Universal, School-Based Social

and Emotional Learning 2

Interventions and Their Potential
to Improve Students’ Mental
Health

Neil Humphrey

 hat Do We Mean When We Talk

W thinking about population mental health, where it
About Mental Health, and Why can capture greater variability than a solely
Does Mental Health Matter? symptom-driven approach (Alexander et al.,
2020).
In this chapter I adopt the ‘complete state’ view The societal significance of complete mental
of mental health (Keyes, 2005), also sometimes health (e.g. high well-being, low symptoms) can-
referred to as the ‘dual factor’ model (Petersen not be overstated. Mortality studies demonstrate
et al., 2020), in which mental health is theorised that well-being is associated with longer life
as simultaneously comprising our experience of (Chida & Steptoe, 2008). In children, higher lev-
symptoms of psychological distress (e.g. anxiety, els of well-being are concurrently and prospec-
depression) and well-being (e.g. life satisfaction, tively associated with better academic attainment
positive affect). These are not proposed to form a (Gutman & Vorhaus, 2012). By contrast, mental
single bipolar dimension, but rather correlated health difficulties lead to reduced quality of life,
unipolar dimensions that together form a com- destabilisation of communities, and higher rates
plete state of mental health. of health, education and social care utilisation
These dimensions share a complex relation- (Humphrey, 2018). The global direct (e.g. health-
ship, with different determinants. For example, it care) and indirect (e.g. productivity and income
is possible to experience elevated symptoms of loss) economic cost of these difficulties is esti-
mental health difficulties alongside high levels of mated at US$2.5 trillion (Trautmann et al., 2016),
well-being. In dual factor nomenclature, this stra- and they account for 13% of disability-adjusted
tum of the population is referred to as the symp- life years (Vigo et al., 2016). The case for invest-
tomatic but content class, whose social ing in prevention is therefore extremely strong,
determinants are distinct from other symptomatic particularly during the school years.
groups (such as those who are troubled, experi- Approximately one in eight children and adoles-
encing high symptoms and low well-being) cents across the world experience clinically sig-
(Petersen et al., 2020). The complete state per- nificant mental health problems (Polanczyk et al.,
spective is a particularly useful model when 2015), and where recent data are available, preva-
lence appears to be increasing over time, particu-
N. Humphrey (*) larly since the COVID-19 pandemic began (e.g.
Sarah Fielden Chair: Psychology of Education, in England; Vizard et al., 2020). Furthermore,
Manchester Institute of Education, University most lifetime cases of mental illness have their
of Manchester, Manchester, UK first onset in adolescence (Jones, 2013). Those
e-mail: Neil.humphrey@manchester.ac.uk

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 9

S. W. Evans et al. (eds.), Handbook of School Mental Health, Issues in Clinical Child Psychology,
https://doi.org/10.1007/978-3-031-20006-9_2
10 N. Humphrey

who experience mental health difficulties during mental health provision includes the requirement
childhood and/or adolescence go on to experi- for every school to have a designated mental
ence poorer physical and mental health, are less health lead, and the creation of mental health
likely to be employed and more likely to incur support teams, managed jointly by schools and
additional societal costs (e.g. criminal justice) as the National Health Service (Department for
adults (D’Amico et al., 2014; Goodman et al., Education/Department of Health, 2017).
2015; Knapp et al., 2011). Finally, from birth to However, such developments place increasing
midlife, less than 20% of people experience demand on the teaching workforce without guar-
‘enduring mental health’ (that is, they never anteeing any additional resources to support this.
experience a period of significant distress), mak- More generally, we know that many teachers feel
ing at least one episode of impairing mental inadequately prepared to engage with mental
health difficulties the norm, rather than the excep- health issues in the classroom, though the extent
tion (Schaefer et al., 2017). of training available at the school level appears to
be related to their perceived capacity in this
regard (Mansfield, et al., 2021). In other words,
 hy Are Schools Important
W the more mental health training available in a
Settings for the Promotion given school, the more teachers within it report
of Mental Health? feeling that they have the capacity to undertake
mental health-related practices as part of their
School is a critical developmental context for role.
children and young people (Bronfenbrenner, The role of school staff in promoting mental
2005), in which many key determinants of mental health can arguably be distilled into four distinct
health are primarily situated (e.g. bullying) but related areas of work. First, they can provide
(Patalay & Fitzsimons, 2016). Schools benefit a nurturing environment in which children and
from very wide reach, a prolonged period of young people feel safe and happy. Second, school
engagement and a central role in most communi- staff can monitor and assess mental health needs
ties (Greenberg, 2010). If parents are concerned in the student population, and identify those with
about their child’s mental health, they are most emerging or established difficulties. Third, they
likely to contact their teacher(s) as a first port of can provide support for mental health needs.
call (Ford et al., 2007). Fourth, where necessary, school staff can refer
Furthermore, children’s mental health and children and young people to external services
their learning are concurrently and temporally and agencies (e.g. child and adolescent mental
related; for example, girls’ academic attainment health services) for more specialised and inten-
in middle childhood predicts later emotional sive intervention than they are able to provide
symptoms, even after accounting for prior symp- themselves (NatCen Social Research & the
tom levels and risk factor exposure (Panayiotou National Children’s Bureau Research and Policy
& Humphrey, 2018). Collectively, these findings Team, 2017). In this chapter, my focus is univer-
support the view of school as an important setting sal interventions, and so the discussion that fol-
for the promotion of well-being and prevention lows pertains primarily to the first and third areas
of the development, maintenance or escalation of noted above. However, it is important to note that
mental health difficulties among children and these interventions do not occur in a vacuum.
young people (Greenberg, 2010). Accordingly, Schools are complex ecological systems, and
there has been an increased policy emphasis on provision in the second and fourth areas, in addi-
this issue in recent years. For example, in tion to the broader socialisation practices and
England, mental health education was made interactions that occur in school (e.g. that which
compulsory in all schools in 2020 (Department is ‘caught’ as opposed to being ‘taught’), will
for Education, 2019); alongside this, an ongoing have a strong bearing on students’ experiences
plan to transform children and young people’s and outcomes relating to their mental health.
2 Universal, School-Based Social and Emotional Learning Interventions and Their Potential to Improve… 11

