Developing a taxonomy to characterise trusted adult support in the lives of adolescents.

INTRODUCTION: A systematic review of trusted adult interventions for adolescents revealed that there was no common terminology, agreed explicit definition, or detail, regarding the personal qualities, functions and roles fulfilled by trusted adults that was used consistently across the papers. To provide clarity, we therefore aimed to produce a taxonomy of trusted adult input, using evidence drawn from our review.
METHODS: Data from the review findings were used to compile the taxonomy, moving from the general to the more specific, in a four stage process. This involved: (1) compilation of elements described in individual papers, (2) grouping of elements derived from stage 1 into categories, (3) examination of context and nature of the relationship, (4) development of a categorisation of trusted adult input.
FINDINGS: The resulting taxonomy encapsulates core essential qualities provided by people acting in trusted adult or mentoring roles, and gives details relating to what a young person might expect from individuals they put their trust in. The taxonomy consists of six categories relating to: delivery context, roles of trusted adults, nature of support, personal qualities, actions/functions, and impact.
CONCLUSIONS: The taxonomy describes key elements that define the trusted adult role, and has the potential to inform the development of policies and guidelines relating to support provision. It may be used as a framework for the reporting of trusted adult interventions within research, and act as a helpful guide if a young person is in doubt about the behaviour or qualities displayed by an adult in their environment.

Introduction

The importance of protective social networks for children has come into sharper focus during recent years National Society for the Prevention of Cruelty to Children (NSPCC) (2016); National Center for Missing and Exploited Children (NCMEC, 2006). Such support may be particularly beneficial during adolescence, due to the health and education choices that are made at this stage. The support of a trusted adult may also be particularly beneficial during a time when increasing independence and physiological maturation may heighten the likelihood of risk taking behaviours (McAteer et al., 2018; Pringle et al., 2017).

We recently conducted a systematic review to identify the influence of trusted adult presence on adolescent health and education outcomes (Pringle, Whitehead, Milne, Scott, & McAteer, 2018). For the purposes of the review, we used the term ‘trusted adult’ as defined by the Scottish Government (Show all authorsLyn Craig2Social Policy Research Centre UNSW Australia Sydney AustraliaSee all articles by this authorScottish Government, 2017): someone who ‘children and young people may turn to for help, and will take them seriously’. This was further defined as support from a specific, dependable (adult) individual, who acts in a responsible manner, rather than (social) support as a more general concept (Orbit, 2017). The methods used to conduct the review (Pringle et al., 2018) are summarised in Table 1, and registered on Prospero (CRD 42017076739).

Table 1

Review methods.

Review Stage	Summary of review process and implementation
Search strategy	Developed using the Medline thesaurus, and adapted for use across all review databases, including: MEDLINE, ERIC, Education Abstracts, Web of Science, ASSIA, Sociological Abstracts, PsycINFO, Social Services Abstracts, IBSS, the Cochrane Library, ProQuest Public Health, and SCIE Social Care Online.In addition: searches in Google/Google Scholar, contact with experts in the field, grey literature searches, screening of reference lists, and citation searches.MeSH headings and key/text words associated with trusted adults and adolescents were sought; alternative terms of reference or evidence of such a role sought within individual papers/reports
Definitions	For this review, a trusted adult was defined as someone who ‘children and young people may turn to for help, and will take them seriously’ [Show all authorsLyn Craig2Social Policy Research Centre UNSW Australia Sydney AustraliaSee all articles by this authorScottish Government, 2017]. This is further defined as support from a specific, dependable [adult] individual, who acts in a responsible manner, rather than [social] support as a more general concept [Orbit, 2017]
Inclusion criteria	• All research designs and types of publication, including unpublished [grey] literature and reports • Adolescents 10–19 years, based on the World Health Organisation's ‘second decade’ definition of adolescence • Where an intervention by [or broader relationship with/contact with] a trusted adult has an influence or impact on an adolescent health or educational outcome • Studies published from 2007 onwards; grey literature from 2012 onwards • English language
Exclusion criteria	• Protocols or studies that did not report outcomes • Populations with specific/specialist health or education needs • English language version not available [but no geographical limitations]
Outcomes	Focus on health [e.g. general wellbeing, health behaviour, mental health etc] or education-related outcomes [e.g. attainment, years in education, educational engagement/attendance etc].
Study selection	Titles and abstracts of identified papers were screened using the broad inclusion criteria of a trusted adult, and adolescent health or education outcomes. Full paper screening of all papers fulfilling these initial criteria. Two reviewers blind screened papers independently at all stages of the screening process, with disagreements being resolved by a 3rd review team member. Duplicate removal and screening was facilitated using Covidence. Searches and screening took place between November 2017 and September 2018.
Data extraction	Data from individual studies were extracted onto a pre-designed table, including year and location of study, study design, study population, type of intervention/support, outcomes, findings, and conclusions; further data was extracted to Excel
Analysis	A descriptive summary of studies was produced.A mind-mapping approach was used to plot, in diagrammatic form, associations and links between individual study outcomes or populations. This helped to identify sub-groups for more in-depth analysis. Further analysis was aligned to study method [s], as appropriate. Limitations and appraised biases were documented
Findings	The Preferred Reporting of Systematic Reviews and Meta-Analysis [PRISMA] chart was used to summarise the processes and findings of the review search and screening process, and detailed results presented

