Typing Fluencies of 12–13-Year-Old Students with Dyslexia and Peers with Typical Development

Abstract

Several factors can impede the writing process of students with dyslexia. One recommended adjustment to help them overcome these writing challenges is the use of personal computers for writing. The research underscores the significance of effective keyboarding skills in optimizing the benefits of computer-based writing for these students, with touch-typing being the frequently recommended typing technique. Although research findings point to various reasons that indicate that students with dyslexia may face difficulties in developing fluent touch-typing skills, we found a lack of research in this area. In order to address this gap, in our study we examined and compared 12–13-year-old Slovenian students with dyslexia and their typically developing peers in terms of their existing, self-taught typing technique fluency, the fluency achieved in learning touch-typing and the fluency achieved in learning a simple finger-typing task. We found a significant difference between handwritten and self-taught typing fluency of students with dyslexia, with handwriting being a more fluent form of transcription. The results of students with dyslexia are significantly lower from their peers in handwriting fluency, self-taught typing fluency, touch-typing fluency and simple finger-typing task fluency. However, the learning trends for touch-typing and the simple finger-typing task do not differ between the groups of students. We comment on the differences found and make suggestions when considering typing as an additional or alternative transcription mode.

Introduction

Fluency in transcription (spelling, handwriting, or typing) affects the quantity and quality of writing of students with and without learning disabilities (Christensen, 2004; 2005; De La Paz & Graham, 1997; MacArthur & Cavalier, 2004; Peverly et al., 2007; Salas & Silvente, 2020). Transcription problems in the writing of students with learning disabilities consume working memory resources and leave little room for other important writing processes (Baker et al., 2003). Thus, it is not surprising that students with learning disabilities who have transcription problems are more likely to forget what they intended to write down, and they tire more quickly and give up writing (Gillespie & Graham, 2014). Research also shows that automaticity of transcription in normative students increases and is reflected in longer language bursts, shorter pauses and greater fluency (Alves & Limpo, 2015), but problems with fluency of transcription in students with dyslexia often persist (Connelly et al., 2012). It is important to note that the problem of students with dyslexia in transcription does not appear to be primarily a fine motor or graphomotor problem. Although students with dyslexia often demonstrate difficulties in handwriting fluency and correct letter formation (Hebert et al., 2018), there is mixed evidence that observed differences in handwriting fluency among children with dyslexia are directly attributable to graphomotor differences (Connelly et al., 2012). There are several other possible explanations for why dyslexic students have slow transcription. Hebert et al. (2018) state that the poor letter production observed in less fluent handwriting is not due to poor handwriting motor function but is likely due to an overload of working memory when simultaneously spelling and writing down letters. Morken & Helland (2013) concluded that cognitive skills that are important for reading also affect the writing of students with dyslexia. Researchers observed that students with dyslexia pause more frequently and for longer periods when writing even in text copying tasks (Sumner et al., 2014). They compose texts in shorter language bursts, have a slower composing rate (Beers et al., 2017) and have shorter final writing products (Beers et al., 2017; Connelly et al., 2012; Sumner et al., 2014) than their neurotypical peers. The transcription problems of these students, evidenced by slow and nonfluent handwriting, appear to be due to spelling problems (Hebert et al., 2018; Sumner et al., 2014; Van Waes et al., 2015). A possible explanation of slower transcription could also be a slow processing speed. Individuals with dyslexia may namely also have problems with tasks that require rapid performance (deficit in rapid automatic naming, da Silva et al., 2020). This is what leads to the suggestion that the poor transcription fluency observed in slow handwriting or typing may be the result of slower speed in recalling letter shapes in memory and integrating these shapes into hand movements (Connelly et al., 2012). Research also suggests that individuals with dyslexia may also have problems with procedural learning and the quality of automation of new, complex sensorimotor skills (Marchand-Krynski et al., 2017; Nicolson & Fawcett, 2000), more so with the learning of procedural motor skills and sequences that involve linguistic features (Gabay et al., 2012; Nicolson et al., 2010). They also appear to have difficulty unlearning mappings of previously learned motor skills (Nicolson & Fawcett, 2018). An important consideration when discussing possible handwriting or typing difficulties of students with dyslexia also lies in the obvious challenges of differential diagnosis and the frequent co-occurrence of reading and writing disorders such as dysgraphia (Berninger, Richards, et al., 2015; Connelly et al., 2006). All of this leads us to the conclusion that it is important to assess students with dyslexia in writing and support them where necessary so that they can develop effective writing fluency.

Writing with computer as an instructional adjustment

One of the commonly suggested adjustments to support students with dyslexia in writing is the use of computers for writing (Beers et al., 2017; Berninger, Nagy, et al., 2015; Mishra & Mohan, 2016; Weigelt-Marom & Weintraub, 2016). There are several potential benefits of using computers for writing: it is easier to revise and quickly add, delete, or move text (Beers et al., 2017), letter production is physically easier (Daiute, 1986) and can therefore free up working memory resources (Beers et al., 2017; Peverly et al., 2007). Word processors also provide additional supports for writing, such as spell-checkers, grammar checkers, and the writing is more legible (Beers et al., 2017). Writing with a computer can be effective if supported by appropriate word processing and keyboarding (typing) skills (Hebert et al., 2018). A frequently recommended method for acquiring proficient keyboarding (typing) skills for students with dyslexia, is systematic instruction in touch-typing (Mishra & Mohan, 2016). In contrast to self-taught typing, learning touch-typing is a systematic process in which the keys are added gradually at the beginning, usually two per learning stage (Weigelt-Marom & Weintraub, 2016). Touch-typing also differs from frequent self-taught techniques of visually led keyboarding: it is a typing technique in which we use both hands, each finger, and its movements in a specific and consistent way, using kinesthetic rather than visual feedback (Freeman et al., 2005). Touch-typing requires a precise connection between fingers and keys, that, to be efficient, must be fast, consistent, and independent of visual tracking. It thus allows for the development of truly automated typing that supports the full potential of computer writing (Connelly et al., 2007). In contrast, self-taught typing techniques are casually, spontaneously developed. They are more visually guided; therefore, writers spend less time looking at the screen while writing. Self-taught typing thus demands a great deal of visual feedback (looking at the keyboard) even in adult self-taught typists. In addition, fingers are also less consistent in terms of finger-key patterns (Feit et al., 2016). It is important to add that an already developed, self-taught typing technique prior to learning touch-typing could influence the learning of touch-typing, as the individual also must unlearn the motor patterns of typing technique they used before (Weigelt Marom and Weintraub, 2015).

