the probability of youth suicides rises by 32 %.

The diagram shows that youth suicide rates go down in the last days before and during the holidays and go up again in the last two days of the holidays and when school starts again.
Figure 2: Increases and decreases in youth suicide probability at the beginning and end of school holidays (findings for the first two school days are statistically significant)

Stress and anxiety take a toll

But what makes school stressful to the point that it enhances suicidal tendencies? As is the case with mental health issues in general, youth suicides can certainly not be traced back to a single factor causing them. However, the researchers suggest that stress and performance anxiety in schools can play a part. Expectations of parents and teachers may lead to significant levels of stress. Bullying could be another factor: In Germany, as many as 8.6 % of pupils report that they have been bullied.

Boys and young men are more at risk

According to the researchers’ data (personal communication), completed suicides occur more frequently in boys and young men than in girls and young women. As a result, the decrease in suicide probability during the holidays is about four times stronger in males than in females. The increase during the first two school days, however, is also stronger in boys and young men: Where figures rise by 24.6 % for girls and young women, they rise by 36.5 % in males. The researchers list several possible reasons for the fact that males seem more affected:

  • Schools might be failing young men more than they fail young women. In other words, the system of education seems to be less suited to boys and young men.
  • Girls and young women might have a better social network to help them cope with stress.
  • A good part of bullying amongst girls takes place online. The stress associated with this might therefore not stop during school holidays, such that suicide incidence will remain more or less the same.
  • Females might display other types of self-harming behaviour, like self-injury or eating disorders.

Suicide prevention needs to become a major concern in schools

The findings reported by Chandler, Heger and Wuckel show that we need to do more than post emergency numbers for Christmas. Schools and educational systems as a whole need to be aware of the mental health problems of young people – especially since some of those problems are bred within the schools themselves. The authors of the study suggest the introduction of hybrid schooling opportunities, which can reduce stress in certain students. Teacher training needs to include mental health education as well as specific information on suicide prevention.

We might not always know who is at risk – but sometimes we do. We therefore need to raise awareness that the first days of school after a holiday are the most dangerous time in terms of youth suicides. We must be aware of what is happening if want to be able to offer support to young people at risk of suicide.

Chandler, V., D. Heger & C. Wuckel (2022). The perils of returning to school – New insights into the impact of school holidays on youth suicides. Economics of Education Review 86, 102205.

Prosodiya – cute little lights help children with dyslexia acquire German spelling

The Prosodiya learning app offers support for pupils with reading and spelling difficulties by training their awareness of syllable stress. After a lengthy break, Prosodiya is now available in app stores again and is free of charge. The research paper detailing the app’s efficacy has been published in the journal Learning and Instruction.

Two of the spherical lights which are the main characters of the Prosodiya app
Figure 1: “Kugellichter” – spherical lights guiding players through the Prosodiya app, designed by Pikomi (Alexander Pierschel)

Many people involved in diagnosing and treating German-speaking children with developmental dyslexia had already sorely missed them: the colourful spherical lights, the cute little blobs of the learning app Prosodiya which help primary and lower secondary school pupils overcome their problems with spelling. In return, the players help the little lights to free their country – Prosodiya – from the gloomy fog that is making its inhabitants feel sad. But not only is the app back in app stores – it is now even available free of charge, giving all families equal access to this support programme. At the same time, the results of the scientific study of the app’s efficacy have been published in the journal Learning and Instruction.

Prosodiya operates on the basis of speech rhythm and syllable stress awareness

Prosodiya is an educational game for mobile devices that was developed and scientifically analysed by a team of linguists, computer scientists and psychologists at the University of Tübingen. In around 100 levels, learners develop their spelling skills on the basis of speech rhythm and syllable stress awareness. This is important as developmental dyslexia is often caused by incorrect perception of stress patterns. For instance, to be able to spell the German word gewinnen correctly, children must recognize that the word has three syllables (ge-win-nen), the second of which is stressed. In addition, the vowel of the second, stressed syllable is short. Pupils need to understand this in order to be able to apply the correct rule: The short vowel i is followed by a double consonant (nn). Primary school pupils can use Prosodiya without the support of parents or teachers: The colourful spherical lights themselves explain the rules and practise tasks with players.

Study assesses feasibility and efficacy of training with the app

The study by Heiko Holz, Manuel Ninaus, Jakob Schwerter, Cora Parrisius, Benedikt Beuttler, Katharina Brandelik and Detmar Meurers detailed in Learning and Instruction centres around three questions:

  1. Feasibility:
    Can primary school pupils use the app without support of caregivers?
  2. Efficacy:
    Does the use of Prosodiya improve reading and spelling skills?
  3. Validity:
    Does the use of the app confirm the assumption on which it is based? In other words, are perception of syllable stress and speech rhythm indeed related to reading and spelling skills?

116 primary school pupils take part in the study

In total, data from 116 second to fourth graders aged 7 to 11 years were analysed. The children were asked to use Prosodiya at home for eight to ten weeks, five days a week, for around 20 minutes a day. The reading and spelling skills of most participants were below average, so training with Prosodiya served a double purpose: It provided data for the study and provided help for families seeking to support their children in improving their spelling skills. All participants attended schools in the Tübingen area and, in addition to training with Prosodiya, underwent three appointments during which their performance was tested. The study followed the procedure shown in Figure 2:

  1. Test 1: Assessment of reading and spelling skills in all participants prior to the beginning of the study.
  2. Training phase 1: Half of the participants (group 1) train their spelling skills with Prosodiya, the other half (group 2) do not yet have access to the app.
  3. Test 2: Assessment of reading and spelling skills in both groups.
  4. Training phase 2: Group 2 now train their spelling skills with Prosodiya, the other half (group 1) no longer have access to the app.
  5. Test 3: Assessment of reading and spelling skills in both groups.
Graphic representation of the design of the study
Figure 2: Design of the study

Study relies on reading and spelling tests as well as in-game data

The study largely relies on standardised reading and spelling tests which are also used in schools and in educational therapy. In addition, several tests and questionnaires were developed specifically for the purpose of the study. In-game data used in the study consists, for example, of the scores achieved by players, as well as the times and repetitions required to complete the various levels.

  1. Standardised reading and spelling tests: DRT 2/3/4 (Diagnostischer Rechtschreibtest für 2./3./4. Klassen – Diagnostic Spelling Test for 2nd/3rd/4th Grade), SLS 2-9 (Salzburger Lese-Screening für die Schulstufen 2-9 – Salzburg Reading Screening for Grades 2-9), SLRT-II (Lese- und Rechtschreibtest – Reading and Spelling Test)
  2. Self-designed tests und questionnaires: Cloze spelling test, questionnaires for parents and children
  3. In-game data: Scores attained, times required for levels, repetitions of levels

Primary school pupils are able to use Prosodiya with support from caregivers

The primary school pupils who took part in the study did so with some stamina, even if they did not strictly follow the defined training plan. On average …

… they trained for 27.9 days during the eight to ten weeks of training.

… they practised for 18.5 minutes a day.

… they reached level 69 of the approximately 100 levels of the game.

… they practised 161.7 levels including repetitions.

… they needed 3.0 minutes for one level.

… they achieved 138.6 of the possible 150 points per level.

… they managed to solve 8.2 tasks on their first attempt.

The children enjoyed the training and stated that the app was easy to use. Training with Prosodiya had a positive effect on the children’s sense of self-efficacy: They experienced that practising with the app actually helped them improve their skills. There was also positive feedback on the game from parents and teachers.

Training with Prosodiya improves spelling skills

The study confirms the positive experiences reported by children, parents and teachers. In the diagnostic spelling test DRT 2/3/4, all learners improved in all phases of the study – even in the phases in which the respective group did not use Prosodiya. But training with the Prosodiya app led to significantly higher learning gains (see Figure 3). The two groups improved by more than three and four points respectively on the T-score scale during the training period.

Graphical representation of study results regarding spelling
Figure 3: Improvement of spelling skills as measured by the standardised spelling test DRT 2/3/4

The T-score scale is used to represent results in many standardised tests: Its median is a T-score of 50, with a standard deviation of 10. In other words, if you move ten points to the left and right of the scale starting from the median, you define the range that is considered to represent average results. 68.2 % of a norm group – in this case 68.2 % of second/third or fourth graders – achieve results which are within the range of one standard deviation to the right and left of the median. A T-score between 40 and 60 is therefore considered to be average. Bearing this in mind, the dimension of an increase by several points on the T-score scale in just eight to ten weeks becomes clear.

In addition, the research group around Heiko Holz separately evaluated those tasks from the standardised test that related to vowel length. These were the tasks most closely related to the training tasks in the Prosodiya app. The benefits of the app were clearly recognisable here as well, with the group that had trained with Prosodiya in phase 1 achieving the greatest gain in competence. All results relating to data from the standardised test, DRT 2/3/4, were significant, which means that it is highly unlikely, even virtually impossible, that these results are due to chance. The fact that spelling skills of the group that had trained with Prosodiya first did not decline in the second phase of the study suggests that the training has a positive long-term effect on spelling.

The spelling test developed specifically for the study largely confirmed the results of the standardised test: With one exception, there were always significant differences between the group that had trained with Prosodiya and the group that had not done so during the phase in question.

Prosodiya improves syllable stress awareness

Learners were also able to significantly improve their perception of stressed and unstressed syllables while training with Prosodiya. In both groups, there were learning gains in all phases of the study, but training with the app led to a greater increase in competence (see Figure 4). Here, too, results were highly significant. In other words, it is virtually impossible for differences between the groups to be due to anything other than the fact that one group trained with Prosodiya and the other did not.

Graphical representation of results regarding syllable stress awareness
Figure 4: Improvement of syllable stress awareness

Study does not show improvement of reading skills

The researchers were not able to conclusively show that the use of Prosodiya leads to an increase in reading skills. With one exception, there were learning gains in both groups in all phases of the study. However, these were not statistically significant, which means that the increase in competence cannot be unequivocally linked to the use of Prosodiya.

Syllable stress awareness is essential

The basic assumption underlying the development of the app was that the correct perception of stressed and unstressed syllables influences spelling skills: Learners with reading/spelling difficulties demonstrate poorer syllable stress awareness than pupils who do not experience such difficulties. For this reason, the participants’ syllable stress awareness scores were correlated with the reading and spelling data collected. In fact, scores for syllable stress awareness correlated with all scores of the reading and spelling tests. The higher the scores for syllable stress awareness, the better the results for spelling, but also for reading fluency and word and pseudo-word reading. There was a negative correlation with the errors occurring when marking vowel length: The higher the score for syllable stress awareness, the fewer errors were made. The assumption on which the development of Prosodiya is based could therefore be confirmed.

Prosodiya is a suitable diagnostic and intervention tool for children with dyslexia

The researchers observed similar correlations between in-game data and reading and spelling skills: The higher the scores and the less time players needed to attain them, the better they performed on the reading and spelling tests. The researchers conclude that the Prosodiya app’s approach is suitable for supporting pupils with poor reading and spelling skills. It may even be possible to use the app for diagnostic purposes.

Free access to all levels for teachers and educational therapists

Since the relaunch of Prosodiya, the app has been available in the app stores free of charge. However, even a free download would mean that teachers and educational therapists would have to progress through the entire training programme in order to gain an insight into all types of tasks. Prosodiya therefore offers so-called “Alles frei” (“all free”) profiles, with which all training tasks can be viewed and played without having solved all previous tasks.

Is Prosodiya a suitable tool for learners without dyslexia?

Prosodiya offers an interesting, motivating and valuable way of supporting children diagnosed with developmental dyslexia, particularly since access to the app is now extremely easy thanks to it being free of charge. Lower secondary school pupils also enjoy practising with the colourful spherical lights. It would be interesting to concentrate on how training with Prosodiya affects learners with average spelling skills: A further study would certainly be worthwhile in this respect. The use of Prosodiya in German as a Foreign Language lessons is also of interest: Particularly with native speakers of languages whose spelling is not based on syllable stress, this might support learning motivation and lead to more effective spelling training.

Holz, H., Ninaus, M., Schwerter, J., Parrisius, C., Beuttler, B., Brandelik, K., & Meurers, D. (2023). A Digital Game-Based Training Improves Spelling in German Primary School Children – A Randomized Controlled Field Trial. Learning and Instruction, 87, 101771.