Hazing in orientation programmes in boys-only secondary schools

Hazing, associated with initiation, aims at taking newcomers from novice status to a status of functional and acknowledged members of a new group. However, the process is often dangerous, injurious, and usually secretive. Hazing may occur as an unauthorised component of institutionally sanctioned orientation programmes commonly held for new students at educational institutions at the beginning of the academic year. This study focuses on the occurrence of hazing elements in orientation programmes (OP) for Grade 8 boys primarily run by Grade 12 learners in boys-only secondary schools in South Africa. A cross-sectional survey was conducted by administering a researcher-designed questionnaire to a non-probabilistic sample of 296 Grade 12 learners enrolled at three boys-only secondary schools in Johannesburg. The computer assisted analysis strategy included frequency distributions, exploratory factor analysis, and analysis of variance. Findings indicated that respondents generally agreed with regard to the structure, aims, and behaviours common to orientation programmes. Respondents strongly disagreed about the occurrence of physical and sexual abuse and activities aimed at discomfort in the OP; however, respondents showed ambivalence about the occurrence of certain activities, which may deteriorate into hazing. Prior experience of an orientation programme when in Grade 8; length of enrolment in the school, and boarder status affected respondents’ perceptions of certain aspects of orientation programmes.


Introduction
Hazing, associated with initiation, is a centuries-old practice aimed at initiating newcomers from novice status to acceptance as functional and acknowledged members of a new group.The process of hazing is dangerous, risky and usually secretive (Hollman, 2002).Different individuals and organisations describe hazing in diverse ways but, broadly speaking, it comprises negative experiences suffered by the newcomer as he/she starts to integrate into the group (Ellsworth, 2005).Injurious practices include humiliation, degradation, psychological, physical or sexual abuse and any other form of endangerment that in some way compromises individual dignity.This could range from a simple act of humiliation to an action which may result in death (Ellsworth, 2006).The role players in the hazing process are the perpetrator or hazers, that is, those who inflict an action; the bystanders, who are a critical part of the group but are not actively involved; and the victims, who are those newcomers at whom the actions are directed as they endeavour to join the new grouping (Lipkins, 2006).No single individual takes responsibility for hazing and any brutality is rationalised as a justifiable means to build group unity (Nuwer, 2000).Victims in hazing may cooperate with the process willingly or under compulsion.Paradoxically, the victim is in favour of the ultimate consequence, namely, group membership, but not necessarily of the process.Further, role players in hazing are usually unwilling to speak about hazing activities and will even lie about personal injury to prevent exposure -to disclose is to betray group trust.For the purpose of this paper, hazing can thus be defined as "any activity expected of someone joining [a group or institution] that humiliates, degrades, abuses or endangers them regardless of their willingness to participate in the activity" (Allen & Madden, 2008:2).Hazing occurs in many contexts, such as the military, secret societies, universities and schools and sports teams (Wegener, 2001).
In the light of ambiguity created by associated terms, the authors' choice of terminology is clarified.The most common term used in South African schools and universities to denote processes whereby new students are admitted to the community is initiation, also popularly known as ontgroening in Afrikaans (Afr.)-mediuminstitutions (transliterated: the process whereby greenhorn status is exchanged for mature status).Initiation is not intended to be harmful but practice suggests that it is difficult to guarantee the neutral nature of the process.Initiation in South African context may also refer to cultural practices whereby boys from certain groups participate in rites to mark entry into the community as adult men (South African Human Rights Commission, 2006).Cultural initiation has on occasion been accompanied by documented cases of abuse (Venter, 2011).Initiation may also refer to religious ceremonies regulated by orthodox ritual.Cultural and religious initiation falls beyond the scope of this paper.All terms were carefully considered during the research design; however, due to the preponderance of North American research identified in the exploratory literature survey, the authors opted for the term hazing.This should be regarded as a term frequently associated with but not synonymous with initiation.