Keyboarding skills formal instruction in education varies globally. In Slovenia and many other European countries, keyboarding is not explicitly included in official curricula. There is no compulsory subject for the use of computer science and technologies in Slovenia for students at the lower secondary level, i.e., ages 11–15 or younger (Bourgeois et al., 2019). In these subjects, but also in all other compulsory subjects, there are no explicit learning objectives for mastering the keyboard or touch-typing. According to the OECD TALIS Survey (Teaching and Learning International Survey, Ainley & Carstens, 2018) Slovenian schools lagged behind OECD countries in the use of Information and communications technology (ICT) in education. For example, according to the TALIS survey, only 37% of Slovenian teachers (the OECD average is 53%) at lower secondary level (ISCED level 2, i. e. lower secondary level, ages 11–15) frequently or always allowed students to use ICT for projects or class work. According to the latest report by the European Commission on ICT in Education for students at ISCED level 2, only 32% of Slovenian students (the fewest among many other EU countries) used a computer (desktop, laptop or notebook) at school at least once a week for learning purposes (p. 42). Thirty-five % of Slovenian students stated that they had never or almost never used a computer at school in the last 3 months (p. 44) (European Commission, 2019). We can conclude that Slovenian students do not systematically start using typing for writing at school, but develop their own technique of visually guided typing, through the occasional use of ICT in learning. The majority do not use computers at school regularly (daily, weekly) for learning.

Assessing handwriting and typing fluency

Proficient typing is important for writing productivity of the writing process and its result, reduced demands on working memory, and the ability to look at the screen while writing (Christensen, 2004; Johansson et al., 2010). Writing competence at the transcription mode level is the degree of individual efficiency, often referred to as fluency or automaticity in typing and handwriting, and measured by speed and accuracy, the number of self-corrections made during the writing task, the amount of attention paid, and the effort required to perform the typing or handwriting movements (Rieger, 2004; Soukoreff, 2010; Tucha et al., 2008). Fluency in writing (handwriting or typing) is often measured by the number of words correctly copied in a limited time (Berninger 1994, as cited in Klein, 2021) or by the number of correct letters or characters in a selected unit of time, e.g., one minute (Feng et al., 2019). The tasks for assessing writing fluency vary in complexity and involve different writing skills and therefore more or less cognitive demands. For example, the text copying task is primarily determined by transcription mode skills (typing or handwriting) and not so much by other writing or language skills, as is the case with text composition tasks (Van Waes et al., 2021). There are also different learning expectations and criteria/norms for assessing typing fluency. One of the common criteria we found in the literature for elementary school students is the ability to type alphanumeric information at least as fast as in handwriting, with touch-typing being the recommended typing technique (Beaton, 2012; Beers et al., 2017; Freeman et al., 2005; MacArthur, 2009). However, speed equivalence should be used with caution with students who are already struggling to achieve adequate handwriting speed (Freeman et al., 2005; Rosenberg-Adler & Weintraub, 2020). So that equivalence alone may not be a sufficient criterion for an appropriate typing fluency, as the existent handwriting fluency may already be too low. When planning keyboarding instruction, more accurate student age-appropriate norms and benchmarks that provide additional support for data-driven goal setting are thus valuable. For the United States, Freeman et al. (2005) presented an earlier comprehensive review of research on keyboarding speed at different grade levels (studies from 1983 to 2004). However, these are more than 20 years old as well authors reported that it was difficult to compare the data collected at that time because the different studies used different technologies and research methodologies. It seems that there is also a lack of newer fluency norms for keyboarding fluency for different age groups, as Morress et al. (2020) recently noted that there are many educational guidelines in the United States that discuss elementary standards for keyboarding (such as the CCSS—Common Core State Standards), but there is still a lack of more specific numerical expectations for keyboarding speed and accuracy that would be based on newer grade-specific fluency and accuracy rates. Nonetheless, we found that the national keyboarding standards in the United States, i.e., Standards for the Production of Written Language for Handwriting and Keyboarding for Grades PreK-8 (HW21 Community, 2014), provide general numerical benchmarks for keyboard fluency when copying tasks using touch-typing. Benchmarks for fluency in WPM (words per minute) are provided for the first time in 3rd grade and standards for accuracy in 5th grade. At copy typing tasks, students in 7th grade should achieve at least 25 words per minute with 90% accuracy, and in 8th grade, at least 30 words per minute with 90% accuracy, using the touch-typing technique.

Typing skills of students with dyslexia and their peers with typical development

There is a reported lack of studies on self-taught typing performance and touch-typing learning in students with dyslexia and their typically developing peers (Abecassis et al., 2023; Weigelt Marom & Weintraub, 2015). However, research data shows that the general student population can type more fluently with increasing age (Freeman et al., 2005; Horne et al., 2011). Typing speed was found to lag behind handwriting speed in 11-year-old British students without explicit keyboarding instruction (Connelly et al., 2007). Half of 13-year-old British students typed faster than they handwrote (Horne et al., 2011). Seventy % of sixth grade students who had previously received keyboarding instruction typed faster than they handwrote on a text copying task (Roger and Case-Smith, 2002, as cited in Horne et al., 2011). Fifteen-year-old Swedish students with reading and writing difficulties typed less fluently than their peers on expressive writing task, with fluency assessed as longer transition times between keystrokes (Wengelin et al., 2014). It is also worth highlighting that students who struggle with handwriting and typing due to problems with linguistic processes (such as reading and spelling) tend to write very slowly in both writing modes (Rosenberg-Adler & Weintraub, 2020), i.e., in handwriting and in typing. Self-taught typing was not entirely adequate in 4th to 9th grade students with specific learning disabilities (SLD) in reading and writing (dyslexia or dysgraphia) and their peers without SLD (Beers et al., 2017): The group of students with dyslexia had comparable handwriting and typing composing fluency (measured as words per minute), but poorer composing fluency compared to their typical peers. They could not touch type and had to look at their hands when typing.

The only study we have found on the process of touch-typing acquisition in students with dyslexia is the study of Hebrew university students with and without specific reading and/or writing difficulties (SLD). In this study (Weigelt-Marom & Weintraub, 2018), all students typed more slowly than handwrote before touch-typing training. Typing of SLD students was statistically significantly slower than that of their peers. The difference between the groups was no longer statistically significant after the delayed posttest of touch-typing training; the students with SLD also became faster at typing than handwriting. Their touch-typing learning pattern was similar to that of their peers. To our best knowledge, there is no comparable study on the acquisition of touch-typing in students aged 11–14 years (lower secondary school students) with dyslexia.