 he Rationale for Universal, School-­

T develop difficulties (Greenberg & Abenavoli,
Based Interventions 2017). Second, the relatively ‘light touch’
approach taken in universal interventions (com-
Universal, school-based interventions are defined pared to targeted/indicated approaches) means
as those that are for all students, regardless of that children who are at risk may not benefit
need. They are therefore distinct from targeted/ (Greenberg, 2010). Though the assumption that
selective (for those at increased risk or with students will not all respond to an intervention in
emergent difficulties) and indicated (for those a uniform manner is sound, we still do not know
with established/diagnosed difficulties) interven- enough about exactly who benefits more or less
tions (Foxcroft, 2014). In the literature discussed from universal interventions (Durlak et al.,
throughout this chapter, reference is variously 2011). Third, if targeted and/or indicated inter-
made to ‘school-wide’, ‘whole-school’ and ventions are always needed as part of a tiered
‘multi-component’ approaches. Although there approach to intervention (as is the case in educa-
are some important distinctions between these tion systems around the world), one might ask
terms (e.g. whole-school is used in Europe to what exactly the universal layer is preventing
describe programmes characterised by work (Humphrey et al., 2013)? Finally, the assumption
across multiple system levels in a school to enact that universal interventions are cost-­ effective
change – hence an intervention may technically remains largely untested (McCabe, 2008). These
be universal but not whole-school; Demkowicz & are issues to which I will return later in the
Humphrey, 2019), they are united by their funda- chapter.
mental emphasis on all rather than some students. Even a cursory glance at the evidence base
The rationale for universal interventions is multi-­ reveals a very wide range of universal, school-­
faceted. First, use of universal interventions based interventions that may influence student
aligns with the public health approach to mental mental health outcomes. A useful distinction to
health promotion (e.g. Embry’s [2011] notion of be made at this point is between those where
‘behavioural vaccines’). Second, they are poten- mental health is the primary focus and those
tially more cost-effective than targeted/indicated where it is a secondary focus. Examples of the
approaches because even though treatment former include those where intervention content
effects are expected to be more modest, universal and processes focus directly on the development
interventions are much less resource-intensive of protective strategies to prevent the emergence
(McLaughlin, 2011). Third, universal approaches of symptoms of anxiety and depression
may serve to reduce stigma (Greenberg, 2010). (Johnstone et al., 2018), and those that focus on
Conversely, targeted/indicated interventions may mental health education/literacy (Wei et al.,
yield unintended negative consequences (e.g. iat- 2013).
rogenic ‘deviancy training’ effects; Evans et al., Examples of the latter include interventions
2015). Finally, universal school-based interven- focused on social and emotional learning (SEL)
tions can influence outcomes for children and (Wigelsworth et al., 2016), substance abuse pre-
young people who would not otherwise access vention (Onrust et al., 2016) and behaviour man-
the support they need through usual care path- agement (Korpershoek et al., 2016). Although the
ways (given that most who experience significant range of available programmes does not neatly
mental health difficulties do not get specialist reside within a single category, the distinction
support; NHS Digital, 2018). remains an important one, not least in terms of
However, the above arguments are counter- expectation management regarding the magni-
balanced by a series of concerns about the pre- tude and timing of intervention effects on mental
dominance of universal provision. First, the low health outcomes. In other words, we would natu-
prevalence of mental health difficulties means rally expect more substantial and immediate
that much of the effort in universal approaches is intervention effects on student mental health in
expended on children who are unlikely to interventions where this is the principal focus,
12 N. Humphrey

compared to those where it is a secondary Kingdom, Ireland, Netherlands, Denmark,

