Positive youth development in Māori and New Zealand European adolescents through an adventure education programme

ABSTRACT

Within Aotearoa New Zealand there is growing interest in positive youth development (PYD). A PYD approach provides balance to narratives surrounding outcomes for youth and broadens our views of what we consider beneficial developmental outcomes. In the current study we used sail training as a method to promote PYD in Māori and New Zealand European adolescents. Specifically, 54 Māori and 37 New Zealand European adolescents completed the 7-day youth-development voyage on-board the gaffed rigged schooner, R. Tucker Thompson. We demonstrate that, for both Māori and New Zealand European adolescents, psychological resilience, self-esteem, and positive outlook on life increased from the first day to the last day of the voyage. In addition, we demonstrate that the increases in psychological resilience were driven by the social/collective identity adolescents formed with their group over the course of the voyage.

KEYWORDS:

• Positive youth development
• adventure education programme
• psychological resilience
• self-esteem

Introduction

Mental disorders affect 10%–20% of children and adolescents worldwide (Kieling et al. 2011) and most adult mental disorders are extensions of those that arose during childhood and adolescence (Rueter et al. 1999). Aotearoa New Zealand is no exception to these general findings (Fergusson et al. 1993; Kim-Cohen et al. 2003; Gluckman 2011; Clark et al. 2013). For example, data from a national survey (Youth’12) of 8,500 adolescents across 91 schools revealed that 16% of female students and 9% of male students reported levels of depression that were likely to be of clinical significance. Further, deliberate acts of self-harm (females 29% and males 18%) and suicide ideation (females 21% and males 10%) were relatively commonplace. With respect to the relationship between mental health during development and mental health in adulthood, data from the Dunedin Multidisciplinary Child Development Study revealed that among adults with a mental disorder, 74% had received a diagnosis before 18 years of age and 50% before 15 years of age, with adult disorders generally preceded by their juvenile equivalent (Kim-Cohen et al. 2003).

The mental health challenges faced by adolescents in Aotearoa New Zealand disproportionately fall on Māori, the tangata whenua (Indigenous people) of this land, and adolescents from low socio-economic contexts (Horwood and Fergusson 1998; Baxter et al. 2006; Baxter, Kokaua, et al. 2006; Clark, Robinson, Crengle, Herd, & Grant, 2008; Clark, Robinson, Crengle, Fleming, et al. 2011). For example, relative to non-Māori, Māori adolescents experience higher rates of depression and anxiety (Marie et al. 2008), are over-represented in admissions to in-patient facilities (van Kessel et al. 2012), and are markedly more likely to take their own life (Beautrais and Fergusson 2006; Snowdon 2017). Although socio-economic status is a key contributing factor to the poorer mental health outcomes experienced by Māori adolescents (Gillies et al. 2017), the impact of this status is exacerbated by a range of factors linked to cultural alienation, living in a society that values individualism rather than collectivism, and a poor sense of identity (Skegg et al. 1995; Te Aho Lawson 1998; Beautrais and Fergusson 2006; Durie et al. 2009).

While a great deal of literature focuses on these adverse experiences and negative outcomes, there is growing interest in positive youth (i.e. adolescent) development (PYD) approaches. In an Aotearoa New Zealand context, a PYD approach would not only provide balance to narratives surrounding outcomes for Māori adolescents, and those from low socio-economic contexts, but also broaden our focus on what we consider positive developmental outcomes (Anae et al. 2002; Lerner et al. 2005; Farruggia and Bullen 2010; Jansen et al. 2010; Ware and Walsh-Tapiata 2010; Simmonds et al. 2014). As Damon (2004) notes,

While the positive youth development approach recognises the existence of adversities and developmental challenges that may affect children in various ways, it resists conceiving of the developmental process mainly as an effort to overcome deficits and risk. Instead, it begins with a vision of a fully able child eager to explore the world, gain competence, and acquire the capacity to contribute importantly to the world. The positive youth development approach aims at understanding, educating, and engaging children in productive activities rather than at correcting, curing, or treating them for maladaptive tendencies or so-called disabilities (p. 15).

Importantly, the general ethos underlying PYD is thought to map well onto Māori views of health and well-being, which tend to be more holistic in nature (Anae et al. 2002; Farruggia and Bullen 2010).

