How a Digital Intervention in Schools Contributed to Students’ Social and Emotional Skills, and Impacted Writing

Published accounts of school interventions often focus on a small ‘slice’ of the complex systems and layers of learning across schools and homes. A series of studies into a digital intervention in low SES schools, is summarised here providing a fuller picture of how students’ engagement in digital contexts contributed to their social and emotional skills; and as a consequence, whether these skills impacted on achievement in writing. Overall, 9 to 13 year olds found self-regulation both harder and different when on line. The intervention was associated with higher levels of self-regulation when online, but had no effect on social skills. These patterns were related to specific features of the schools’ practices and instructional foci. Students’ use at home added to the general effects, with social media usages reducing ratings of self-regulation (e.g. chatting), others increasing social skills (e.g. commenting on blog posts), but high levels of parental monitoring positively impacted both. Parents’ own digital experience and their beliefs were related to their guidance. Higher levels of social emotional skills (specifically, response inhibition and cognitive empathy) were associated with gains in writing achievement. Putting the slices together enabled a more complete picture of the intervention, including how: (a) it supported the development of self-regulation online but needed more focus on social skills; (b) higher levels of self-regulation and social skills impacted writing; and (c) the coherence between settings contributed to these skills, underlining the need to provide support and advice to some parents.


Introduction
This paper is an account of how an innovative digital approach in a cluster of schools contributed to students' intra-individual (self-regulation) and inter-individual (perspective taking, sociability and empathy) skills.The digital intervention involved an ongoing research and development partnership between the researchers and a cluster of schools in which all students from Year 4 (9-yearolds) to Year 13 (18-year-olds) used their own devices (Chromebook, Netbook) on a daily basis and across most lessons.A charitable trust provided whānau / families with the devices on a 'rent to buy' basis.The schools served mostly low socio-economic communities, and aimed to change the inequitable profile of achievement of their mostly Māori and Pacific heritage students.
The overarching aim of this article is to describe whether and in what ways the school-wide digital intervention, with ubiquitous access to digital tools in and out of school, contributed to students' self-regulation and social skills, and whether and how these skills were, in turn, associated with the students' achievement in writing.
We had addressed different components of this overall aim in a programme of research which has produced multiple studies.This paper brings them together to provide a fuller account across the multiple components, thereby contributing to a broader understanding of the multi-directional relationships between intra-and inter-individual skills, writing, school and home.In what follows we draw on the evidence in these published studies (see Table 1), hereafter referred to as Papers A through Paper H.