consideration. France) and the world (e.g. the United States,
In this chapter, my focus is primarily on those Chile, India), Zippy’s Friends (ZF) aims to equip
classified as social and emotional learning inter- children with the social and emotional skills that
ventions; accordingly, the following section pro- enable effective coping in difficult circumstances.
vides an overview of this body of work. This intervention is characterised by eight key
principles, as follows: (i) children choose their
own solutions; (ii) positive skills are reinforced;
What Are Social and Emotional (iii) repetition and continuity are essential for
Learning Interventions? learning; (iv) abilities are developed in different
settings; (v) children are active participants; (vi)
Universal, school-based social and emotional children help each other; (vii) children evaluate
learning (SEL) interventions aim to develop the their own success; and (viii) teachers are open to
social and emotional skills (e.g. self-awareness, listening to children (Partnership for Children,
self-management, social awareness, relationship 2016). The intervention follows a modular
skills, responsible decision-making) of students approach built around six stories about Zippy, a
through explicit instruction in the context of stick insect and his friends (a group of children).
learning environments that are safe, caring, well-­ The stories focus on feelings, communication,
managed and participatory (Humphrey, 2013; making and breaking relationships, conflict reso-
Weissberg et al., 2015). These skills have consid- lution, dealing with change and loss, and coping.
erable utility. They help children to effectively Each story is explored over the course of four
navigate the social world and promote resilience weekly sessions, wherein part of the story is read
to bullying and victimisation, violence and a by the teacher and children then participate in a
wide range of other negative processes and out- range of activities including games, drawing and
comes (Sklad et al., 2012). Crucially, SEL skills discussion. Sessions follow a common format
also facilitate learning in the classroom (Durlak that begins with a review of previous learning
et al., 2011). Learning is a social process and it and ends with each child providing feedback to
stands to reason that improved social and emo- reflect their feelings (Partnership for Children,
tional competence will facilitate academic suc- 2016).
cess. Furthermore, longitudinal studies highlight
the predictive utility of childhood social–emo-
tional competencies for mental health and labour Promoting Alternative Thinking
market outcomes in later life (Goodman et al., Strategies
2015). Accordingly, effective promotion of SEL
skills has emerged as a policy priority in educa- The Promoting Alternative Thinking Strategies
tion systems around the world (Marcelino Botin (PATHS) curriculum aims to help children aged
Foundation, 2015). Below I provide two brief 4–11 to manage their behaviour, understand their
case examples of SEL interventions. The inter- emotions and work well with others (Greenberg
ested reader can find further examples in the & Kusche, 1993). It has been implemented in a
Collaborative for Academic, Social and variety of countries around the world, including
Emotional Learning’s (CASEL) programme the United States, the United Kingdom,
guide (CASEL, 2013). Switzerland and Croatia. PATHS is delivered by
class teachers and includes a series of lessons on
topics such as identifying and labelling feelings,
Zippy’s Friends controlling impulses and understanding other
people’s perspectives, with associated physical
Implemented in early elementary education (ages resources and artefacts (e.g. Feelings Face cards,
5–7) settings across Europe (e.g. the United Feelings Dictionaries and posters relating to
2 Universal, School-Based Social and Emotional Learning Interventions and Their Potential to Improve… 13

PATHS concepts and strategies). Lessons are can need regulating when they threaten to over-
intended to be delivered approximately twice a whelm or need to be amplified… these [social–
week throughout the year. These are supported emotional] skills help them to experience more
by generalisation activities and techniques that well-being and maintain satisfying relationships
support the application of new skills during the with others’ (Denham, 2006, p. 70). Research on
school day, and parent materials that aim to the determinants of well-being provides support
extend learning to the home environment. In for these propositions. For example, our own
addition to this, a daily procedure of compliment-­ research has demonstrated an inverse relationship
giving is encouraged using the ‘Kid of the Day’ between SEL skills and mental health difficul-
system, in which children are randomly selected ties, both concurrently (Humphrey &
and wear a badge or identifier to be recognisable Wigelsworth, 2012) and longitudinally
to other pupils and staff around the school. The (Panayiotou et al., 2019b).
Kid of the Day may be assigned special roles and
responsibilities, and other pupils and staff com-
plete a compliment sheet for them. Teachers in  hat Does the Evidence Base Tell
W
PATHS schools usually are aided by trained Us About the Potential of Universal
external coaches, who offer ongoing technical SEL Interventions to Improve
support and assistance (e.g. lesson modelling, Students’ Mental Health?
observation and feedback) throughout the school
year as a means to optimise implementation The SEL evidence base has grown exponentially
(Humphrey et al., 2018). in the last three decades. Unsurprisingly, this has
resulted in the publication of multiple meta-­
analyses (Corcoran et al., 2018; Durlak et al.,
 ow and Why Might SEL
H 2011; Sklad et al., 2012; Taylor et al., 2017;
Interventions Improve Students’ Wigelsworth et al., 2016). These provide rigor-
Mental Health? ous evidence that illustrates the impact of SEL
interventions on a range of outcomes, including
Before we examine evidence on the efficacy of social and emotional skills, school attitudes, aca-
SEL interventions in improving students’ mental demic performance and, importantly, mental
health, it is important to first consider underpin- health. In terms of the latter, aggregated effect
ning theory. SEL theory (e.g. the SEL logic sizes (ES) observed in relation to internalising
model; CASEL, 2007) and models of risk and problems (e.g. anxiety) range from 0.19 (Sklad
resilience processes in human development (e.g. et al., 2012; Wigelsworth et al., 2016) to 0.24
Wright et al., 2013) both highlight the impor- (Durlak et al., 2011). Larger but more variable
tance of social– emotional competence in serving ES are reported for externalising difficulties (e.g.
important promotive and protective functions, conduct problems), ranging from 0.22 (Durlak
and accordingly, they have been described as, et al., 2011) to 0.43 (Sklad et al., 2012). Meta-­
‘the skills and competencies that underlie mental analyses of longer-term follow-up studies indi-
health’ (Weare & Markham, 2005, p. 14). As pre- cate that intervention effects are still evident, but
viously noted, SEL skills help children and young attenuate somewhat over time. Thus, Sklad et al.
people to navigate their social environment suc- (2012) reported average intervention ES of 0.1
cessfully, particularly in difficult or challenging and 0.2 (for internalising and externalising diffi-
circumstances. Students who are able to under- culties, respectively) in studies where measures
stand, articulate and manage their emotions, were taken at least seven months after a given
while also being better equipped to develop and intervention was concluded.
maintain positive social relationships (including Analysing studies with a follow-up period of
social problem solving), are more likely to expe- at least 24 weeks post-intervention, Taylor et al.
rience greater levels of positive affect: ‘Emotions (2017) reported average intervention ES of 0.16
14 N. Humphrey