By way of example, a popular western PYD model is the Five Cs model (Lerner et al. 2000; Roth and Brooks-Gunn 2003; Lerner et al. 2005). The Five Cs model holds that the presence of (1) competence in social, cognitive, and vocational settings, (2) a feeling of personal self-worth and efficacy (i.e. confidence), (3) positive bonds with people and institutions (i.e. connection), (4) respect and integrity (i.e. character) and, (5) empathy for others (i.e. caring) are integral to positive development. Similarly, Durie’s (1994) Te Whare Tapa Whā model of health is holistic, suggesting wairua (spiritual), hinengaro (thoughts and feelings), tinana (physical), and whānau (family) are integral to good health and well-being. Although both models are holistic, an important distinction can be drawn between how western models of PYD and Māori models of health conceptualise identity. Indeed, as Macfarlane et al. (2008) note

In New Zealand, and in many other western societies, the self is largely constructed within the context of individualism, and individual achievement … in a Māori worldview qualities such as personal autonomy, independence, leadership, and prestige are all learned and exercised within a social context in which people share a powerful collective identity. Personal autonomy, strength and leadership are always exercised within the context of whanaungatanga, of nurturing and caring relationships. The self is therefore conceptualised within the context of the collective or community, and not as a totally autonomous and separate entity (pp. 118–119).

The individualistic nature of many western societies is exemplified by the fact that, in the Five Cs model, identity is situated within the construct of confidence, while one could argue a Māori worldview would situate it within the construct of connection.

For PYD in Aotearoa New Zealand, the potential importance of whānau and whanaungatanga becomes even more salient because they are central concepts in Māori views of human development. This is because whānau and whanaungatanga ‘ … indicate both a sense of belonging to and a sense of relating to others, within a context of collective identity and responsibility’ (Macfarlane et al. 2008, p. 107). One category of PYD programmes that may help to foster both a powerful collective identity and sense of belonging are Adventure Education Programmes (AEPs) (Larson 2000; Breunig et al. 2010; Jostad et al. 2015; O’Connell et al. 2016; Scarf, Moradi, et al. 2016; Scarf, Hayhurst, et al. 2017). Of course, AEPs come in many different forms, some of which may focus on individualism and neglect Te Ao Māori (a Māori worldview). Therefore, it is important to emphasise that our focus is on AEPs that take a place-based approach (Brown 2008, 2009; Townsend 2011; Wattchow and Brown 2011; Cosgriff et al. 2012). A place-based approach revolves around the AEP taking place in a locale that has meaning for the youth involved and an appreciation of the cultural meaning and significance of place. To quote Brown (2008)

 … one’s sense of belonging and connection with the land, one’s sense of place, from both a Māori and non-Māori perspective is an issue that is central to being a New Zealander … we could do well to understand our own and others values in relation to land and it’s meanings for our identity (p. 14).

The specific focus of the current study are the 7-day youth development voyages run on the R. Tucker Thompson, a traditional gaff-rigged schooner based in the Bay of Islands, Northland, Aotearoa New Zealand. The voyages draw youth exclusively from the Northland region of Aotearoa New Zealand, an area characterised by a large Māori population and a number of low socio-economic indicators (e.g. low educational achievement, high unemployment, etc.). The voyages run from April to November each year and involve youth aged 13 to 18 years. The youth who take part are predominantly Māori. With respect to the voyage, there are a number of important aspects that link with the concepts of whanaungatanga and collective identity (Macfarlane et al. 2008) and place (Brown 2008). For example, there is a particular emphasis on linking the sea and voyaging tradition back to youth’s ancestors, noting that their tupuna (ancestors) would have sailed on schooners similar to the R. Tucker Thompson. Before youth step on board the ship there is first a mihi (welcome) and then whanaungatanga is established by way of a young crew member facilitating a set of games and activities that provide the opportunity for the youth to talk and become comfortable with one another. The games and activities are run until any apprehension is seen to disappear. Establishing whanaungatanga before youth step on board the ship ensures that when they do step on board, they do so as a collective or group. Once on board, youth are encouraged to ‘self-discipline’ through a ‘tātou’ (all of us) process, whereby they establish as a group what constitutes acceptable behaviour. If a member of the group pushes the boundaries during the voyage, the group revisits the rules they established during this process. Throughout the voyage, a number of historical pā sites (villages or defensive settlements) are passed and their stories are inevitably shared. The specific sites that are passed vary between voyages but all involve the Northland region of Aotearoa New Zealand. Although not a component of every voyage, aspects of celestial navigation are included when the opportunity to have specific guests on board (e.g. Ngatiwai Kaumatua) or visit specific places (e.g. Hekenukumai (Hec) Busby’s purpose built star compass) arise.