The results of the review are reported elsewhere (Whitehead, Pringle, Scott, Milne, & McAteer, 2019). In brief, the review concluded that it is often difficult to specify the unique impact of the trusted adult role on adolescent health and/or education outcomes. This is in part due to a lack of clarity around the role's operationalisation, particularly in quantitative studies. However, qualitative studies, which have the ability to probe participants' experiences, beliefs and understanding, predominantly indicated that the trusted adult role is positive and that it can support positive health behaviour and educational engagement outcomes.

While the value of mentoring has been recognised for decades (Sipe, 2002), during the course of the review it became apparent that there was no common terminology, agreed explicit definition, or detail regarding the personal qualities, function and roles fulfilled by trusted adults that was used consistently across the papers (Whitehead et al., 2019). In such circumstances, where terms are used inconsistently, and with differing meanings, the development of a taxonomy can be helpful (Ring et al., 2011). A taxonomy, as the Greek root ‘taxis’ implies, provides a means of classifying such information into an organised form, to aid understanding (Hunter, 2018).

Many of the papers in the original review referred to people in the trusted adult role as mentors, but did not give specific detail of what this entailed. The term mentor is derived from Greek mythology, and refers to a ‘wise and trusted counselor or teacher’, or an influential senior sponsor or supporter (Roberts, 1999). Explicit or implicit within the mentoring role, is that it is provided without expectation of personal gain (Thomas, Lorenzetti, & Spragins, 2011). In order to fully explore those acting in these roles, we put no restrictions on who might fulfil the functions and responsibilities (e.g. family members, teachers, adults within the community etc.).

Aim and objectives

Aim: To produce an evidence-based taxonomy defining the key characteristics of trusted adult support.

Objectives: To use evidence to identify the key characteristics of trusted adult support.

To develop a practical resources for a range of stakeholders.

Taxonomy development

The taxonomy builds on the work of Miranda-Chan, Fruiht, Dubon, and Wray-Lake (2016), who focussed on non-familial natural mentors as trusted adults. We incorporated both structured, formal, mentoring (e.g. mentoring programmes), and naturally occurring mentoring, including familial mentoring (e.g. provided by parents, grandparents, aunts and uncles, step-parents) in order to cover as wide a range of potential inputs as possible. Miranda-Chan and colleagues also focussed on mentoring functions, with the qualities of the mentor being more implicit. In our taxonomy, we explicitly consider the characteristics that a trusted adult might bring to the role, in order to inform guidelines around the individual traits that may be fostered or enhanced to best function in the role. We distinguish between social and functional roles to further differentiate between the formal (social) position the adult holds, and the function they fulfil for the young person (Darling, Hamilton, Toyokawa, & Matsuda, 2002; Hamilton, Hamilton, DuBois, & Sellers, 2016).

Methods

The methods used to devise the taxonomy were developed from a number of sources (Burns & Buser, 2015; Hunter, 2018; Khan, 2017), and were based on the over-riding principles of Information Architecture, as a means of making large quantities of information more understandable (Covert, 2014). The process involved compiling and classifying extracted data from the systematic review papers, with definitions or descriptions, context and content of delivery, being organised from the more general to the more specific, in a four stage process, as detailed in Table 2.

Table 2

Stages in taxonomy development.