Theoretical Framework
The overall programme of research was based on the model shown in Fig. 1.It represents children's learning as multi-layered, multi factorial and dynamic, as proposed by models such as socio-ecological systems (Bronfenbrenner, 1979) and those focused on the socio-cultural bases of development and learning, mediated by language and tools (Rogoff, 2003).Such models propose that what is learned reflects the beliefs, experience, guidance and activities provided by families, and the pedagogy and instruction provided by educators (who also hold beliefs and have experience); and the interdependencies between these.In both settings children and their teachers, as well as children with their whānau, are embedded in communities of practice using tools for particular social purposes (see Fig. 1).
The model does not unpack the manner and modes of children's learning, other than to propose that self-regulation and other social and emotional skills are central to learning and have a bidirectional relationship with ongoing academic learning (Peng & Kievit, 2020), in our studies specifically in the domain of writing (Graham, 2018).Various predictions were made in each of the individual studies about the relationships between the components (Papers A through H, see Table 1).Here we integrate across the studies to provide an overall summary of predicted relationships.Any effects of adopting digital tools, with ubiquitous access across school and home settings, on self-regulation, perspective taking, sociability and empathy would be conditional.Drawing on the non-digital evidence base, we expected impacts would depend on conditions at the school level, classroom level and at home.The conditions would include how the school context created communities of practice in which norms, values and practices would enable the development of these skills and link to those practiced at home.But impact would also depend on whether the classroom instruction provided opportunities to learn and practice these skills.Evidence available at the time suggested that providing digital devices to all students for their learning could impact aspects of agency in the classroom, particularly the behavioural features of attending and being on task (Karich et al., 2014).But there was no evidence for whether this impact on behavioural self-regulation would generalise to the executive control functions of Attention Focus and Inhibitory Control.However, we hypothesised that impacts on these might be possible, especially given appropriate conditions provided by the available guidance, the communities of practice and specific ways that the digital tools were used.
There was less evidence to draw on for possible impacts of digital interventions on inter-personal skills.Nevertheless, studies of digital games, accounts of the affordances of the tools, and what is known about development also suggested that under appropriate conditions these skills could be impacted (Paper A).
An additional set of predictions was made.We knew that the intra-individual and inter-individual skills, at least in non-digital school contexts, have bidirectional relationships with children's academic skills (Peng & Kievit, 2020).That is, learning to write could be impacted by levels of self-regulation and social skills, but in turn, learning in writing could feed back into these skills.Once again, this impact would depend on certain conditions.These would be the quality of the instruction for writing and the degree to which the instruction enabled further learning in writing.The interdependencies could in fact be negative as much as positive, depending on the conditions supporting progress in writing skills.
We also knew that school was not the only context within which digital devices would be used, and within which opportunities to either develop these skills or have these skills undermined might happen.Young people in Aotearoa New Zealand are high users of the Internet and social media (OECD, 2021).There are complex positive and negative associations between the frequency of use and the content of that use at home with the social and emotional skills (Belanger et al., 2011).Despite the paucity of evidence for parents' roles through digital practices, it was predicted that parents' practices would reflect their beliefs and experience, and impact children's skills, again either positively or negatively in addition to what teachers, classroom practices and communities in schools provided.

Approach
Researchers have a range of quantitative and qualitative tools for making sense of the relationships represented by the model.Because educational systems are complex, open and dynamic, whatever lenses and methods we adopt, ultimately, we need to put interconnected pieces together to create as full a picture of a whole as is possible for a specific time and place.Being able to provide a full account of a programme of research with a cluster of schools and their communities which relies on multiple specific components is challenging, although there are elegant examples, for example by using the vehicle of book length accounts to bring the pieces together (e.g., Bryk et al., 2010).
Here, we consider the contribution of the combined programme of work, which drew on multiple methods to understand associations across differing aspects of the digital intervention.To describe the contribution, we adopt a 'slicing' analogy which signals the way that analyses of different sections of the model represent 'slices' over a particular period of time.Within the overall study, different slices of the whole have been studied, and are now considered together in this narrative account.The slices differ in terms of their granularity and scope, however together they enable us to present an account that is more than the sum of the disparate parts.
Addressing the overarching aim involves four steps.The first step is to describe the conditions for learning with digital tools in the school and home contexts.The second is to establish whether or not the overall digital intervention in the schooling context impacted the inter-and intra-individual skills on the one hand and, separately, writing achievement on the other.The third step is to connect whatever ways the intervention impacted on the skills to the conditions of the school context and the home context established in the first step.The final step, dependent on what the previous steps show, is to establish links between the intra-and inter-individual skills and writing achievement.
The research questions which address these steps are: (1) What were the features of the schooling and home contexts which provide conditions for learning with the digital tools?(2) Did the overall intervention in schools impact self-regulation and social skills?(3) How did contexts of school and home together relate to self-regulation and social skills?(4) Were there bidirectional links between self-regulation and social skills with writing achievement?
The numbers of schools, students, teachers, and parents vary depending on the slice, but included 162 parents, 186 students from ages 9 years to 13 years and 46 teachers.The participants came from nine suburban schools in Auckland serving culturally and linguistically diverse communities which, according to the decile ratings at the time, had lowest levels of income and employment (low SES).The study of intra-and inter-personal skills took place within an ongoing Research Practice Partnership (Snow, 2015) approach, using Design-Based Research (Anderson & Shattuck, 2012) to jointly address an agreed problem (Jesson et al., 2018a).This process involved iterative cycles of detailed evidence gathering enabling the ongoing design and redesign of the overall intervention, working together with practitioners with the shared aims of solving the challenge of inequitable achievement in valued outcomes.The research included detailed classroom observations and writing achievement measures assessed using the e-asTTle assessment tool (http://e-asttle.tki.org.nz/), with up to 835 students.
Alongside the achievement outcomes, we also took slices of the students' self-regulation and social skills, ranging from measures of those that were very general to those very specific to the digital context.At the most general level, this meant students' ratings on measures of personality using the Big Five Inventory (BFI, John et al., 2008).The five dimensions are predispositions reflecting New Zealand Journal of Educational Studies (2023) 58:361-377 specific skills, such as Conscientiousness (e.g., 'I am someone who does a thorough job') reflecting self-regulation (e.g., 'I can stick with my plans and goals').We then delved down into more specific measures using established tools for rating self-regulation (Inhibitory Control and Attention Focus) on the one hand, and on the other those for rating empathy, perspective taking and prosocial skills.These were ratings for contexts not specific to digital tools and Internet use, so we further probed the same skills at an even more specific level, as rated when using the Internet and digital tools [e.g., Inhibitory control: 'I can get off task easily by different online sites (i.e., emails, chat, games) or devices (e.g., mobile phone) while doing my schoolwork on the computer'].The measures had satisfactory psychometric properties (see Papers B-D).
Yet another slice was made across two clusters of schools, one was the original cluster and the second had implemented the same digital intervention more recently.These clusters differed on how much time the intervention had been in place.This was technically a 'dosage' analysis, which added to the statistical techniques we employed to check the relationships.This helped us to determine whether we could attribute patterns in the target skills to the intervention.
The interdependencies between the two major socialisation contexts were considered by a slice that looked at digital use across schools and homes of the same children.In both the school and the home context the focus was on the socialisation patterns to see how the teachers and parents may have supported or constrained the development of the target skills.

