Soc Psychol Educ (2009) 12:443–459 DOI 10.1007/s11218-008-9088-5
Increases in global and domain specific self-esteem following a 10 day developmental voyage Andrew C. Grocott · John A. Hunter
Received: 12 March 2008 / Accepted: 19 December 2008 / Published online: 28 January 2009 © Springer Science+Business Media B.V. 2009
Abstract Although positive effects are often reported, research assessing the impact of Adventure Education and Outward Bound programmes on self-esteem is fraught with methodological weaknesses pertaining to an emphasis on scales assessing global self-esteem, a lack of follow-up measures to assess the potential long-term benefits of such programmes and inadequate pre-intervention measures. In a far less researched area of structured outdoor programmes, research using the sea and sailing ships as a medium, have suffered from the same weaknesses as many of their land-based counterparts. The current study sought to improve and extend on previous research in this broad area. In doing this global and multiple dimensions of self-esteem were assessed 3 weeks prior to a 10 day Spirit of New Zealand developmental voyage, on the first day of the voyage, the last day of the voyage, and finally, 3 months following the voyage. Two hypotheses were subsequently tested. The first stated that participants would experience increases in global and domain specific self-esteem following the voyage. The second stated that increases in global and specific domains of self-esteem would have a long-term effect (i.e. they will be in evidence 3 months following the last day of the voyage). Support was found for both hypotheses. Keywords Domain specific self-esteem · Youth development · Intervention · Sailing 1 Introduction The idea that people are motivated to achieve and maintain positive self-esteem is a central feature in a diverse array of psychological theories (Leary and Downs 1995; Rogers 1961; Greenberg et al. 1997; Tajfel and Turner 1979; Tesser 1988). Some
A. C. Grocott · J. A. Hunter (B) Department of Psychology, University of Otago, P.O. Box 56, Dunedin, New Zealand e-mail:
[email protected]
123
444
A. C. Grocott, J. A. Hunter
argue that attempts to feel good about oneself are fundamental to the human condition (e.g. Brown 1998; Sedikides et al. 2003; cf. Heine et al. 1999). Others remind us that it is “difficult to conceive of an area of behaviour that has not been linked in some way to a need for self-esteem” (Solomon et al. 1991, p. 107). Thus, in his recent evaluation of the research in this area, Emler (2001) concludes that low selfesteem is linked to phenomena such as depression, teenage pregnancy, suicide, eating disorders, victimization, long-term unemployment, and relationship difficulties (see also Brown 1998; Harter 1993). It is perhaps unsurprising to note therefore that, in large part, it has generally been assumed that positive or high self-esteem is a ‘good thing’ for both society and the individual (Brown 1998; Mecca et al. 1989). Not all though would necessarily agree with this view (e.g. Seligman et al. 1995). Certainly there is evidence which shows that people with higher levels of selfesteem may sometimes respond with aggression when threatened, engage in risky behaviour and manifest more pronounced patterns of prejudice (Aberson et al. 2000; Baumeister et al. 1996; Emler 2001). An important and far reaching review conducted by Crocker and Wolf (2001) has, however, demonstrated that whilst the association between self-esteem and social problems is more complex than has often been assumed (e.g. Baumeister et al. 1996; cf. Staub 1989; Turner 1999), self-esteem is powerfully associated with numerous indices of psychological well-being. As such, interventions which promote positive self-esteem (if not proving to be the panacea for all societal ills) are nevertheless likely to provide important benefits for the individuals involved. There are a wide variety of interventions, which purport to enhance self-esteem (Branden 1994; Emler 2001; Ewart 1983; Hattie et al. 1997). When empirically assessed, it is clear though that many of these interventions fail to achieve the intended effect (see Emler 2001; Dennison 2000; Hurry and McGurk 1997; Lynam et al. 2000; Wylie 1979). One apparent exception to these findings may be found in those interventions, which use the outdoor environment as a medium for enhancing self-esteem. Typically referred to as Adventure Education and/or Outward Bound programmes, outdoor interventions are specifically designed to be both physically and mentally demanding. Conducted in wilderness or backcountry areas, those taking part engage in activities (such as rock climbing, cross-county running, white water rafting, camping and solo wilderness experiences), which force them to confront and overcome their related doubts and fears (Marsh et al. 1986a; Moore and Russell 2002). In spite (or perhaps because) of the demanding nature of these interventions evidence documenting their benefits have been reported by a number of authors (e.g. Cason and Gillis 1994; Ewart 1983; Marsh et al. 1986a, b; Purdie et al. 2002). Moreover although much of the work in this area (see Marsh et al. 1986a) is beset with methodological problems, Hattie et al. (1997) in a world wide meta-analysis of 97 of the more recent (and methodologically sophisticated) programmes revealed that such interventions largely tend to be successful in their ability to improve both global and domain specific self-esteem. As such, contrary to the concerns expressed by some authors (see Emler 2001; Wylie 1979) it is clear that certain types of intervention do have the potential to enhance self-esteem. Unfortunately, of the wide range of interventions, which do purport to enhance selfesteem through their programmes, few have provided anything other than anecdotal
123
Self-esteem and sailing
445
evidence for their claims (Branden 1994; Emler 2001; Hunter et al. 2002; Moore and Russell 2002). The current study in attempting to rectify this state of affairs thereby sought to empirically assess one such intervention: a structured 10-day developmental voyage aboard the sailing ship ‘Spirit of New Zealand’. Owned by the Spirit of Adventure Trust (SOAT), one of the primary aims of this programme is to help young people develop positive self-esteem (Grocott 2000). Developmental voyages, typically on traditionally rigged sailing vessels, are presently operating all over the world (Hunter et al. 2002). Each year tens of thousands of young people (most in late adolescence and early adulthood) take part in such programmes. Though many of these organisations believe that their programmes do have a positive impact on participant’s self-esteem and other areas of their life (Hunter et al. 2002), empirical evidence in support of their assertions is virtually non-existent (e.g. Crane et al. 1997; Neill and Richards 1994). Moreover, to date there has only been one study, which has examined the consequences of such voyages on participants’ self-esteem. The study in question was conducted by Norris and Weinman (1996). These authors, arguing that shorter interventions would have little impact on outcomes, assessed the effects of a 3-month transatlantic voyage, from the United Kingdom to the Caribbean, on (amongst other things) participants’ levels of global self-esteem. Significant improvements in both global self-esteem and one coping strategy (positive reinterpretation and growth) immediately following the voyage (i.e. when compared to pre-voyage measures and a control group) were found. Norris and Weinman’s (1996) research is not, however, without criticism. Several of these criticisms have been outlined by Marsh and his colleagues (Hattie et al. 1997; Marsh et al. 1986a, b) who have documented a number of methodological problems associated with the research that sought to assess the impact of (land based) outdoor programmes. These researchers have suggested that in order to accurately