Stage	Process
1	Compilation of elements described in individual papers
2	Grouping of elements derived from stage 1 into categories
3	Examination of context and nature of relationships
4	Development of a categorisation of trusted adult input, ranging from the general mode of input from, or contact with, through to the potential impact at individual level

All stages of the process were cross referenced by two researchers (JP and JMcA), both of whom had worked on the original systematic review, and were familiar with the topic and the papers included in the review.

Results

The four stage process described above facilitated an in-depth examination of the way in which trusted adult support was detailed across 192 different papers. For the purposes of the taxonomy development, data from these studies were extrapolated into a table detailing the lead author, date, geographical location of study, adolescent population, person fulfilling the trusted adult social role, and any definitions or descriptive elements of their functional role (Appendix 1).

Each stage is described in greater detail in the following sections.

Stage 1: Compilation of elements described in individual papers:

All papers included in the review were screened for references to trusted adult input, what it consisted of, and any definitions of the trusted adult as a person, or the qualities they brought to the role. In the vast majority of papers, there was no explicit definition of a trusted adult, although mentors were most frequently referred to. Implicit definitions could be extrapolated by examining the tools (e.g. questionnaires, or standardised measures) used in the studies. For example, participants may have been asked to identify someone to whom they might trust to go to with a problem, or who made an important difference to their life. As described above, these details were extracted and tabulated (Appendix 1: Definitions and social/functional roles table), as a first stage in the taxonomy development. These details were further refined and reduced by noting (in list form) the individual qualities of a trusted adult and the actions or functions fulfilled by the trusted adult. See Appendix 2 for these lists.

With regard to the qualities mentioned in the studies, 47 individual qualities were identified. In terms of the actions or functions of a trusted adult, 60 individual actions/functions were identified. These were not measured in terms of frequency of mention across the papers, since this might have given an unrepresentative emphasis or ‘weighting’ to the elements. Further details are summarised in stage 2, below.

Stage 2: Grouping of categories derived from Stage 1: individual qualities and role details

The individual qualities and role details identified in stage 1 were then examined to identify overarching categories. These were agreed as relating to the categories of Access, Emotion, and Functions, as shown in Table 3. We combined both qualities and functions, when it became apparent that these were inextricably linked to each other, with many incorporating the action associated with the quality (e.g. quality = inspiring, reasonable; function = inspires, reasons). To cross reference these against the original lists in stage 1, each category in stage 2 was given a number (1–3). All elements from stage 1 were then assigned to the closest matching number. This was cross-checked by the 2 reviewers, until category agreement was reached. A third person from the systematic review team acted as an arbitrator in the case of non-agreement.

Table 3

Categories of trusted adult support.

Overarching description/category	Designation	Description	Qualities or skills [examples derived from review evidence]
Access	1	Elements relating to the presence of the trusted adult in the young person's life	Approachable
			Attentive
			Available
			Engaged
			Interactive
			Involved
			Reliable
			Responsive
Emotion	2	Emotional qualities that the person acting as the trusted adult might display	Authentic
			Caring
			Empathetic
			Nurturing
			Sensitive
			Supportive
			Understanding
			Warm
Functions	3	Functions that are fulfilled by trusted adult support	Advisor
			Coach
			Confidant
			Inspirer
			Listener
			Motivator
			Non-judgemental
			Honest
			Trustworthy
			Responsible
			Persuasive
			Role model

Stage 3: Grouping of categories derived from Stage 1: delivery context and nature of the relationships

In stage 3, the means and mode of delivery were examined. In addition to the qualities and functions of the trusted adult, the type of intervention was classified by how it was structured or delivered. This included such elements as whether the input from, or contact with, a trusted adult was formally structured (e.g. a mentor provided by an institution), or a naturally occurring mentor (e.g. someone from an existing social circle, or within their local community, who was chosen by the young person as someone they might trust to confide in). Since the taxonomy development was not concerned with assessing effectiveness of differing types of intervention, these were noted as variations, rather than analysed for effect. Categories of support were also extrapolated (e.g. material/functional support, emotional support etc; outcome classifications were also noted) (e.g. potential effects on adolescent well-being or education). This information was carried over to stage 4, where input was structured into a categorisation.

Stage 4: Taxonomy refinement

In this stage, all categories were drawn together to form a refined taxonomy. The taxonomy consists of 6 categories, ranging from the general context in which the relationship develops or is provided, through to the potential individual effect on the young person, in similarity to the Ecological Systems Theory of Bronfenbrenner (Bronfenbrenner, 1979). In this respect the taxonomy moves from the broad to the specific, and includes information gathered in the first three stages relating to context, qualities and actions, and ending with potential impact on the young person. See Table 4.