Question One: What Were the Features of the Schooling and Home Contexts Which Provide Conditions for Learning with the Digital Tools? (Papers B-H)
Our approach to understanding the contexts was to use existing descriptions of the teaching and classrooms in research papers, document analyses and our own slices of the school context to determine the school and classroom level conditions.This included specific analyses of Digital Learning Objects (DLOs) created by students and patterns of discourse with digital tools that could make a difference to aspects of the skills, such as perspective taking.

The School Context
The ongoing Design-based Research Partnership used evidence in the form of detailed and repeated measures of many aspects of the intervention such as classroom phenomena, as well as valued student outcomes, in iterative cycles of design and redesign.Full descriptions of the school intervention are in Papers B through H, with additional details in publications about the overall research and development partnership (Jesson et al., 2018a).The school context is characterised by five themes.
1 3 (1) A purpose-built pedagogy.'Learn-Create-Share' had been implemented consistently across schools, as a core pedagogical framework within the overall intervention.It had specific instructional features and was demonstrated to be the proximal cause of gains in writing achievement.The Learn phase often involved topic-based research, followed by the Creation of DLOs and then Sharing of these occurred within and across classes and beyond.(2) High use of digital devices.The devices were used in almost 90% of class time, with students engaging directly in activities with these about half of the class time.A quarter of the time was spent in online collaboration and the creative design of the DLOs, within which multimodal creation and student co-authorship were prevalent features.Face-to-face collaboration occurred in more than half the lesson time.
(3) School-wide and classroom level communities of practice.In these communities, norms, values and practices for using digital devices were socialised.There was also an instructional focus on the skills, particularly on self-regulation, seen across multiple observations in many lessons and through teacher directed and peer interactions.These two components were designed to develop: (a) 'Digital Citizenship' through being 'cybersmart'.This meant taking personal responsibility, being in the right place at the right time when online, respecting both the tools and their impact on others' use online, and appropriate activity undertaken with the device.Students signed agreements (as did teachers and parents) which covered very specific online behaviour such as not sending or receiving inappropriate pictures, how online behaviour can affect future job prospects and further education, alerting parents to inappropriate content received including harassment and not using social media without seeking parental permission.(b) Social and emotional skills.The majority of teachers we observed (in English/literacy maths and science lessons) indicated that they used opportunities in content teaching to promote both sets of skills either weekly or daily, using a range of methods.More teachers said they were aware of opportunities to teach self-regulation than the social skills.Observations backed this up.Direct forms of guidance for self-regulation by the teachers were mostly directed at being independent (attention focus) and not being distracted (inhibitory control), and occurred in nearly half the observed time; less of the time was spent focused on social skills.However, frequent feedback for collaboration and joint and independent use of the tools was observed.
(4) Significant impact overall on students' writing.The Research-Practice Partnership demonstrated that a Year 4-10 student present over three years, on average made one more year's progress nett of the nationally expected rate, per year.One way to put the significance of this impact is that in a group of 1000 students, 120 more students would be at or above average national levels.This high quality impact was not limited to writing.For example students' DLOs were being used 1 3 New Zealand Journal of Educational Studies (2023) 58:361-377 as mechanisms for communicating mathematical understanding and promoting interest, through which students' showed heightened audience awareness, use of multimodality and personalisation.(5) Explicit intervention principles.Seven dimensions were identified as the core features of treatment integrity in ongoing scaling from the original cluster of schools who designed the intervention.These were dimensions of: Pedagogy, Community engagement, Technology infrastructure and support, Administrative support, Research and Development (and innovation), Investment, and Leadership.The most difficult to fully implement in scaling was found to be engagement with the community.Apart from this, these features were rapidly implemented to high levels in the second cluster of schools who contributed data to the slices (Jesson et al., 2018b).