and adequately assess the effect of an intervention on self-esteem we must move beyond the simple practice of using measures of global self-esteem. One of the major reasons for this is that the self is multidimensional. The self may be evaluatedly generally (i.e. in terms of global self-esteem) and/or with respect to specific domains (i.e. in terms of the esteem related to particular attributes such as attractiveness, honesty or intelligence). Because various domains of self-esteem are orthogonal to one another and global self-esteem (Marsh et al. 1986a, b; Marsh and O’Neill 1984) scales assessing global self-esteem are unlikely to detect the effects of interventions, which target particular behaviours, attributes or self-esteem domains (Hunter et al. 1996; Hunter et al. 1997). Research supporting the importance of this perspective has been reported in a series of studies conducted by Marsh et al. (1986a, b; Marsh 1993; Marsh and Richards 1988). These researchers found that just as participation in academic learning programmes had the greatest impact on academic self-esteem, participation in physical programmes had the greatest impact on physical self-esteem. This point has been further elucidated by Rosenberg et al. (1995) who in extending the findings reported by the Marsh group found that although academic success was related to academic but not global selfesteem, psychological well-being was linked to global but not academic self-esteem is not (Rosenberg et al. 1995). Taken together, the main implication of these findings is therefore that global and domain specific self-esteem are related to different outcomes (i.e. mental health and particular behavioural outcomes, respectively).
123
446
A. C. Grocott, J. A. Hunter
As such, research which does not incorporate multiple measures of (i.e. global and specific) self-esteem will fail to detect many of the potential benefits bestowed during the course of the intervention. In addition to the aforementioned issues, Marsh et al. (1986a) also argued that follow-up measures were necessary to (a) establish the long term ability of an intervention to enhance self-esteem, and (b) to avoid a potential bias they termed ‘postgroup euphoria’ (PGE)—the feelings of elation experienced at the completion of a demanding course. The PGE these researchers suggested was conceptually similar to the Hawthorne, placebo, or cognitive dissonance effect (where participants responses are influenced by the time, effort, and money they have put into the course). The presence of such factors in studies (such as that conducted by Norris and Weinman) do of course, constitute a genuine and serious threat to the validity of the results. A further effect not considered by Marsh et al., but very much related to the current intervention relates to the process of group identification. According to social identity theory (SIT, Tajfel and Turner 1979) if participants identify with a particular social category (e.g. the ship’s crew) and then internalise this as part of their self-concept, then the group’s outcomes can have consequences for each person’s self-esteem (see Hunter et al. 2004). Moreover on a sailing ship where people have to live and work together in order to complete their various goals (e.g. keep watch, climb aloft, cook, clean) and ultimately sail the vessel we might expect this to be especially so. Consequently if these shipboard tasks are successfully achieved, by the respective social group, participants may experience temporary increases in self-esteem. From the perspective of Marsh’s work on the PGE and SIT there are therefore at least two ways in which participants who take part in the current intervention would manifest high levels of self-esteem at the end of the voyage. However, if the increases in self-esteem at the end of the voyage are attributable to either a group identification or PGE effect, then self-esteem should decrease again soon after the voyage is over and participants are in their usual home environment. For this reason, in order to assess whether the effects of an intervention are not just an artefact of group identification, PGE or some other related factor, it is necessary that follow-up measures be incorporated. Thus, if self-esteem is elevated from the first to the last day of the voyage and these effects are still apparent some months after the completion of the intervention, then alternative explanations for the results (such as those based on group identification and/or PGE) are seriously undermined. The importance of including follow up assessments to establish both the long term impact of an intervention and to help rule out alternative explanations for any discerned effects, although often ignored, cannot be doubted. It would be a mistake though to assume that the inclusion of follow up measures are, on their own, enough to establish the veracity of an intervention. This is because there is evidence to show that participants can manifest decreased levels of self-esteem on the first day of an intervention (Marsh et al. 1986a, b) as they experience doubts or apprehension as regards the undertaking of a potentially difficult and challenging endeavour. Studies (such as those conducted by Marsh et al.) which compare the self-esteem scores of participants on the first day of the programme with those on the last day and/or at some period following the completion of the programme are therefore problematic. To put it another way, because self-esteem may be lower on the first day of the intervention,
123
Self-esteem and sailing
447
the elevation in self-esteem seemingly apparent on the last day and/or following the completion of the programme may not actually be a true reflection of enhanced selfesteem but rather an instance of a return to pre-existing levels of self-esteem after the dissipation of threat. 1.1 Summary Research assessing the impact of Adventure Education and Outward Bound programmes on self-esteem is fraught with methodological weaknesses pertaining to an emphasis on scales assessing global self-esteem, a lack of follow-up measures to assess the potential long-term benefits of such programmes and inadequate pre-intervention measures. Studies by Marsh et al. (1986a, b) overcame many of these shortcomings and suggested that both global and specific domains of self-esteem could be enhanced by participating in such courses. In a far less researched area of structured outdoor programmes, research using the sea and sailing ships as a medium, (e.g. Norris and Weinman 1996) have suffered from the same weaknesses as many of their land-based counterparts. The current study sought to improve and extend on previous research in this area. In doing this global and multiple dimensions of self-esteem were assessed 3 weeks prior to a 10 day Spirit of New Zealand developmental voyage, on the first day of the voyage, the last day of the voyage, and finally, 3 months following the voyage. Following the work of Marsh and others (e.g. Hattie et al. 1997) two hypotheses were subsequently generated. The first stated that participants would experience increases in global and domain specific self-esteem following a 10-day developmental voyage. The second stated that increases in global and specific domains of self-esteem would have a long-term effect (i.e. they will be in evidence 3 months following the last day of the voyage). 2 Method 2.1 Participants One hundred and ninety-nine trainees from five voyages on the sailing ship ‘Spirit of New Zealand’ took part in this study.1 The data from five participants were not included as they failed to complete the study properly. A further participant withdrew as a consequence of having reading difficulties. Thus, the final sample comprised 193 participants. One hundred and nineteen participants were female, 74 were male. All participants were aged between 15 and 18. Four voyages were of mixed sex (20 males, 20 females), and one was a single sex voyage (40 females).