for internalising symptoms and 0.14 for external- Education, 2014), and the likelihood of later
ising problems. None of the SEL meta-analyses escalation of such problems and the huge societal
published to date has reported aggregated effects costs that can accrue as a result if they are not
on well-being, probably owing to a lack of pri- effectively addressed at an early stage (e.g. Scott
mary studies. However, findings from individual et al., 2001), the effects of SEL interventions
studies are promising. For example, Panayiotou must be considered very promising indeed.
et al. (2019b) found that the aforementioned However, it is important to remember that such
PATHS curriculum produced an intervention ES effects are not uniform. The next step in this
of 0.17 in relation to children’s well-being. chapter, therefore, is to consider some common
To what extent can these intervention effects intervention effect modifiers.
be considered meaningful? A preliminary caution
here is to resist the temptation to reflexively
resort to the effect size thresholds outlined by I ntervention Effect Modifier 1:
Cohen (1992), since these are completely devoid Implementation Variability
of context and are misaligned with empirically
derived intervention effect sizes in prevention Implementation is, ‘the process of putting a prac-
science (Tanner-Smith et al., 2018). Instead, we tice or program into place’ (Forman, 2015, p. 10).
might start by asking how the magnitude of SEL Dimensions of implementation include behav-
intervention effects on mental health outcomes iours of the implementer, such as fidelity (whether
compares to those observed in the broader field prescribed procedures were followed), adapta-
of universal school-based interventions (which tions (what changes were made to an interven-
includes, for example, those designed to promote tion), dosage (how much of an intervention was
healthy eating, prevent substance abuse or man- delivered) and quality (how well an intervention
age behaviour in the classroom). Here, there is was delivered), and those of recipients, such as
reason for optimism; even when one adopts a reach (whether intended recipients were present
conservative approach (e.g. the smallest average when the intervention was delivered) and respon-
ES noted above for internalising [0.19] and exter- siveness (the extent to which recipients engaged
nalising [0.22] problems), one places SEL inter- with an intervention) (Berkel et al., 2011).
ventions above the 50th percentile in the Increasingly, contextual factors such as pro-
distribution of effect sizes for these outcomes gramme differentiation (the extent to which an
among all universal school-based interventions intervention is distinct from existing practice) are
(Tanner-Smith et al., 2018). An alternative per- also considered under the implementation rubric.
spective is to consider what these intervention It is now widely accepted that these dimensions
effects mean in practical terms – in other words, are likely to vary when SEL interventions are
do they translate to genuine, noticeable effects in implemented in schools. Thus, in studies where
daily life? This is, of course, highly subjective, implementation data are recorded, nearly 40%
but Durlak et al. (2011) argue that the kinds of report problems relating to one or more of the
gains evidenced for SEL interventions would be dimensions noted above (Durlak et al., 2011;
noticeable in typical classroom contexts. For Wigelsworth et al., 2016). Research has demon-
example, the most conservative estimate for the strated clearly that this variability influences the
impact of SEL on externalising problems noted achievement of intended outcomes (Durlak,
above translates to a 9-percentile point improve- 2016). For example, in Durlak et al.’s (2011) SEL
ment (Durlak, 2009). Given the fact that even meta-analysis, the average intervention ES on
very modest decreases in externalising problems emotional symptoms in studies reporting no
can have positive consequences for the broader implementation problems was 0.35, compared to
school environment (e.g. up to an hour of learn- 0.15 in studies where implementation problems
ing a day may be lost as a consequence of persis- were noted. Early evidence indicates a similar
tent disruptive behaviour; Office for Standards in pattern in relation to well-being.
2 Universal, School-Based Social and Emotional Learning Interventions and Their Potential to Improve… 15

In their aforementioned trial of the PATHS The compensatory effects hypothesis predicts
curriculum, Panayiotou et al. (2019a) observed that at-risk children will benefit more from SEL
that the magnitude of intervention ES on well-­ interventions because they are at greater risk and
being grew from 0.17 in their intent-to-treat anal- have more room for improvement (McClelland
ysis to 0.43 when complier average causal effect et al., 2017). Thus, SEL can offset the significant
estimation (CACE) was employed in order to disruption of developmental processes brought
take account of variability in dosage. about by risk exposure. Several studies have pro-
vided support for the compensatory effects
hypothesis. For example, the Conduct Problems
Intervention Effect Modifier 2: Prevention Research Group (2010) reported
Subgroup Effects greater benefits of the Fast Track intervention
(which combines the PATHS curriculum with
As noted earlier, it stands to reason that children parent training and other supports) among chil-
and young people will not respond uniformly to dren with higher baseline levels of aggression.
SEL interventions. However, we still know rela- Similarly, Low et al.’s (2015) trial of Second Step
tively little about exactly who benefits more or found that this SEL intervention primarily pro-
less from them (Durlak et al., 2011). An initial duced significant improvements in social skills
problem here is how to robustly investigate indi- and mental health among children who started
vidual differences in responsiveness to interven- the school year with skill deficits relative to their
tion while avoiding ‘data dredging’ (that is, peers. We know that these results are not
systematically searching through a dataset in the explained by regression to the mean because sim-
hope of finding a significant intervention effect; ilar trends were not evident in the trial control
Keller, 2019). It is therefore recommended that group. The findings of such studies are therefore
subgroup analyses are specified in advance, encouraging because they indicate that SEL
informed by theory and/or research, and include interventions do indeed benefit those most in
clear specification of the expected direction of need of support.
effects and population subgroup(s) of interest In contrast, the accumulated advantages
(using characteristics measured pre-­hypothesis (also known as the ‘rich get richer’
randomisation in trials, e.g. demographic charac- model) predicts that children from more advan-
teristics, individual differences at baseline and/or taged, lower-risk backgrounds will benefit more
family factors) (Farrell et al., 2013). from SEL interventions because they are better
I focus here on subgroup moderator effects equipped to take advantage of learning opportu-
among students deemed to be ‘at risk’ by virtue nities and more capable of consolidating and
of their existing levels of need (e.g. elevated building on their existing skills (McClelland
symptoms of distress at baseline in a given study) et al., 2017). This prediction was borne out in a
and/or socio-economic and other circumstances trial of the PATHS curriculum in Croatia, where
(e.g. those from more deprived backgrounds) the researchers reported significant improve-
because these are central to the issues noted ear- ments in SEL skills and reductions in mental
lier (see section The Rationale for Universal, health problems only among those students
School-Based Interventions). Furthermore, classed as ‘above average’(low risk) in pre-­
although common, subgroup analyses relating to intervention assessments (Novak et al., 2016).
demographic characteristics such as sex and age Though fewer in number, studies like this set a
tend to be poorly theorised, if at all (in other challenging precedent because they indicate that
words, while researchers frequently test to see if the benefits of SEL go to those who are already in
interventions affect boys and girls differently, positions of relative advantage.
they usually do not explain their justification for
doing so).
16 N. Humphrey