The Home Context
The home context can be characterised with five themes too.
(1) Māori and Pacific communities (eight out of ten parents).Nearly half had a postsecondary qualification, but more than one in ten had not completed secondary education.They were in a wide range of occupations, with very few unemployed.
(2) Digitally experienced whānau/families.Broadband had been provided through the philanthropic funding.Nine out of ten parents used a device either at home or at work, or in both contexts, and seven out of ten reported a high level of familiarity.But a quarter felt they had a low level of expertise.
(3) Access to classroom learning sites.The children could access classroom learning sites and use the Internet.Six out of ten students reported high usage, on most days of the week, although a fifth reported less than or about once a week.A typical online session was 1.7 h for school-related work and 2.7 h for non-schoolrelated work.(4) Support for children's usage.Nine out of ten parents reported some support for social and emotional skills with digital usage, with a third providing that on a 'daily' or 'weekly' basis.Parents described use of four strategies to support their children (monitoring, discussing, using devices alongside and directly teaching) with half employing all four.There was a marked shift with older children.Lower frequencies of support were provided for 10-13 year olds, but parents of older children had greater awareness (than parents of younger children) of providing that guidance.Parents in general were keen to know how best to support their children at home.(5) Varied beliefs and practices relating to Internet use.Most parents were familiar and experienced with digital tools, and used a variety of guidance strategies.They differed in how frequently they provided support and how confident they were.They were mostly neutral about whether the Internet had positive or negative effects on their children's skills (four out of ten reported no effects either way), but over a third believed there were positive effects.Parents of 10-13 year olds were much more likely to identify negative effects than parents with 8-9 year olds.Their beliefs mattered, as parents who were positive about the Internet's effects had a higher frequency of providing support.However, parents who were neutral, were less aware of providing any support.Their own digital self-efficacy mattered too.Parents who saw themselves as expert with digital tools had lower frequencies of providing support, perhaps reflecting being more confident.Four groups of parents were identifiable.A group Highly Engaged with their children around using and learning with the devices; a Confident group; a group who were Less Engaged; and a small Novice Group (around 5%) who were unfamiliar with digital tools with no or little experience at home or at work.These Novice parents provided the lowest frequency of guidance and used fewer than three strategies.Not surprisingly, they had the greatest concern about how best to help.

The Connection Between Home and School
There was extensive information flow between schools and homes, accompanied by free courses or workshops which ranged from foundational computer skills through to how to help support and protect their children's online presence and skills.The agreement protocols parents signed explicitly stated that if they were unsure about the device use, or how to support the values and norms they would attend a parent training session.All parents were expected to complete a basic parent training session before the device went home.

