Printer Friendly

Measuring masculinity in an Afro-Caribbean context/Como medir la masculinidad en un contecto Afro caribeno/Mesurer la Masculinite Dans le Contexte Afro-caribeen.

ABSTRACT

This paper proposes a new scale for the measurement of masculinity, with special reference to Afro-Caribbean males. The scale which is labeled the Macho Scale focuses on masculine identity and gender relations. The macho scale achieved an alpha level of .82 based on a survey of 1,141 fathers in four Jamaican communities, and is shown to have a high level of convergent validity. The paper seeks to situate this analysis within the wider literature on the Afro-Caribbean family and gender relations, and current policy concerns in regard to gender socialization, domestic violence and crime.

Este trabajo propone una nueva escala para medir la masculinidad, con especial referencia a los hombres afro caribenos. Dicha escala, que seria nombrada la Escala de Macho, se centra en la identidad masculina y las relaciones de genero. La escala de macho lograda a partir de un nivel alfa de .82 se basa en una encuesta en la que participaron 1,141 padres en cuatro comunidades jamaicanas, y muestra un alto nivel de validez convergente. El trabajo trata de ubicar este analisis dentro de la existente literatura mas amplia sobre la familia afro caribena y las relaciones de genero, y tambien se refiere a las preocupaciones concernientes a la politica actual con respecto a la socializacion de los generos, la violencia domestica y el delito.

Cet article propose une nouvelle echelle, a l'attention toute particuliere des hommes afro-caribeens, pour mesurer la masculinite. Cette echelle, baptisee Echelle Macho, se focalise sur l'identite masculine et les relations entre les genres. L'echelle macho a atteint un niveau alpha de 0,82 lors d'une enquete menee aupres de 1 141 peres de famille dans 4 communautes jamaicaines, et manifeste egalement un niveau eleve de validite convergente. L'article tente de replacer cette analyse dans la documentation plus generale sur la famille afro-caribeenne et les relations entre les genres, et sur les questions actuelles de politique publique en matiere de socialisation entre les genres, la violence domestique et la criminalite.

**********

The Afro-Caribbean Context

Despite the efforts of researchers to capture and to measure the essential qualities that inhere in masculinity, the goal continues to be an elusive one; since arguably, scales measure only what they agree to explore. Nonetheless, the issue of what constitutes masculine self-definitions assumes particular cogency in societies and cultures where patterns of gender socialization create pressures for young males to assert dominance and to engage in risk-taking, and where the society itself exhibits a high level of interpersonal violence and group conflict. Jamaica is an example of such a society.

Jamaica is a small Afro-Caribbean country which has achieved international recognition through its achievements in music and sports, and because of its history of struggle for social and economic justice. This was linked with a spirit of assertiveness which, from the 1970s, earned it a place of leadership among Third World nations. It is these same qualities which seem to be expressed in definitions of gender roles for both men and women. However, for males, these definitions are intertwined with patriarchal expectations and a strong emphasis on sexuality. How these definitions are shaped by social class, or constitute core values, is a central question. In addition, as observed by Alexander (1977), there are persistent cultural representations of the Jamaican husband-father which do not always match the reality of men's domestic behaviour. This paper presents the findings of a study of Jamaican fathers across a range of socio-economic levels, and proposes a new measure of masculinity, the Macho Scale. The study, which was conducted between 2005 and 2006, sought to explore men's attitudes towards their role as fathers; the extent of their actual involvement in fathering activities; their view of women's roles; and their sense of themselves as men. The analysis is based on household surveys in four communities, which were selected so as to represent a range of social class levels and urban-rural contexts. After a brief review of the theoretical underpinnings for the study, the methodology is outlined for the construction of the macho scale, and assessing its reliability. The paper then demonstrates that macho levels do vary with community context, age and education level, and presents findings on the relationship between macho levels and number of baby-mothers, patterns of domestic violence, and homophobia.

Foundation Work on Family Structure

The analysis of male gender attitudes in the Caribbean finds its roots in the early studies of the Afro-Caribbean family undertaken by social anthropologists, where the major concern was to identify the structure of the family. These investigations took place within the context of a wider debate on the Afro-American family, which argued either that the Black family was fragmented and disorganized as a result of slavery (Frazier 1939) or that it represented important continuities with the African past (Herskovits 1941). In the Caribbean, the foundation work on family structure can be traced to the 1950s, and was conducted by Edith Clarke, George Cumper, Raymond Smith, M.G. Smith, George Roberts and Lloyd Braithwaite. Clarke and Cumper each separately established that the structure of the family varied with the type of economic organization of the community (Clarke 1957; Cumper 1958), while Raymond Smith traced the cycle of conjugal unions through which women moved over their child-bearing period. On this basis, he pointed to the likelihood that at some point men would become marginal to the family as it became increasingly matrifocal (Smith 1956, 1973; McKenzie and McKenzie 1971).

The contributions of M.G. Smith were advanced within his theory of social and cultural pluralism, which identified cultural sections rather than social classes as the basis for social organization, linking this to the resulting variations in family structure (Smith 1962). The consistent variation in family forms received empirical support from the work of Roberts and Braithwaite (1961), whose demographic analyses showed that child-bearing occurred within a range of conjugal unions, legal and non-legal, coresidential and non-residential. It should be noted that in the early work advanced by Caribbean anthropologists, the major goal was to establish the structure of the family and to delineate the roles of men and women. However, an inadequate reading of the work often served to provide a basis for later accusations of male marginality and irresponsibility.

The growth of women's studies in the Caribbean in the 1970s and '80s widened considerably the understanding of the structures which limited developmental opportunities for women, locking them into positions of disadvantage (Anderson 1986; Barriteau 1998). Within the English-speaking Caribbean, this body of research was pioneered by academics at the University of the West Indies through the Women in the Caribbean (WIC) project (Massiah 1986), also serving to expand analysis beyond the realm of kinship and the family. However, the project paid little attention to the expressed attitudes of men, beyond examining male perceptions of women. Nonetheless, these early studies served to highlight some of the basis for male-female conflict, given the negative expectations of female behaviour expressed by men (Barrow 1986).

The direct exploration with Caribbean men regarding their own gender attitudes has developed both in response to growing policy concerns about male sexual risk-taking, fathering inadequacies and increasing violence at the community and societal level, as well as in relation to the burgeoning field of masculinity studies (Barrow 1998; Branche 2001; Chevannes 2002; Lewis 2003; Reddock 2004). Since the current study has developed out of the research on men's family roles and attitudes, the policy-oriented stream is briefly reviewed here.

One of the first studies with Jamaican men was undertaken by Barry Chevannes at the request of the National Family Planning Board, where the explicit concern was to explore the impact of male chauvinism on contraceptive behavior (National Family Planning Board 1986). This research project utilized a survey with a large national sample, followed by in-depth interviews, and established that men expected to head their households, although they endorsed women's employment outside the home. Close to a half of the sample admitted to having more than one sexual partner, although support for "outside" women declined with age, and was lower among those legally married. A later study, sharing some similar objectives, was conducted in 1999 by the Research in Fertility Unit at the University of the West Indies (Bailey et al. 1999). This study also utilized mixed methods and observed that among young males, and among those at lower income levels, there was pressure to demonstrate masculinity by having many girlfriends, and by fathering children at an early age. The need for more positive models of masculinity was among the conclusions from this study.

In addition, detailed analysis conducted by Norman and Uche of the 1997 Reproductive Health Survey documented the impact of "machismo attitudes" among young men aged 15 to 24 years, and showed that in this national Jamaican sample, such attitudes were strong predictors of sexual risk-taking (Norman and Uche 2003). The machismo attitudes included the belief that boys should have sex to demonstrate their manhood, and that there are benefits from having early sex. The risk-taking behaviours included multiple sexual partnerships, and inconsistent condom use.

Collaborative research between the University of the West Indies (UWI) and the University of California (UCLA) in the early nineties generated rich qualitative data on male understandings of sexuality and the factors that contributed to high-risk sexual behaviour (Chevannes and Mitchell-Kernan 1992). The study pointed to a strong impetus among Jamaicans to establish multiple relationships, explaining this both in terms of male ideology as well as the material gains which both genders derive from such relationships. The nineties also saw the UWI embarking on a study of the role of men in the family, which sought to explore male perspectives and behavior. This project, titled "The Contribution of Caribbean Men to the Family", was pioneered by Janet Brown of the Caribbean Child Development Centre, with support from the International Development Research Centre (IDRC). It combined community surveys with in-depth community-based animation sessions. The findings established beyond doubt that fathering was an essential part of male identity, and that men were more greatly involved with "father work" than was generally recognized (Brown, Anderson and Chevannes 1993). This path-breaking study provided the basis for the subsequent research on gender socialization in the Caribbean, which served to document the fact that males were socialized towards being tough, aggressive and to restrict emotional expressiveness (Brown and Chevannes 1998; Chevannes 2001). Other research, focusing on young persons, provided further support for these findings (Bailey et al. 1998; Branche 1998, 2001; Gayle et al. 2004).