 re Universal SEL Interventions

A assumptions relating to the costing approach
‘Worth It’? taken and time horizon adopted for a given analy-
sis. Third, these analyses routinely use an intent-­
In the preceding sections, I hope to have con- to-­
treat approach that does not account for
vinced the reader that universal SEL interven- variability in levels of implementation. As a
tions can produce meaningful improvements to result, cost-effectiveness estimates based on
students’ mental health outcomes. Despite this, it moderate or high levels of compliance (as in the
is important to note that just because SEL inter- aforementioned CACE models) are currently
ventions are effective, this does not necessarily lacking.
mean that they are cost-effective. It is here where
economic analyses (e.g. basic cost, cost-­
effectiveness, cost-utility, cost-consequence and Current and Future Directions
benefit-cost analyses) have great utility, as they
provide critical information that can help inform At the time of writing, there is an accumulated
decision-making about how best to deploy scarce body of robust evidence to support the proposi-
resources by examining intervention effects in tion that universal SEL interventions can improve
the context of the costs that were required to gen- students’ mental health. However, there is still
erate them. However, economic research on SEL much that we do not know. First, more economic
interventions is in its infancy (McClelland et al., analyses are required; indeed, a rigorous cost-­
2017). effectiveness analysis (or equivalent) should
Indeed, an early review of universal, school-­ become a fundamental component of future trials
based mental health interventions (including in this area (Schmidt et al., 2020). Second, given
SEL) found no published studies (McCabe, what we know about the inevitability of imple-
2008), and a recent systematic review only iden- mentation variability, CACE or related instru-
tified nine (Schmidt et al., 2020). Those studies mental variable approaches should also be
that have been published yield tentative promise. undertaken as standard (Peugh & Toland, 2017).
Analyses reported by Turner et al. (2020) deter- Third, an increased emphasis on the factors that
mined that the PATHS curriculum was likely to facilitate or inhibit effective implementation is
be cost-effective under most, but not all scenarios warranted, as this can inform future programme
(e.g. different costing approaches and time hori- training and implementation support activities.
zons). Hunter et al.’s (2018) examination of the Fourth, since a key purpose of universal SEL
cost-effectiveness of the Social Skills interventions is to alter developmental trajecto-
Improvement System Classwide Intervention ries, it is important that this is reflected in the
Programme drew similar conclusions. analytical techniques adopted by researchers;
In interpreting the results of such studies, sev- hence, the use of growth curve models is recom-
eral issues need to be borne in mind. First, the mended (Greenberg & Abenavoli, 2017). The
economic perspective adopted should be taken field is currently limited by a reliance on ‘point-­
into account. For example, in the Turner et al. in-­time’ estimates that do not analyse the devel-
(2020) study, a UK Health Service perspective opmental process of growth (although there are a
was adopted. In other words, the economic ben- couple of notable exceptions, e.g. Nix et al.,
efit was quantified based on improvements to 2016). Finally, a shift away from programmatic
health-related quality of life, for which the UK approaches is underway, with a parallel increase
Health Service has a ‘willingness to pay’ thresh- in research on the constituent components that
old per quality adjusted life year. Second, cost-­ drive improvement in outcomes (Jones &
effectiveness estimates are sensitive to key Bouffard, 2012).
2 Universal, School-Based Social and Emotional Learning Interventions and Their Potential to Improve… 17