Hazing frequently occurs as an unauthorised component of the hidden 'curriculum' of institutionally sanctioned OPs commonly held for new students at the beginning of the academic year at universities and secondary schools (Hollman, 2002).The official intention of such orientation programmes is to familiarise new students with both the unfamiliar physical environment and the history and ethos of the educational institution; to introduce new students to the senior student body and its leaders; and to create a sense of unity and belonging among new students (Cambron-McCabe & Bueschel, 2005).Notwithstanding, OPs, however well planned and well intended, frequently degenerate into hazing or, at least, accommodate elements of hazing.The practice of hazing during orientation-type programmes in educational institutions, particularly universities, has been well-documented in the United States of America (USA) (Allen & Madden, 2008) and in European, Asian and Australasian higher education and schooling systems (Finkel, 2002).Hazing practices are thus a concern for educational administrators who have a duty of care towards students and should be proactive in this regard if legal charges of dangerous hazing activities made by parents or students are to be avoided (Crow & Rosner, 2002).
OPs for new Grade (Gr.) 8 learners are frequently conducted at South African secondary schools.Many activities in OPs are devised and run by Gr. 12 learners, who may or may not hold leadership positions in the school.Certain activities, which commonly form part of an OP and which are overtly aimed at having innocuous 'fun' and achieving affiliation within the new group, have the potential to deteriorate into hazing.Hazing practices during OPs are seldom critically questioned by the participants as the process is generally accepted as part of school 'tradition' (Huysamer, 2012).
Parents may also tend to reinforce the idea that school-based rituals, which have the potential to accommodate hazing activities, are acceptable as they recall similar events from their school days, which, in their view, did them no harm (Gibson, 2009) or which form part of the strict school discipline favoured by many South African parents (Morrell, 2001a).
Thus understood, it can be argued that hazing has occurred in South African boys' schools to some degree for more than a century (Morrell, 2001b), although little formal research has been conducted in this regard (Serrao, 2009).While both boys and girls are subjected to hazing, the degree of hazing that boys experience is more physically tangible and widely publicised.Further, hazing is more likely to occur to a greater extent in boys-only schools than in coeducational schools.This is linked to the stereotype that the male gender is strong, able to cope with physical hardships and can bear emotional pressure with no apparent signs of difficulty (Morrell, 2001c;Reichert, 2007).In a local context, hazing at boys-only schools could also be connected to the conscription into military service which many white South African males complied with between 1952 and 1992 (Callister, 2007).Boys' schools with boarding facilities for learners are particularly prone to including elements of hazing during OPs as activities in school hostels after school hours are more difficult to supervise adequately (Huysamer, 2012).The controversial and injurious nature of hazing in South African education has regularly been brought to public attention when allegations have been made of physical, sexual or emotional abuse suffered by first-year university students (beyond the scope of this paper) (Soudien, 2010) or by Gr. 8 learners during OPs in secondary schools (Ritchie, 2009).
Against this background, this paper reports on a study which investigated the views of Gr. 12 boys regarding the occurrence of hazing embedded in OP programmes held for Gr. 8 learners in selected boys-only schools located in the Johannesburg area.The main research question: What are the views of Gr. 12 learners of the occurrence of hazing in OPs presented for Gr. 8 learners in boys-only secondary schools?was addressed by a literature review and survey research.

Theoretical framework
To provide a theoretical framework for the phenomenon of hazing, with particular reference to hazing in male groups, this paper draws on gender theory and the psychological theories of cognitive dissonance, severity-attraction hypothesis and severityaffiliation-attraction theory.Nuwer (2004) argues that an understanding of hazing is powerfully shaped by social norms related to gender.According to gender theory, masculinity is a socially constructed, ideologically driven concept which problematises a proposed set of qualities, characteristics or roles generally considered typical of, or appropriate to, a man (Connell, 2005).Versions of masculinity are culturally and historically embedded and learnt during socialisation rather than determined by biology (Kimmel, 2000).Crucial to understanding masculinity is the notion of patriarchy which supports a male-female gender dichotomy which privileges masculinity over femininity and stresses physical prowess, established codes of 'male' honour and qualities such as composure in the face of trial, courage, 'gameness', self-reliance and integrity (Johnson, 2000).The stereotype of masculinity influences what is considered a 'regular' boy, is entrenched in school culture and determines how boys bond in competitive, peer-based, age-graded, single-sex groups and how they transmit notions of ideal masculine behaviours (Martino, 2005).Against this theoretical background, the connection between masculinity and hazing is easily discerned (Nuwer, 2004).