Question Two: Did the Overall Intervention Impact Self-regulation and Social Skills (Papers B, D)?
Yes-for self-regulation, unclear-for social skills, but possibly an effect on overall self-confidence.The 9-year-olds to 13 year olds' developmental patterns were no different from typical patterns described in studies of development not in ubiquitous digital contexts.Like others, they rated their social skills higher than those of self-regulation, and the personality dimension of conscientiousness, as well as ratings of self-regulation, trended downwards over time.There were few gender or ethnicity differences.
But there was a clear effect of digital use on their ratings of self-regulation.Students were increasingly aware of difficulties of self-regulating in digital contexts such that by 13 years old, they rated themselves as marginally unable to maintain self-regulation, and markedly 'less able' than their younger peers; and across all ages less able than in non-digital contexts [the Effect Size (ES) for this difference with digital contexts was g = − 0.54 for Inhibitory Control; g = − 0.30 for Attention Focus].Moreover, their judgements of themselves in the different contexts were unrelated to each other (correlations were around zero) meaning not just lower or higher in the digital context, but their self-regulation was different in digital contexts.

3
New Zealand Journal of Educational Studies (2023) 58:361-377 In contrast, social skills did not appear to be as consistently sensitive to the digital context (ES differences ranged from negative to positive depending on the measure).But judgements about their social skills were closely related (r = 0.30 to 0.53).All the personality ratings were higher than the ratings of social skills and self-regulation in digital contexts.But they were unrelated to capabilities of self-regulating in digital contexts (correlations were all zero, even those that were conceptually similar such as 'Conscientiousness').In contrast, how students rated their personality was consistently positively related to judgments about social skills in digital contexts (correlations ranged from r = 0.26 to 0.48).Taken together, the students' social skills in digital contexts reflected their more general predispositions, but being able to self-regulate was independent of those predispositions and self-regulation in nondigital contexts.
Despite this picture, the digital intervention had a detectable positive impact on self-regulation, partly mitigating the general downward trend, and providing a generalised effect.Evidence for this came from comparing two clusters of schools which implemented the digital intervention at different times.These groups provided a natural experiment contrasting how long the intervention had been in place in schools and how much exposure students had (a 'dosage' analysis).Being in the digital intervention for 5 or more years, compared with less than a year meant higher average self-regulation in digital contexts (d = 0.64).This positive difference also was detectable on ratings in more general non-digital contexts too (ES Inhibitory Control d = 0.20; and Attention Focus d = 0.16).
Again, in contrast, dosage did not matter for social skills, either in digital contexts (d = − 0.09) and for three of the social skills rated in non-digital contexts (the ESs were close to zero with the exception being Cognitive Empathy which was higher in the lower dosage schools).However, students in high dosage schools were more confident, rating themselves considerably higher on all 5 dimensions of the BFI (the ES estimates ranged from d = 0.37 for Openness to d = 0.89 for Conscientiousness).
Question Three: Were These Patterns Related to the Contexts of School and Home? (Papers A, B, D and E) School: Yes, three features stand out in the school context.
Evidence from the answer to the second question support an effect of the overall intervention on self-regulation.The schools' communities of practice was a central feature of the digital intervention.This, together with the second feature of classroom instruction having a focus on self-regulation, provided the core components of omnibus programmes known to increase levels of social and emotional skills (Durlak et al., 2011).In addition, the increased awareness of self-regulating in digital contexts use is consistent with the evidence about heightened agency using devices but being sensitive to the 'impulsogenic' properties of digital tools and their uses (Duckworth & Sternberg, 2015).
The greater effect on self-regulation compared with the social skills is the third feature.It is likely due to the lower focus on these latter skills at the whole school level and through classroom instruction.When an instructional focus was apparent, it was mostly expressed as a concern for student difficulties, rather than deliberate teaching of the skills.In addition, the cybersmart protocols adopted within and across schools and with whānau had more of a focus on personal responsibility, than on specific interpersonal skills.Finally, it appears that the collaboration and sharing parts of the pedagogy were not sufficient in and of themselves to further boost the component skills.
Home: Yes, aspects of home usage were associated with a positive effect on social skills, a negative effect on self-regulation, but parent monitoring helped both.
These effects were quite specific to the context of being online, and additional to the overall impact of the school context.The effects were also additional to other variables such as gender, ethnicity, age and personality.But they were against the backdrop of their digital citizenship embedded at school and the pedagogy and instruction taking place in classrooms.