1 A 5 (voyage)×13 (self-esteem domains)×4 (time: T1, T2, T3, T4) mixed model analysis of variance (ANOVA) was conducted to assess whether self-esteem was affected as a consequence of the particular voyage undertaken by participants. A marginally significant interaction was found between voyage and selfesteem domain (F(48, 388) = 1.53, p < .05), however post hoc comparisons (using Dunn’s Bonferroni t) failed to reveal any significant effects.
123
448
A. C. Grocott, J. A. Hunter
2.2 Design A 2 (sex: male/female)×13 (self-esteem domains: maths, general school work, religion/spirituality, global self, honesty, opposite sex relations, same sex relations, parental relations, physical appearance, physical ability, verbal ability, emotional stability, problem solving ability)×4 (time of voyage: T1, T2, T3, T4) mixed model (time series design) was utilised. Although this design does not contain a formal control, the present study, following Marsh et al. (1986a), has two control measures incorporated into its procedure. The first is based on the idea that the intervention should impact only on those domains of self-esteem likely to be affected by the voyage. Thus, domains of self-esteem unrelated to the voyage should not be affected by the intervention and can therefore serve as control dimensions. The second control measure relates to the stability of responses from T1 to T2, and T3 to T4. According to this view, provided that there are no pre-voyage anxiety (i.e. concerns about the undertaking which may lead to temporary decreases in some domain of self-evaluation), post-voyage euphoria or group cohesion effects, comparisons between T1–T2 and T3–T4, should not be significantly different. The lack of such effects at each of these times would therefore tend to suggest that any effects (i.e. self-esteem increases) found between T2 and T3 are a function of the intervention (and not the result of some other factor which threatens the validity of our results). 2.3 The 10-day youth development voyage programme The intervention for the current study was a 10-day youth development voyage aboard the ‘Spirit of New Zealand’, a 148 barquentine rigged sailing ship. A 10-day voyage programme comprises the trainees working in four groups (or watches) of 10. On a mixed sex voyage there are five males and five females in each watch. Various activities are completed during each voyage. There is no pre-organised sequence of activities, although some activities are more suited for the beginning of a voyage (such as team building exercises) while others are best completed later in a voyage (such as learning to successfully manoeuvre the ship under sail). Moreover, many activities are dependent on the weather conditions at any given time. The ship is divided up into four areas, called sail stations. These are ‘foredeck’, ‘midships’, ‘main’, and ‘specials’. Each watch moves to a new station each day. ‘Foredeck’ sail station tends the most forward sails, while ‘midships’ and ‘main’ look after sails in the middle of the vessel. ‘Specials’ tend the sails at the rear of the vessel, take turns at steering and navigation, and assist the cook in meal preparation and clean-up. The rigging of the vessel has been purposely kept as simple as possible to ensure that participants, most of whom have never done any sailing, can learn the correct methods of raising, lowering, or manoeuvring the sails with a high chance of success. Indeed, ensuring that activities, whether setting a sail or other watch exercises, have every chance of success is an important part of the voyage programme, in so far as the successful completion of such activities has the potential to aid feelings of achievement amongst participants. Each day a different trainee will be leader of their watch. The trainee’s job is to lead the watch in the various activities throughout the day, from cleaning the ship in
123
Self-esteem and sailing
449
the morning, to the final debriefing session in the evening. The trainee watch leader is encouraged by the crew to ensure every member of the watch has a job to do when sailing or participating in an activity. After each activity a crew member debriefs the trainees by discussing the activity just completed. Each evening the full day’s events and activities are debriefed. Debriefing is part of the ‘Adventure Wave’ which involves an introductory discussion (to prepare participants for any given activity), a discussion of what the activity itself entails, followed by debriefing to process the experience of the activity and what participants may gain from it. In addition to a member of the crew facilitating a final debrief at the conclusion of the day’s activities, trainee watch leaders of the day each take a turn at talking about what they personally gained from the day, how well their watch worked together, what they liked or disliked, and pass on suggestions to others about ways of succeeding at the tasks set. Discussion about leadership skills and what makes an effective leader is also part of the voyage programme. This is also discussed during the evening’s final debriefing session where issues of leadership often arise. For example, if a watch worked well together, the discussion will focus on why this occurred, particularly to see whether the trainee watch leader of the day played an important role in the watches’ success. Likewise, if a watch did not function together well, the discussion is directed towards learning why not. A typical voyage sequence is as follows. Day one includes safety briefs, ship familiarisation, routine for going aloft (climbing the rigging), introduction to ship board routine and the sail station system in place to sail the vessel. Days two to four normally consists of activities set up to facilitate teamwork within the respective watches (and indeed the trainees as a whole group). This is achieved by a series of activities that will only be successful if the members of the watch cooperate. Thus for example in one activity called ‘Spiders web’ a ‘web’ of rope is tied into the rigging of the ship. This web contains 10 gaps. Each member of the watch must pass through a different gap in the web without touching the ropes. Some of the gaps are above head height to add challenge to the exercise, and to encourage teamwork for successful completion of the task. During these first few days of the voyage general safety features of the vessel are discussed along with correct response in an emergency situation. For this part of the voyage there is a ‘hands on hands’ approach from the crew. That is, the crew are actively involved in demonstrating and assisting with sail handling, and are fully involved with other watch activities. During days five to eight there is a continuation of teamwork and leadership development. Here the emphasis is on achievable tasks within watches, and as a complete crew. Activities are varied, ranging from tramping, sailing small boats, barbeques ashore, and inflatable raft paddling races, as well as sailing the ship. For the first part of this phase the crew adopt a ‘hands on’ approach where direction and assistance are given when needed. This changes to a ‘hands off’ approach around day seven where trainees are encouraged to use their watch as a resource to solve problems, rather than relying on crew input, while at the same time handling sail and other watch activities. The 8 days of fostering success, teamwork, and leadership (by positive reinforcement, facilitating success by having achievable activities, teamwork activities, and discussing effective leadership) leads to what is termed ‘trainee day’ on day nine. Trainee day is where the trainees take over all aspects of running the ship for the day. The previous night the trainees elect their own Captain, Mate, Navigators, Engineers,
123
450
A. C. Grocott, J. A. Hunter
Cooks, and Watch Leaders. This is an integral part of the voyage programme, and is one which is emphasised as a goal during the earlier part of the voyage. It is an opportunity for the trainees to use newly acquired skills and pool their knowledge of the ship before asking for assistance from crew members. In the evening after trainee day awards and certificates are presented to each trainee. The vessel is usually alongside the wharf by 7 am the following morning where trainees depart and travel back to their homes. 2.4 Materials The Self-Description Questionnaire (SDQ III, Marsh and O’Neill 1984) was used to assess global and domain specific self-esteem. Developed in order to measure global and 12 distinct components of the self-concept and thus self-esteem (see Breakwell 1987; Greenwald et al. 1988; Marsh et al. 1983; for reviews) the SDQ III was developed for use with teenagers and adults. Comprised of 136 items in total, each domain of selfesteem measured by the SDQ III is tapped by means of a series of subscales consisting of 10 or 12 items. All answers are recorded on eight-point Likert scales (1 = definitely false; 8 = definitely true). Approximately half of the 136 items are worded negatively to avoid response bias. These scores are reversed for analysis. Higher scores on the scale indicate higher levels of self-esteem. Components of the self assessed by this instrument were developed on the basis of theoretical and empirical work conducted by Shavelson et al. (1976) and Marsh and O’Neill (1984). As a consequence the SDQ III measures components of the self based broadly on academic, moral, social, emotional, and global dimensions. The latter global dimension is largely derived from the Rosenberg (1965) global self-esteem scale, and thus provides an excellent measure of global self-esteem (see Burns 1979, for a review). An example of the content of each subscale is as follows: mathematics, ‘I find many mathematical problems interesting and challenging’; verbal, ‘I am good at expressing myself’; academic, ‘I hate studying for many academic subjects’; problem solving, ‘I am never able to think up answers to problems that haven’t already been figured out’; physical ability, ‘I am a good athlete’; physical appearance, ‘I have good body build’; same sex relations, ‘I have few friends of the same sex that I can really count on’; opposite sex relations; ‘I get a lot of attention from members of the opposite sex’; parental relations, ‘I still have many unresolved conflicts with my parents’; religion/spirituality, ‘my spiritual/religious beliefs provide the guidelines by which I conduct my life’; honesty/trustworthiness, ‘I have never stolen anything of consequence’; emotional stability, ‘I tend to be highly strung, tense, and restless’; general self, ‘overall, I have a lot of respect for myself’. 2.4.1 Reliability and validity of the Self-Description Questionnaire III The reliability and validity of the SDQ III is exceptionally well documented (e.g. Byrne 1988a, b, 1996; Marsh 1987a,b; Marsh and O’Neill 1984; Marsh et al. 1986a, b). Indeed, on the basis of her review, Byrne points out that the SDQ III is the “most extensively validated self-concept measure available” (Byrne 1996, p. 204). The SDQ III has proven to be reliable across cultures. These include Northern Ireland (Hunter et al.