Hazing activities in male groups test the masculine traits that the dominant heterosexual male wishes to pass down through the implementation of 'ordeals' designed to achieve an acceptable level of 'manliness' within the group.Components of a masculine ideology entrenched in hazing often include the marginalisation of homosexuals, the sanctioned use of aggression, use of social isolation, testing of pain tolerance and the devaluation of women (Johnson, 2000).Activities which would be otherwise condemned are regarded as "somehow 'natural', 'normal' or just 'boys being boys'" (Martino, 2005:2).Groups of young males are the most volatile and at risk as they lack the level of self-consciousness to appropriately monitor the tone and nature of activities which occur under the umbrella of hazing.Within an ideology of masculinity, the removal of hazing actually threatens opportunities for males to prove their manhood and, thus, their heterosexuality.Further, the masculinity of men/boys who avoid hazing is likely to be subject to suspicion and scrutiny among their peers (Nuwer, 2004).
Taken with the above discussion of gender theory, the theories of cognitive dissonance, the severity-attraction hypothesis and the severity-affiliation-attraction theory further contribute to an understanding of hazing.Cognitive dissonance comes about when a person simultaneously holds two opposing views.This causes a tension within the individual and, as a result, the individual adapts or changes one of their views in an attempt to reduce the internal tension caused by the opposing views (Cooper, 2007).For example, when a victim who considers himself to be intelligent is forced into a de-grading act during hazing, he softens the experience by convincing himself that 'it was not so bad' and so reduces the conflicting tension between his identity as an intelligent person and his participation in a demeaning act.Consequently, he can regard the group with which he seeks affiliation, or the perpetrator, positively (Lipkins, 2006).Cognitive dissonance as explanatory theory for hazing can be fruitfully related to the severity-attraction hypothesis.
The severity-attraction hypothesis proposed by Aronson and Mills (in Aronson, Wilson & Akert, 2010) states that the more effort an individual exerts to reach a goal, the more the individual will rationalise that the effort is worthwhile no matter how rigorous or injurious.When a young man strongly desires to join a group but hears that the hazing requirements are extreme, he rationalises that there is greater merit in joining the particular group than another group which demands a milder degree of hazing for affiliation in an attempt to reduce his own level of cognitive dissonance.
The severity-affiliation-attraction theory is similar to the severity-attraction hypothesis (Lodewijkx & Syroit, 2001).When someone is placed in a threatening or stressful situation, the individual will identify with other individuals, especially those who have experienced a similar situation.As the level of threat or stress increases, so does the bond between those who experience the situation and those who have previously experienced a similar one.Thus, a bond is formed between victim and hazer or hazer and bystanders who are already seasoned members of the group.
The above-mentioned psychological theories in conjunction with gender theory are useful in understanding why it is so difficult for authorities to stamp out hazing practices in boys schools and why young men involved in hazing are so reluctant to speak out against it, even after serious injury is incurred.

Method
This study made use of a cross-sectional survey design and was conducted by administering a researcher-designed questionnaire to 296 Gr. 12 learners, 18 years and older, enrolled at three boys-only secondary schools located in the Johannesburg area.A nonprobabilistic convenience sampling procedure was used (McMillan & Schumacher, 2006).The schools were selected purposefully against the following criteria: wellestablished boys-only schools (in existence for 50 years or more) which implement an OP for new Gr.8 learners, which is administered primarily by Gr. 12 learners as part of the annual school programme.Participation in the study was voluntary and the protocol, including the questionnaire, was approved by the Gauteng Department of Education and the respective school principals.School access was facilitated by the chief researcher's reputation as a long-time member of the teaching corps in the area and her undertaking to give principals a copy of the completed report and to discuss findings with school staff on request.This commitment has since been honoured.All respondents were 18 years of age or over and their written consent was lent through the completion of the questionnaire.A covering letter stipulated the identity of the chief researcher, purpose of study, protection afforded the respondents by keeping identities confidential, instructions for completion, opportunity to obtain results and thanks.