With respect to more general aspects of the voyage, it is widely known that sail training AEPs are especially demanding. In addition to being away from home, a challenge common to most AEPs, youth must adjust to the tight quarters of a ship and it is not uncommon for youth to have to deal with sea sickness. Against this back drop, youth must learn how to sail the ship, complete a range of daily tasks required to keep the ship running, and take part in a number of group-based activities. Importantly, youth are not allowed to take their mobile phones on board, meaning that they must look to their fellow voyage members for support when things become difficult.

The aim of the current study was to assess changes in three PYD outcomes (i.e. psychological resilience, self-esteem, and positive outlook) as a result of the voyage. In addition, to determine the contribution of whanaungatanga to these outcomes, on the final day of the voyage we measured the extent to which collective identification contributed to changes in the three PYD outcomes listed above. Drawing on recent theorising derived from the social identity perspective (Tajfel and Turner 1979), we measured collective identity using four items, that drew from the sense of connection that one can attain through group membership (Scarf et al. 2016). We had two primary hypotheses. First, we hypothesised that psychological resilience, self-esteem, and positive outlook would all increase from the first day to the last day of the voyage. Second, we hypothesised that collective identity would be associated with increases in each of the PYD outcomes.

Method

Participants

Ninety-one adolescents (55 females) participated in the current study. Of those, 54 (59.3%; 28 females; Mean age = 15 years 4 months; Range = 13 years 3 months to 17 years 11 months) self-identified as Māori and 37 (40.7%; 27 females; Mean age = 15 years 2 months; Range = 13 years 2 months to 17 years 2 months) as New Zealand European. Within the Māori group, 41 adolescents identified solely as Māori, 5 as Māori and New Zealand European, 3 as Māori, New Zealand European, and Pacific Island, 2 as Māori and Pacific Island, 2 as Māori and other, and 1 as Māori, New Zealand European, and other. Within the New Zealand European group, 32 adolescents identified solely as New Zealand European and 5 as New Zealand European and other. Although prioritising participants into Māori or New Zealand European categories is far from ideal (Cormack and Robson 2010; Didham and Callister 2012), the current data set was limited by a) how ethnicity was recorded and, b) the relatively small sample size.

All participants had taken part in one of eight 7-day youth voyages on the R. Tucker Thompson. With respect to the participants in the current study, using the Aotearoa New Zealand secondary school decile rating categories to represent socioeconomic status (Wikaire et al. 2017), 35 (64.8%) of Māori participants were in the lowest category (i.e. high deprivation), 9 (16.7%) in the middle category (i.e. medium deprivation), and 1 (1.9%) in the highest category (i.e. low deprivation). The remaining Māori adolescents were in partnership schools (n = 3, 5.6%) or other education providers for which decile was not available (n = 1, 1.9%), in a support service (n = 3, 5.6%), or did not provide a response (n = 2, 3.7%). For New Zealand European adolescents 10 (27.0%) were in the lowest category, 20 (54.1%) were in the middle category, 2 (5.4) were in the highest category, and the remaining 5 (13.5%) were in correspondence schools for which decile was not available.

Consultation was undertaken with the Ngāi Tahu (principal Māori tribe of Aotearoa New Zealand’s southern region) Māori Consultation Committee and the study was approved by the University of Otago Human Ethics committee.

Design and analysis

The resilience, self-esteem, and positive outlook of voyage participants was assessed at two time points; on the first day of the voyage (Time 1) and the last day of the voyage (Time 2). Social identity and social support were assessed only at Time 2. With respect to analysis, changes in resilience, self-esteem, and outlook over the course of the voyage were assessed using repeated measures Analysis of Variance (ANOVA). To assess the contribution of social identity and social support to this change, bivariate correlations and standard multiple regression were conducted.

Measures

Resilience was assessed using 10-items drawn from Wagnild and Young’s (1993) Resilience Scale. The 10-items were based on those used by Neill and Dias’ (2001) to measure changes in resilience brought about by outdoor interventions (e.g. ‘I usually manage one way or another’, Cronbrach’s α = Māori: .903, New Zealand European: .832). Self-esteem was assessed using Robins et al.’s (2001) Single Item Self-Esteem Scale (i.e. ‘I have high self-esteem’). Outlook was assessed using 5 of the positively worded items from Beck et al.’s (1974) Hopelessness Scale (e.g. ‘When I look ahead to the future, I expect I will be happier than I am now’, Cronbrach’s α = Māori: .909, New Zealand European: .823). Social identity was measured using a single item from Ellemers et al. (1999) self-categorisation subscale (i.e. ‘I identify with other members of my voyage group’) and all three items from Sheldon and Bettencourt’s (2002) inclusion scale (e.g. ‘I feel a sense of belonging with this voyage group’, Cronbrach’s α = Māori: .875, New Zealand European: .959). Finally, a single item drawn from Neill and Dias’ (2001) Social Support scale was used to assess how supportive individuals judged their group to be (i.e. ‘How supportive, to you personally, have you found the other members of your voyage group to be throughout the voyage?’). Responses on all scales were scored on 7-point Likert scales. Resilience, self-esteem, outlook, and belonging were rated from Strongly Disagree (1) to Strongly Agree (7), while the social support question was rated from Not at All (1) to A Great Deal (7). All measures were completed with respect to how participants felt ‘right now’. Finally, it is important to note that Cronbach’s alpha was high for both the outlook and social identity scales for Māori and New Zealand European youth, supporting the reliability of the scales for both groups.

Results

Resilience

Four outliers were identified whereby two youth had an extremely low score at Time 1 and two had an extremely low score at Time 2. Following Tabachnik and Fidell (2007), each value was transformed so that the value outside the mean value ±3 SD was equal to the next closest value (i.e. a value within 3 SD) +1. For Resilience Scale scores, we conducted a 2 (Ethnicity: Māori and New Zealand European) × 2 (Time of measurement: Time 1 and Time 2) mixed model ANOVA. Cell means and standard deviations are presented in Table 1. There was no main effect of Ethnicity, F(1, 89) = 3.816, p = .054, η² = .041, but a significant effect of Time, F(1, 89) = 63.093, p < .001, η² = .415, and a significant Ethnicity by Time interaction, F(1, 89) = 4.043, p = .047, η² = .043. Post-hoc planned comparisons revealed Māori, t(53) = 7.412, p < .001, and New Zealand European youth, t(36) = 4.202, p < .001, displayed a significant increase in resilience from Time 1 to Time 2. When directly compared, Māori youth reported significantly lower levels of resilience than New Zealand European youth at Time 1, t(89) = 2.501, p = .014, but a comparable level of resilience at Time 2, t(89) = .822, p = .414.

Table 1. Resilience scale means and standard deviations (SD) for Māori and New Zealand (NZ) European adolescents.

	Māori		NZ European		1	2	3	4
	M	SD	M	SD
1. Resilience at Time 1	53.79	9.28	58.26	6.81	–	.482***	−.091	.039
2. Resilience at Time 2	61.70	7.64	62.97	6.61	.585***	–	.488***	.449***
3. Social Identity	25.68	3.80	25.70	4.12	.245	.547***	–	.854***
4. Social Support	6.24	1.35	6.34	1.03	−.102	.122	.649***	–

**Correlation is significant at the 0.05 level (2-tailed).

***Correlation is significant at the 0.01 level (2-tailed).

Also shown are the correlational analyses for social identity and social support. Correlations for Māori participants appear below the diagonal; correlations for NZ European participants appear above the diagonal.

With respect to correlational analyses for social identity and social support, while all participants completed the group belonging questions, 8 Māori (14.8%) and 2 New Zealand European (5.4%) youth did not complete the social support question. This is likely due to the fact the social support question was the final question of the questionnaire and in a different format, due to the change in the scale from Strongly Disagree (1) and Strongly Agree (7) to Not at All (1) and A Great Deal (7). Correlations between variables are shown in Table 1. For both Māori and New Zealand European youth there was a strong relationship between resilience at Time 1 and resilience at Time 2, resilience at Time 2 and social identity, and social identity and social support. The sole difference between groups was a relationship between resilience at Time 2 and social support for New Zealand European, but not Māori, youth.

To determine the extent to which social identity and social support contributed to increased resilience at Time 2 we conducted standard multiple regression. Due to the relatively low sample size for the current regression model, and those that follow, we did not conduct separate regressions for each ethnicity or add an interaction term. With respect to resilience, resilience at Time 2 was entered as the dependent variable. Resilience at Time 1, ethnicity, social identity, and social support were entered as predictor variables. The overall model was significant, R² = .351, F(4, 75) = 10.142, p < .001. Inspection of beta weights revealed that only resilience at Time 1, β = +.482, p < .001, and social identity, β = +.384, p = .005, but not social support β = +.006, p = .964 or ethnicity β = −.107, p = .278, made a unique contributions to resilience at Time 2.

Self-Esteem

Three outliers were identified whereby one youth had an extremely low score at Time 1 and two had an extremely low score at Time 2. Following Tabachnik and Fidell (2007), each value was transformed so that the value outside the mean value ±3 SD was equal to the next closest value (i.e. a value within 3 SD) +1. For self-esteem, we conducted a 2 (Ethnicity: Māori and New Zealand European) × 2 (Time of measurement: Time 1 and Time 2) mixed model ANOVA. Cell means and standard deviations are presented in Table 2. There was no main effect for Ethnicity, F(1, 89) = 3.323, p = .072, η²= .036, but a significant main effect of Time, F(1, 89) = 29.600, p < .001, η² = .250, and a significant Ethnicity by Time interaction, F(1, 89) = 7.609, p = .007, η² = .079. Post-hoc planned comparisons revealed Māori youth displayed a significant increase in self-esteem from Time 1 to Time 2, t(53) = 5.936, p < .001, while New Zealand European youth displayed a moderate increase, t(36) = 2.017, p = .051 (Table 2). When directly compared, Māori youth reported significantly lower level of self-esteem than New Zealand European youth at Time 1, t(89) = 2.673, p = .009, but a comparable level of self-esteem at Time 2, t(89) = .501, p = .081.

Table 2. Single-item self-esteem scale means and standard deviations (SD) for Māori and New Zealand (NZ) European adolescents.

	Māori		NZ European		1	2	3	4
	M	SD	M	SD
1. Self-esteem at Time 1	4.96	1.37	5.70	1.18	–	.517**	−.088	.003
2. Self-esteem at Time 2	6.04	1.01	6.05	0.94	.410**	–	.305	.305
3. Social Identity	25.68	3.80	25.70	4.12	.023	.205	–	.854***
4. Social Support	6.24	1.35	6.34	1.03	−.241	.142	.649***	–

**Correlation is significant at the 0.05 level (2-tailed).

***Correlation is significant at the 0.01 level (2-tailed).

Also shown are the correlational analyses for social identity and social support. Correlations for Māori participants appear below the diagonal; correlations for NZ European participants appear above the diagonal.

Correlations between variables for both groups are shown in Table 2. Similar to resilience, there was a strong relationship between self-esteem at Time 1 and self-esteem at Time 2. Surprisingly, there was no relationship between self-esteem at Time 2 and social identity. To investigate whether social identity and social support contributed to increased self-esteem at Time 2 we conducted standard multiple regression. Self-esteem at Time 2 was entered as the dependent variable. Self-esteem at Time 1, ethnicity, social identity, and social support were entered as predictor variables. The overall model was significant, R² = .254, F(4, 75) = 6.400, p < .001. Consistent with the simple correlations between variables, inspection of beta weights revealed that only self-esteem at Time 1, β = +.481, p < .001, but not social identity, β = +.069, p = .631, social support β = +.227, p = .117 or ethnicity β = −.172, p = .108, made a unique contributions to self-esteem at Time 2.

Outlook

Three outliers were identified whereby two youth had an extremely low score at Time 1 and one had an extremely low score at Time 2. Following Tabachnik and Fidell (2007), each value was transformed so that the value outside the mean value ±3 SD was equal to the next closest value (i.e. a value within 3 SD) +1. For outlook, we conducted a 2 (Ethnicity: Māori and New Zealand European) × 2 (Time of measurement: Time 1 and Time 2) mixed model ANOVA. Cell means and standard deviations are presented in Table 3. There was a significant effect of Time, F(1, 89) = 22.298, p < .001, η² = .200, but no effect of Ethnicity, F(1, 89) = .184, p = .669, η² = .002, or Ethnicity by Time interaction, F(1, 89) = 1.476, p = .228, η² = .016. Post-hoc planned comparisons revealed Māori, t(53) = 4.418, p < .001, and New Zealand European youth, t(36) = 2.488, p = .018, displayed a significant increase in positive outlook from Time 1 to Time 2. When directly compared, Māori and New Zealand European youth displayed a comparable level of positive outlook at Time 1, t(89) = .923, p = .359, and Time 2, t(89) = .404, p = .687.

Table 3. Positive outlook scale means and standard deviations (SD) for Māori and New Zealand (NZ) European adolescents.

	Māori		NZ European		1	2	3	4
	M	SD	M	SD
1. Outlook at Time 1	28.50	5.33	29.46	4.11	–	.345**	−.130	−.075
2. Outlook at Time 2	31.61	3.57	31.30	3.73	.378**	–	.540***	.470***
3. Social Identity	25.68	3.80	25.70	4.12	.238	.599***	–	.854***
4. Social Support	6.24	1.35	6.34	1.03	−.036	.245	.649***	–

**Correlation is significant at the 0.05 level (2-tailed).

***Correlation is significant at the 0.01 level (2-tailed).

Also shown are the correlational analyses for social identity and social support. Correlations for Māori participants appear below the diagonal; correlations for NZ European participants appear above the diagonal.

Correlations between variables for both groups are shown in Table 3. For both Māori and New Zealand European youth there was a significant relationship between outlook at Time 1 and outlook at Time 2 and outlook at Time 2 and social identity. To investigate whether social identity and social support contributed to positive outlook at Time 2 we conducted standard multiple regression. Outlook at Time 2 was entered as the dependent variable. Outlook at Time 1, ethnicity, social identity, and social support were entered as predictor variables. The overall model was significant, R² = .314, F(4, 75) = 8.580, p < .001. Inspection of beta weights revealed that outlook at Time 1, β = +.295, p = .003, and social identity, β = +.454, p = .001, but not social support β = +.019, p = .890 or ethnicity β = −.143, p = .149, made a unique contributions to outlook at Time 2.

Discussion

The current study investigated the relationship between the social identity youth develop with their voyage group over the course of a 7-day voyage on the R. Tucker Thompson and three PYD outcomes (i.e. psychological resilience, self-esteem, and positive outlook). Our findings for both Māori and New Zealand European youth support our first hypothesis, with psychological resilience, self-esteem, and positive outlook increasing from the first day to the last day of the voyage. Support for our second hypothesis was mixed. For both Māori and New Zealand European youth, social/collective identity correlated with increases in psychological resilience and positive outlook from the first day to the last day of the voyage. However, social/collective identity did not predict the changes observed for self-esteem.

An unexpected finding in the current study was the fact Māori youth displayed lower levels of resilience and self-esteem than New Zealand European youth at Time 1. Although the specific factors underlying this difference are beyond the scope of the current study, the findings are broadly consistent with the literature demonstrating Māori adolescents face more mental health challenges than New Zealand European adolescents (Marie et al. 2008). Moreover, when using the Aotearoa New Zealand secondary school decile rating categories to represent the socioeconomic status of participants (Wikaire et al. 2017), it is clear that markedly more Māori adolescents were in the lowest category (i.e. high deprivation) compared to New Zealand European adolescents (64.8% vs. 27%). Thus, are findings are also consistent with the links between socio-economic status and mental health (Gillies et al. 2017). However, more important than the differences between Māori and New Zealand European youth, in terms of resilience and self-esteem at Time 1, is the absence of differences at Time 2. This finding was driven by the fact that the degree of improvement was more marked for Māori relative to New Zealand European, by a factor of 2 to 3.

Our findings for psychological resilience are consistent with a growing body of work demonstrating that, rather than being a property of the individual and static in nature, psychological resilience is built on a foundation of collective/social identity (Ungar 2005; Scarf, Moradi et al. 2016; Scarf, Hayhurst et al. 2017). With respect to the marked increases in resilience displayed by Māori youth, our findings are consistent with recent work involving Aboriginal youth in Canada (Ritchie, Wabano, Russell, et al. 2014; Ritchie, Wabano, Corbiere, et al. 2015). For example, Ritchie et al. (2014) developed and implemented a 10-day wilderness canoe expedition with youth from Wikwemikong Unceded Indian Reserve in northern Ontario, Canada. Much like Māori youth in Aotearoa New Zealand, Aboriginal youth in Canada experience higher rates of mental illness (NCCAH 2012) and suicide (Kirmayer et al. 2007) compared to non-Aboriginal populations. Using a slightly longer version of the resilience scale employed in the current study, Ritchie et al. (2014) reported a significant increase in resilience from one day before the expedition to one month after.

Moving beyond the quantitative approach employed in the current study, Ritchie et al. (2015) collected qualitative data in the field, through participant interviews, journals, focus groups, and talking circles, with the aim of understanding how the expedition promoted resilience. Although using different methods, similar to the present study, connection was identified as a central concept that helps to promote resilience. Their qualitative approach, however, provided a much richer view of connection, with the expedition proving ‘ … opportunities for connecting with their ancestors, each other, their culture, their community and, ultimately, with their own sense of identity’ (p. 9). The expedition was also said to result

 … in a spiritual realization for many youth that could also be described as re-connecting or awakening. In other words, many of the connections were already there, but were not necessarily noticed or realized until the experience reached a threshold level, where the connections became apparent (p. 9).

Limitations

An important limitation of the current study was the absence of a long-term follow-up assessment. Previously, we have demonstrated that the increases in self-esteem and resilience observed following a voyage on the Spirit of New Zealand are maintained several months after the voyage (Hunter et al. 2013; Kafka et al. 2016; Scarf, Moradi, et al., 2016; Scarf, Kafka, et al., 2018). With respect to the Tucker, a long-term follow-up is challenging due to the fact youth require considerable support to complete the Time 1 and Time 2 questionnaires. To address this issue, we are currently exploring the possibility of developing new measures in collaboration with youth that have taken part in a voyage on the Tucker.

A second limitation was the use of a single-item self-esteem scale. Indeed, while single item scales can provide a reliable and valid assessment of self-esteem, they often lack the explanatory power of multi-item measures (Gosling et al. 2003). This may, in part, explain the absence of a relationship between social identity and self-esteem, a finding that conflicts with our earlier work demonstrating a clear relationship between these two factors following a similar developmental voyage (Scarf et al. 2018).

A third limitation was the absence of culturally specific measures of social/collective identity. These measures may have stronger predictive validity than the more general scales used in the current study. Indeed, as noted above, the emphasis on linking the sea and voyaging tradition back to youth’s tupuna (ancestors), stories regarding pā sites, etc. may positively influence Māori identity. Thus, future studies may also include measures of Māori identity, such as the Multidimensional Model of Māori Identity and Cultural Engagement (Houkamau & Sibley, 2010, 2015; Sibley and Houkamau 2013).

Summary and Conclusion

Ritchie, Wabano, Russell, et al.’s (2014; Ritchie, Wabano, Corbiere, et al. 2015) research and methods provide a promising path for extending the current study. Indeed, although the quantitative approach taken to the concepts of social identity allows us to integrate it to a much wider literature investigating the link between social identity and multiple aspects of health (Haslam et al. 2009; Jetten et al. 2012; Cruwys et al. 2014), it provides a limited view of concepts such as identity, connection, and belonging. A fruitful next step will be to employ qualitative measures and broaden our focus to fully capture all four dimensions of Durie’s (1994) Te Whare Tapa Whā model. Indeed, the current study captured aspects of hinengaro (e.g. resilience, self-esteem, etc.), tinana (e.g. physical changes that result from sailing the ship), and whānau (e.g. connection with fellow voyage members) but did not capture wairua; the potential spiritual changes youth experience through the voyage. Finally, in addition to qualitative measures, ethnographic methods could be employed to provide a much more detailed and thorough understanding of youth in the context of their voyage on the R. Tucker Thompson. Combined, these approaches will provide us with a much richer insight to the mechanisms at play on this particular voyage.

Acknowledgements

The authors would like to thank the R. Tucker Thompson Sail Training Trust. A special thank you to Jane Hindle, Pauline Moretti, and the crew on board the ship. We would also like to thank the adolescents that participated.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Hitaua Arahanga-Doyle http://orcid.org/0000-0002-5416-1008

Saleh Moradi http://orcid.org/0000-0003-0668-5891

Tia Neha http://orcid.org/0000-0003-4335-2383

Damian Scarf http://orcid.org/0000-0002-9486-8059