123
Self-esteem and sailing
451
1996), Canada (Byrne 1988a), Australia (Marsh and O’Neill 1984), Portugal (Faria 1996), and New Zealand (Hunter et al. 1997; Mitchell and Mitchell 1989). Each of the 13 subscales of the SDQ III have excellent levels of internal (median r = .90) and retest reliability (median r = .87). Correlations between each of the 13 subscales of SDQ III are low (median r = .10). These findings reveal clear and distinct differences between each of the domains of self-esteem measured by the SDQ III. 2.5 Procedure The SDQ III was administered to participants on four separate occasions over a 4 month period. The first questionnaire (T1), an information sheet, informed consent form, and explanatory letter was posted to potential participants by the staff of the Spirit of Adventure Trust (SOAT) approximately 3–4 weeks before the voyage. This was included with the information the SOAT sends to trainees (e.g. travel arrangements, equipment lists, and medical forms). The time two (T2) questionnaire was administered on the first day of the voyage prior to leaving the wharf. The time three (T3) questionnaire was administered at the end of the voyage after the completion of trainee day prior to the presentation of awards and certificates. The time four (T4) questionnaire was posted out to participants approximately 3 months after the last day of their voyage. In order to ascertain which domains of self-esteem were likely to elevate during the intervention, the senior master for the SOAT, Captain Paul Leppington (a man with 20 years experience working with youth on sea based developmental programmes) was asked to specify a priori which of the 13 dimensions measured by the SDQ III would and would not be related to activities untaken during the voyage. Because so many of the available dimensions could potentially be affected by the processes involved on the voyage, this task proved somewhat difficult. Following Marsh et al. (1986a) we therefore decided to focus on those dimensions which could be judged as being clearly less relevant to the intervention. Using this procedure the self-esteem domains of religion/spirituality, parental relations, physical appearance, verbal ability, general school work, and honesty were so identified. Global self-esteem, and the esteem in which mathematical ability, same sex relations, opposite sex relations, same sex relations, problem solving ability and physical ability were all assumed to be ‘potentially’ related to the intervention. 3 Results Data from participants of each of the five voyages was pooled together (see footnote 1). The final analysis consisted of 158 participants (54 males, 104 females) who completed three or more questionnaires. One hundred and thirty-one participants completed the SDQ III at T1, 157 participants at T2, 156 participants at T3, and 141 participants at T4. One hundred and eleven participants completed the questionnaire on all four occasions. Of the 27 participants who did not complete the questionnaire at T1, approximately 18 did not receive it as a result of being a ‘last minute’ allocation to the voyage, had a recent change of address, or supplied a home address while attending boarding school. Missing item values (if .82; T3, M = 765.77 (SD = 89.61); T4, M = 767.01 (SD = 89.19), t(138) = .30, p > .75). In overall terms, such findings indicate a broad degree of support for our hypotheses. A number of relevant two-way interactions were also discerned. A first 2-way interaction (approaching significance) was found between sex and self-esteem domain (F(12, 93) = 1.71, p < .08). To analyse this effect further a series of between subjects’ ANOVA’s were conducted. This analysis revealed a number of significant effects. Males were significantly higher than females in the self-esteem domains of physical appearance (males M = 218.21, females M = 193.23, F(1, 109) = 6.38, p < .02), problem solving ability (males M = 218.47, females M = 203.60, F(1, 109) = 4.65, p < .05) and emotional stability (males M = 240.42, females M = 225.67, F(1, 109) = 4.09, p < .05). The esteem in which honesty is held was higher for females (M = 301.96) than males (M = 285.74; F(1, 109) = 6.29, p < .02). Two other domains of self-esteem approached significance. Females (M = 230.42) were higher than males (M = 215.29) in the domain of verbal ability (F(1, 109) = 3.45, p < .07). Males (M = 218.74) were higher than females (M = 196.40) in the domain of mathematics (F(1, 109) = 3.28,
123
454
A. C. Grocott, J. A. Hunter
p < .08). A second 2-way interaction was found between sex and time of self-esteem measurement (F(3, 102) = 3.34, p < .05). To assess this effect further a series of 2 (sex)×2 (time of self-esteem measurement: T1 T2, T3, T4) ANOVA’s were conducted. No significant effects were, however, found. As expected a third two-way interaction was found between time of self-esteem measurement and self-esteem domain (F(36, 69) = 19.47, p < .001). To assess this effect further a series planned comparisons between each time of self-esteem measurement (T1–T2, T2–T3, T3–T4 and T1–T4) were carried out on each domain to identify where the changes occurred. In order to reduce the likelihood of a type I error occurring (e.g. Howell 1987) the more conservative significance level of p < .01 was adopted for these analyses. With respect to T1–T2 comparisons (i.e. 3 weeks prior to the beginning of the voyage to the first day of the voyage) only one relevant effect emerged. The esteem in which participants held their physical appearance esteem which was found to decrease from T1 to T2 (T1, M = 49.28; T2, M = 48.11, t(129) = 2.51, p < .02). This effect did not, however, reach the required ( p < .01) level of significance and thereby confirmed the use of T1–T2 comparisons as an indice of self-concept stability. T2–T3 comparisons (i.e. from the first day of the voyage to the last day of the voyage) revealed a number of significant effects. The esteem in which participants held their feelings of global self-worth (t(153) = 3.40, p < .002), opposite sex relations (t(154) = 3.08, p < .003), physical appearance t(154) = 6.63, p < .001), emotional stability (t(154) = 2.84, p < .006) and mathematical ability (t(154) = 3.33, p < .002) all increased over the course of the voyage. A similar trend was also found for parental relations (t(154) = 2.38, p < .02). However, this effect failed to reach the required level of significance. As expected, and in keeping with the results of the T1–T2 comparisons, analysis of T3–T4 comparisons failed to reveal any differences. These findings therefore suggest two main things. The first is that the T2–T3 global and domain specific self-esteem increases from the first to the last day of the voyage remained (or did not dissipate) 3 months following its conclusion. The second is that T3–T4 comparisons serve as a further indice of self-concept stability. Nevertheless, as a final check on the veracity of our findings we reasoned that if the intervention truly did have positive consequences for participants’ self-esteem then the effects found from T2 to T3 should also be apparent from T1 to T4 (i.e. 3 weeks before the voyage to 3 months after the voyage). Analysis conducted to assess this eventuality successfully replicated the effects found for global (t(113) = 3.44, p < .002), opposite sex (t(154) = 3.08, p < .003), physical appearance (t(154) = 6.63, p < .001), and the esteem in which emotional stability was held (t(113) = 2.65, p < .01). The effects for mathematical self-esteem were not replicated. Thus, although mathematical esteem increased from T2 to T3 and did not subsequently decrease between T3 and T4, there no significant difference between T1 and T4 scores (t(113) = .61, p > .55).