The more often and longer students spent time online, the lower their ratings of self-regulation in digital contexts, estimated to be close to half a point, dropping ratings from a mid-point on a 5 point scale to being negative about being able to selfregulate.It was social media that was the problem, and chatting directly or through games and posting contributed to their increased awareness of finding it difficult to be in control.But high levels of parental monitoring counteracted this negative impact, by about the same amount (half a point).This association between parental monitoring and self-regulation was specific to the digital context.
There was no direct positive or negative effect of frequency and duration of overall Internet use on the ratings of their social skills.But what students did online did impact social skills.More often posting or emailing photos, and commenting on blog posts was associated with more positive ratings of social skills, again primarily in digital contexts (increasing ratings by around a quarter to a third of a point, to being notably positive).This association was not attributable to differences in personality dimensions, but was likely the result of the increased perspective taking when being online.This impact of social media is attributable to social media being a tool for socialising, so it affords a platform for knowing about and taking others' perspectives.Interestingly, students who expressed high levels of concern for what others do in the online environment (e.g.Do you worry about things your friends (or others) post or comment about (such as in emails, or game chat such as Minecraft") had rated their social skills higher.
An important finding was that high levels of parental monitoring also mattered for social skills.Students who reported 'always being monitored' had higher ratings not only of self-regulation but also of social skills in digital contexts (0.49 of a point difference).
Question Four: Were There Links Between Self-regulation and Social Skills with Writing Achievement?(Paper C) Yes, but in very specific areas of self-regulation and social skills.
Inhibitory Control (e.g."I can stick with my plans and goals") was the important component in self-regulation related to writing gains.An increase of 1 point in 1 3 New Zealand Journal of Educational Studies (2023) 58:361-377 ratings of being better able to ignore or control competing responses was associated with a gain in writing achievement of about half a year.This was not attributable to the students' ethnicity or gender, nor to their personality, or whether their school was a very experienced one or not.The relationship was not found with the second component we examined, Attention Focus, nor was it found in the specific measures of self-regulation in digital contexts.It was the general capability of managing distractions that mattered.This is only the second study to report the impact of Inhibitory Control with writing (c.f.Drijbooms et al., 2017).It has been found more often studies with reading achievement (Dodell-Feder & Tamir, 2018).
The second relationship was between Cognitive Empathy (e.g."I can often understand how people are feeling even before they tell me") and writing, a connection that had not been reported in the literature before.An increase of one point in Cognitive Empathy meant an achievement advantage of more than a school term.Again, only Cognitive Empathy was related to achievement.The other measures of Affective Empathy, Prosocial Skills and Perspective Taking, as well as these social skills in digital contexts, were not.
Relationships between some forms of reading and empathy have been found previously (Dodell-Feder & Tamir, 2018).This new finding for writing likely reflects how writing is taught and assessed, where there is an emphasis on knowing and writing in different ways for different audiences (Paper C).

Implications
Along with many other jurisdictions, in Aotearoa-New Zealand there is a pressing need to develop children's resilience and well-being, a need that is exacerbated in the digital worlds, with existential risks such as cyber bullying and victimisation (Zhu et al., 2021), and vulnerability to mis-dis-and mal-information (Kahne & Bowyer, 2018;Mair et al., 2019).Greater self-regulation and more positive social skills contribute to mitigating these (Nascimbeni & Vosloo, 2019;Zhu et al., 2021) and also to more general areas of well-being (Durlak et al., 2022;Smith et al., 2022).They do this by enabling awareness and control over behaviours, emotions and cognitions (including the critical thinking skills) associated with the risks, and by building the prosocial skills needed to interact positively within online communities.Collective resilience is especially important in Aotearoa New Zealand (Penehira et al., 2014).More positively, the digital tools have affordances which, like other cognitive and social tools might enable us to develop these skills more effectively.
In essence, the whole made up of these slices, depicted in Fig. 1, demonstrates that it is possible to design contexts for schools and homes which promote these skills and contribute to resilience and well-being.But just providing 1:1 access to digital devices will not guarantee that.Having a curriculum that celebrates these key competencies, without access to detailed pedagogical and instructional models and parental guidance for how to promote them, will not be sufficient.

Students' Skills: Learning and Development
One of the most important findings is how students' awareness of the need to selfregulate is distinctly different in the digital contexts and distinctly impacted by using digital tools.Even being highly self-regulating in general, or having general predispositions to be conscientious, will not guarantee confidence in having the skills and strategies to self-regulate with the digital tools.This means that our students need help to develop specific strategies and knowledge as well as the awareness of conditions to do this well.Social media is particularly problematic and warrants particular attention when providing guidance to teachers and parents and for them to students.
Students' social skills (perspective taking, sociability and empathy), and more general predispositions appear to be more generalizable to their use of social media and the Internet and not particularly impacted by the school-based intervention.But it is also the case that there was less focus on these skills in the overall intervention.However, use of social media at home was associated with higher ratings.The implication is that more deliberate attention to these skills could be influential at school and at home.We know they can be developed further as shown by studies into how the pedagogy could be augmented to increase skills such as argumentation and collaborative reasoning which have both cognitive and social skills components (such as perspective taking) (Papers G and H).

School Design: Communities, Pedagogy and Instruction
These slices within and across the schooling and home context show that the innovative intervention had an important effect not just on writing achievement but also on intra-and inter-personal skills.Importantly, it was scalable to a new group of schools.The evidence from the various slices also suggests the following three general conditions are required if we are to build a broad spectrum of social and emotional skills for and through digital tools at school and at home.
(1) An overall design for ubiquitous digital use in schools.Dimensions to that design include: providing a progressive evidence-based pedagogy which provides common practices and supports local adaptations; appropriate resourcing of technological infrastructure and support; the availability of investors especially for communities with compromised access to funding; pedagogical and strategic leadership; access to research and development through research partnerships; strong reciprocal and dynamic relationships with local communities; and the administrative capability to bring these together.(2) Deliberate design for promoting social and emotional skill though and with learning areas such as writing.The twin vehicles for this are school-wide communities of practice and classroom level pedagogy and instruction.Students need agency in this and they need specific strategies and knowledge.Also, specific teaching of the skills for the digital context, as well as developing generalizable capability.It may require a greater focus on the social skills.
1 3 New Zealand Journal of Educational Studies (2023) 58:361-377 (3) Mutually supportive guidance and support between schools and whānau/families.Home usage adds to the impact schools might have on the valued skills.The impact of whānau/families on these skills can be positive or negative and depends considerably on the knowledge and confidence parents have to support their children.That support is likely to be dynamic, and not just focused on mitigating the risks through monitoring and surveillance, but also on developing their children's positive skills of self-regulation and sociability.But, because of the benefits of coherence, the best sources for ongoing development of the specific parenting support is likely to be their schools.

Parents and Whānau
How parents think about digital tools and the risks and benefits, and their approach to supporting their children's learning and development with digital tools varies.Knowing about the characteristics of different parents might enable more precision in the provision of the advice.The 'Novice' group for example, was unfamiliar with digital tools, and had no or little experience at home or at work.They had the greatest concern about the Internet, provided the lowest frequency of guidance and used fewer than three strategies.This means that targeted advice about strategies would be important.But the novice group also needed greater access to and familiarity with the tools and the Internet themselves as conditions for being able to act on that advice.
et al., (2022a, 2022b) In school and out of school digital use effects on self-regulation and social skills Paper C McNaughton et al., (2022a, 2022b) Relationships between self-regulation, social skills and writing achievement Paper D McNaughton et al. (2023) Intervention effects on self-regulation and social skills Paper E McNaughton et al. (in press) Parenting beliefs and strategies for self-regulation and social

Fig. 1
Fig. 1 Predicted relationships between contexts, skills and writing achievement