References Department for Education. (2019, February 25). All pupils

will be taught about mental and physical wellbeing.
https://www.gov.uk/government/news/all-­pupils-­will-­
Alexander, L. M., Salum, G. A., Swanson, J. M., & be-­taught-­about-­mental-­and-­physical-­wellbeing
Milham, M. P. (2020). Measuring strengths and weak- Durlak, J. A. (2016). Programme implementation in social
nesses in dimensional psychiatry. Journal of Child and emotional learning: basic issues and research find-
Psychology and Psychiatry and Allied Disciplines, ings. Cambridge Journal of Education, 46(3), 1–13.
61(1), 40–50. https://doi.org/10.1111/jcpp.13104 https://doi.org/10.1080/0305764X.2016.1142504
Berkel, C., Mauricio, A. M., Schoenfelder, E., & Sandler, Durlak, J. A. (2009). How to select, calculate, and inter-
I. N. (2011). Putting the pieces together: An inte- pret effect sizes. Journal of Pediatric Psychology,
grated model of program implementation. Prevention 34(9), 917–928. https://doi.org/10.1093/jpepsy/jsp004
Science, 12(1), 23–33. https://doi.org/10.1007/ Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor,
s11121-­010-­0186-­1 R. D., & Schellinger, K. B. (2011). The impact of
Bronfenbrenner, U. (2005). Making human beings human: enhancing students’ social and emotional learning:
Bioecological perspectives on human development. a meta- analysis of school-based universal interven-
Sage Publications. tions. Child Development, 82(1), 405–432. https://doi.
CASEL. (2007). How evidence-based SEL programs org/10.1111/j.1467-­8624.2010.01564.x
work to produce greater student success in school and Embry, D. D. (2011). Behavioral vaccines and evidence-­
life. CASEL. based kernels: Nonpharmaceutical approaches for the
CASEL. (2013). Effective social and emotional learning prevention of mental, emotional, and behavioral dis-
programs: preschool and elementary school edition. orders. Psychiatric Clinics of North America, 34(1),
CASEL. 1–34. https://doi.org/10.1016/j.psc.2010.11.003
Chida, Y., & Steptoe, A. (2008). Positive psychologi- Evans, R., Scourfield, J., & Murphy, S. (2015). The unin-
cal well-being and mortality: A quantitative review tended consequences of targeting: Young people’s
of prospective observational studies. Psychosomatic lived experiences of social and emotional learning
Medicine, 70(7), 741–756. https://doi.org/10.1097/ interventions. British Educational Research Journal,
PSY.0b013e31818105ba 41(3), 381–397. https://doi.org/10.1002/berj.3155
Cohen, J. (1992). A power primer. Psychological Bulletin, Farrell, A. D., Henry, D. B., & Bettencourt, A. (2013).
112(1), 155–159. Methodological challenges examining subgroup dif-
Conduct Problems Prevention Research Group. (2010). ferences: Examples from universal school-based youth
The effects of a multiyear universal social-emotional violence prevention trials. Prevention Science, 14(2),
learning program: The role of student and school 121–133. https://doi.org/10.1007/s11121-­011-­0200-­2
characteristics. Journal of Consulting and Clinical Ford, T., Hamilton, H., Meltzer, H., & Goodman, R.
Psychology, 78(2), 156–168. https://doi.org/10.1037/ (2007). Child mental health is everybody’s busi-
a0018607 ness: The prevalence of contact with public sec-
Corcoran, R. P., Cheung, A., Kim, E., & Xie, C. (2018). tor services by type of disorder among British
Effective universal school-based social and emo- school children in a three-year period. Child and
tional learning programs for improving academic Adolescent Mental Health, 12(1), 13–20. https://doi.
achievement: A systematic review and meta-­analysis org/10.1111/j.1475-­3588.2006.00414.x
of 50 years of research. Educational Research Forman, S. G. (2015). Implementation of mental health
Review, 25(1), 56–72. https://doi.org/10.1016/J. programs in schools: A change agent’s guide.
EDUREV.2017.12.001 American Psychological Association.
D’Amico, F., Knapp, M., Beecham, J., Sandberg, S., Foxcroft, D. R. (2014). Can prevention classification
Taylor, E., & Sayal, K. (2014). Use of services and be improved by considering the function of preven-
associated costs for young adults with childhood tion? Prevention Science, 15(6), 818–822. https://doi.
hyperactivity/ conduct problems: 20-year follow- org/10.1007/s11121-­013-­0435-­1
­up. The British Journal of Psychiatry : The Journal Goodman, A., Joshi, H., Nasim, B., & Tyler, C. (2015).
of Mental Science, 204(6), 441–447. https://doi. Social and emotional skills in childhood and their
org/10.1192/bjp.bp.113.131367 long-term effects on adult life. Early Intervention
Demkowicz, O., & Humphrey, N. (2019). Whole school Foundation.
approaches to promoting mental health: what does the Greenberg, M. T. (2010). School-based prevention: cur-
evidence say? Anna Freud Centre. rent status and future challenges. Effective Education,
Denham, S. A. (2006). Social-emotional competence as 2(1), 27–52.
support for school readiness: What is it and how do we Greenberg, M. T., & Kusche, C. A. (1993). Promoting
assess it? Early Education and Development, 17(1), social and emotional development in deaf children:
57–89. https://doi.org/10.1207/s15566935eed1701_4 The PATHS Project. University of Washington Press.
Department for Education/Department of Health. (2017). Greenberg, M. T., & Abenavoli, R. (2017). Universal
Transforming children and young people’s men- interventions: Fully exploring their impacts and
tal health provision: a green paper. Department of ­potential to produce population-level impacts. Journal
Health. of Research on Educational Effectiveness, 10(1),
18 N. Humphrey

40–67. https://doi.org/10.1080/19345747.2016.1246 of classroom management strategies and classroom