Recent data by Le Franc et al. (2008) based on a survey of persons aged 15 to 30 years has served to document the high level of interpersonal violence within Jamaica, and has estimated that violence at the hands of a partner was experienced by 45 percent of women and 40 percent of men. Sexual coercion was reported by 72 percent of women and by 57 percent of men in this survey. The authors concluded that "the arena for most violence and injury is indeed the partner or intimate relationship setting" (Le Franc et al. 2008, 9). It should be noted that the emphasis on aggressive definitions of masculinity has generally been linked to the escalating levels of crime among young males, although no empirical relationship has yet been established. However, the link between macho attitudes and domestic violence is supported by some brief qualitative research conducted as part of a United Nations Population Fund (UNFPA) Baseline Assessment of Domestic Violence in Jamaica (UNFPA 2009). Young males interviewed in focus groups indicated that their notions of masculinity were characterized by early sexual initiation, having many partners, displaying aggressive attitudes, and the pressure to confirm their heterosexual orientation (UNFPA 2009, 40). While most young men had witnessed gender-based violence, such violence was seen as wrong; nonetheless they felt that women needed to be disciplined if they disrespected them by cheating on them, by failing to obey them, or by not fulfilling their home duties (UNFPA 2009, 44).

The recent analysis of the 2008 Reproductive Health Survey by Priestley is therefore of considerable significance, as it has provided evidence of a clear link between the attempts by men to control the behavior of their female partners and higher levels of intimate partner violence (Priestley 2010). It was noted that a third of this national sample of Jamaican women aged 15 to 49 years said that their partner insisted on always knowing where they were, and among these women, the risk of having experienced physical violence over the previous year was four times higher than among other women. Similarly, the risk of being subjected to sexual violence was five times higher.

Research on gender socialization has consistently highlighted the fact that among Caribbean parents, there is deep concern to discourage any homosexual orientation among young males, and that the establishment of manhood is closely tied to sexual activity with women (Brown and Chevannes 1998). It has also been observed that in Jamaica, homosexuality was often viewed as a male "disease", leading to very extreme reactions towards persons believed to be homosexual (Chevannes 2002). This attitude has resulted in negative international publicity for the country, especially when these homophobic views have been promulgated in popular songs (Hope 2006).

In summary, there is considerable agreement that among Afro-Caribbean males there is a fairly entrenched value system which defines masculinity in terms of dominance, virility and sexual freedom, although there is now evidence of some change. These earlier studies have either used in-depth qualitative methods to explore these issues, primarily in low-income communities, or have approached the constituent aspects of masculinity as separate dimensions and sought to relate these to sexual risk-taking. Where the machismo attitudes have been combined into a single indicator (Norman and Uche 2003), the attitudes explored and their correlates have been necessarily dependent on those questions already available from large-scale surveys. The present study builds on the agreements from existing Caribbean research, and specifically sought to replicate the earlier study of the Contribution of Caribbean Men to the Family (generally referred to as the Fathers Study), conducted between 1991 and 1993 by the Caribbean Child Development Centre. However, the present work is distinguished by the attempt to design an integrated measure to examine the impact of masculinity constructs on a wide range of gender attitudes and behaviour, across different social classes in Jamaica.

Formative Research

This study, like its predecessor, is situated squarely within a social constructionist framework, in which it is argued that men and women learn their expected social roles, but that change is also possible through resocialization and personal growth. The possibility of role conflict and role strain is very real, since men's identities represent a complex of different roles, derived from different domains. The definition of these roles, and the strength of attachment, will also vary with a range of factors, among which social class and education are salient. Kilmartin (2000) summarizes the social constructionist perspective on masculinity by observing that "Masculinity is embedded within social structures that are constructed from a variety of environmental forces: interpersonal relationships, economics, history, culture, group dynamics, reward, punishment and social learning" (Kilmartin 2000, 48). Since from this perspective, masculinity is seen as a set of social arrangements which change in response to changes in these forces and contexts, it is essential to adopt research designs which provide room for these variations to be expressed. This was achieved through the research design, which selected communities at different socio-economic levels in which to conduct in-depth qualitative investigations and household surveys.

The extensive community animation activities undertaken during the qualitative exploration of fathering in the earlier Fathers Study served to highlight the fact that having children had intrinsic value for men. Children were identified as having three powerful meanings:

"... a) they signal a man's prowess with women and prove his manhood

b) they lay the man's claim to a woman he wants, even if only temporarily

c) they serve to affirm a man's maturity through acceptance of responsibility for another person's welfare and development" (Brown, Anderson and Chevannes 1993, 70).

These definitions of manhood which centered on being able to father children were reinforced by comments made in response to the survey questionnaire. Here the large majority of men stated unequivocally that they would not feel like men if they had not been able to have children; they would not consider marrying a woman who was infertile; and that fathering had changed them significantly (ibid., 129). As part of the present study focus groups were held with fathers representing different social classes, and it was observed that for the working-class participants, the father role seemed to be that of a master status. Once they had decided to publicly declare this role, they were now recognized as "a good man" and as "someone you can trust". They also pointed out that they are regarded differently by the society, and they may receive slightly better treatment, once they were recognized as having family responsibilities.

The group discussions held during the course of the first Fathers Study pointed to a general agreement that a man was the rightful head of the household, but that this was linked to his maintaining his position as breadwinner, provider and protector of the household. The traditional gender division of domestic activities was frequently described, both for adults and children. Boys were expected to sweep the yard while girls tidied the house and did the dishes. It was also agreed that boys were generally given more freedom than girls, especially when they approached puberty.

The Bible was often quoted as justification for the idea that a woman does not have the right to refuse her husband sex, but there were very mixed views as to whether men had stronger sexual urges than women. The report noted that throughout the community discussions, there was repeated reference to the sexual double standard, citing the comment of a Mavis Bank woman: "The man is expected to have several women, who in turn are each expected to be faithful to him" (Brown, Anderson and Chevannes 1993, 84). Domestic violence was generally not endorsed, although there was some agreement with the comment "but some women deserve it" (ibid., 91). Such reasons included infidelity and talking back "feistiness".

In the present study, attitudes towards fathering, relationships with women and gender roles were explored through sets of statements with which respondents were asked to agree or disagree. The phrasing of these items was often based on the remarks made by participants in the original Fathers Study, either in the community discussions or in response to the survey questionnaire.

METHODOLOGY

(a) Community surveys

This study is based on a survey of four communities, three of which were situated within the Kingston Metropolitan Area. The fourth was a semi-urban community in the hills of St. Andrew. Because the study was designed as an attempt to replicate and extend the earlier Fathers study, the survey was conducted in two of the communities originally surveyed in 1991 (Braeton and Mavis Bank), with the addition of a middle-income community (Havendale). Given the high level of intra-community conflict within Seivright Gardens at the time of the field work, Denham Town was used instead as a low-income community, which is also generally classified as an inner-city area.

Community Profiles

The three urban communities represent different social class levels, based on education, occupation and subjective social class identification. Braeton represented the working-class community, and is part of the huge Portmore development, initiated in the seventies in response to the unmet housing demand in the Kingston Metropolitan area. Accordingly, Braeton developed as a dormitory community with many residents traveling to pursue employment within Kingston. This entailed a journey across the Causeway, a bridge constructed to link the two parishes. The 1991 study noted that within Braeton, the bulk of occupations fell into the category of artisan and blue-collar. The present study found that the basic blue-collar character of the community had remained stable, as among fathers interviewed, 60 percent were classified as having craftsmen and operative jobs, or working in service and sales jobs. However, considerable infrastructural improvement could be observed as houses had been extended over time.

Mavis Bank represented the semi-rural community, with its social class ranking being generally equivalent to lower middle/ working-class. In 1991, this hillside community was dependent on agriculture, with significant agro-industry made possible by a coffee processing plant and a large egg-producing enterprise. In addition, the proximity of the Mavis Bank community to Kingston allowed some residents to commute to pursue urban employment. This was reflected in the relatively large proportion of residents who reported artisan and blue-collar jobs. The return survey in 2005 found a similar range of economic activities in the community, with the addition of some horticulture; although there was more discussion of youth unemployment and the limited options for youth.

Denham Town has for several years been classified as a garrison community based on the formulation of this concept by Carl Stone (Stone 1985; Chevannes 1992). It is part of the West Kingston complex which includes Tivoli Gardens. Entry to this community was facilitated by UWI's earlier research involvement with the Inner City Housing Scheme (Anderson et al. 2003). The 2005 survey found that the community continued to be quite closed to non-residents, but with the residents expressing a high level of community identification. Economic activity centered on self-employment, both through involvement in the nearby Coronation market and through trades such as auto mechanics, construction and tailoring. More formal employment was generated through the nearby Kingston Public Hospital and the Kingston Wharves.

Havendale is one of the earliest middle-income communities, and has seen a considerable degree of change over time. It is built on the pattern of single-family homes on separate lots, and the area climbs upwards in the direction of the Red Hills. At the upper end are expensive gated homes, while the lower end borders a crowded informal settlement. The location of a major secondary school has lent stability to the area. Residents report a fairly wide range of occupations, with several entrepreneurs who have achieved upward mobility through entertainment or construction gaining a foothold there, despite lower education levels.