The survey instrument consisted of a three-page questionnaire with 75 items intended to obtain descriptive data.The questionnaire was researcher-designed and informed by literature, lengthy professional experience, observation and preliminary discussions with teachers and former pupils at boys' schools undertaken in the principal researcher's professional capacity.Expert peer critique and pre-testing of the questionnaire preceded data collection.Eight closed form questions were used to elicit information on biographical attributes without disclosure of identity.Perceptions of the current structure, aims and the kind of activities included in current OPs at the respective schools according to three domains: administration of the OP (20 items); Gr. 8 behaviours common to an OP programme (14 items); and activities expected of Gr. 8 learners by Gr. 12 learners (32 items), ranging from supportive activities to hazing, were measured using a five-point Likert scale, where 1 = Strongly disagree; 2 = Disagree; 3 = Unsure; 4 = Agree; 5 = Strongly agree.A non-applicable option was also included.An open-ended item allowed the respondent to express any individual concerns.Data were collected by the principal researcher after school hours at a venue and time determined by the school principal in August 2011.A response rate of 100% was achieved.
Data were coded and analysed by using the Statistical Analysis System (SAS) version 9.2.The analysis strategy adopted is described as follows: One-way frequency tables were compiled on the biographical attributes of respondents, which allowed for a description of the sample.These have not been included in the paper to conserve space.Composite frequency tables (Tables 1 to 3) on the questionnaire items that described each of the three domains listed provided detailed information of the response pattern of individual items within each domain and gave an overall view of respondent perceptions.Chi-square tests performed indicated where the response patterns of certain items differed significantly from other item response patterns on the 0.01 level of significance.Thereafter, to provide a more compact presentation of the data results contained in the composite frequency tables, a dimensionality reduction technique, that is, exploratory factors analysis using varimax rotation, was carried out.
The analyses investigated the underlying data structure of the domains (constructs) to determine factors present within each domain.The rotated factor pattern indicated the questionnaire items which weighed on a specific factor and factors were labelled to best describe the issue addressed by the grouping of questionnaire items.Scale reliability analyses (Table 4) were carried out on each construct and set of factors by calculating Cronbach's alpha coefficient.A score of 0.7 and higher indicated acceptable reliability of all constructs measured.The mean rating value of each of the factors was calculated and used to describe respondents' general perceptions of each factor.Mean scores were interpreted on the same agreement rating scale as the original questionnaire item responses (1 = Strongly disagree to 5 = Strongly agree).Finally, the contribution of four biographical attributes to respondent perceptions was investigated by means of analysis of variance (ANOVA) (Table 5).Once statistically significant biographical attributes had been identified, the nature of the effect of the biographical attribute on respondent perceptions was determined by the Bonferroni multiple comparison of means scores.Open-ended responses were coded to identify recurring themes.The statistical findings represent the combined views of respondents in all three schools; the scope of the study did not allow for an inter-school comparison.

Biographical profile of the respondents
One-way frequency distributions on the biographical attributes of respondents assisted in describing the sampled learner population.The frequency distribution of home language indicated that 61% of the respondents were English (Eng.)speaking; a further 20% were speakers of African languages; and 20% had Afr. or another language as home language.Christianity was the most common religious affiliation (78%).A substantial percentage of respondents (77%) had participated in an OP when they were in Gr. 8. Almost a quarter of the respondents (24%) were boarders and the majority of respondents (61%) had been enrolled at the same school for their entire school education inclusive of primary schooling.In most cases (82%), schools fees were paid by parents.In total, 335 leadership positions were filled by the 296 respondents indicating a sample with a very strong leadership element whereby certain respondents held more than one leadership position in the school.In general the biographical profile confirmed that respondents would have valid experience of OP implementation as a leader, a bystander, or both.