Table 4

Categorisation of trusted adult support: taxonomy formation.

Taxonomy of trusted adult support
Item	Detail
1. Context of provision/engagement	i. Formal/informal ii. Structured/naturally occurring iii. Provided to/chosen by YP iv. Time or length of input v. Other
2. Social role of trusted adult	i. Parent/person acting in parental roleVii. Wider family iii. Non-parental adult (may or may not have a formal role in YP's life) iv. Other
3. Nature of support	i. Instrumental/material/pragmatic (e.g. financial support, goal setting, academic support) ii. Emotional (e.g. affection, rapport, caring) iii. Instructional (e.g. guidance, counselling) iv. Other
4. Qualities of trusted adult	i. Details relating to the Emotional qualities that person acting as a trusted adult might bring to role (may be specified from adolescent feedback or evaluation) ii. Elements relating to the presence of person in the young person's life (may be specified from adolescent feedback or evaluation)
5. Actions or functions of the trusted adult	i. Skills that support the ability to function in the trusted adult role
a. Intended areas of impact on young personb. Demonstrated areas of impact	i. Self-esteem and/or confidence ii. Mental health iii. Health behaviour outcomes e.g. smoking, alcohol, drugs iv. Educational engagement v. Other

YP = young person.

Discussion

This paper describes the development of a taxonomy of trusted adult interventions for young people in the formative second decade of their lives. The evidence that informed the taxonomy was derived from a comprehensive systematic review, examining a broad range of international papers.

The categorisation detailed in the 6 items within Table 4 provides a practical means of describing any input that might be referred to as trusted adult support or mentorship, in any given context. For example, a youth support service may be able to populate the specific details of the service they provide within column 2, against elements 1–6 in column 1. In this way more explicit information can be given, in a structured form, about the extent of their contribution to youth development and support, as well as potentially facilitating service evaluation.

The terms trusted adult or mentor may be used without sufficient thought being given to the meaning of the terms, or the service/input provided. This is problematic in terms of building the trusted adult evidence base, since specific intervention detail would assist in the understanding of what works best. As our review has shown, the functional role fulfilled by such a person or service may vary widely. Many of the papers in the original review referred to people in the trusted adult role as mentors, but did not give specific detail of what this entailed. The taxonomy seeks to encapsulate the core essential qualities provided by the role, and gives details relating to what might be expected of an individual who is entrusted by a young person. It may also be a helpful reference point if a young person is in doubt about the behaviour or qualities displayed by an adult in their environment, and the taxonomy is able to provide a useful starting point for further research in this respect.

Within the review results, while the majority of included studies (70%) examined informal or natural mentoring, a reasonable number also considered the effect of formal mentoring (n = 57, or 30% of total). While there may be freedom of choice benefits relating to naturally occurring mentoring relationships, young people living in more deprived circumstances may not have access to positive role models in their immediate environment, emphasising the potential importance of more formal, structured programs in such circumstances (e.g. Akrimi, Raynor, Johnson, & Wylie, 2008; Slater, Mitschke, & Douthit, 2011). As an example of structured program health benefits, Black et al. (2010) reported sustained improvements in body fat percentage and healthier dietary intake at 2 year follow up of black youth from low-income urban communities. Avery (2011) also reported greater formal mentor program benefits for ‘at risk’ youth relating to educational outcomes, psychological well-being, and problem behaviour.

A review of school-based mentoring programs by Curran and Wexler (2017) concluded that mixed groups of ‘at risk’ and not ‘at risk’ young people encouraged a more positive peer-to-peer climate, taking into account the importance and influence of peer relationships for this age group (McAteer et al., 2018). However, while peer support may be helpful, it cannot necessarily compensate for adult support (Buchanan & Bowen, 2008; Chang, Greenberger, Chen, Heckhausen, & Farruggia, 2010), which may have a uniquely protective effect beyond peer support. It can therefore be seen that identifying the context of delivery, and type of person providing support, are important areas to specify, as can be seen in the taxonomy (Sections 1, 2).