In this digital age of writing, it is important that teaching encourages the development of hybrid writers who are proficient in the use of different writing tools, including keyboards, and explore the advantages and weaknesses of different writing tools (James et al., 2016). This is particularly important for learners who struggle with learning impairments and who can benefit from computer writing supports in the ways mentioned earlier. Touch-typing is one of the commonly recommended techniques for promoting transcription fluency in students with dyslexia. It is a complex procedural motor skill that involves a linguistic component and requires the unlearning of previously learned motor patterns for students who have not been taught to touch-type from an early age. Based on the aforementioned difficulties of students with dyslexia and specifics of touch-typing, we can assume that touch-typing is not necessarily a quick and easy solution for these students. It needs to be considered whether learning adjustments may be required when learning touch-typing and whether additional/alternative aids are needed to support fluent transcription on the computer. Our research questions in this study are therefore:

1. Is there a difference in handwriting and self-taught typing fluency between students with dyslexia and their typically developing peers?

2. Is there a difference between students with dyslexia and their typically developing peers in terms of achieved touch-typing fluency and trend of learning touch-typing?

3. Is there a difference between students with dyslexia and their typically developing peers in simple finger-typing task fluency and the trend of task learning?

Materials and methods

Our study uses quantitative Quasi-Experimental Group Comparison Design.

Participants

After approval by the Ethics Committee of the Faculty of Education at the University of Ljubljana, Slovenia, essential information and the possibility to register for a two-week touch-typing course were made available on a custom web portal and on a Facebook page. Information about the possibility of participation was sent to the school counseling services for students in 7th and 8th grade of lower secondary school (12–13 year old students) in the capital Ljubljana, as well as to the Faculty of Education in Ljubljana and the Counseling Center for Children, Youth and Parents in Ljubljana. Our sampling method was a combination of voluntary response and purposive selection. From the students whose parents expressed an interest in their child’s participation, we selected students who met the inclusion criteria. All students had to be in 7th or 8th grade and had to have no previous experience with learning touch-typing (which was confirmed by parents and students). Students without dyslexia (typically developing peers) had to meet the following criteria: Parents had to confirm that they did not have a learning impairment or other special needs or were gifted. Students with dyslexia had to meet the following criteria: They had to provide a state-verified certificate identifying a specific developmental impairment in the area of reading, as determined by prior in-depth expert assessments. All certificates were examined by the first author of this study, a special education teacher. The formal procedure for identifying a specific developmental disorder in reading in Slovenia is a complex, detailed assessment that includes a series of cognitive tests for specific dyslexia indicators (such as phonological awareness, rapid automatic naming of words, reading, spelling, working memory tests and intelligence tests). It also includes the information collected through questionnaires and interviews with teachers, children, and parents, as well as through the analysis of work samples and tests of children’s reading and writing skills (Košak-Babuder et al., 2019). The Slovenian criteria for defining a specific learning disorder with impairment in reading include the criterion of IQ - achievement discrepancy (general intellectual ability above 85) as well as criteria of cognitive processing deficits. All students in our sample had received state-verified certification of their disability status in previous school years based on the procedures described. Parents whose children were included in the sample had to confirm the validity of the basic inclusion conditions and sign a consent form for their child’s participation, which also included all information on data collection and processing according to our data management plan during and after the experimental sessions. All students were Slovenian, their mother tongue was Slovene.

From the group of 46 students who met all inclusion criteria and who had been taught touch-typing, we selected 42 students who had complete data for at least five of six measurement points of touch-typing learning (see Table 3 for measurement points definition) as unfortunately, not all typing learning data were successfully recorded due to database problems. Of the 42 students whose data were included in the following data analysis, half met the dyslexia criteria and half of the students were peers who also met the required criteria. The even number of participants (21/21) is random and not planned. The final participant sample consisted of 21 7th and 8th grade students (mean age = 13.5) with specific learning disorder with impairment in reading (developmental dyslexia, DD) with documented concurrent difficulties in writing, not specified, and 21 7th and 8th grade students with typical development (mean age = 13.2) without dyslexia, other special needs, or gifted status (TD). There was no statistically significant difference in age between the two groups (t = 0.082, p = 0.935). The structure of the sample is shown in Table 1.

Table 1. Sample structure by student group, grade, and gender.

	Grade 7 students		Grade 8 students
Group of students	Male (N)	Female (N)	Male (N)	Female (N)
DD	10	4	3	4
TD	10	6	1	4

Table 3. Touch-typing cumulative practice time for each of the six measurement points (M) according to the learning stages of touch-typing.

Touch-typing learning stage	Cumulative touch-typing practice time	Touch-typing fluency measurement point (M)
1st Stage AFJČ	30 min	M1
	90 min	M2
2nd Stage AFJČDK	20 min	M3
	110 min	M4
3rd Stage AFJČDKSL	20 min	M5
	110 min	M6

Students participated in two weeks of touch-typing training on a QWERTZ keyboard middle row (ASDFJKLČ) during summer (N = 36) or winter (N = 6) breaks in 2020/2021 or 2021/2022. To collect such a sample, we organized and conducted a total of 13 two-week touch-typing learning groups, six in 2020/2021 and seven in 2021/2022. The size of the learning groups varied from a minimum of 2 to a maximum of 6 students per group. We limited the number of participants in the learning groups to a maximum of 6 students to ensure the quality of the teaching and testing procedures.