4 Discussion This study measured the global and domain specific self-esteem of participants 3 weeks prior to a SOAT 10-day voyage (T1), the first day of the voyage (T2), the last day of the voyage (T3), and 3 months following the voyage (T4). Using this methodology two
123
Self-esteem and sailing
455
hypotheses were tested. The first stated that participants would experience increases in global and domain specific self-esteem following a 10-day developmental voyage (i.e. from T1 to T2). The second stated that increases in global and specific domains of self-esteem would have a long-term effect (i.e. they will be in evidence 3 months following the last day of the voyage). Support was found for each hypothesis. From the first to the last day of the voyage (T2–T3) participants experienced increases in the esteem in which they held their feelings of global self-worth, opposite sex relations, physical appearance, emotional stability and mathematical self-esteem. Subsequent analysis revealed that each of these effects remained 3 months following the voyage. There is evidence to show that participants can manifest decreased levels of selfesteem on the first day of an intervention (Marsh et al. 1986a, b). Presumably this is because of the evaluative concerns associated with undertaking a new and potentially difficult task. As such, it might therefore be argued that the self-esteem levels, seemingly apparent on the last day and/or following the completion of the programme, may not actually reflect genuine effects but rather an instance of a return to pre-existing levels of self-esteem after the dissipation of threat. To guard against this eventuality, it was reasoned that if the intervention truly did have long-term positive consequences for self-esteem then the effects found from T2 to T3 should also be apparent from T1 to T4 (i.e. 3 weeks before the voyage to 3 months after the voyage). Analysis conducted to assess this question successfully replicated each effect except that relating to mathematical self-esteem. Participants’ mathematical self-esteem increased from T2 to T3 and did not subsequently decrease between T3 and T4. The implication being that the voyage had a positive long-term effect on the mathematical domain of self-esteem. However, in light of the fact that there was no significant difference between T1 and T4 such an interpretation should be treated with some caution. This qualification not withstanding, the results of the present study would, nevertheless, tend to indicate that the effects of the intervention on global and domain specific self-esteem were maintained 3 months following the completion of the voyage. Overall our findings are broadly consistent with those of Marsh and others (e.g. Marsh et al. 1986a, b; Hattie et al. 1997; Purdie et al. 2002), who have reported similar sorts of effects among those who undertake land based Adventure Education/Outward Bound programmes. In this respect, our results both replicate and extend the work of Norris and Weinman (1996). These researchers found high levels of global selfesteem among those who had just completed a 3-month voyage in the Caribbean. As is the case with many of the land-based programmes, Norris and Weinman’s work was open to a number of criticisms. Methodologically, their study failed to (a) assess the multidimensional nature of self-esteem, (b) establish the long term impact of the intervention, (c) consider alternative explanations (such as those pertaining to PGE or SIT), and (d) inadequate pre-intervention measures. The present study in utilising measures of global and domain specific self-esteem together with (3 month) follow up assessments overcame each of these criticisms. Subsequently, it was demonstrated that the benefits of the current intervention were apparent on scales assessing both global and domain specific self-esteem, not likely to be caused by other extraneous factors (such as PGE or social identification) and maintained over time. Moreover contrary to the concerns expressed by Norris and Weinman the results of the current study reveal
123
456
A. C. Grocott, J. A. Hunter
that interventions considerably shorter than 3 months (in this case 10 days) are capable of producing benefits that are maintained over long periods. A potential shortcoming of the current study was our inability to specify (a priori) which domains of self-esteem were most relevant to the voyage. Of the seven domains deigned to be potentially affected by the intervention (e.g. global, mathematical, same sex relations, opposite sex relations, same sex relations, problem solving ability and physical ability) corresponding effects were found in only four (or 57% of) cases. Predicting which domains that were less relevant to the intervention proved to be a more successful undertaking. Those aspects of the self judged a priori as being less relevant to the voyage (and thus less likely to increase as a result of the intervention) were religion/spirituality, parental relations, physical appearance, verbal ability, general schoolwork, and honesty. These predictions were correct in five out of six instances (or in over 83% of cases). In spite of the relative degree of accuracy on this latter task, the fact we were unable to specify exactly which domains of self-esteem would and would not be affected by the voyage undermines the idea that those domains not relevant to the intervention can serve as controls. Indeed, from this perspective, it might subsequently be argued that the global and domain specific self-esteem increases found over the duration of the voyage (i.e. from T2 to T3) may have been caused not by the intervention but by PGE or group identification. There are two main sources of evidence, which undermine this view. One is based on the assumption that any benefits conferred by PGE and/or group identification would be transitory. The elevated global and domain specific selfesteem levels found as a function of the intervention were, however, found to be long lasting (i.e. of at least 3 months duration). For this reason explanations which attempt to account for these effects through reference to short term processes are, therefore, seriously flawed. A second is based on stability scores. These analyses in finding no reliable differences between T1–T2 and T3–T4 comparisons indicate that the T2–T3 global and domain specific self-esteem increases are a function of the intervention and not the result of some other factor (related to PGE, SIT or pre-voyage anxiety), which threatens the validity of our results. The results of the current study illustrate the importance of assessing multiple dimensions (i.e. global and domain specific) self-esteem when considering the effects of a given intervention. Insofar as global and domain self-esteem are linked to diverse outcomes it would seem therefore that the intervention will likely bestow a range of psychological and behavioural benefits upon participants (see Brown 1998; Crocker and Wolf 2001; Emler 2001; Harter 1993; Rosenberg et al. 1995). This is not to say that high self-esteem is always beneficial for people or will serve as a social vaccine that will inoculate against all society’s ills. Clearly, their are social problems largely unrelated to self-esteem. There is also a down side to overly inflated self-esteem. People with high self-esteem this do not necessarily act in ways that benefit others in society. Sometimes such individuals will respond with aggression when their self-esteem is threatened, engage in dubious acts and display increased levels of prejudice (e.g. Aberson et al. 2000; Baumeister et al. 1996; Emler 2001). That said the nature of the current intervention might well have the potential to impact on such outcomes. Thus, in its apparent ability to produce stable (long lasting) increases in self-esteem—because there is evidence to suggest that persons whose self-esteem is both positive and stable
123
Self-esteem and sailing
457
are actually the least likely to show hostility (Kernis et al. 1989)—the programme may lead to decreased aggression. Similar possibilities are apparent with respect to intergroup discrimination. The circumstances experienced in undertaking a SOAT voyage encourage cooperation so that common goals may be achieved, equal status (i.e. all participants are in the same position), acquaintance potential and recategorisation (as members of a the ship’s crew) appear to be almost uniquely suited to fulfill the criteria required to facilitate successful intergroup contact (e.g. Pettigrew 1998). As such, the current (sort of) intervention may prove to produce positive outcomes not just for those involved but for others as well. Future research in this exciting new area needs to empirically examine such possibilities and indeed investigate exactly which aspects of the voyage experience are likely to predict positive outcomes for both the individual and society.