632 management programs on students academic, behav-
Gutman, L. M., & Vorhaus, J. (2012). The impact of pupil ioral, emotional, and motivational outcomes. Review
behaviour and wellbeing on educational outcomes. of Educational Research, 86(3), 1–38. https://doi.
Department for Education. org/10.3102/0034654315626799
Humphrey, N. (2013). Social and emotional learning: A Low, S., Cook, C. R., Smolkowski, K., & Buntain-­
critical appraisal. Sage Publications. Ricklefs, J. (2015). Promoting social– emotional com-
Humphrey, N., & Wigelsworth, M. (2012). Modelling the petence: An evaluation of the elementary version of
factors associated with children’s mental health diffi- Second Step®. Journal of School Psychology, 53(6),
culties in Primary school: A multilevel study. School 463–477. https://doi.org/10.1016/J.JSP.2015.09.002
Psychology Review, 41(3), 326–341. Mansfield, R., Humphrey, N., & Patalay, P. (2021).
Humphrey, N. (2018). Are the kids alright? Examining the Educators’ perceived mental health literacy and capac-
intersection between educational and mental health. ity to support students’ mental health: associations
Psychology of Education Review, 42(3), 4–12. with school-level characteristics and provision. Health
Humphrey, N., Lendrum, A., & Wigelsworth, M. (2013). Promotion International, 36(6), 1621–1632.
Making the most of school-based prevention: learn- Marcelino Botin Foundation. (2015). Social and emo-
ing from the SEAL programme. Emotional and tional education: An international analysis. Marcelino
Behavioural Difficulties, 18(3), 248–260. Botin Foundation.
Humphrey, N., Hennessey, A., Lendrum, A., Wigelsworth, McCabe, C. (2008). Estimating the short term cost effec-
M., Turner, A., Panayiotou, M., Joyce, C., Pert, K., tiveness of a mental health promotion intervention
Stephens, E., Wo, L., Squies, G., Woods, K., Harrison, in primary schools. National Institute for Health and
M., & Calam, R. (2018). Evaluating the e­ fficacy of the Care Excellence.
Promoting Alternative Thinking Strategies (PATHS) McClelland, M. M., Tominey, S. L., Schmitt, S. A., &
curriculum in promoting social and emotional well Duncan, R. (2017). SEL interventions in early child-
being among children in primary school: a cluster ran- hood. The Future of Children, 27(1), 33–47.
domised controlled trial, process evaluation and eco- McLaughlin, K. A. (2011). The public health impact of
nomic analysis. Public Health Research, 6(10), 1–116. major depression: A call for interdisciplinary preven-
Hunter, L. J., DiPerna, J. C., Hart, S. C., & Crowley, M. tion efforts. Prevention Science, 12(2), 361–371.
(2018). At what cost? Examining the cost effective- NatCen Social Research & the National Children’s
ness of a universal social–emotional learning program. Bureau Research and Policy Team. (2017). Supporting
School Psychology Quarterly, 33(1), 147–154. mental health in schools and colleges. Department for
Johnstone, K. M., Kemps, E., & Chen, J. (2018). A meta-­ Education.
analysis of universal school- based prevention pro- NHS Digital. (2018). Mental health of children and young
grams for anxiety and depression in children. Clinical people in England, 2017. NHS Digital.
Child and Family Psychology Review, 21(4), 466–481. Nix, R., Bierman, K. L., Heinrichs, B. S., Gest, S. D.,
https://doi.org/10.1007/s10567-­018-­0266-­5 Welsh, J., & Domitrovich, C. E. (2016). The
Jones, P. B. (2013). Adult mental health disorders and their randomized-­ controlled trial of Head Start REDI:
age at onset. British Journal of Psychiatry, 202(s54), Sustained effects on developmental trajectories of
s5–s10. https://doi.org/10.1192/bjp.bp.112.119164 social-emotional functioning. Journal of Consulting
Jones, S. M., & Bouffard, S. M. (2012). Social and emo- and Clinical Psychology, 84(4), 310–322. https://doi.
tional learning in schools: From programs to strate- org/10.1037/a0039937
gies. Social Policy Report, 26(4), 1–33. Novak, M., Mihić, J., Bašić, J., & Nix, R. L. (2016).
Keller, F. (2019). Subgroup analysis: “What works best PATHS in Croatia: A school-based randomised-­
for whom and why?”. In Z. Sloboda, H. Petras, controlled trial of a social and emotional learning cur-
E. Robertson, & R. Hingson (Eds.), Prevention of sub- riculum. International Journal of Psychology, 52(2),
stance use (pp. 247–261). Springer. 87–95. https://doi.org/10.1002/ijop.12262
Keyes, C. L. M. (2005). Mental illness and/or men- Office for Standards in Education. (2014). Below the
tal health? Investigating axioms of the complete radar: Low-level disruption in the country’s class-
state model of health. Journal of Consulting and rooms. Ofsted.
Clinical Psychology, 73(3), 539–548. https://doi. Onrust, S. A., Otten, R., Lammers, J., & Smit, F. (2016).
org/10.1037/0022-­006X.73.3.539 School-based programmes to reduce and prevent sub-
Knapp, M., King, D., Healey, A., & Thomas, C. (2011). stance use in different age groups: What works for
Economic outcomes in adulthood and their asso- whom? Systematic review and meta-regression analy-
ciations with antisocial conduct, attention deficit sis. Clinical Psychology Review, 44, 45–59. https://
and anxiety problems in childhood. The Journal of doi.org/10.1016/j.cpr.2015.11.002
Mental Health Policy and Economics, 14(3), 137– Panayiotou, M., Humphrey, N., & Hennessey, A. (2019a).
147. Retrieved from http://www.ncbi.nlm.nih.gov/ Implementation matters: Using Complier Average
pubmed/22116171 Causal Effect Estimation to determine the impact
Korpershoek, H., Harms, T., de Boer, H., van Kuijk, M., of the Promoting Alternative Thinking Strategies
& Doolaard, S. (2016). A meta- analysis of the effects (PATHS) curriculum on children’s quality of life.
2 Universal, School-Based Social and Emotional Learning Interventions and Their Potential to Improve… 19