DeWit et al. (2016) emphasise that the length and consistency of mentor presence is a central factor in successful outcomes, a view supported by Johnson, Pryce, and Martinovich (2011). The latter argue that limited input may be worse than no input for vulnerable youth, with 12–18 months, as a minimum, showing better results. If using the taxonomy to describe input, length of contact or delivery can therefore be specified within the context section, as detailed above.

The nature of adult support within the included studies was quite varied, with a number concentrating on specific areas: e.g. educational support (Deutsch, Reitz-Krueger, Henneberger, Futch, & Lawrence, 2017; Engels et al., 2016) or sexual health (Kogan, Simons, Chen, Burwell, & Brody, 2013; McQuestion, Ahiadeke, Posner, & Williams, 2012; de Graaf, Vanwesenbeeck, & Woertman, 2010). Similarly, the qualities the adult brought to the relationship were specified in some studies as being uniquely beneficial. In the study by Chang et al. (2010), for example, having a mentor with good educational attainment was helpful for progression after graduation from school; in other research, Mihalas, Witherspoon, Harper, and Sovran (2012) emphasised the value of emotional support from teachers, beyond their teaching capacity, in relation to victimisation prevention. Categories of support, qualities and actions of mentors are therefore valuable to record and describe (sections 3, 4, 5 in the taxonomy). These categories can also be used indicate the broad range of support that may come from one or more trusted adults, according to individual adolescent need, which may also vary with time. While the review did not differentiate between family and unrelated adult support, and how the qualities of the two might differ - for example, in terms of affective or instrumental support, and potential consequence (Darling et al., 2002; Hamilton et al., 2016) - this would be a theme worth pursuing in further studies.

In our review, a broad range of adolescent outcomes were measured in the included studies. Beneficial effects were varied, with some studies reporting significant benefits in one area, but not in others. For example, the systematic review of out-of-school youth development interventions (Ciocanel, Power, Eriksen, & Gillings, 2017), reported benefits relating to educational achievement and psychological adjustment, but not to problem behaviour or sexual risk behaviour. Such variations may be down to limitations within individual study tools or the study methods utilised (Whitehead et al., 2019), but are nevertheless important to note. Details of outcome areas that will be measured as part of, for example, intervention evaluation, can be itemised in section 6 of the taxonomy.

The examples given above demonstrate why each element of the taxonomy is important to specify, when describing trusted adult input. It cannot be assumed that a mentor in one given context is fulfilling the same role, and providing the same support, as input in another context. The taxonomy therefore has a function to fulfil as a structure to more fully detail the context and content of trusted adult support, beyond the use of limited terminology or description. The taxonomy has the potential to act as a foundation upon which to build a consensus within the field of adolescent support.

Conclusions

The taxonomy seeks to describe key elements of the trusted adult role that need to be given due consideration in any given environment where support is provided. The taxonomy table can act as a prompt for those who support young people to detail the specific nature of their input, and the areas of youth development they are seeking to influence.

With regard to research, the taxonomy is of potential benefit to compare differing inputs, or to ensure intervention delivery is consistent across locations of provision. The value of the taxonomy, as suggested above, has yet to be proven in practice, but we present it here as a useful tool, where no explicit detail or consistency of definition was previously available.

Summarising the taxonomy

It was recognised that, while useful for giving a detailed explanation of trusted adult input, the key messages from the taxonomy might be summarised into a shortened form, as a day-to-day reminder. We therefore sought to reduce the taxonomy into an accessible abbreviation, and one that might be memorable for young people and those seeking to act as trusted adults. Key terms and qualities were derived from the detail of the taxonomy, and built into an acronym to encapsulate central features of trusted adult input, as follows:

‘On TAP’:

Trustworthy, Time-generous.

Accessible, Approachable, Attentive, Authentic, Available Adult.

Provides Persuasive, Protective, Positive example, and promotes positive development.

It was felt that this acronym might help to deliver a fundamental explanation of some of the essential roles and qualities of trusted adult input. As such, it might help young people identify a trusted adult, and also remind those seeking to be, or already in the role, of assets that are important to achieve and maintain in that position.

Contributors

Conceptualisation JP; taxonomy development: JP, JMcA; manuscript draft and review: JP, JMcA, RW, ES, DM, RJ.

Funding sources

MRC (MR/KO 023209/1), CSO and NHS Health Scotland. The views expressed in this publication are those of the authors and not necessarily those of the funders.

Conflicts of interest

The authors declare no competing interests.