Community Samples

Although it was originally expected that respondents would be randomly selected based on an initial listing of fathers in each household, this approach did not prove to be feasible, given the particular characteristics of different communities. The dormitory nature of Braeton and the involvement of many males in self-employed activities such as operating taxis meant that fathers who were listed and selected for interviewing could not be found at home, despite repeated visits and the fact that interviews were conducted in the evenings and on weekends. The project therefore revised its approach and opted to collect a convenience sample, based on all fathers who were available and willing to be interviewed. Following the procedure adopted in the first Fathers Study, interviews were limited to men under 60 years. In all communities, the survey encountered a strong positive response, with many men indicating that they enjoyed the opportunity to express their views on family life, given the previous neglect of male participants in such studies. Since the method of sample selection became equivalent to quota sampling, this unfortunately limits the type of statistics which are used in the present analysis. The estimated coverage for the three urban communities ranged between 50 and 66 percent of all fathers who were listed as residing in households. Specifically, the coverage rate was 59 percent of all fathers in Havendale, 59 percent in Braeton and 66 percent in Denham Town. In Mavis Bank, given the hilly terrain, it was agreed that interviewers would visit all households in addition to interviewing men at community meeting spots. In addition, on one Saturday, the project received unexpected support from a fairly large employer in Mavis Bank who suggested that workers could be interviewed on the job. Interviewing continued until it was estimated that the pool of all available fathers was exhausted.

It is possible to observe the differences between the socioeconomic levels of these communities by noting that within Havendale, 60 percent of all respondents had tertiary education, in comparison to 22.4 percent in the blue-collar community of Braeton. In the low-income community of Denham Town, only 2.9 percent reported having any tertiary-level training. This was fairly similar to the level of 6.5 percent reported by fathers in the hill-side farming community of Mavis Bank.

Given the lack of a random sample, it is useful to compare the profiles of these male respondents with data from the 2001 census in order to provide some perspective on how representative were the fathers samples. (1) It should be noted that any such comparison is limited by the fact that the census does not distinguish fathers at any point, except in relation to men who head households and who reside with their children. Since such a comparison would exclude the men who do not live with any of their children, it was judged preferable to provide a comparison with census data on men aged 25 to 59 years in the four communities. It is expected that in a low-income community, there should be a closer level of similarity, since the majority of men in these areas would have fathered a child by that age. Comparisons based on median age and on educational attainment are shown in Table I and Table 2.

From Table 1, it is apparent that there is a fairly close match between the medianage of the four community samples and that derived from the census distribution, despite the fact that not all men in the census data were known to be fathers. In general the divergence between these two sources ranges between three and six years, and is in the expected direction, with the Fathers samples being slightly older.

The comparison of both sets of education profiles, derived from different sources, shows greater congruence for Braeton and Denham Town than for Havendale and Mavis Bank. In the case of Havendale, there is agreement on the high representation of persons with tertiary education, although the census recorded a relatively larger proportion with secondary education. When distributions for Mavis Bank are compared, it is evident that the census identified a larger proportion with secondary education and lower shares with only primary schooling than among those fathers interviewed in the community survey. In each case, this may be explained by the fact that in these communities men embark on fatherhood at a later age. In the rural community this leads to a survey sample which has a lower education level than that for all men in the area. The opposite is true in the middle-income community of Havendale, where more advanced age is associated with greater educational attainment.

(b) Scale Development

This section provides an account of the development of the macho scale, which was designed to measure specific aspects of masculinity, namely attitudes to male identity and gender relations. A second scale, which is not presented here, measured attitudes related to male parenting (Anderson 2009). As indicated earlier, the design of the macho scale was based on the body of accumulated qualitative research on male gender attitudes and gender relations in the Caribbean, and it sought in particular to advance the findings conducted in the first Fathers Study (Brown et al. 1993) as well as the subsequent work on gender socialization.

It is important to make clear from the outset that the macho scale is being advanced as a measure of masculine identity and gender relations, but it should not be interpreted that it is a total measure of masculinity. There are other significant aspects of masculinity reported in the wider literature, such as the need for status, which this new scale has not attempted to capture. This has been referred to as "the Big Wheel" syndrome in the scale developed by Brannon in 1976 based on American samples (cited in Thompson and Pleck 1987). Three other clusters of norms comprise Brannon's scale, and these include the avoidance of femininity ("no sissy stuff"); cultivating independence and self-confidence (the "sturdy oak"); and the endorsement of aggressiveness ("give them hell"). The concerns which men may have to establish their status vis-a-vis other men are partly expressed in the Jamaican society through the macho attributes, but we acknowledge that these do not exhaust all of the bases on which status is accorded.

Other important work that should be acknowledged is the development of the Gender Equitable Men (GEM) scale by the Promundo organization in Brazil (Pulerwitz and Barker 2008). The GEM scale was constructed to measure gender-equitable norms and behaviors, and to use this as a basis to design intervention programmes among young men to reduce violence and sexual risk-taking. This has led to the innovative Programme H Alliance H ('H' referring to 'hombre' or 'homens'), which received significant support from the Population Council and from the UNFPA, and which has been credited with considerable success in effecting attitude change among young men. The items in the GEM scale relate to domestic activities, sexual relationships, sexual and reproductive health, homophobia and intimate partner violence.

There are both parallels and differences between the GEM scale and the macho scale. One of the most important differences is the purpose for which the Jamaican scale has been designed, namely the measurement of attitudes related both to gender relations and to fathering, as an aspect of male identity. In contrast, the GEM scale was constructed in order to establish a baseline for active intervention with young men to promote more gender equitable attitudes.

In regard to male identity and gender relations, the attitudes which were agreed to be of special interest related to the following aspects:

* male dominance and authority

* physical dominance of women

* virility and sexual freedom

* the right to have children outside of a current union

* restrictive emotionality and toughness

* aggression

* homophobia

* gender-linked domestic roles/domestic freedom

On this basis, items were compiled to tap these attitudes, using the Likert-type format of Agree strongly/Agree/Uncertain/Disagree/ Disagree strongly. The procedure for scale development followed the steps outlined by Spector (1992), with principal components analysis being used for exploratory factor analysis. The process for developing and testing the scales entailed the drafting of an instrument with 42 items, distributed as follows:

Dominance: 14 items

Virility including homophobia: 8 items

Aggression: 6 items

Restrictive emotionality: 9 items

Need to have children: 5 items

The complete list of items, indicating those that were successively eliminated, is shown in Appendix A. The questionnaire with the draft scale was first administered on a non-random basis to 146 persons from two different social classes, distinguished by area of residence. On the basis of the pilot test, it became apparent that any existing tendencies towards aggression were not readily tapped through this approach. In general, respondents declared themselves to be models of good behaviour and reasonableness, even where an item was linked to the defense of their children's rights within a school context. The only item designed to tap aggression which received much support was the statement that one should object strongly to being bossed around. Six of these aggression items were accordingly omitted after the pre-test, leaving only the statement that a father needed to toughen his son so that he could deal with life.

In the same vein, most of the questions designed to probe restrictive emotionality did not seem useful to distinguish attitudes as the large majority of respondents agreed on the importance of expressing love, and discussing problems. Seven of the items originally included in the pre-test were therefore omitted from the final questionnaire. The 27 items which were incorporated into the final questionnaire are shown in the appendix. The exploratory factor analysis conducted on this set of items led to the deletion of nine items, based on their significance levels in the inter-item correlation matrix. The remaining 18 items were again subjected to factor analysis, using principal components analysis, and the results are shown in Appendix B. This solution included three components which were not theoretically consistent, and accounted for 41.7 percent of the variance. This was rejected in favor of the more consistent 13-item scale which contained two clearly distinguished components, using orthogonal rotation, and which accounted for 41.4 percent of the variance. The two components are labeled:

* sexual dominance and virility

* the primordial need to beget children. (2)

The first component included eight items and accounted for 22.5 percent of the variance, while the second component included five items and contributed 18.9 percent. Both factors were selected based on eigenvalues greater than 1, and both were clearly indicated based on the graphical scree test. The final factor loadings for the 13-item macho scale are shown in Table 3.

Given the very low levels of non-response to the attitude statements [usually below 2 percent] the decision was taken to impute the mean for any missing cases. In each case, the mean for the specific community was used so as not to dilute the community variations.

As reported, aggressiveness and restrictive emotionality did not prove to be part of this complex of attitudes, at least to the extent that these are tapped through self-descriptions. Similarly, homophobia was not tied into these two factors. The characteristics of this macho scale, its two sub-scales and the community patterns are discussed below.

(c) Data Analysis

In conducting the analysis, the macho scale was utilized in its continuous form; in order to provide a broad overview, it was also divided into categories. Accordingly, the scale was trichotomized and scores were labeled low, medium and high on this basis. Individuals in each community were graded in relation to the total sample. The Kruskal-Wallis statistic is used to assess associations between macho values and community of residence, age-group and education level. In assessing correlations between the two subscales, very similar results were obtained when using Spearman's rho as with Pearson's r. Beyond that, a comparative assessment is provided based on tabular analysis in order to demonstrate construct validity.

FINDINGS

(a) Macho Values

The individual items which comprise the macho scale serve to tap attitudes towards masculinity based on two related constructs: (i) dominance and virility as measured by the belief in dominance over females, and the right to sexual and domestic freedom (ii) the felt need to produce children. The responses to these statements which were obtained from the combined four-community sample are shown in Table 4.