Composite frequency tables (Tables 1 to 3) calculated on questionnaire items dealing with the three domains gave an overall view of respondents' perceptions of the implementation of the OP, including possible hazing elements.Column 1 refers to questionnaire item; columns 2 to 5 give the agreement rating (the first line of each row gives the frequency count; the second line the frequency percentage).The last column gives the totals.The ÷ test results are included at the end of each table.In the 2 discussion, percentages for strongly disagree and disagree and strongly agree and agree have been combined.
Table 1 indicates that most respondents agreed on the aspects dealing with the structure and content of current OPs as follows: commencement is at the beginning of the school year (89.1%); participation is required of Gr. 8's (89.7%);OP is run in part by Gr. 12 learners (92.1%);Gr. 12 learners are assigned mentorship responsibility (58.1%);Gr. 8's are required to participate in sports events (84.2%); and Gr.8's are required to attend a camp (51.6%).Most respondents also agreed on the aims of the OP as follows: to instil a sense of belonging in the school (80.6%); to instil pride in the school (65.9%); and to build school tradition (74.9%).In designing items dealing with aims of OP, we argued that although these aims may not be stipulated in a curriculum document, aims are effectively communicated to and across learner cohorts through informal discourse (e.g.speeches, impromptu remarks by seniors, staff and fathers who are 'old boys').This is especially likely in schools with a lengthy history.Accordingly, aims become part of school ethos and institutional memory; although it is possible for an individual boy to resist institutional aims.Thus, we argued that the meaning of items dealing with aims as integral part of the OP would have been generally understood and responses would yield valid results.Further, most respondents agreed that Gr. 8's have to perform chores for Gr.12's (79.5%) and have to stand back for Gr.12's as a token of respect (86.2%) during the OP.Importantly, a significant percentage of respondents disagreed that a member of staff is always present at the OP (39%) or were unsure on the issue (21.6%).Teachers act to ensure a safe environment and thus teacher invigilation is essential to ensure that OP activities do not violate the physical or psychological wellbeing of juniors.A note of ambivalence emerged around respondents' responses regarding the treatment of Gr. 8's: a significant percentage disagreed that Gr. 12's humiliate Gr. 8's (46.9%); however, a similar percentage of respondents agreed that Gr. 12 learners intimidate Gr. 8's (48.9%).
Table 2 indicates that most respondents agreed that the OP included Grade 8 behaviours that could be regarded as neutral, such as learning general information about the school (70.1%), school rules (95.2%) and about the Gr. 12 learners (80.3%); supporting school events (93.9%) and performing simple duties (78.1%).A greater proportion of unsure responses (20% +) was reported for the items: taking an oath of loyalty to school; keeping compulsory silence; wearing a large name tag; and carrying 'silly' objects around.This indicates some uncertainty among respondents about the occurrence of activities which may embarrass Gr. 8's and could thus be regarded as hazing.The totals row in Table 3 indicates that the most responses fell within the disagree and strongly disagree categories (3245 + 2046) indicating that respondents generally disagreed that such activities formed part of the OP, including items which comprised hazing (items 9 -32).For example, respondents disagreed that the following took place: striking a Gr. 8 with the fist (70.4%); sleep deprivation (64.4%); consuming strange food (69.2%) or alcohol (80.4%); exposure of genitals (92.1%); encouraging group masturbation (92.7%) and exposure to pornography (82.4%).However, respondents agreed that Gr. 8's are required by the boys leading the OP to do excessive exercise (68.2%) and that Gr. 12's shouted at Gr. 8's (58.5%).A significant proportion of unsure responses (20% +) were reported for the occurrence of activities: swearing at Gr. 8's and disturbing Gr. 8s' sleep This suggests ambivalence among respondents as to certain Gr. 12 behaviours injurious to Gr. 8's.