Instruments and data collection

Assessment of fluency in handwriting and self-taught typing

To assess handwriting and typing fluency (in terms of speed and accuracy) we chose a text copying task. We selected three different age-appropriate narrative prose texts (no pictures, no dialog in the text, comparable length, and difficulty) from the age-appropriate literature textbook. From each of the three texts, we selected two non-adjacent passages of approximately about 100 words each. We used one for the handwriting task and one for the typing task. In total, we prepared three text passages for the handwriting task and three for the typing task. In each writing mode students copied three text passages, each for one minute. We introduced all text passages in the same font (Arial) and size (18). In each writing mode, the students copied all three text passages one after the other, each for one minute. For the data collection of the typing task, we prepared the task in the self-created software application for touch-typing learning (described in the section The Touch-typing Software Application and Touch-typing Instruction). During this copy typing task students could correct their typing errors but were not able to use spell-check, word prediction or other support. The instruction to all students were to write fast and accurate and at the same time in handwriting maintain the legibility of their transcription. We assessed and compared students’ fluency in handwriting and self-taught typing with the same measure by the number of correct characters per minute (CPM), including all letters and punctuation marks as suggested by Weigelt Marom & Weintraub (2015), excluding spaces as in Feng et al. (2019) and excluding capitalization. In handwriting, a letter was considered correct if it was legible and in the correct place in the sequence, whereas in typing, as in Connelly et al. (2007), only the criterion of the correct place in the sequence was used. The handwriting and typing tasks were conducted and scored by the first author of this study, an experienced special education teacher. For the typing task, she used the automatic counting of characters in Microsoft Word. For the handwriting task, she counted the legible letters. There were no major problems with the legibility of the students’ letters, so we did not opt for an additional scorer. Such of writing fluency assessments are thought to have good inter-rater reliability and discriminate well between students with and without learning difficulties in handwriting (e.g., Rosenberg-Adler and Weintraub 2020). Nevertheless, we calculated Pearson coefficients to examine the alternate-form reliability of the three alternative text passages per writing mode. All correlation coefficients were high, ranging from 0.81 − 0.93. We also documented each student’s use of self-taught typing technique in a one-minute individual observation while performing previously described copy-typing task. For this purpose, we used the Keyboard Observation Scale as described in Weigelt Marom and Weintraub (2015). All students typed with two hands, used two to four fingers in each hand, and used repeated visual feedback. Handwriting and self-taught typing fluency tasks were carried out one after the other. Half of the randomly selected students performed a handwriting task first, immediately followed by the typing task, and the other half of the randomly selected students performed the task simultaneously in reverse order. Each student performed the tasks on the first day of the study procedure.

Assessment of indicators of specific learning disorders with impairment in reading

To further corroborate the data from the state-verified certificate, we administered a series of subtests from the Slovenian validated version of the SNAP - Special Needs Assessment Profile (Weedon & Reid, 2008) designed to capture various indicators of dyslexia. The SNAP contains cutoff points for each of the tasks against which students’ scores are compared. A score of less than 1.5 or 2 SD of the mean validated score in the subtest that measures typical dyslexia indicators (e.g. reading fluency, reading non-words) means that students are at risk of dyslexia (Košak-Babuder, et al., 2019). It is one of the most commonly used instruments in Slovenia to assess the risk of dyslexia and has proven to be valid in studies by Slovenian researchers (e.g. Košak-Babuder et al., 2019). We used the following subtests as dyslexia indicators: Timed Reading Test (the number of words read correctly in 30 seconds), Phonological Awareness Test (a phoneme deletion task in which students have to delete certain phonemes in non-words they hear), Non-Word Reading Test (the number of non-words read correctly in one minute), Timed Spelling Test (the number of words spelled correctly in a two-minute dictation), Backward Span Test (the number of successful repetitions of the backward word sequence) and Picture Naming Test (the seconds taken to name pictures). Our analysis (see Table 2) using the independent t-test or the Mann-Whitney U-test confirmed that the students with dyslexia performed significantly worse in all subtests, except in the phonological test. We think that this task might be too easy for the students, as Slovenian is a transparent language, or that the task is already too easy for this age group. We do not discuss this difference as it is beyond the scope of this study, but research shows that phonological awareness as a predictor of reading difficulties decreases in more transparent languages (e.g., Furnes & Samuelsson, 2010). We can conclude that the groups of students differ in all other important dyslexia indicators, which further confirms the validity of our sample. SNAP was administered individually and in the same testing sequence by three certified and experienced special education teachers on the last day of the study procedure.

Table 2. Comparison in SNAP between the groups of students in SNAP.

SNAP subtest	Timed reading	Phonol. Awareness	Non-Word Reading	Timed Spelling	Backward Span	Picture Naming
Test Statistic, p-value	U = 15	U = 151.5	U = 26	t = −2.955	U = 124	t = 3.341
	Z = −5.171	Z = −1.860	Z = −4.901	df = 40	Z = −2.464	df = 40
	p < 0.001	p = 0.063	p < 0.001	p = 0.005	p = 0.014	p = 0.002
Effect size	ES = 1.596	ES = 0.574	ES =1.512	ES = 0.912	ES = 0.760	ES = 1.031

Finger-typing task

We investigated the ability of fine motor fluency performance also with a finger-typing task (FTT) that required repeating a sequence of 41324 on a computer keyboard with fingers of the dominant hand. We modeled our task after the examples of Cellini (2017) and Biotteau et all. (2017). In our study students used the following keyboard keys when their right hand was dominant in handwriting (N = 38): V = 1 (index finger), B = 2 (middle finger), N = 3 (ring finger), M = 4 (little finger). Students (N = 4) whose handwriting was left hand dominant repeated the sequence as V = 4 (little finger), B = 3 (ring finger), N = 2 (middle finger), M = 1 (index finger). We used the VBNM keys instead of the number keys on the keyboard to minimize accidental pressing of other keys. Students learned the sequence every other day for two weeks (on-site practice days, see Table 4 for study procedure). Students performed three trials per week and six trials in total. The first trial was guided and used to learn the rules and procedure of the task and to check individually if all students understood the rules. For the analysis of student results, we thus used the second trial as a baseline and the sixth trial as the final learning trial. Each trial consisted of four consecutive blocks. During each block, students had to repeat the sequence as quickly and accurately as possible without pauses, for a total of 30 seconds, followed by a 20-second pause. Before each trial, students were tested individually, by the first author of this study, on correct memorization and repetition of the sequence. They had to perform three consecutive correct finger repetitions with fingers on the tabletop and three consecutive correct repetitions with their fingers on the keyboard keys. They had to perform the exercise so that all fingers touched the keys in a resting position and the wrist was low without touching the table. We measured keystrokes for all learning trials in the self-created touch-typing application (described in the next section). For each block, students were shown two screens: one for repeating the sequence with a dynamic green circular time counter of 30 seconds and one for pause time with a dynamic red circular time counter of 20 seconds. Students were observed performing the task but were not given feedback on their correct or incorrect repetitions. Mastery of the task was measured by the number of correct repetitions of the whole 41324 sequence (CSR) in 30 seconds, as in Cellini (2017). The simple finger-typing task was performed in according to the instructions of the first author of this study. Each student performed the task individually in the touch-typing application according to the study procedure.