An overview of the responses in this table points to the fact that the men in this survey tended to place very great importance on having their own biological children. This was indicated by the proportions who agreed that they would not feel like a man if they had no children (51.3 percent); who admitted that in such a situation they would feel jealous of other men who had children (63.5 percent); who were reluctant to marry a woman who could not bear children (51.1 percent); and who said that in such a situation, they would probably seek to get children outside of their union (60.6 percent). This was accompanied by the belief that a man could tell when a child was really his, with nearly two-thirds (65.4 percent) expressing support for this statement.

There was less open support for having multiple partners, with only 13.1 percent agreeing that men who had several partners (who were "girly-girly") received more respect as men. In the same vein, there was limited agreement that a man's "nature" (his sexual drive) was stronger than a woman's, and that this justified his having more than one sexual partner; only a quarter of the respondents agreed with this statement. However, women's rights in a relationship were not uniformly respected, as 22.2 percent said that a man had the right to physically discipline his partner if she stepped out of line, and 28.1 percent said that a woman did not have the right to refuse to have sex with her partner.

The obligations of the man towards maintaining his own male friendship network were accorded priority over domestic demands by roughly a quarter of the respondents as 26.8 percent agreed that a man was entitled to be able to buy drinks for his friends, even if the housekeeping money was short. This privileging of the male sphere, together with the belief that the man should not wash the clothes for the family (endorsed by 28.8 percent) provided evidence of a continuing pattern of segregated sex roles in some of these households.

When the 13 macho items are examined by community, it was evident that men in each community varied in the level of support for these statements. The items which dealt with child-bearing (items 2, 9, 10 and 11) received the strongest support in both Denham Town and in Mavis Bank, while the support for multiple partners (items 1 and 4) was greatest in Denham Town. The view that men should be able to physically discipline their partners was not widely endorsed, although Denham Town fathers were most likely to express approval for beating women. Furthermore, in both Denham Town and Mavis Bank, a third of all fathers agreed that women should not be able to refuse to have sex with their partners, in contrast to 15 percent in Havendale and 20 percent in Braeton.

Disapproval of men doing the household laundry was not consistent across communities, and it was mainly in Denham Town that this traditional definition of gender roles lingered. In this low-income community, 44.7 percent of fathers agreed that the man should not wash the clothes for the family, as compared with 14.2 percent in Havendale, 22.6 percent in Braeton and 24.7 percent in Mavis Bank. Similarly, it was mainly in Denham Town and Mavis Bank that there was significant support for the argument that the man should still be free to buy drinks for his friends even if the household money was not in place. This view was supported by roughly a third of the Denham Town respondents (37.5 percent) and by 31 percent of those in Mavis Bank. This may be compared with 15.5 percent of Havendale fathers and 16.9 percent of the Braeton sample who accorded this much priority to male bonding.

Given the community differentials which have emerged in regard to the levels of agreement with the individual statements espousing traditional gender attitudes, it may be expected that there will also be significant variation by community in regard to scores on the composite macho scale. The validity of the scale, as well as community variations, are evident in the analyses presented below.

The Macho Scale

The range for the macho scale is from 13 to 65, with higher values indicating a stronger attachment to these sets of attitudes. For the entire sample, the mean value was 35.99, with a standard deviation of 8.2. At the community level, the mean values were 29.86 for Havendale (S.D. = 7.13), 33.06 for Braeton (S.D. = 6.9), 40.65 for Denham Town (S.D. = 7.08) and 37.52 for Mavis Bank (S.D. = 7.18). The distribution approximated normality for the entire sample (skewness was .05) as well as for individual communities. This was .23 for Havendale, .06 for Braeton, .11 for Denham Town and .07 for Mavis Bank.

The macho scale was found to be internally consistent for the entire sample and for each of the four communities. Table 5 presents the internal consistency coefficient (Cronbach's alpha) and related analyses for the entire sample, and the matching statistics are presented for the four communities in Table 6. While the scale achieved an overall alpha of .82, the corresponding values were .79 for Havendale, .75 for Braeton, .72 for Denham Town and .77 for Mavis Bank.

Macho Subscales

The subscales which comprised the Macho Scale also proved to be reliable when examined separately. The Dominance and Virility subscale achieved an internal consistency coefficient of .75, with this value ranging from .66 to .73 in individual communities. The Primordial Need to have Children subscale achieved an alpha of .72, and the value ranged from .62 to .71 in the four communities. These analyses are shown in Tables 7 and 8, while the summary statistics for the communities are shown in Table 9. It should also be observed that the two subscales showed a clear correlation, with this standing at .53, based on Spearman's r (significant at .01). At the community level, this correlation ranged from .35 in Denham Town to .49 in Havendale.

Variations in Macho Levels

The basic argument which emerges from this study is that among men of Afro-Jamaican descent, there is a complex of attitudes which serves to define their view of themselves, weaving together a sense of entitlement to dominance, both sexually and domestically, with a strong need to beget children as expression of their masculinity. This complex is similar to some of the qualities which are generally labeled hyper-masculinity, and is distinct from the commitment to parenting, which has also been shown to be deeply rooted among Jamaican fathers. In this section, the opportunity is taken to examine whether there are consistent patterns of attachment to macho values which vary with social class (as reflected in community of residence), and with key socio-demographic variables. To the extent that machismo exists, the question must also be raised as to what are its impacts. This is explored by a brief look at some of its correlates, such as numbers of partners, intimate partner violence and homophobia.

(a) Community Patterns

The distribution of macho scores of fathers in Havendale and in Denham Town represent opposite ends of the continuum. When the scale is trichotomized this is readily apparent, as shown in Table 10. In Havendale, slightly more than three-fifths (64.9%) of all fathers were classified as low on the macho scale; while in Denham Town, 54.7 percent were classified as high. Braeton's profile showed important differences from that of Havendale, as a lower proportion (46.1 percent) was classified as low macho and 18 percent were located at the high end. Mavis Bank respondents were distributed more evenly over the range, with 25 percent being classified as low macho, 35.2 percent in the medium category, and 39.8 percent being high-macho. These differences were shown to be highly significant using the Kruskal-Wallis test, which resulted in a chi-square value of 293.0 (3, N = 1141) p < .000.

These community variations are of considerable theoretical interest, as the four communities were purposively selected to represent different social class levels. This is supported by the findings based on education, discussed below, which make clear that despite broad areas of agreements regarding masculine ideologies among men at different social class levels, education and community contexts do exert an impact on attachment to traditional definitions of masculinity.

(b) Socio-Demographic Factors

When the macho scale is examined in relation to two selected socio-demographic characteristics of the sample--age and education--significant associations were also observed. These are summarized in Table 11, which provides a comparison based on the grouped data.

Age

For the sample as a whole, age was correlated with macho levels, with older men showing reduced macho levels. This relationship was indicated by the Kruskal-Wallis statistic which achieved a value of 6(1140) = 58.2, p <.000, and it is evident from the proportions of men in the different age-groups who ranked low, medium and high on the macho scale (Table 11). When respondents were classified into these macho categories, it may be seen that while only 16.1 percent of the men below 25 years were classified as low macho, the corresponding proportion for men above age 49 was 46.4 percent. At the upper end of the range, a half of young men (under 25 years) expressed views that represented high macho values in comparison with 23.2 percent of older men. These age-related variations are worthy of note, and demonstrate that although increasing age, representing greater maturity, serves to temper these traditional attitudes, they do not disappear.

Education

Education, and particularly tertiary-level education, is the factor which appears to have the greatest effect in depressing levels of machismo, based on the relationships which were observed for the combined sample, as well as for individual communities. This may be inferred on the basis of the patterns shown in Table 11. The percentages of fathers who were ranked as high-macho were fairly similar at the level of primary education (49 percent) and at the level of low secondary (46.8 percent), but declined to 33.3 percent for those with higher secondary (grades 10 or higher). Among those who had attained a tertiary education, only 3.3 percent were classified as high-macho. At the other end of the distribution, where fathers were classified as low-macho, it may be seen that those with tertiary training were almost four times as likely as those with only primary schooling to be low in their macho attitudes. The strength of this relationship is evident from the value of the Kruskal-Wallis statistic which had a chi-square value of 3(1098) = 267, p < .000.

(c) The Correlates of Macho

While education, age and community of residence may serve to predict macho levels, adherence to this value system also has clear outcomes for gender-related attitudes and behaviour. This may be briefly illustrated by examining the relationship between macho values and the numbers of baby-mothers reported by males, as well as men's earlier involvement in domestic violence. The relationship of macho attitudes to homophobia is of special interest, given that machismo attitudes are often reported to be linked to a rigid heterosexual identification. These patterns serve to provide important construct validation for the scale, particularly as they are based on actual behaviour as well as attitudes which were not already tapped by the instrument.