Journal of Educational Psychology, 112(2), 236–253. Tanner-Smith, E. E., Durlak, J. A., & Marx, R. A. (2018).
https://doi.org/10.1037/edu0000360 Empirically based mean effect size distributions for uni-
Panayiotou, M., Humphrey, N., & Wigelsworth, M. versal prevention programs targeting school-­aged youth:
(2019b). An empirical basis for linking social A review of meta-analyses. Prevention Science, 19(8),
and emotional learning to academic performance. 1091–1101. https://doi.org/10.1007/s11121-­018-­0942-­1
Contemporary Educational Psychology, 56(1), 193– Taylor, R. D., Oberle, E., Durlak, J. A., & Weissberg,
204. https://doi.org/10.1016/j.cedpsych.2019.01.009 R. P. (2017). Promoting Positive Youth Development
Panayiotou, M., & Humphrey, N. (2018). Mental health through school-based social and emotional learning
difficulties and academic attainment: Evidence for interventions: A meta-analysis of follow-up effects.
gender-specific developmental cascades in middle Child Development, 88(4), 1156–1171. https://doi.
childhood. Development and Psychopathology, org/10.1111/cdev.12864
30(2), 523–538. https://doi.org/10.1017/ Trautmann, S., Rehm, J., & Wittchen, H. U. (2016). The
S095457941700102X economic costs of mental disorders: Do our societ-
Partnership for Children. (2016). Zippy’s Friends. http:// ies react appropriately to the burden of mental disor-
www.partnershipforchildren.org.uk/teachers/zippy-­s-­ ders? EMBO Reports, 17(9), 1245–1249. https://doi.
friends-­teachers.html org/10.15252/embr.201642951
Patalay, P., & Fitzsimons, E. (2016). Correlates of men- Turner, A. J., Sutton, M., Harrison, M., Hennessey, A.,
tal illness and wellbeing in children: are they the & Humphrey, N. (2020). Cost- effectiveness of a
same? Results from the UK Millennium Cohort school-based social and emotional learning interven-
Study. Journal of the American Academy of Child & tion: Evidence from a cluster-randomised controlled
Adolescent Psychiatry, 55(9), 771–783. https://doi. trial of the Promoting Alternative Thinking Strategies
org/10.1016/j.jaac.2016.05.019 Curriculum. Applied Health Economics and Health
Petersen, K. J., Humphrey, N., & Qualter, P. (2020). Latent Policy, 18(2), 271–285. https://doi.org/10.1007/
class analysis of mental health in middle childhood: s40258-­019-­00498-­z
Evidence for the dual-factor model. School Mental Vigo, D., Thornicroft, G., & Atun, R. (2016). Estimating
Health, 12(4), 786–800. https://doi.org/10.1007/ the true global burden of mental illness. The Lancet
s12310-­020-­09384-­9 Psychiatry, 3(2), 171–178. https://doi.org/10.1016/
Peugh, J. L., & Toland, M. D. (2017). Psychometric and S2215-­0366(15)00505-­2
quantitative methods for school psychology. Journal of Vizard, T., Sadler, K., Ford, T., Newlove-Delgado, T.,
School Psychology, 60, 5–6. https://doi.org/10.1016/j. McManus, S., Marcheselli, F., et al. (2020). Mental
jsp.2017.01.001 health of children and young people in England, 2020.
Polanczyk, G. V., Salum, G. A., Sugaya, L. S., Caye, A., NHS Digital.
& Rohde, L. A. (2015). Annual research review: A Weare, K., & Markham, W. (2005). What do we know
meta-analysis of the worldwide prevalence of mental about promoting mental health through schools?
disorders in children and adolescents. Journal of Child Global Health Promotion, 12(3), 4–8. https://doi.org
Psychology and Psychiatry and Allied Disciplines, /10.1177/10253823050120030104
56(3), 345–365. https://doi.org/10.1111/jcpp.12381 Wei, Y., Hayden, J. A., Kutcher, S., Zygmunt, A., &
Schaefer, J. D., Caspi, A., Belsky, D. W., Harrington, H., McGrath, P. (2013). The effectiveness of school
Houts, R., Horwood, L. J., et al. (2017). Enduring mental health literacy programs to address knowl-
mental health: Prevalence and prediction. Journal of edge, attitudes and help seeking among youth. Early
Abnormal Psychology, 126(2), 212–224. https://doi. Intervention in Psychiatry, 7(2), 109–121. https://doi.
org/10.1037/abn0000232 org/10.1111/eip.12010
Schmidt, M., Werbrouck, A., Verhaeghe, N., Putman, Weissberg, R. P., Durlak, J. A., Domitrovich, C. E., &
K., Simoens, S., & Annemans, L. (2020). Universal Gullotta, T. P. (2015). Social and emotional learn-
mental health interventions for children and ado- ing: Past, present and future. In J. A. Durlak, C. E.
lescents: A systematic review of health economic Domitrovich, R. P. Weissberg, & T. P. Gullotta
evaluations. Applied Health Economics and Health (Eds.), Handbook of social and emotional learning
Policy, 18(2), 155–175. https://doi.org/10.1007/ (pp. 3–19). Guilford Press.
s40258-­019-­00524-­0 Wigelsworth, M., Lendrum, A., Oldfield, J., Scott, A., Ten-­
Scott, S., Knapp, M., Henderson, J., & Maughan, B. Bokkel, I., Tate, K., & Emery, C. (2016). The influence
(2001). Financial cost of social exclusion: Follow of trial stage, developer involvement and international
up study of antisocial children into adulthood. BMJ transferability on the outcomes of universal social
(Clinical Research Ed.), 323(7306), 191. and emotional learning outcomes: A meta-analysis.
Sklad, M., Diekstra, R., De Ritter, M., Ben, J., & Cambridge Journal of Education, 46(3), 347–376.
Gravesteijn, C. (2012). Effectiveness of school- https://doi.org/10.1080/0305764X.2016.1195791
based universal social, emotional, and behavioral Wright, M., Masten, A. S., & Narayan, A. (2013).
programs: Do they enhance students’ development Resilience processes in development: Four waves of
in the area of skills, behavior and adjustment? research on positive adaptation in the context of adver-
Psychology in the Schools, 49(9), 892–909. https:// sity. In S. Goldstein & R. B. Brooks (Eds.), Handbook
doi.org/10.1002/pits of resilience in children (pp. 15–37). Springer.