Table 4. Study procedure.

Week/Day (form of learning)	Monday (on site)	Tuesday (distance, online)	Wednesday (on site)		Thursday (distance, online)	Friday (on site)
1st Week	TIIIM1F	IIII	IIM2F	IIM3I	IIII	IIIIM4FI
2nd Week	IIM5FIII	IIII	IIM6FIII		IIII	IIIFS

T - test of handwriting and self-taught typing; I - every 10 - minute touch-typing learning interval; M - measurement points for touch-typing fluency; F - finger-typing task; S - SNAP tests. Light gray – 1st typing stage; medium gray – 2nd typing stage; dark gray – 3rd typing stage.

The touch-typing software application and touch-typing instruction

The self-constructed touch-typing online software application was developed in JavaScript platform and MySQL database by the authors of this study and Pavle Gartner from L2 tech d.o.o for the purposes of this study and was thoroughly tested a year before the main study began. The development of such an application was necessary because there is no comparable application for touch-typing in Slovenian with age- and research-appropriate lessons and would allow access to students’ key log data and its ethical security and storage. We made the visual design of the application dyslexia-friendly (see Figure 1), similar to the example of a high-quality web-based application for touch-typing, Typing Club (EdClub, 2011). In addition to the touch-typing lessons, the application also contains the self-taught typing task and the finger-typing task, both of which have already been described.

Figure 1. Visual design of the screen of a self-created touch-typing software application.

The course content (lesson text) was developed based on the Slovenian typing textbook (Degen, 1997), which we adapted in an age-appropriate way for the purposes of this study. Our typing course was divided into learning stages according to the keyboard characters (letters) included. Each subsequent level includes all previously learned letters in a ratio of at least 1:6 compared to the newly added letters. Each learning stage includes practicing typing in 10-minute intervals. During the typing exercises, every correct character is immediately highlighted in green and every error in red. The application does not block typing errors and thus does not require errors to be corrected. It also provides feedback on typing fluency (correct characters per minute) and accuracy (% correct characters) at the end of each 10-minute practice interval. Each student used a Chrome browser program to run the touch-typing application and used an identical model of external keyboard.

The touch-typing instruction was conducted by the first author of this study. She drew on the knowledge gained of experienced teachers teaching touch-typing to Slovenian lower secondary school students, on foreign literature on touch-typing (Beaton, 2012), and completed 30 hours of formal training in level 1 touch-typing herself a year before the study began. The total duration of the touch-typing training in our study was divided into two weeks of daily practice at the first three stages of the typing course, i.e., AFJČ at the first stage, adding DK at the second stage and SL at the last stage. The daily practice included 50 minutes (five 10-minute intervals) of on-site learning at the Faculty of Education in Ljubljana (Mondays, Wednesdays, and Fridays) and 40 minutes (four 10-minute intervals) of practice during online distance learning via ZOOM sessions (Tuesdays and Thursdays). Each day, the learning was divided into two typing sessions with a break between the two sessions. The practice time and order were the same for all students. The instructions for practicing and the order of test tasks were standardized for all students, whether on-site or during the online distance learning sessions. When necessary, the instructor gave students brief instructions (prompts) on their body, hand and finger posture during the training sessions, but not during the assessment sessions for which the collected data were analyzed. Student typing fluency data was collected via the typing application. In total, there were six assessment sessions, each defines one measurement point. In total we have thus six measurement points, two per each learning stage (see Table 3 for the measurement points and Table 4 for the study procedure). The measurement points were defined according to the cumulative typing practice time of the stage. At these measurement points, students were given a short copying task (included in the described touch-typing application) with 42 words and non-words on the screen, which they had to type as fast and accurately (90–95% accurately) as possible. The tasks were carried out while on-site training and had the same visual form as the regular typing task but contained only two- to three-syllable words (four to six letters) and non-words (readable in Slovenian but without meaning) with no word or non-word repetitions. Using data from the computer database keystroke logs, we calculated students’ average typing fluency in CPM (the average number of correct characters per minute, including spaces) and accuracy (average percentage of correct characters per minute, including spaces, % CPM). Students’ pause times longer than 3 seconds are not included in the calculation of fluency and accuracy.

Study procedure

The study procedure for each touch-typing learning group lasted two consecutive weeks. On the first touch-typing training day, before beginning to learn touch-typing, students were tested on their handwriting and self-taught typing fluency. Touch-typing was then practiced on the first and all subsequent days, for a total of 10 days. Only on the on-site days (Mondays, Wednesdays and Fridays) were typing skills and simple finger-typing task assessed at the assigned measurement points (M). On every other day (Tuesdays and Thursdays), the typing training was conducted online as distance learning. On the last, 10th day, the SNAP test was also administered individually. The study procedure is shown in Table 4.

Data analysis

We used Python ver. 3.10.10 for data preprocessing and trend estimation. The program IBM SPSS Statistics 28 was used for the statistical hypothesis tests. The distributions of the variables were tested using the Kolmogorov-Smirnov test with Lilliefors correction, and a t-test for independent samples or a Mann-Whitney U-test, Pearson (r) or Spearman (r_s) correlation were used depending on the results. We also calculated Cohen’s d or Hedge’s effect size, or an estimate of Cohen’s d based on the MW-U Z-statistic as d = Z√ (1/N1 + 1/N2) (Gignac, 2019). The risk level is α = 0.05. If necessary, we use the Bonferroni correction, where the risk level α = 0.05 is divided by the number of comparisons. If necessary, we perform comparisons without outliers, defining outliers using the Tukey method in SPSS. Trends were estimated and compared using general linear models. We calculate post-hoc power using G*Power by Faul et al. (2009).

Results

We firstly investigated the differences between girls and boys in the main research variables of transcription fluencies in self-taught typing, touch-typing and handwriting. We made this comparison to exclude gender as a confounding factor, as research shows that girls perform better than boys in transcription fluency (e.g. Cordeiro et al., 2018). For the comparisons, we used the t-test for independent samples or the Mann-Whitney U-test. No differences were found between girls and boys in self-taught typing fluency (t = −1.880, df = 40, p = 0.067, ES = 0.575). There were also no statistically significant differences between boys and girls in touch-typing fluencies measured at all six touch-typing measurement points M1 (U = 135, Z = −1.826, p = 0.068, ES = 0.579), M2 (t = −1.951, df = 38, p = 0.058, ES = 0.612), M3 (t = −1.812, df = 35, p = 0.079; p = 0.104, ES = 0.594), M4 (t = −0.948, df = 38, p = 0.349, ES = 0.295), M5 (t = −0.805, df = 36, p = 0.426, ES = 0.259), M6 (U = 162, Z = −0.484, p = 0.628, ES = 0.158). There is also no difference between girls and boys in the fluency of the simple finger-typing task (t = −0.703, df = 37, p = 0.486, ES = 0.221). However, we found statistically significant differences in handwriting, with girls performing better than boys (t = −4.878, df = 40, p < 0.001, ES = 1.492), confirming Cordeiro et al. (2018). The results of differences are the same when outliers are excluded. We can’t say that gender explains the differences between groups of students with and without dyslexia observed in the forthcoming analysis of typing fluencies.

Students’ handwriting and self-taught typing fluency in a text copying task

In accordance with our first study’s research question, we compared the students’ average score for handwriting and the average score for self-taught typing fluency, calculated from the scores of the three alternative passages per writing mode for each student in CPM (number of correct characters per minute, including all letters and punctuation marks, excluding spaces and capitalization). For the results see Table 5.

Table 5. Comparison of self-taught typing (TY) and handwriting (HW) fluency in correct characters per minute (CPM), between the group of students with dyslexia (DD) and the group of typically developing peers (TD).

Variable	Students with Dyslexia (DD)			Typically developing peers (TD)			Test Statistic, p-value	Effect size	Post hoc Power
	N	Mean	SD	N	Mean	SD
HW	21	67.6	18.6	21	81.7	20.6	t = −2.337 df = 40 p = 0.025	ES = 0.721	t = 2.02 df = 40 PWR = 0.626
TY	21	48.8	14.7	21	82.0	26.2	t = −5.055 df = 40 p < 0.001	ES = 1.560	t = 2.02 df = 40 PWR = 0.998
TY/HW %	21	74.1	19.9	21	104.2	39.7	MW U = 68 Z = −3.836 p < 0.001	ES = 1.183	t = 2.02 df = 38.11 PWR = 0.870

All but one student with dyslexia (DD) and more than 60% of their typical developing peers (TD) lagged behind their handwriting fluency in a copy typing task. Upon independent-samples t-test, students in the DD group had significantly lower transcription fluency than TD students in both writing modes i.e., in handwriting (t = −2.337, df = 40, p = 0.025, ES = 0.721) and in self-taught typing (t = −5.055, df = 40, p < 0.001, ES = 1.560). The differences were significant also when outliers were excluded (two for handwriting and three for self-taught typing in total). We also calculated the percentage relationship i.e., (TY/HW) × 100%, between self-taught typing fluency and handwriting fluency to assess whether and to what extent the students’ typing fluency corresponds to their handwriting fluency. Using Mann-Whitney U-test we found a statistically significant difference between the groups of students, with students from the DD group showing a significantly lower ratio (U = 68, Z = −3.836, p < 0.001, ES = 1.183). The differences are significant also if two outliers are excluded.

Using the paired-samples t-test, the difference in average fluency between the two writing modes is significant for students with dyslexia (t = 5.695, df = 20, p < 0.001, ES = 1.243), but not for their typically developing peers (t = −0.044, df = 20, p = 0.966, ES = 0.010). There are four outliers in the TD group. If these are excluded the difference is still not significant (t = 0.867, df = 16, p = 0.399 and EF = 0.210).

A moderate Pearson correlation coefficient between handwriting and self-taught typing fluency was found for students with dyslexia (r = 0.613, p = 0.003, N = 21) and their typically developing peers (r = 0.523, p = 0.015, N = 21). There were four outliers in the TD group (according to the Tukey method in SPSS). After excluding these, we obtained r = 0.670 (p = 0.003, N = 17).

Students’ learning performance in learning to touch type

For the second research question, we compared students’ performance in learning touch-typing in terms of fluency as CPM (number of correct characters per minute, including spaces) and accuracy as % CPM (% of correct characters per minute, including spaces, % CPM) at all six measurement points. Using an independent-samples t-test or a Mann-Whitney U-test, students in the DD group lagged statistically significantly behind their peers in fluency of touch-typing at all six measurement points (see Table 6). The calculated effect sizes are high. If we exclude outliers (from one to a maximum of four in total at each measurement point), the significance level at all six measurement points is still p < 0.001, with effect sizes 1.493, 1.296, 1.288, 1.435, 1.199, 1.021. In both cases (with or without outliers), the differences are significant, even with the Bonferroni corrected level 0.05/6 = 0.008.

Table 6. Comparison of touch-typing fluency in CPM (correct characters per minute), from first to the sixth measurement point (M), between the group of students with dyslexia (DD) and the group of typically developing peers (TD).

Touch-typing fluency measurement point	Students with Dyslexia (DD)			Typically developing peers (TD)			Test Statistic, p-value	Effect size	Post hoc Power
	N	Mean	SD	N	Mean	SD
M1	20	83.7	19.7	21	112.4	28.2	MW U = 72 Z= −3.599 p < 0.001	ES = 1.124	t = 2.03 df = 37.15 PWR = 0.928
M2	20	111.8	26.7	20	162	32.4	MW U = 46 Z = −4.166 p < 0.001	ES = 1.317	t = 2.03 df = 36.20 PWR = 0.977
M3	20	92.5	18.6	17	122.6	27	MW U = 49 Z = −3.688 p < 0.001	ES = 1.217	t = 2.03 df = 33.33 PWR = 0.938
M4	21	115	31.7	19	167.7	38.6	MW U = 55 Z = −3.914 p < 0.001	ES = 1.239	t = 2.03 df = 36.20 PWR = 0.960
M5	21	93	18	17	128.4	36.6	t = −3,896 df = 36 p < 0.001	ES = 1.244	t = 2.03 df = 35.24 PWR = 0.957
M6	20	127,3	50,7	18	158,3	41.4	MW U = 84 Z = −2.807 p = 0.005	ES = 0.912	t = 2.03 df = 34.29 PWR = 0.760

At some measurement points in both groups, some students’ data was lost due to technical problems with the application.

Students differ in accuracy (% CPM) of touch-typing in all but one measurement point, with outliers excluded or included (all but one p are under α = 0.05). Upon using Bonferroni correction 0.05/6 = 0.008, only one out of six remains significant (for M2 outliers included: p < 0.001, ES = 1.202, outliers excluded: p < 0.001, ES = 1.175). Thus, the student groups do differ significantly in typing accuracy. Since we found that some students typed with very low accuracy, we compared the fluency differences with the control of touch-typing at 70% accuracy (as evidence of a minimum learning effort level), to see if the differences in fluency of touch-typing persisted. This threshold was determined using the histogram of accuracy values. In this case, the N of the DD group is at least 15 and that of the TD group is at least 17. The differences in fluency are still significant at all six measurement points, with or without outliers, using the Bonferroni correction, all p are below 0.05/8 = 0.008 and the effect sizes are 1.128, 1.300, 1.221, 1.208, 1.222, 1.124 with outliers included and 1.526, 1.279, 1.309, 1.401, 1.156, 1.407 with outliers excluded. The difference in fluency between the groups is therefore not due to the extremely low accuracies of some students.

We also compared the learning trends of the student groups for the three learning intervals that we defined for each learning stage. The first interval at the first learning stage is defined by M1 to M2, the second at the second learning stage by M3 to M4, and the third at the third learning stage by M5 to M6. All three trends are positive and significant in both groups (all p < 0.017). When applying the Bonferroni correction, all trends are significant for each student group (all p-values are less than the corrected level 0.05/3 = 0.0167), but the third trend (M5 to M6) for the typical developing peers (p = 0.03). The trends are not significantly different between the student groups (all p-values are between 0.075 and 0.803). The linear regression trend models are shown in Figure 2.

Figure 2. TD group (blue solid line) and DD group (orange dotted line) linear regression trend models for touch-typing fluency in correct characters per minute (CPM) for the three learning stages.

To see if there were correlations between transcription modes, we calculated the correlation coefficients between handwriting fluency, self-taught typing fluency, and the touch-typing fluency scores at the six measurement points for each student group, applying the Bonferroni corrected risk level 0.05/6 = 0.008.

For the DD group, we found moderate to strong significant Pearson (r) and Spearman (r_s) correlations for four out of six touch-typing fluency measurement points and self-taught typing fluency (r_s = 0.717, p < 0.001, N = 20; r_s = 0.674, p = 0.001, N = 20; r = 0.717, p < 0.001, N = 20; r = 0.542, p = 0.011, N = 21). For the TD group, there are moderate to strong significant correlations between five of six measurement points of touch-typing fluency and self-taught typing fluency r_s = 0.751, p < 0.001, N = 21; r_s = 0.785, p < 0.001, N = 20; r = 0.694, p = 0.002, N = 17; r = 0.631, p = 0.004, N = 19; r = 0.638, p = 0.006, N = 17). We can say that self-taught typing fluency is moderately to strongly correlated with touch-typing fluency in both groups of students.

Students’ learning performance in learning finger-typing task

To be able to answer our third and final study research question, we examined and compared groups of students performing the finger-typing task (FTT). This task is less complex and more pure motor learning measure compared to touch-typing as it does not include a linguistic component for reading or spelling, it is motorically simpler (unimanual). We assessed student outcomes by comparing group means on the total number of correct sequence repetitions in 30 seconds for each learning trial. Using the t-test for independent samples or the Mann-Whitney U-test, the significant differences are for all but the fifth learning trial, with students with dyslexia showing significantly lower fluency measured as number of repetitions per 30 seconds (see Table 7). We also excluded the outliers in each trial using the Tukey method in SPSS (there is a total of one to a maximum of three in each trial) and used Bonferroni’s corrected risk level 0.05/5 = 0.01. As a result, there are three significant differences, in the second (t = −3.082, df = 27, p = 0.005, ES = 1.113), fourth (U = 64.5, Z = −2.925, p = 0.003, ES = 0.989) and last, sixth trial (t = −2.921, df = 36, p = 0.006, ES = 0.928) and two insignificant differences in the third (U = 97.5, Z = −2.043, p = 0.041, ES = 0.681) and fifth (t = −1.254, df = 35, p = 0.218, ES = 0.412) trials. The effect size is still large for the last trial, which measures final performance on the task. From this we conclude that there is a statistically significant overall difference between the groups in fluency on the finger-typing task.

Table 7. Comparison of the number of correct sequence repetitions (CSR) in the finger-typing task (FTT) from the second to the sixth trial, between the group of students with dyslexia (DD) and the group of typically developing peers (TD).

FTT trial	Typically developing peers (TD)			Students with Dyslexia (DD)			Test Statistic, p-value	Effect size	Post hoc Power
	N	Mean	SD	N	Mean	SD
2nd	16	11.6	6.4	16	8.6	4.4	MW U = 74 Z= −2.042 p = 0.043	ES = 0.722	t = 2.04 df = 28.56 PWR = 0.487
3^rd	20	13.9	6.3	18	10.4	4.4	MW U = 97.5 Z = −2.414 p = 0.016	ES = 0.784	t = 2.03 df = 34.29 PWR = 0.630
4^th	19	14.79	5.9	19	11.7	6.3	MW U = 101.5 Z = −2.309 p = 0.021	ES = 0.749	t = 2.03 df = 34.29 PWR = 0.592
5^th	19	16.8	5.7	19	13.5	5.3	t = −1.610 df = 36 p = 0.116	ES = 0.522	t = 2.03 df = 36 PWR = 0.347
6th	20	18.4	7.0	19	13.5	4.5	t = −2.961 df = 37 p = 0.005	ES = 0.929	t = 2.03 df = 37 PWR = 0.806

At some trials in both groups, the data of some students was lost due to technical problems with application.

The four sequential trends of learning the finger-typing task, at Bonferroni-corrected risk level 0.05/4 = 0.0125, are not significant in both groups of students (smallest p = 0.246), so the task is truly a simple motor learning task. The trend difference between groups of students is not significant as well (smallest p = 0.535).

We estimated the Pearson (r) and Spearman (r_s) correlations between fluency of touch-typing at the touch-typing stages beginning measurement points (M1, M3 and M5) and the outcome of the last trial of the finger- typing task. For the TD group, correlation coefficients are r_s = 0.498, p = 0.025; r_s = 0.638, p = 0.008; r_s = 0.602, p = 0.014, and for TD group r_s = 0.609, p = 0.007; r = 0.654, p = 0.003; r = 0.491, p = 0.033. After Bonferroni corrected level 0.05/3 = 0.0167, two correlations are significant in the TD group and two in the DD group. We estimated Pearson (r) and Spearman (r_s) correlations between fluency of touch-typing at the touch-typing stages final (second) measurement points (M2, M4 and M6) and the outcome of the last trial of the finger-typing task. For the TD group, the correlation coefficients are r_s = 0.579, p = 0.009; r_s = 0.654, p = 0.003; r_s = 0.578, p = 0.015. For the DD group r = 0.437, p = 0.070; r = 0.556, p = 0.013; r_s = 0.473, p = 0.047. According to the Bonferroni-corrected risk level 0.05/3 = 0.0167, three correlations are significant in the TD group and one in the DD group. Our results show that students with dyslexia have significantly poorer results in both typing tasks and that there is a medium correlation between touch-typing fluency and simple finger-typing fluency.

Discussion and conclusion

In this study, we aimed to examine typing fluency performance in terms of self-taught typing fluency, in touch-typing fluency, and fluency in a simple finger-typing task. We found that students with dyslexia were significantly less fluent in all three typing tasks. In the context of our first research question, we examined and compared students’ handwriting and self-taught typing fluency. We found statistically significant differences in average handwriting and self-taught typing fluency between the groups of students. Students with dyslexia lagged behind their peers in both transcription modes. For students with dyslexia, the difference between the two modes is significant, with handwriting being the more fluent writing mode. For their typically developing peers, the difference in average fluency between the two modes is not significant. Only one student from the DD group and only 8 (38%) students from the TD group exceeded their handwriting fluency in typing. As Beers et al. (2017), we also documented that all our sample students had to look at their hands when using self-taught typing in the text copying task and did not use all fingers when typing. A greater proportion of students in our sample typed less fluently in a text copying task during self-taught typing than, for example, sixth graders in the study of Rogers & Case-Smith (2002), 70% of whom exceeded their handwriting speed in the text copying task after typing instruction (Rogers & Case-Smith, 2002). This comparison suggests that even today, in an increasingly digitized society, the occasional use of computers for writing was not sufficient for the majority of students in our sample to develop efficient (at least handwriting comparable) typing fluency. The correlation between handwriting and self-taught typing fluency is moderate for both groups of students, similar to Connelly et al. (2007) who report r = 0.7 in the sample of students aged 4–11 years. In our study, students with dyslexia (who have a language-related impairment) tended to be significantly slower in handwriting and slower in typing, which is consistent with Rosenberg-Adler & Weintraub (2020). As the literature also suggests, the results show that students with dyslexia need additional, systemic support to become more efficient at typing (Beers et al., 2017). In line with previous research (e.g. Grabowski, 2008; Weigelt-Marom & Weintraub, 2018).

To address our second research question, we examined the fluency of touch-typing during the first three learning stages of touch-typing. We found that students in the DD group differed statistically significantly from their peers in all three observed learning stages of learning touch-typing. Students with dyslexia are less fluent than their peers in learning to touch type. However, their learning trend is significant and positive and similar to the results of Weigelt-Marom and Weintraub (2016), the learning pattern does not differ significantly from that of their peers. This shows us that students with dyslexia and writing difficulties can learn to touch type, but their initial and final fluency is lower than that of their peers. It is important to point out that we observed these differences in the early stages of learning, which are less complex (in motor and linguistic terms) than the later stages. We conclude that students with dyslexia may need more practice to achieve touch-typing fluency and accuracy comparable to their peers. However, it is also possible that their ultimate achievable typing fluency is lower than that of their neurotypical peers. It is not within the scope of our research to clarify these important questions. An important observation regarding the correlation between self-taught typing fluency and achieved touch-typing fluency is that students who were more fluent in self-taught typing were also more fluent in touch-typing. Thus, also for students with dyslexia who have low self-taught typing skills, it’s crucial to know that touch-typing for them may not always the quick solution to better typing fluency.

Addressing our third and the final research question we examined students’ performance on a simple finger-typing task. Our results indicate that dyslexia students performed the task less fluently than their peers. Lower speed and accuracy in learning a motor task in students with dyslexia was found, for example, by Nicolson & Fawcett (2000). The finger-typing task fluency like self-taught typing fluency correlated to the touch-typing fluency. This is an indicator that those who are less fluent in it are probably also less fluent also in touch-typing. As the task of finger-typing itself is less motorically complex compared to touch-typing and does not include a linguistic component, this could indicate that there are differences in the fluency of fine motor typing skills in students with dyslexia that could influence also their touch-typing fluency. Research on dyslexia-specific motor differences is still inconclusive (Connelly et al., 2012), with observed difference between learning simple and complex motor tasks in students with dyslexia (Marchand-Krynski, et al. 2017). Based on the ES on group differences in both tasks (touch-typing and simple finger motor task), we hypothesize that this difference may exist, but it would need to be validated using a larger sample of students and additional analyzes.

It seems that for some students with dyslexia, touch-typing may not be a quick and sufficient solution. If students with dyslexia are slow in both handwriting and self-taught typing, serious consideration should be given if their learning touch-typing fluency tends to be low as well. Although this was not the subject of our research, future research should therefore further investigate whether it is necessary to adapt the learning of touch-typing, such as the learning objectives and the required practice time, or to use additional tools, such as the use of speech-to-text tools, word prediction tools or simple AI solutions, when these students write with the computer although they have not yet developed an efficient touch-typing fluency. The basis of such data would provide a clearer idea of how dyslexic students can be supported to achieve better computer writing transcription fluency and thus use computers more effectively for writing. We believe that finding solutions to support efficient transcription fluency is a wise investment in unlocking the full writing potential of young writers with dyslexia that may be overlooked due to lack of understanding of their possible transcription fluency differences.

Limitations and implications for future research

There are several limitations of this study. The first is in the small sample. Most students in the dyslexia group also have co-occurrent problems with writing, so the results of our study must be used with caution when referring to individuals with dyslexia only. It would also be important to consider and control for the co-occurrence of impairments and developmental disorders. It would be necessary to also collect data on the entire training of touch-typing to see the overall learning achievements and to assess the role of touch-typing fluency efficiency in various writing tasks.

Acknowledgments

The authors thank Pavle Gartner from L2 tech d.o.o. for his exceptional work and dedication in programming the touch-typing software application. This work was supported in part by the Slovenian Research Agency under the research program ICT4QoL — Information and Communications Technologies for Quality of Life, grant number P2-0246.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.