Baby-Mothers

It has been long observed that for the large majority of Jamaicans, child-bearing does not always occur within the framework of legal marital unions, and children within a family are often the products of different unions (Roberts and Braithwaite 1961; Blake 1961; Smith 1973; Roberts 1975). Given this type of family structure, which is linked closely to social class, women may have had more than one baby-father, and men may have had more than one baby-mother. In the present sample of Jamaican fathers, it was found that men had an average of 1.7 baby-mothers. The mean for older men (35 years and older) ranged from 1.8 in Havendale to 2.6 in Denham Town. In order to assess whether this variation is linked to macho levels, Table 12 displays the relationship between macho categories and number of baby-mothers. The correlation between the two variables was shown to be a modest one, based on the value of Spearman's r which was .15 (p < .000). Among low-macho men, 12.7 percent reported having more than two baby-mothers, in contrast to 27.3 percent of high-macho men.

Violence towards Women

Given the importance of assessing patterns of physical abuse between partners, the questionnaire included questions about the sources of conflict between men and their female partners, and the typical responses of each. In general, the main reasons which men gave for their partners becoming angry related to infidelity or suspected infidelity. In contrast, the reasons which they gave for their own annoyance tended to hinge around their perceived neglect of household duties by partners, authority issues and interpersonal relations. This set of questions was followed by a specific question on violence. This question, although sensitive, was fairly neutral in that it was based on a longer time frame and did not focus only on their current partner.

Accordingly fathers were asked: "If you look back at your dealings with women, have there been times when you hit the woman?"

This question elicited responses which showed that physical violence was indeed part of the repertoire of responses which men employed in situations of conflict with women. In reflecting on their overall behavior, a half of all men acknowledged that in the past they had resorted to violence. As shown in Table 13, the proportions who said that they had hit the woman varied markedly by community of residence and by macho level.

The tendency to resort to violence in situations of conflict with women was evident in all communities, although it was greater among Denham Town fathers, and among men with higher macho scores. In Denham Town, more than seven out of every ten respondents (72.5 percent) acknowledged that they had hit a female partner at some point, in comparison with roughly 40 percent of respondents in each of the other three communities. Again it seems fair to infer that there are strong community influences which propel men into particular types of behavior. Men with high macho scores proved to be twice as likely to admit to intimate partner violence as those at the low-macho end of the scale. These proportions were 65 percent as compared with 34 percent.

Homophobia

Although items related to homophobia were not included in the scale, there was nonetheless an apparent relationship between adherence to macho values and the willingness to admit to homophobia. This was explored through the familiar scale item, also based on the agree/disagree format, and phrased as follows:

If I discovered that one of my friends was a homosexual, I would stop being friends with him.

The responses for the entire sample are as shown:
Strongly agree   Agree    Undecided    Disagree    Strongly
                                                   disagree

29.7%            34.0%    6.9%         26.1%       3.3%


In this case, close to two-thirds (63.7 percent) of the combined sample either agreed or strongly agreed that they would discontinue a friendship if they discovered the male friend to be a homosexual. These attitudes varied with both the community of residence and the macho level, as shown in Table 14. Residents of Denham Town and Mavis Bank were considerably more likely to express this negative attitude when compared with the middle-income community of Havendale, although even in this community where education levels were significantly higher, close to half of all respondents endorsed this sentiment. Similarly, high-macho men were twice as likely as their low-macho counterparts to say they would sever the friendship. This may be observed in the comparative proportions who endorsed homophobic behaviour, which was 40.5 percent for low-macho men, and 85.4 percent for high-macho males.

SUMMARY

In overview, this attempt to explore male gender attitudes across social classes by the use of a scale to measure traditional macho attitudes has produced findings that are both unequivocal and realistic. The observed variation by community, representing different social classes, was expected, given the accumulated body of Caribbean research, as well as the data available from popular observation. When these fairly negative attitudes are set against the other findings from this study, which documented a deep commitment among men to parenting roles (Anderson 2009), it is possible to understand why debate continues to hinge around "what men really believe". The use of the macho scale has demonstrated that as men mature, they abandon some of the more stereotypical attitudes associated with hyper-masculinity. This transition occurs within all social classes. However, community influences are also important, so that where there are high-fertility norms within a community, the number of children fathered and the number considered desirable will remain above replacement levels.

It has also been seen that where men subscribe strongly to macho values, the fathering urge becomes expressed in an exaggerated manner, so that regardless of community of residence, high-macho men accumulate more baby-mothers and inevitably, more children. The continuing involvement in multiple unions and the support for male dominance create a context within which gender conflict flourishes, and interpersonal violence is not uncommon. The implications of these findings are discussed in the following section.

There were also some clear limitations which the study encountered. The most striking is undoubtedly the lack of available data on men's fertility patterns, since at no point is it possible to obtain nationally representative data on the number of children that men have fathered. This is the direct result both of the traditional bias of demographic studies towards examining the fertility of women in the child-bearing span, as well as the fact that the household is the basis for analysis in census counts. The Afro-Jamaican family therefore presents considerable challenges for sociologists and demographers, and nowhere is this more apparent than when the spotlight shifts to men and their fathering of children, a process which may distribute offspring among several households. Although the present study sought to compile random samples of fathers in different communities, this goal could not be satisfactorily achieved given the high spatial mobility of many fathers as they pursued occupations which often entailed spending long periods away from their homes.

Nonetheless, the study succeeded in obtaining community samples which have demonstrated considerable representativeness. Other methodological limitations relate to the unsuitability of many traditional measures to capture attitudes reflective of aggression. It seems clear that socially desirable responding (SDR) does operate in this regard so that endorsement of aggressive behavior could not be adequately assessed among the community samples. In light of the findings from this study that age, education and community of residence are important in shaping commitment to machismo, further work will seek both to assess the relative weight of these factors in these samples through multivariate analysis, as well as to explore whether the macho scale has wider applicability to other populations.

DISCUSSION

There are two pillars on which this study of masculinity has been constructed. One is the extensive body of research on the Afro-Caribbean family, and the other is the complex of findings from gender studies, and in particular, the newer field of masculinity studies. More recent studies of masculinities seek to show that there is not any monolithic construct of manhood, and that there are divisions of power and status which limit choices for men (Thompson and Pleck 1995). By examining the perspectives of Jamaican fathers across a range of social classes, these variations in self-definition have been highlighted; although in this paper the explicit focus has been on masculine ideology, gender relations and fathering. These issues were explored substantively through interviews with fathers, and they have been assessed quantitatively through the design of a scale to measure masculinity/macho attitudes. Fathering/parenting commitment has been measured separately, as reported elsewhere. Through this approach, the wider study has sought to disentangle men's attitudes towards parenting from their attitudes regarding masculinity and gender relations.

The study established that despite extensive commitment to the parenting role, there remains a coherent set of attitudes which define masculinity in terms of dominance, virility and domestic freedom, and which bind men together to a greater or lesser degree, but which vary considerably with social class. This normative system is a driving force for child-bearing and male-female interaction, although at the same time it creates behavioral patterns which are often in direct conflict with men's family-building goals. Community influences have also been shown to play a critical role in the exaggeration of definitions of masculinity along the macho trajectory, and in many inner-city communities, popular culture often serves to reinforce these negative attitudes, and contributes to expressed homophobia. On the other hand, religious participation has been shown to have a dampening effect on machismo (Anderson 2009).

In recent years, the debate on the construction of masculinity and the situation of men in the Jamaican society has moved beyond academic research into the arena of national policy. These questions have rapidly gained currency since the apparent links between male socialization and issues of crime and violence, academic underachievement and sexual risk-taking appear to be generating societal problems of huge significance (Harriott 2003; Gray 2004; Figueroa 2004; Le Franc et al. 2008). This concern has been reflected in the rush towards solutions that address parenting education, conflict resolution and sexual and reproductive health, with the basic premise and hope that behavior can be changed, if knowledge is increased and if attitudes can be altered. Analysis of the underlying social and economic structures which create differential opportunities for different groups in the society, and specifically for different groups of men, has generally not been advanced. However, these structural factors shape each community context, and are strongly aligned with adherence to machismo. The problem remains.

Appendix A

ITEMS TESTED FOR THE MACHO SCALE

Items retained in the Final Scale [13 items]

A man's nature is stronger than a woman's so it is okay for him to have more than one woman.

Even if a man is living with his partner, it is okay for him to have outside children if he looks after them.

A man has the right to physically discipline his partner if she steps out of line.

If a man has a lot of girlfriends, he is seen as more of a man than if he sticks with one woman.

A man should never let a woman know that he really loves her.

Even if the housekeeping money is short, a man is allowed to buy a drink for his friends.

A woman does not have the right to refuse to have sex with her partner.

Even if a man helps the woman with the housework, he should not wash the clothes for the family.

If I did not have children, I would feel jealous of other men who have. I would not feel like a man if I did not have children.

I would never marry a woman who could not have children.

If my partner could not have children, I would seek to get children elsewhere.

A man can tell when a child is really his.

Items subject to factor analysis but deleted from scale in favour of more consistent factors [5 items]

There would be less problems in relationships if the woman would listen to the man.

A father has to toughen up his son so that the son can deal with life. You can always look at a man and tell if he is a homosexual.

If I discovered that one of my friends was a homosexual, I would stop being friends with him.

The responsibility for birth control really belongs to the woman.

Item retained in questionnaire but deleted on basis of significance levels for Item-Correlation matrix for preliminary factor analysis [9 items]

In my view, the man should be the head of the household.

A father's most important responsibility is to maintain his children.

A man should still spend time regularly with his friends even if his family wants him to stay at home.