Results of data reduction through factor analysis and reliability testing
Exploratory factor analyses were conducted separately on the response data to obtain a more compact presentation of results.The analyses determined the factor(s) underlying each domain and allocated questionnaire items to each factor.Summary results are presented as follows: For the domain Administration of OP, four factors were identified: i) structure of OP; ii) instilling tradition and belonging; iii) risky elements in OP; and iv) Gr. 8 participation required in school events.The domain, Gr. 8 behaviours common to OP, was described by the factor, General activities expected of Gr. 8's.For the domain, Activities expected of Gr. 8's administered by Gr. 12's in OP, four factors were identified and labelled: i) support received from Gr. 12; ii) physical abuse activities; iii) activities aimed to discomfort; and iv) sexual abuse activities.Activities aimed at discomfort included sleep deprivation and sleeping without cover outside a building to distinguish them qualitatively from overt abuse.As mentioned earlier, scale reliability testing using Cronbach's alpha coefficient was performed to establish the internal consistency reliability of each factor in order to test whether the grouping of items represented an underlying construct.Subsequently, the mean rating value of each of the factors was calculated to describe respondents' general perceptions regarding each factor (see Table 4).
Table 4 indicates that the Cronbach's alpha values calculated were all above 0.7; therefore the constructs and concomitantly the questionnaire could be deemed reliable.The mean values (interpreted on the same agreement rating scale as the questionnaire item responses) indicated that respondents agreed strongly about the structure of the OP (mean score of 4.33) and the Gr. 8 behaviours common to OP (mean score of 4.08).Respondents disagreed very strongly about the occurrence of sexual abuse activities (mean score of 1.35) and were disagreed, albeit to a lesser degree, about the occurrence of physical abuse (mean score of 2.58) and activities aimed at discomfort (means score of 2.13).
Thereafter, the question was asked whether biographical attributes of respondents also affected their perceptions of the OP.To address this question as set out in the analysis strategy, ANOVA were performed on each set of factors to determine the significance of the effect of biographical attributes on perceptions.Properties investigated were boarder status; prior OP participation when in Gr. 8; length of current school attendance; and home language.Moreover, as the analysis of variance identifies biographical attributes that significantly affect perceptions, but does not indicate the nature of the effect on perceptions, Bonferroni multiple comparisons of means tests were also performed on the data to investigate how perceptions were affected by the biographical attribute.Table 5 presents the results of the analysis of variance and associated Bonferroni results.Table 5 indicates that prior participation in OP when in Gr. 8 affected respondents' perceptions of the structure of the OP and its aims, instilling tradition and belonging.Thus, respondents who had prior experience of an OP when they themselves were in Gr. 8 held a significantly more positive perception regarding the factors mentioned.For example, respondents with this attribute agreed significantly more strongly than respondents who did not have this attribute (compare mean scores of 3.78 and 3.61) that participation in an OP contributed to instilling tradition and a sense of belonging.Table 5 also indicates that boarder status affects respondents' agreement concerning the occurrence of: risky elements in the OP; general activities expected of Gr. 8's; support received by Gr. 12; physical abuse; and activities aimed at discomfort.
The category mean scores for respondents who are boarders were significantly higher than those of respondents who were day scholars.However, perceptions of the aforementioned factors were not necessarily positive.For example, boarders agreed more strongly about the occurrence of risky elements in the OP (a mean score of 3.74, which rounds to 4) as opposed to the perception rating of day scholars (a mean score Agreement rating scale level for score interpretation: 1 = strongly disagree; 2 = disagree, 3 = Unsure; 4 = agree; 5 = strongly agree of 3.42, which rounds to 3).Day scholars were more inclined to disagree that physical abuse occurred in the OP (mean score of 2.49) in comparison with respondents with the attribute 'boarder' (a mean score of 2.82 approximates 3).Table 5 also indicates that length of time at current school affected respondent perceptions of the occurrence of: risky elements involved in OP; required Grade 8 participation in OP; and support received from Grade 12. Respondents who had undergone only their secondary schooling at the respective school (i.e. 5 years) responded with unsure, whereas respondents who had spent more years at the school inclusive of primary schooling (i.e.> 5 years) held significantly more positive perceptions of the occurrence of risky elements in the OP and of support received from Gr. 12's.A significant percentage of respondents (39%, or 115 out of the total of 296) completed the open comment.Comments fell into three broad categories: endorsing aims of the OP; ambivalence around hazing; and conceptions of masculinity linked to hazing.Examples of comments endorsing the aims of the OP were: It is done to keep the school tradition just so the Gr.8's should keep the tradition when they are in Gr. 12; and Gr. 8 OP makes the learners feel like they are a part of the school 'clan' and gives them good memories about high school in the long run; and OP is something that each grade can only share with each other and stories that only they will understand.Comments which illustrated ambivalence around hazing were: It should not be taken overboard.I think it is best to do small things like wearing abnormally large boards with a joke or carrying a brick; and Taking Gr. 8's out of the boarding school at midnight for a compulsory swim is one of the best ways of pulling them together.Comments which alluded to popularly held views of masculinity were: It is part of growing up.Suck it up and be a man; Gr. 8's…must not be babied; Things that happen in boys' schools…are for boys and boys understand why these things are done; Pain is our friend.We should use any means of discipline necessary to correct behaviour.

Discussion
The results of this exploratory study which made use of a non-probabilistic sampling only allow for broad generalisations to be made about the views of the questionnaire respondents, that is, Gr. 12 boys in three selected boys-only schools in a specific location.However, this limitation does not rule out the usefulness of the results which raise certain points for discussion.Firstly, it was positive to find that the majority of respondents in the survey strongly disagreed about the occurrence of physical and sexual abuse and traditional injurious hazing activities.However, respondents were generally in agreement about the structure and the aims of OP programmes.In particular, prior experience in an OP when in Gr. 8 was associated with strong agreement about the aims of the OP.However, positive agreement about the idealistic goals of building school tradition and school spirit requires interrogation.Endorsement of these goals of the OP may lead participants, both junior and senior boys, to tolerate demeaning elements and to justify risky behaviours in an OP as a means to meet the ends embedded in camaraderie and esprit de corps.Furthermore, the general agreement among respondents that OPs inculcate information about the school and its rules suggests the effectiveness of OPs as means of socialisation into the school.Borderline activities, such as shouting orders and excessive exercise, may not be violent but act as tools of control whereby seniors exercise ascendancy over juniors.Herein power is conceived as male on male dominance whereby the weaker (the junior boy) is persuaded to conform to codes of behaviour by the stronger (the senior boy) (Connell & Messerschmidt, 2005).This may also support another implicit mechanism: activities which intimidate to prepare juniors to bend to teacher control, predominantly by male teachers in the boys' school context.Thus, a 'regime' of masculinity is constructed daily through unquestioned routine practices (Connell, 2005).This would also explain why some adults in the school may turn a blind eye to and tacitly support hazing when it does occur (Nuwer, 2004).
Interestingly, Vigar-Ellis (2013) found that rigorous 'rites of passage' enacted in South African boys' boarding schools as part of a 'toughening up' process are approved of and even required by parents if it serves school tradition.Where such beliefs among learners (and their parents) prevail, risky activities are more likely to be perpetuated and masked within the general structure and content of OPs.Several studies on hazing (Ellsworth, 2005;2006;Nuwer, 2004) stress that popular acceptance of the stated goals of initiation programmes makes it difficult to uncover and eliminate the hazing elements which frequently infiltrate such programmes.Indeed, respondents agreed that the complementary objective of engendering group belonging was part of the OP.Further, respondents who had spent the longest time in a particular school agreed more strongly about the occurrence of risky elements in the OP, possibly because of their lengthy experience in the school had exposed them to these negative behaviours, which had been implemented in the interests of eventual group belonging.However, the price exacted for the achievement of group belonging can be high.Martino and Pallotta-Chiarolli (2001) found that boys may sanction and endure humiliating or even physically injurious activities simply to achieve the highly desirable goal of group belonging in a male school culture.In line with the tenets of cognitive dissonance and severity-attraction theory, it is possible that senior boys who administer OP, the bystanders and the junior boys may endorse humiliating or even physically injurious activities simply to gain or retain the respect and acceptance of the group.The general agreement of respondents that the OP includes Gr. 8 acts of servitude and demonstration of respect for Gr.12's is also a point which deserves comment.These activities, together with those which might have been merely intended to be silly or fun, such as wearing a name tag, being forced to carry an object or to do compulsory exercise, can easily deteriorate into harmful activities and the boundaries between such activities can easily be blurred (Nuwer, 2004).In this case hazing then becomes difficult to untangle from standard components of an OP.The latter is confirmed, at least to some degree, by anecdotal evidence given by adults who had been 'victims' of so-called initiation in South African schools where the topic has been debated in the media (Clark, 2009;Initiation (no boundaries) -what are boarding schools thinking !, 2009).This confusion as to what is horseplay and what is injurious is suggested by the relatively high percentage of 'Unsure' responses among respondents in this survey regarding the occurrence of risky activities in the OP.Further, the statistically significant association between boarder status of respondents and their perception of the occurrence of physical abuse found in the study also corroborates that boarding schools for boys are more likely to be places where learners are most vulnerable to hazing activities.
The open comment item gave respondents the opportunity to express personal views around hazing and their experience thereof which was not possible in survey items.The comments hinted that many respondents approved of the more hazardous OP elements in the light of a culture of masculinity.Comments also suggested the existence of hazing practices which had not been captured in the survey items.A supporting example follows: I know that from shouting and talking and writing out punishment given to juniors, they do not learn.They are young and sometimes giving them a hiding is the only way to teach them.A hiding for the right reason.This kind of comment suggests that boys who identify with the predominant cultural construction of masculinity, hard to avoid in an all-boys school (Martino & Pallotta-Chiarolli, 2005), will be less likely to recognise or report borderline activities to authorities.This is an issue that should be further researched in subsequent studies.
A further significant observation made by most respondents was that the presence of a member of teaching staff could not always be counted on during the administration of OP programmes.Educators are expected to care for and ensure the physical, psycho-logical and spiritual well-being of the learner (Woolman & Fleisch, 2009) and this injunction is of particular importance if the structure and content of OPs are to avoid violating the constitutional rights of learners to an educational environment free from harm and degrading treatment (Masitsa, 2011).In particular, the legal implications of hazing are articulated in the South African Schools Act 84 of 1996: Regulations to prohibit initiation practices in schools (Republic of South Africa [RSA], 2002).

Conclusion
Hazing activities in educational institutions may range from the subtle harassment of the vulnerable to activities which are illegal and may result in permanent harm and even fatalities.While OPs do not set out to serve the ends of hazing, they constitute a vulnerable part of school practice, particularly in the environment of boys-only schools.While taking into consideration the limitations of this small-scale study, it is recommended that adult invigilation of all activities constituting an OP is essential both to protect participants and to avoid litigation against individuals and the school.Potentially hazardous activities in an OP should be omitted and replaced with positive behaviours to build affiliation and school loyalty.A written guideline which defines hazing activities (e.g.physical abuse, sexual abuse, and activities aimed at discomfort as described in the questionnaire) and the consequences for infringement should form part of the code of conduct for learners and be disseminated to the entire school community, including parents, every year to inform every new cohort of role players.In particular OP programmes in schools with boarding facilities must be managed with utmost care and activities after school hours in boarding hostels should be strictly supervised.

Table 1
Respondents' responses relating to administration of OP as currently implemented at their school

Table 2
Respondents' responses relating to Gr. 8 behaviours which commonly form part of the OP at their school

Table 3
Respondents' responses relating to the occurrence of activities expected of Grade 8 learners by Grade 12 learners at their school

Table 4
Results of scale reliability testing and mean scores