References Aberson, C. L., Healy, M., & Romero, V. (2000). Ingroup bias and self-esteem: A meta-analysis. Personality and Social Psychology Review, 4, 157–173. doi:10.1207/S15327957PSPR0402_04. Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened egotism to violence and aggression: The dark side of high self-esteem. Psychological Review, 103, 5–33 . doi:10.1037/0033-295X. 103.1.5. Branden, N. (1994). The six pillars of self-esteem. New York: Bantam. Breakwell, G. M. (1987). Identity. In H. Beloff & A. M. Coleman (Eds.), Psychology survey 7 (pp. 94–114). Leicester: British Psychological Society. Brown, J. D. (1998). The self. Boston: McGraw-Hill. Burns, R. B. (1979). The self-concept in theory, measurement, development and behaviour. London: Longman. Byrne, B. (1988a). Measuring adolescent self-concept: Factorial validity and equivalency of the SDQ III across gender. Multivariate Behavioral Research, 23, 361–375. doi:10.1207/s15327906mbr2303_5. Byrne, B. (1988b). The Self Description Questionnaire III: Testing for equivalent factorial validity across ability. Educational and Psychological Measurement, 48, 397–406. doi:10.1177/0013164488482012. Byrne, B. M. (1996). Measuring self-concept across the lifespan: Issues and instrumentation. Washington: American Psychological Association. Cason, D., & Gillis, H. L. (1994). A meta-analysis of outdoor adventure programming with adolescents. Journal of Experiential Education, 17, 40–47. Crane, D., Hattie, J., & Houghton, S. (1997). Goal setting and the adventure experience. Australian Journal of Psychology, 49, 6–13. doi:10.1080/00049539708259844. Crocker, J., & Wolf, C. (2001). Contingencies of self-worth. Psychological Review, 108, 593–623. doi:10. 1037/0033-295X.108.3.593. Dennison, S. (2000). A win-win peermentoring tutoring program: A colaborative model. The Journal of Primary Prevention, 20, 161–174. doi:10.1023/A:1021385817106. Emler, N. (2001). Self-esteem: The causes and consequences of low self-esteem. York: Joseph Rowntree Foundation. Ewart, A. (1983). Outdoor adventure and self-concept: A research analysis. Eugene: University of Oregon, Centre of Leisure Studies. Faria, L. (1996). Marsh’s Self Description Questionnaire III (SDQ III): Adaptation study with Portuguese college students. Social Behavior and Personality, 24, 343–350. doi:10.2224/sbp.1996.24.4.343. Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory of self-esteem and cultural worldviews: Empirical assessments and conceptual refinements. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 29, pp. 61–139). London: Academic Press. Greenwald, A. G., Bellezza, F. S., & Banaji, M. R. (1988). Is self-esteem the central ingredient of the self-concept? Personality and Social Psychology Bulletin, 14, 34–45. doi:10.1177/0146167288141004. Grocott, A. C. (2000). Sailing and self-esteem. Unpublished Masters Thesis, Department of Psychology, University of Otago. Dunedin, New Zealand.
123
458
A. C. Grocott, J. A. Hunter
Harter, S. (1993). Causes and consequences of low self-esteem in children and adolescents. In R. F. Baumeister (Ed.), Self-esteem: The puzzle of low self-regard (pp. 87–116). New York: Plenum Press. Hattie, J., Marsh, H. W., Neill, J. T., & Richards, G. E. (1997). Adventure Education and Outward Bound: Out-of-class experiences that make a lasting difference. Review of Educational Research, 67, 43–87. Heine, S. J., Lehman, D. R., Markus, H. R., & Kitiyama, S. (1999). Is there a universal need for positive self-regard? Psychological Review, 106, 766–795 . doi:10.1037/0033-295X.106.4.766. Howell, D. C. (1987). Statistics for psychology (2nd ed.). Boston: PWS- Kent. Hunter, J. A., Boyes, M., Maunsell, S., & O’Hare, D. (2002). Sail training in the international arena: The value of sail training to young people today. Final Report for Phase One of a Project Commissioned by the International Sail Training Association (ISTA). Dunedin: University of Otago. Hunter, J. A., Kypri, K., Boyes, M., Stokell, N. M., O’Brien, K. S., & McMenamin, K. (2004). Social identity, self-evaluation and ingroup bias: The relative importance of particular domains of self-esteem to the ingroup. The British Journal of Social Psychology, 43, 59–81. Hunter, J. A., Platow, M. J., Bell, L. M., Kypri, K., & Lewis, C. A. (1997). Intergroup bias and self-evaluation: Domain specific self-esteem, threats to identity and dimensional importance. The British Journal of Social Psychology, 36, 405–426. Hunter, J. A., Platow, M. J., Howard, M. L., & Stringer, M. (1996). Social identity and intergroup evaluative bias: Realistic categories and domain specific self-esteem in a conflict setting. European Journal of Social Psychology, 26, 631, 647. Hurry, J., & McGurk, H. (1997). An evaluation of a primary prevention programme for schools. Addiction Research, 5, 23–38. doi:10.3109/16066359709005579. Kernis, M. H., Gannemann, B. D., & Barclay, L. C. (1989). Stability and level of self-esteem as predictors of anger arousal and hostility. Journal of Personality and Social Psychology, 56, 1013–1022 . doi:10. 1037/0022-3514.56.6.1013. Leary, M. R., & Downs, D. L. (1995). Interpersonal functions of the self-esteem motive: The self-esteem sysem as a sociometer. In M. H. Kernis (Ed.), Efficacy, agency and self-esteem. New York: Plenum. Lynam, D. R., Milich, R., Zimmerman, R., Novak, S. P., Logan, T. K., Martin, C., et al. (2000). Project DARE: No effects at 10 year follow up. Journal of Consulting and Clinical Psychology, 67, 590–593. doi:10.1037/0022-006X.67.4.590. Marsh, H. W. (1987a). The factorial invariance of responses by males and females to a multidimensional self-concept instrument: Substantive and methodological issues. Multivariate Behavioral Research, 22, 457–480. doi:10.1207/s15327906mbr2204_5. Marsh, H. W. (1987b). The hierarchical structure of self-concept and the application of hierarchical confirmatory factor analysis. Journal of Educational Measurement, 24, 17–39. doi:10.1111/j.1745-3984. 1987.tb00259.x. Marsh, H. W. (1993). Academic self-concept: Theory and measurement research. In J. Suls (Ed.), Psychological perspectives on the self (Vol. 4, pp. 59–98). Hillsdale: Erlbaum. Marsh, H. W., & O’Neill, R. (1984). Self Description Questionnaire III: The construct validity of multidimensional self-concept ratings by late adolescents. Journal of Educational Measurement, 21, 153–174. doi:10.1111/j.1745--3984.1984.tb00227.x. Marsh, H. W., & Richards, G. E. (1988). The Outward Bound Bridging course for low achieving high school males: Effect on academic achievement and multidimensional self-concepts. Australian Journal of Psychology, 40, 281–298. doi:10.1080/00049538808260049. Marsh, H. W., Relich, J. D., & Smith, I. D. (1983). Self-concept: The construct validity of interpretations based upon the SDQ. Journal of Personality and Social Psychology, 45, 173–187. doi:10.1037/ 0022-3514.45.1.173. Marsh, H. W., Richards, G. E., & Barnes, J. (1986a). Multidimensional self-concepts: The effect of participation in an Outward Bound program. Journal of Personality and Social Psychology, 50, 195–204. doi:10.1037/0022-3514.50.1.195. Marsh, H. W., Richards, G. E., & Barnes, J. (1986b). Multidimensional self-concepts: A long term followup of the effect of participation in an Outward Bound program. Personality and Social Psychology Bulletin, 12, 475–492. doi:10.1177/0146167286124011. Mecca, A. M., Smelser, N. J., & Vasconcellos, J. (1989). The social importance of self-esteem. Berkeley: University of California Press. Mitchell, H. A., & Mitchell, M. J. (1989). A study of self-concept over a two year period: Possible effects of an intervening Outward Bound course. Nelson: Mitchell Research.
123
Self-esteem and sailing
459
Moore, T., & Russell, K. C. (2002). Studies of the use of wilderness for personal growth, therapy education and leadership development: An annotation and evaluation. Moscow: University of Idaho, Wilderness Research Centre. Neill, J. T., & Richards, G. E. (1994). A research evaluation of the developmental outcomes of Young Endeavour’s ten day sail training voyages during 1993. Canberra: Australian Outward Bound Foundation. Norris, R. M., & Weinman, J. A. (1996). Psychological change following a long sail training voyage. Personality and Individual Differences, 21, 189–194. doi:10.1016/0191-8869(96)00069-4. Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49, 65–85 . doi:10.1146/ annurev.psych.49.1.65. Purdie, N., Neill, J. T., & Richards, G. E. (2002). Australian identity and the effect of an outdoor program. Australian Journal of Psychology, 54, 32–39. doi:10.1080/00049530210001706493. Rogers, C. R. (1961). On becoming a person. Boston: Houghton Mifflin. Rosenberg, M. (1965). Society and the adolescent self-image. Princeton: Princeton University Press. Rosenberg, M., Schooler, C., Schoenbach, C., & Rosenberg, F. (1995). Global self-esteem and specific self-esteem: Different concepts, different outcomes. American Sociological Review, 60, 141–156. doi: 10.2307/2096350. Sedikides, C., Gartner, L., & Toguchi, Y. (2003). Pancultural self-enhancement. Journal of Personality and Social Psychology, 81, 60–79. doi:10.1037/0022-3514.84.1.60. Seligman, M. E. P., Reivich, K., Jaycox, L., & Gillham, J. (1995). The optimistic child. New York: Houghton-Mifflin. Shavelson, R. J., Hubner, J. J., & Stanton, G. C. (1976). Self-concept: Validation of construct interpretations. Review of Educational Research, 46, 407–441. Solomon, S., Greenberg, J., & Pyszczynski, T. (1991). A terror management theory of social behaviour: The psychological functions of self-esteem and cultural world views. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 24). London, Academic Press. Staub, E. (1989). The roots of evil: The origins of genocide and other group violence. Cambridge: Cambridge University Press. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The Social Psychology of Intergroup Relations. Monterey: Brooks/Cole. Turner, J. C. (1999). Some current issues in research on social identity and self-categorization theories. In N. Ellemers, R. Spears, & B. Doojse (Eds.), Social identity: Context, commitment, content. Oxford: Blackwell. Tesser, A. (1988). Toward a self-evaluation maintenance model of social behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 181–227). New York: Academic Press. Wylie, R. C. (1979). The self-concept. (Vol. 2). Lincoln: University of Nebraska.
Author Biographies Andrew C. Grocott (MA) is the skipper of a luxury yacht and sails all over the world. The research reported in this article is derived from his masters thesis. John A. Hunter (DPhil) is a senior lecturer in the Department of Psychology at the University of Otago. His research interests focus on intergroup discrimination, social motives and youth development.
123