A woman has the right to know how much money her partner makes. A man should feel like a king in his home.

A man should not have to tell his partner everywhere that he is going.

Women usually prefer men who can spend money on them, even if they know the man will not be faithful to them.

The worst thing that a woman can do to her partner is to go with another man (give him 'bun').

A man should make the important decisions for his family.

Note: Answer choices for items were on a five-point Likert-type scale and included: Strongly agree, Agree, Uncertain, Disagree, Strongly disagree and No opinion

Items deleted after the Pilot Test [15 items]

It is easier for women to succeed in Jamaican society than it is for men.

As long as a man looks after his family, it is okay for him to have another woman on the outside.

It is important for a father to openly express love towards his children.

It is important for a father to let his children know that their mother is an important and a special person.

If a man is 'dissed' openly, he should be prepared to fight.

A father should be able to listen to his children, and advise them as well as the mother does.

Even if a man is in the wrong, it is better not to apologize as people may think he is weak

If a man has to choose between giving his last dollar to his baby-mother or to his own mother, it is his mother who should get it.

If I thought that my child was not being treated fairly at school, I would go to the school and make a big fuss and go on bad.

If a man is having problems, it is better for him to discuss them with his friends than to tell his partner.

If someone is being disrespectful towards you, it is better to ignore them than to become aggressive with them.

Children should be encouraged to express their opinions.

If a woman earns more money than her partner, she is really the head of the household.

I think that it is better for a man to keep his problems to himself.

If someone tries to boss you around, you should object strongly.

Note: Answer choices for items were on a five-point Likert-type scale and included: Strongly agree, Agree, Uncertain, Disagree, Strongly disagree and No opinion

Appendix B
Table B1
Factor loadings for Items contained in the 18-item Macho Scale
(N =1141)

                                                    Component

Items                                         1        2        3

A man's nature is stronger than a
woman's so it is okay for him to have
more than one woman                         .715     .205     .135

It is okay for a man to have outside
children if he looks after them             .697     .217     .084

If a man has a lot of girlfriends, he is
seen as more of a man than if he sticks
with one woman                              .630     .057     .016

A man should never let a woman
know that he really loves her               .515     .265     .249

A man has the right to physically
discipline his partner if she steps
out of line                                 .502     .122     .380

Even if the housekeeping money is
short, a man is allowed to buy a drink
for his friends                             .430     -.027    .331

Even if a man helps the woman with
the housework, he should not wash the
clothes for the family                      .408     .123     .188

If I did not have children, I would feel
jealous of other men who have               .098     .743     .003

I would not feel like a man if I did
not have children                           .239     .680     .190

If my partner could not have children,

I would seek to get children elsewhere      .363     .631     .070

I would never marry a woman who
could not have children                     .136     .577     .273

A father has to toughen up his son so
that the son can deal with life             .011     .511     .258

You can always look at a man and tell
if he is a homosexual                       .127     .042     .630

A man can tell when a child is              .066     .278     .628
really his

If I discovered that one of my friends
was a homosexual, I would stop being
friends with him                            .163     .240     .489

A woman does not have the right to
refuse to have sex with her partner         .356     .015     .481

There would be less problems in
relationships if the woman would
listen to the man                           .095     .304     .479

The responsibility for birth control
really belongs to the woman                 .369     .140     .447


REFERENCES

Alexander, Jack. 1977. "The Role of the Male in the Middle Class Jamaican Family". Journal of Comparative Family Studies 8 (3): 369-389.

Anderson, Patricia. 1986. "Conclusion: Women in the Caribbean Project". Social and Economic Studies 35(2): 291-322.

--. 2009. "The Changing Roles of Fathers in Jamaican Family Life." The Planning Institute of Jamaica Working Paper No.10. Kingston: PIOJ.

Anderson, Patricia, Arlene Bailey, Julian Devonish and Roy Russell. 2003. "Report on the 2003 NHT/UWI Survey of Denham Town and Midtown." Mona: University of the West Indies, Department of Sociology, Psychology and Social Work.

Bailey, Wilma, Clement Branche, Gail McGarrity and Sheila Stuart. 1998. Family and the Quality of Gender Relations. Kingston: Institute of Social and Economic Research.

Bailey, Wilma, Hugh Wynter, Jean Jackson, Amy Lee and Pansy Hamilton. 1999. "Male Attitudes to Relationships and Pregnancies in Jamaica." Mona: Research in Fertility Unit, University of the West Indies.

Barriteau, Eudine. 1998. "Theorizing Gender Systems and the Project of Modernity in the Twentieth-Century Caribbean." Feminist Review 59 (Summer): 186-210.

Barrow, Christine. 1986. "Male Images of Women in Barbados." Social and Economic Studies 35 (3): 51-64.

--. ed. 2002. "Caribbean Masculinity and Family: Revisiting 'Marginality' and 'Reputation'." In Caribbean Portraits. Essays on Gender Ideologies and Identities. Kingston: Ian Randle Publishers.

Blake, Judith. 1961. Family Structure in Jamaica. New York: Free Press of Glencoe Inc.

Branche, Clement. 1998. "Boys in Conflict." In Gender and the Family in the Caribbean, edited by Wilma Bailey. Kingston: ISER.

--. 2001. "Ambivalence, Sexuality and Violence in the Construction of Caribbean Masculinity." In Children's Rights, Caribbean Realities, edited by Christine Barrow. Kingston: Ian Randle Publishers.

Brown, Janet, Patricia Anderson and Barry Chevannes. 1993. "Report on the Contribution of Caribbean Men to the Family." Submitted to the IDRC. Kingston: Caribbean Child Development Centre.

Brown, Janet and Barry Chevannes. 1998. Why Man Stay So. Mona: University of the West Indies.

Chevannes, Barry. 1992. "The Formation of Garrison Communities." Paper presented at a symposium in honor of Carl Stone, University of the West Indies, 16-17 November 1992.

--. 2001. Learning to be a Man. Culture, Socialization and Gender Identity in Five Caribbean Countries. Kingston: UWI Press.

--. 2002. "Gender and Adult Sexuality." In Gendered Realities: Essays in Caribbean Feminist Thought, edited by Patricia Mohammed. Kingston: UWI Press.

Chevannes, Barry and Claudia Mitchell-Kernan. 1992. "How we were Grown: Cultural Aspects of High-risk Sexual Behaviour in Jamaica." Kingston: ISER.

Clarke, Edith. 1957. My Mother Who Fathered Me. London: George Allen and Unwin.

Cumper, George. 1958. "The Jamaican Family: Village and Estate." Social and Economic Studies 7 (1): 76-108.

Figueroa, Mark. 2004. "Male Privileging and Academic Underperformance in Jamaica." In Interrogating Caribbean Masculinities, edited by Rhoda Reddock. Kingston: UWI Press.

Frazier, E. Franklyn. 1939. The Negro Family in the United States* Chicago: University of Chicago Press*

Gayle, Herbert with A. Grant, P. Bryan, M. Yee Shui and C. Taylor. 2004. "The Adolescents of Urban St. Catherine: A Study of their Reproductive Health and Survivability." Spanish Town: Children First Agency.

Gray, Obika. 2004. Demeaned but Empowered: The Social Power of the Urban Poor in Jamaica. Kingston: UWI Press.

Harriott, Anthony, ed. 2003. Understanding Crime in Jamaica: New Challenges for Public Policy. Kingston: UWI Press.

Herskovits, Melville. 1941. The Myth of the Negro Past. Boston: Beacon Press.

Hope, Donna. 2006. Inna di Dancehall: Popular Culture and the Politics of Identity in Jamaica. Kingston: UWI Press.

Kilmartin, Christopher. 2000. The Masculine Self. Boston: McGraw Hill.

Le Franc, Elsie, Maureen Samms-Vaughan, Ian Hambleton, Kristin Fox and Dennis Brown. 2008. "Interpersonal Violence in Three Caribbean Countries: Barbados, Jamaica and Trinidad and Tobago." Pan American Journal of Public Health 26 (6): 409-20.

Lewis, Linden. 2003. The Culture of Gender and Sexuality in the Caribbean. Gainesville: University of Florida Press.

Marsiglio, William, Sally Hutchinson and Mark Cohan. 2001. "Young Men's Procreative Identity: Becoming Aware, Being Aware and Responsible." Journal of Marriage and the Family 63 (1): 123-135.

Massiah, Joycelin. 1986. "Postscript: The Utility of WICP Research in Social Policy Formation." Social and Economic Studies 35(3): 157-181.

McKenzie, Herman and Hermione McKenzie. 1971. Sociology and the Caribbean Family. Unpublished Paper. Mona: Department of Sociology, University of the West Indies.

National Family Planning Board. 1986. "Male Sexual Beliefs and Behaviour in Jamaica 1984." Unpublished report, National Family Planning Board.

Norman, Lisa and Chukwudum, Uche. 2003. "The Machismo Culture in Jamaica: Implications for HIV Prevention." Roundtable presentation at the Annual Sociological Association Meeting, Atlanta, Georgia.

Priestley, Sharon. 2010. "Intimate Partner Violence in Jamaica." Unpublished paper, Department of Sociology, Psychology and Social Work, University of the West Indies, Mona.

Pulerwitz, Julie and Gary Barker. 2008. "Measuring Attitudes Toward Gender Norms Among Young Males in Brazil: Development and Psychometric Evaluation of the GEM Scale." Men and Masculinities 10 (3): 322-38.

Reddock, Rhoda, ed. 2004. Interrogating Caribbean Masculinities: Theoretical and Empirical Analyses. Kingston: UWI Press.

Roberts, George. 1975. Fertility and Mating in Four West Indian Populations. Kingston: ISER.

Roberts, George and Lloyd Braithwaite. 1961. "A Gross Mating Table for a West Indian Population." Population Studies XIV (3).

Smith, Michael G. 1962. West Indian Family Structure. Seattle and London: University of Washington Press.

Smith, Raymond. 1956. The Negro Family in British Guiana. London: Routledge and Kegan Paul.

--. 1973. "The Matrifocal Family." In The Character of Kinship, edited by Jack Goody. Cambridge: Cambridge University Press.

Spector, Paul. 1992. Summated Rating Scale Construction: An Introduction. Newbury Park, CA: Sage Publications.

Stone, Carl. 1985. Class, State and Democracy in Jamaica. Kingston: Blackett Publishers.

Thompson, Edward and Joseph Pleck. 1987. "The Structure of Male Role Norms." In Changing Men: New Directions in Research on Men and Masculinity, edited by Michael Kimmel. Beverly Hills, CA: Sage Publications.

1995. "Masculinity Ideologies: A Review of Research Instrumentation on Men and Masculinities." In A New Psychology of Men, edited by R.F. Levant and W.S. Pollock. New York: Basic Books.

UNFPA. 2009. "Country Baseline Assessment of Gender-Based Violence." Report done in collaboration with the Institute for Gender and Development, University of the West Indies. Kingston: United Nations Population Fund.

(1) Assistance with access to census data was kindly provided by Statin.

(2) This component bears some similarities to the concept of Procreative Identity advanced by Marsiglio and colleagues (2001), but does not distinguish between procreative consciousness and procreative responsibility.

Patricia Anderson, This study was made possible through the support of the Campus Research Fellowship Programme, UWI, Mona, and through a grant from the Planning Institute of Jamaica, which provided support for fieldwork. Special appreciation is expressed to colleagues who provided technical advice at different stages of the data analysis. These include Garth Lipps, Colin Williams, Michelle Hindin and Chukwudum Uche. In addition, the comments of an anonymous reviewer were of immense value.
Table 1: Median Age of Men 25 to 59 years in the 2001 census,
and Median Age  of Survey Respondents by Community

Median age            Havendale     Braeton
of men 25-59 years

Census 2001              39            35

Survey Data *           42.5           41

Median age           Denham Town   Mavis Bank
of men 25-59 years

Census 2001              36            36

Survey Data *            39            40

* Also limited to respondents 25- 59 years but based on fathers

Table 2: Education level of Men 25 to 59 years in the 2001 census,
and among Survey Respondents by Community

Education             Havendale         Braeton
Level
                  Census   Survey   Census   Survey

Primary or less    4.9     10.8     13.9     16.3
Secondary         50.6     29.2     59.9     61.4
Tertiary          44.5     59.9     26.1     22.4

Total
  percent         100      100      100      100
  number          1522     220      589      246

Education           Denham Town       Mavis Bank
Level
                  Census   Survey   Census   Survey

Primary or less   13.6     19.7     17.8     44.0
Secondary         80.5     77.7     74.9     49.6
Tertiary           5.9      2.9      7.3      6.5

Total
  percent          100      100      100      100
  number           1321     310      452      232

Table 3: Final Factor loadings for Items contained in
the 13-item Macho Scale  (N = 1141)

                                                       Component

Scale Items                                           1        2

A man's nature is stronger than a woman's so it     0.692    0.230
is okay for him to have more than one woman.

It is okay for a man to have outside children if    0.665    0.214
he looks after them.

A man has the right to physically discipline his    0.602    0.223
partner if she steps out of line.

If a man has a lot of girlfriends, he is seen as    0.567    0.046
more of a man than if he sticks with one woman.

A man should never let a woman know that he         0.539    0.327
really loves her.

Even if the housekeeping money is short, a man      0.520    0.068
is allowed to buy a drink for his friends.

A woman does not have the right to refuse to        0.528    0.144
have sex with her partner.

Even if a man helps the woman with the housework,   0.458    0.143
he should not wash the clothes for the family.

If I did not have children, I would feel jealous    -0.013   0.780
of other men who have.

I would not feel like a man if I did not have       0.210    0.733
children.

I would never marry a woman who could not have      0.179    0.656
children.

If my partner could not have children, I would      0.305    0.621
seek to get children elsewhere.

A man can tell when a child is really his.          0.298    0.434

Note: Extraction method: principal components
analysis; rotation method: varimax which
converged in 3 iterations; items are listed in
order of their final factor loadings.

Table 4: Levels of Agreement and Disagreement with Macho Views in all
Communities [percent]

                                      All Communities

Items in                       Strongly
Macho Scale                     Agree     Agree    Uncertain

1. A man's nature is             3.2       21.7       3.6
stronger than a woman's

2. Okay to have                  4.8       28.8       4.1
outside children

3. A man has the right           3.3       18.9       1.9
to discipline his partner

4. Men with more                 1.6       11.5       2.7
girlfriends are rated highly

5. Should not tell a             5.6       34.1       2.9
woman he loves her

6. Man should be able            2.2       24.6       1.4
to buy drinks

7. Woman cannot                  4.0       24.1       3.2
refuse sex

8. Man should not                4.4       24.4       3.8
wash clothes

9. Would feel jealous of         13.0      50.5       4.1
other men who have kids

10. Would not feel               10.5      40.8       3.5
like a man without kids

11. Would never marry an         16.5      34.6      10.9
infertile woman

12. Would get outside            11.3      49.3      10.7
children

13. A man can tell               11.6      53.8       5.0
when a child is his

                                        All Communities

Items in                                   Strongly
Macho Scale                    Disagree    Disagree       Total

1. A man's nature is             57.0        14.5          100
stronger than a woman's                                (n = 1132)

2. Okay to have                  45.2        17.0          100
outside children                                       (n = 1134)

3. A man has the right           42.8        33.2          100
to discipline his partner                              (n = 1132)

4. Men with more                 66.7        17.5          100
girlfriends are rated highly                           (n = 1132)

5. Should not tell a             42.8        14.6          100
woman he loves her                                     (n = 1130)

6. Man should be able            50.4        21.5          100
to buy drinks                                          (n = 1132)

7. Woman cannot                  54.4        14.3          100
refuse sex                                              (n =1131)

8. Man should not                56.8        10.6          100
wash clothes                                            (n =1129)

9. Would feel jealous of         28.0         4.4          100
other men who have kids                                (n = 1121)

10. Would not feel               38.1         7.0          100
like a man without kids                                (n = 1130)

11. Would never marry an         32.9         5.1          100
infertile woman                                        (n = 1129)

12. Would get outside            23.5         5.2          100
children                                               (n = 1124)

13. A man can tell               26.2         3.5          100
when a child is his                                    (n = 1128)

Table 5: Macho Scale for Total Sample: Corrected Item-Total
Correlations, Alpha Coefficients if Items Deleted and Scale
Statistics

                                                    Corrected
                                                   Item-Total
                                                  Correlations
Scale Items                                       [[r.sub.i-t]]

A man's nature is stronger than a woman's             .566
so it is okay for him to have more than one
woman

Even if a man is living with his partner,             .528
it is okay for him to have outside children
if he looks after them

A man has the right to physically discipline          .499
his partner if she steps out of line

If a man has a lot of girlfriends, he is seen         .361
as more of a man than if he sticks with one
woman

A man should never let a woman know                   .515
that he really loves her

Even if the housekeeping money is short,              .351
a man is allowed to buy a drink for his
friends

A woman does not have the right to refuse             .398
to have sex with her partner

Even if a man helps the woman with the                .345
housework, he should not wash the
clothes for the family

If I did not have children, I would feel              .389
jealous of other men who have

I would not feel like a man if I did not              .533
have children

I would never marry a woman who                       .455
could not have children

If my partner could not have children,                .525
I would seek to get children elsewhere

A man can tell when a child is really his             .407

                                                      Alpha
                                                   Coefficient
                                                 if Item deleted
Scale Items                                         [[alpha]d]

A man's nature is stronger than a woman's              .795
so it is okay for him to have more than one
woman

Even if a man is living with his partner,              .797
it is okay for him to have outside children
if he looks after them

A man has the right to physically discipline           .799
his partner if she steps out of line

If a man has a lot of girlfriends, he is seen          .810
as more of a man than if he sticks with one
woman

A man should never let a woman know                    .798
that he really loves her

Even if the housekeeping money is short,               .811
a man is allowed to buy a drink for his
friends

A woman does not have the right to refuse              .808
to have sex with her partner

Even if a man helps the woman with the                 .812
housework, he should not wash the
clothes for the family

If I did not have children, I would feel               .808
jealous of other men who have

I would not feel like a man if I did not               .796
have children

I would never marry a woman who                        .803
could not have children

If my partner could not have children,                 .798
I would seek to get children elsewhere

A man can tell when a child is really his              .807

N = 1141

Scale alpha =.816

Mean = 35.99

S.D. = 8.20

Table 6: Macho Scale for Community Samples: Corrected Item-Total
Correlations  [[r.sub.i-t]], Alpha if Items Deleted
[[[alpha].sub.d]] and Scale Statistics

                                     Community

Macho Items               Havendale             Braeton

                      [r.sub   [[alpha]   [r.sub    [[alpha]
                      .i-t]    .sub.d]    .i-t]     .sub.d]

1.  Man's Nature       .493      .772      .528       .724
2.  Outside kids       .548      .765      .401       .736
3.  Discipline         .418      .777      .391       .738
    women
4.  Many               .317      .786      .307       .746
    girlfriends
5.  Don't tell         .487      .770      .459       .729
6.  Have drinks        .287      .788      .299       .747
7.  Cant refuse        .439      .775      .273       .750
    sex
8.  No clothes         .327      .785      .189       .758
    washing
9.  Feel Jealous       .321      .788      .415       .734
10. Not feel like      .549      .764      .422       .734
    a Man
11. Not marry          .483      .771      .350       .742
12. Get outside        .438      .775      .548       .720
    kids
13. Man can tell       .365      .783      .322       .745

N                           228                   267
Scale alpha                .791                  .754
Mean                       29.86                 33.06
S.D.                       7.13                  6.90

                                    Community

Macho Items              Denham Town           Mavis Bank

                      [r.sub   [[alpha]   [r.sub   [[alpha]
                      .i-t]    .sub.d]    .i-t]    .sub.d]

1.  Man's Nature       .482      .688      .536      .742
2.  Outside kids       .400      .698      .472      .748
3.  Discipline         .371      .702      .489      .746
    women
4.  Many               .324      .708      .188      .772
    girlfriends
5.  Don't tell         .396      .699      .343      .762
6.  Have drinks        .296      .712      .255      .770
7.  Cant refuse        .322      .708      .333      .762
    sex
8.  No clothes         .266      .716      .278      .767
    washing
9.  Feel Jealous       .166      .726      .430      .753
10. Not feel like      .402      .698      .478      .747
    a Man
11. Not marry          .323      .708      .473      .747
12. Get outside        .381      .702      .435      .752
    kids
13. Man can tell       .344      .706      .393      .756

N                           382                  264
Scale alpha                .722                 .771
Mean                       40.65                37.51
S.D.                       7.08                 7.18

Table 7: Subscale for Dominance and Virility for Total Sample:
Corrected Item-Total Correlations, Alpha Coefficients if Items
Deleted and Scale Statistics

Items for Dominance and Virility Scale            Corrected
                                                  Item-Total
                                                 Correlations
                                                [[r.sub.i-t]]

A man's nature is stronger than a
woman's so it is okay for him to have
more than one woman                                  .571

It is okay for a man to have outside
children if he looks after them                      .530

A man has the right to physically discipline
his partner if she steps out of line                 .495

If a man has a lot of girlfriends, he is
seen as more of a man than if he sticks
with one woman                                       .386

A man should never let a woman know
that he really loves her                             .482

Even if the housekeeping money is short,
a man is allowed to buy a drink for his
friends                                              .371

A woman does not have the right to
refuse to have sex with her partner                  .394
Even if a man helps the woman with the
housework, he should not wash the
clothes for the family                               .341

Items for Dominance and Virility Scale          Alpha Coefficient
                                                 if Item deleted
                                                [[[alpha].sub.d]]

A man's nature is stronger than a
woman's so it is okay for him to have
more than one woman                                    .701

It is okay for a man to have outside
children if he looks after them                        .707

A man has the right to physically discipline
his partner if she steps out of line                   .715

If a man has a lot of girlfriends, he is
seen as more of a man than if he sticks
with one woman                                         .735

A man should never let a woman know
that he really loves her                               .717

Even if the housekeeping money is short,
a man is allowed to buy a drink for his
friends                                                .738

A woman does not have the right to
refuse to have sex with her partner                    .734
Even if a man helps the woman with the
housework, he should not wash the
clothes for the family                                 .743

N = 1141

Scale alpha = .75

Skewness = .34

Mean = 19.44

S.D. = 5.38

Table 8: Subscale for Primordial Need to Have Children for
Total Sample: Corrected Item-Total Correlations, Alpha
Coefficients if Items Deleted and Scale Statistics

Items for Need to Have children                Corrected
                                               Item-Total
                                              Correlations
                                             [[r.sub.i-t]]

If I did not have children, I would feel
jealous of other men who have                     .476

I would not feel like a man if I did not
have children                                     .357

I would never marry a woman who
could not have children                           .486

If my partner could not have children, I
would seek to get children elsewhere              .549

A man can tell when a child is really his         .496

Items for Need to Have children              Alpha Coefficient
                                              if Item deleted
                                             [[[alpha].sub.d]]

If I did not have children, I would feel
jealous of other men who have                      .666

I would not feel like a man if I did not
have children                                      .710

I would never marry a woman who
could not have children                            .662

If my partner could not have children, I
would seek to get children elsewhere               .634

A man can tell when a child is really his          .658

N = 1141

Scale alpha = .715

Skewness = -.16

Mean = 16.56

S.D. = 3.95

Table 9: Summary Statistics for Subscales by Community

                                     Community

Scale Statistics   Havendale   Braeton    Denham Town   Mavis Bank

                           Dominance and Virility Subscale

N                     228        267          382          264
Cronbach's Alpha     .726        .656        .678          .669
Mean                 15.76      17.62        22.52         20.0
S.D.                 4.41        4.35        5.18          4.65

                         Subscale for Need to Have Children

N                     228        267          382          264
Cronbach's Alpha     .671        .658        .613          .714
Mean                 14.1       15.44        18.14        17.51
S.D.                 3.79        3.68        3.38          3.71

Table 10: Level of Commitment to Macho Values by Community

Macho                               Denham     Mavis      All
Category    Havendale    Braeton     Town      Bank     Fathers

Low           64.9%       46.1%      13.1%     25.0%     33.9%
Medium         26.8        36.0      32.2      35.2      32.7
High           8.3         18.0      54.7      39.8      33.4
Total
  Percent      100         100        100       100       100
  Number       228         267        382       264      1141

Table 11: Commitment to Macho Values by Age Group and Education level

Socio-                          Macho Level
Demographic
Characteristics      Low    Medium    High     Total

Age-Group
                                                100
Under 25 years      16.1%    33.3%   50.5%    [n = 93]
                                                100
25-29 years          25.9    30.6     43.5    [n =147]
                                                100
30-34 years          31.7    34.8     33.5   [n = 164]
                                                100
35-39 years          36.1    29.9     34.0   [n = 194]
                                                100
40-44 years          35.6    33.0     31.4   [n = 191]
                                                100
45-49 years          32.9    38.6     28.6   [n = 140]
                                                100
50 and over          46.4    30.3     23.2   [n = 211]

Education level

                                                100
Primary or less      19.9    30.9     49.2   [n = 236]

Secondary            19.1    34.0     46.8      100
Grades 7-9                                   [n = 188]

Secondary            27.4    39.3     33.3      100
Grades [10.sub.+]                            [n = 463]

                                                100
Tertiary             78.2    18.5     3.3    [n = 211]

Table 12: Number of Baby-Mothers reported by Fathers by
Commitment to Macho Values

Macho Category
Number of
Baby-Mothers      Low    Medium    High   All Fathers

One              63.6%    53.4%   43.8%      53.6%
Two               23.7    25.4     28.9      26.0
Three             8.8      9.9     16.2      11.6
Four or more      3.9     11.2     11.1       8.8
Total
  Percent         100      100     100        100
  Number          363      393     370       1126

Table 13: The Proportions of Men who Admit to Having Hit a Woman
by Community and Macho Level

Characteristics    Percent who Admit
                  to Hitting a woman

All Fathers              50.5%

Community
   Havendale             39.2
   Braeton               39.3
   Denham Town           72.5
   Mavis Bank            39.7

Macho level
   Low                   33.9
   Medium                52.8
   High                  65.1

Table 14: The Proportions of Men Who Agree That They Would
Discontinue a Friendship with a Homosexual by Community
and by Macho Level

Characteristics    Percent who Agree they
                  would stop being friends

All Fathers           63.7% [n = 1129]

Community
   Havendale           46.0 [n = 224]
   Braeton             53.0 [n = 264]
   Denham Town         73.2 [n = 380]
   Mavis Bank          75.9 [n = 261]

Macho level
   Low                 40.5 [n = 380]
   Medium              65.1 [n = 370]
   High                85.4 [n = 379]
COPYRIGHT 2012 University of the West Indies, Sir Arthur Lewis Institute of Social and Economic Studies
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Anderson, Patricia
Publication:Social and Economic Studies
Date:Mar 1, 2012
Words:14395
Previous Article:The Turks and Caicos Islands: why does the cloud still hang?/Islas Turcas y Caicos: por que Continua Nublado?/Pourquoi cette Persistante Epee de...
Next Article:The appropriateness of intermediate monetary targets adopted by Caribbean economies/Eficacia de las metas monetarias intermedias adoptadas por las...
Topics:

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |