Gender stereotyping in Bangladesh; the development of the Strength of Gender Stereotyping Scale (SGSS)

Introduction

In Western society the study of gender stereotypes is a very familiar topic emerging from broader stereotyping research during the rise of feminism (Mills et al., 2012). At the heart of gender stereotyping is a set of beliefs about what characteristics can be attributed to males and females (Martin & Dinella, 2001), and such orthodox, culturally prevalent beliefs develop early in the child’s life (Powell & Stewart, 1978). In tandem with the growth of descriptive research regarding gender stereotyping there has been an equal interest in how such gender stereotypes can be changed. For example, Social Role Theory (Eagly, 1987) suggests that the pigeonholing of the respective genders into certain social roles leads to stereotypical beliefs about gender, constituting a powerful perpetuating factor in gender stereotyping (Koenig & Eagly, 2014). However, it follows that the reverse of the process can hold true, that by changing social roles it is possible to change stereotypical beliefs about men and women (Eagly, Wood & Johannesen-Schmidt, 2004). Martin and Dinella (2001) state that ‘gender stereotypes are culturally defined expectations about the sexes’ but the tandem growth just described is a reminder that there are in fact two aspects of stereotyping to consider. First, there are the specifics about the content of stereotypes, what characteristics are attributed to what sexes, and second, there are general processes in stereotyping, such as the relationship between social role occupancy and the development and perpetuation of beliefs about male and female characteristics (Stewart, Powell & Tutton, 1979). Therefore, when it comes to cross-cultural considerations, the extent to which Western findings might pertain to a non-Western society, in this case an Asian society, there are some aspects of Western research that may be quite culture specific, such as the specific content of stereotypes, whereas other aspects of research may be highly generalisable, such as how stereotypes arise and how they change and might be modified. This paper concerns the development of a scale that measures the strength of belief in specific gender stereotypes, stereotypes that have arisen specifically in Bangladesh. From a broader perspective, once developed the scale will be available to help gauge the success or otherwise of initiatives aimed at encouraging positive change in stereotypes, initiatives that encourage social understanding and equality.

Why would or should there be initiatives aimed at encouraging positive change in stereotypes? Because stereotypes are potent determinants of behaviour and how we see not just others but how we see ourselves. Encouraging change in stereotypes in turn encourages better social understanding, tolerance, positive self-development, pro-social behaviour and, ultimately, equality. There are numerous examples of research from the West that identifies the detrimental effects of culturally determined gender stereotypes. Of particular potential importance to Bangladesh is the research on occupational and social role. Fogliati and Bussey (2013) in the USA found that exposure to negative stereotypes resulted in women’s worse performance in a mathematics assessment and decreased motivation to do anything about it. Ertl, Luttenberger and Paechter (2017) noted that in most European countries, the proportion of females pursuing a career in STEM (Science, Technology, Engineering, Mathematics) was still alarmingly low. In a study of German university students reading STEM subjects, Froehlich et al. (2022) found that ‘even though the (female) students participating in the study presumably had good grades in STEM (because they were reading a STEM subject), stereotypes still corrupted their self-concept’. Both studies relate to a process that has become known as ‘stereotype threat’ (Shapiro & Williams, 2012). These two studies are examples of how stereotypes can become internalised and shape performance, behaviour, and self concept with respect to education. The effects of stereotypes persist into the workplace. For example, Hoyt (2010) explores the gender gap in senior roles and ‘how stereotypes, prejudice, and discrimination contribute to women’s under-representation in elite leadership roles by both impacting perceptions of and responses to women as well as impacting the experiences of women themselves’.

Also of particular relevance to Bangladesh (and indeed to any country) is research on the impact of gender stereotypes on interpersonal relationships. In developing the Gender Role Attitudes Scale in young Spanish people, García-Cueto et al. (2015) note how treating the sexes unequally raises the probability of aggressive behaviour in adolescent dating relationships, how certain attitudes (stereotypes) about relationships in couples ‘are the precursor to acceptance of violent behavior’. There is a range of studies concluding that stereotypes shape the nature and perception of domestic violence (Hassouneh & Glass, 2008; Seelau & Seelau, 2005; Gerber, 1991).

The current content of Western gender stereotypes is to be found in a large-scale study of 628 U.S. participants who rated men and women on a set of 74 attributes taken from previous studies (Hentschel et al., 2019). Confirmatory factor analysis found that 15 items formed an ‘Agency’ scale, and 11 items formed a ‘Communality’ scale, both with a high Cronbach’s Alpha of 0.93. The authors note that in the past agency has been termed ‘masculinity’ and that communality has been termed ‘femininity’. Agency items were competent, effective, productive, task-oriented, leadership ability, achievement-oriented, skilled in business matters, dominant, bold, assertive, competitive, independent, desires responsibility, emotionally stable, and self-reliant. Communality items were understanding, kind, compassionate, sympathetic, communicative, collaborative, relationship-orientated, likeable, emotional, intuitive, and sentimental. The Agency scale itself comprised four dimensions (instrumental competence, leadership competence, assertiveness, and independence), and the Communality scale comprised three dimensions (concern for others, sociability, and emotional sensitivity). (We note that no overall Cronbach’s alpha is given for the overall 26 items, but that of the 12 correlations between the Agency and Communality dimensions, 11 were significant, 10 at the.001 level, indicating that all seven dimensions are polarised in the same direction, suggesting that all dimensions are seen in a positive light and as desirable qualities). The study found that stereotypes about communality persisted and were equally prevalent for male and female raters. Agency characterizations, on the other hand, were more complex in that male raters described women as being less agentic than men, but female raters differentiated among the agency dimensions, describing women as less assertive than men but as equally independent and leadership competent. The authors note how historically the persistence of traditional gender stereotypes had been fueled by skewed gender distribution into social roles, but that this pattern was now changing, with, for example, women now holding almost 40% of management positions in the United States (Bureau of Labor Statistics, 2017). This social change is in line with there now being weaker gender stereotyping of Agency. However, gender stereotyping of Communality remains very strong, the authors reporting that ‘All participants rated women higher than men on the three communality dimensions’. The authors conclude that ‘despite dramatic societal changes many aspects of traditional gender stereotypes endure’, and describe how this translates into self-characterizations, especially in women, who tended to characterize themselves in more stereotypic terms than did the men.

Turning specifically to Bangladesh, the research literature on gender stereotyping is sparse but interesting. Studies tend to be exploratory, and the findings indicative rather than definitive.

Historically Bangladesh has had an agricultural economy. As recently as 2006 most women, 80%, still lived in rural areas of Bangladesh (BBS, 2006). Parveen and Leonhäuser (2008) note how although they were the backbone of the rural economy they were ‘handicapped by entrenched gender hierarchies, religious discrimination, and a disproportionate allocation of resources’. Parveen and Leonhäuser set out ‘to identify the ongoing gender stereotypes existing in the rural community’. The participants were 159 farmers’ wives from three villages in Mymensingh District, and data was collected using surveys, interviews, and discussion groups. They found that women were discriminated against in a range of ways, not just in the division of labour but also ‘in access to education, food, property, freedom of mobility, and economic opportunity’. Stereotypes included, for example, ‘Wives must be obedient to husbands because a woman’s paradise is at the feet of her husband’, and ‘Domestic work is women’s obligatory work and responsibility’. The authors note how such attitudes and norms can foster violence, especially domestic violence, against women, and discuss what can be done to promote women’s economic empowerment, including vocational education in livelihood skills, the provision of resources such as micro-credit and the promotion of platforms to challenge traditional beliefs that perpetuate women’s subordination.

Since 2006 there has been an enormous, exponential growth in Bangladesh of manufacturing, particularly in the garment industry. According to Islam et al. (2018), as at 2016 there were more than 4,000 ready-made garment (RMG) organisations and more than 4.2 million workers, 85% of whom are female. Islam and colleagues questioned whether this has led to female progression in leadership positions; apparently not as only 5.4% of middle and senior management positions were occupied by females, and they sought explanations for this. They looked at one RMG organisation and in detail considered the accounts of barriers to career progression faced by 8 female employees who held Masters or Bachelor degrees and who had been with the organisation for 6-14 years. Barriers included the sheer lack of female managers, an organisation that did not allow for a work-home lifestyle, a hiring policy dominated by men, and leadership styles that were patriarchal as per custom. Of concern is that this sample actually expressed typical gender stereotypes themselves, that they seemed to have internalised to some extent the barriers to progression. Quotes include ‘Both power and ambition I believe are much more attractive and suitable to men’, ‘Men are more tolerance (sic) to take pressure from multiple stakeholders; therefore, they tend to take the leadership roles’ and ‘Both male and female feel most comfortable when it is male leadership due to the gender stereotypes in out [r] mind-sets’.

The issue arises as to whether the educational process in Bangladesh is challenging or perpetuating gender stereotypes, whether it does or does not encourage a change of mind-set in the upcoming generation. Islam and Asadullah (2018) undertook a comparative content analysis of Malaysian, Indonesian, Pakistani and Bangladeshi government approved secondary school textbooks to look for any gender stereotyping in them. It is noted that ‘In Bangladesh, while textbook contents have often been manipulated by successive governments for political purposes…. gender bias in the curriculum and learning materials remains an overlooked topic’. The overall finding, first of all, was that females were simply underrepresented; 62.7% of characters in Bangladeshi textbooks were male. Further, when females were depicted, female occupations were mostly traditional, low wage and less prestigious and the female characters were predominantly introverted and passive in terms of personality traits. Women were also shown to be mostly involved in domestic and in-door activities while men had a higher presence in professional roles. Females were presented in domestic roles four times more than their male counterparts. However, it was noted that when Bangladeshi textbooks depicted females in non-domestic roles, they depicted a wider range of professional roles and the professional roles depicted were more prestigious and demanding than in any of the other countries’ textbooks (such as a lawyer, social scientist or TV anchor). However, in short, ‘our analysis confirms that textbooks disseminate a hidden curriculum’.

The issue of gender bias in textbooks in Bangladesh is taken up by Asadullah, Islam and Wahhaj (2018). Despite opposition from conservative religious authorities, coeducational schools have mushroomed throughout the country and the government also implemented programs to encourage girls’ education at all levels, such that gender differences in enrollment at primary level disappeared nearly two decades ago and today an equal number of girls attend secondary schools. But the country’s remarkable achievement on women’s education has recently been undermined by some authorities ‘silently rewriting school textbooks in a way that can do harm to the country’s journey toward gender equality’. ‘These changes not only reflect certain religious biases, they also encourage stereotypes regarding what women should do in Bangladeshi society’. However, textbooks from 1990 were examined and all Bangladeshi school textbooks suffered from a pro-male bias regardless of whether they were based on a secular or religious curriculum. Gender bias was considered to be widespread in government recognized textbooks long before radical groups demanded reforms of the secular school curriculum. It was a general problem to address, and indeed a global problem to address, for example in the USA (Mccabe et al., 2011). Going forward it would be important to tackle structural deficits in the educational system to restore gender balance in textbook content.

Gender stereotyping regarding the lesser role of women is carried forward from school into the workplace. Islam and Akter (2018) considered gender stereotypes in Bangladeshi business firms, taking the Women as Managers Scale (WAMS) that was developed in the USA some years ago (Peters et al., 1974), translating it into Bangla, adapting some of the items (in unspecified ways), and administering it to 260 male and female employees from different firms. It is a 21-item scale, each item rated 1-7, with higher scores reflecting a favourable attitude towards women in managerial roles. It was found that ‘the practice of gender stereotypes in the Bangladeshi working environment still works as a vital factor’ and ‘female employees’ managerial roles are still plagued with stereotypical issues’. There were issues with overall acceptance of women as managers, issues of organisational barriers, and issues arising from supposed personality traits. However, relevant to this paper, the authors also expressed the view that measures developed in the West are not ideal for use in Bangladesh, stating that ‘the sociocultural contexts in Western countries for which WAMS is originally developed [was] found to be a misfit for a more traditional country like Bangladesh’ and ‘there is a considerable probability that some of negative attitudes and stereotypes present in the Bangladeshi culture might not be covered in WAMS’.

It is not just textbooks that show gender bias, but the media, too, though this is to some extent now being changed. Shafi (2021) illustrates this period of change by examining three advertisements that were currently on television. The adverts for Pears soap depicted a ‘fantasy world created around the little girl in which she will follow her mother to be another “beauty-icon” herself’. The authors point out that soap adverts in general hardly ever show women of dark complexion (one of the stereotypes evident in the Parveen and Leonhäuser study, above, is ‘A glowing bride must have a good physical appearance (especially a fair complexion)’). In contrast, the second type of advert was from BRAC (an NGO, Building Resources Across Communities) specifically designed to promote women’s empowerment. It highlights the fact that women own less than 4% of land in Bangladesh (a further stereotype evident in the Parveen and Leonhäuser study, above, was ‘Girls are not allowed to transfer land from biological families after marriage’ so they cannot inherit their parent’s land after marriage). The mother in this case had no right to inherit the land owned by her husband when he died, was left with no resources. She gallantly looks after her child on her own, earning their livelihood. The message is that ‘we can make women empowerment possible by restoring their rights on lands and all other assets’. The third advert depicts a change in roles available to women. It is from the Bangladesh Army, which shows ‘both men and women who are young, energetic and brave’. The advert shows how ‘21^st-century women are no longer vulnerable to the “stereotypical” gender norms’.

There are, then pressures towards change, but change is slow. Upoma (2021) analyzed gender stereotypes in the media, interested in how Bangladeshi teenagers, 50 male and 50 female adolescents aged 14 to 20 years, felt about women being portrayed. The media in Bangladesh, including movies, dramas, and commercials, perpetuated sexist stereotypes on multiple levels, reinforcing ‘old patriarchal gender notions’. The teenagers were conscious of how women were being portrayed, 85.7% saying the roles depicted were stereotypic, but only 54% thought that females were actually being stereotyped in real life and only 35% contemplated rejecting a product on the basis of what was being portrayed. Further, disappointingly, ‘one of the findings showed female students were more likely to support traditional stereotypes about women’, in the sense that the females were less likely to perceive the passivity in the roles being portrayed.

Turning to the present study, it is clear from the above review that here is a need for measures of gender stereotyping but that they have to be more than a translation of outdated Western scales. The objective of this study is to develop a psychometrically sound Bengali language scale that measures the strength of belief in culturally pervasive gender stereotypes, with a view to gauging where social change is needed and to assessing the effects of interventions intended to break down stereotypic prejudice.

Methods

In general terms the study involved the development of a draft scale followed by its administration to participants in order to derive a final scale via item analysis.

Results

The demographic data (as above) were analyzed using descriptive analyses.

Internal consistency was assessed by Cronbach’s alpha.

Test-retest reliability and criterion reliability were assessed by Pearson correlation.

Principal Axis Factoring (PCA) with Promax rotation method was used to perform exploratory factor analyses (EFA). A number of indices were considered for model fit in the confirmatory factor analysis (CFA). Chi-square (χ²), ratio of Chi-square to df (χ²/df), root mean square error of approximation (RMSEA), and comparative fit index (CFI) were used to assess the adequacy of model fit while the criteria for model fit were, χ² with p≥0.01, χ²/df≤2, RMSEA≤0.06, CFI≥0.95, SRMR≤0.08 (see Faruk et al., 2021).

SPSS 24 and AMOS 22 (Arbuckle, 2011) were used to analyze the data.

Discussion

The aim of the study to develop a strength of gender stereotyping scale for use in Bangladesh has been achieved. To the best of our knowledge, this is the first psychometric measure on gender stereotypes taking Bangladeshi culture into account. It is an 11-item scale, each item rated on a 4-point scale, hence a scale range of 11-44, with higher scores indicating stronger belief in stereotypes. The scale has acceptable internal consistency (Cronbach’s alpha of .87) and acceptable test-retest reliability (r=.94), and the factor structure meets accepted statistical requirements. It is a short scale, easy to administer, and in this regard, it has been found that in Bangladesh it can be useful to have psychometric tools with fewer items (Faruk et al., 2021).

Using statistical procedures similar to that employed by Hentschel, Heilman & Peus (2019), the scale has a two-factor structure, factors that we have termed Expression and Authority, but which are to some extent similar to the factors found by Hentschel et al. (2019) namely Communality and Agency. For example, Expression involves boys not crying and the expression of sexual needs while Communality also involves emotional aspects, and Authority involves not showing weakness and men controlling their wife while Agency also involves aspects of dominance. Therefore, there is something of a conceptual similarity between stereotypes in the U.S.A. and Bangladesh even if the way of wording them is different.

Turning to the content of gender stereotypes in Bangladesh, the most strongly held stereotype in the sample as a whole, by some margin, was that ‘Weakness is expressed when males behave like females’. Such a belief potentially denigrates both men and women. It suggests that some men are not strong enough to be a ‘real man’, not strong enough to show manly traits. It denigrates women by suggesting that a core feature of the female is weakness. In essence it devalues females in comparison to males and is liable to perpetuate inequality. It highlights an issue for society and education to address.

The item ‘Weakness is expressed when males behave like females’ loads on the Authority factor. It is perhaps disappointing to see that for both men and women this is the strongest held belief on that factor. The belief is not just a belief held by men; it is a belief held equally by men and women; there is no significant sex difference in the item mean. There are other related examples on the scale; females believe just as much as males that ‘A man should have control over his wife’, that ‘I don’t think a real man should have any feminine traits’ and that ‘A man’s decision is final in the house’. Indeed, the females believed even more strongly than the men that ‘A man’s decision is final in the house’. The ‘male is strength’ stereotype is not just a product of male thinking, it is part of female thinking, too; education has to address this in both men and women. However, there may be more potential for change in females rather than males, because females less strongly endorse the item ‘Boys need to be stronger even if they are young’, which perhaps represents the female questioning whether males inherently need to be ‘stronger’. But there is something of a mixed message here because although males may not need to be strong they must not show it by appearing feminine.

The question arises as to whether current educational programmes and exposure to media debates has had an impact on younger people, whether younger people less strongly hold stereotypic beliefs. The answer provided by this study is a firm ‘no’. There was no significant correlation with age for Expression or Authority or the Total scale. Indeed if anything the opposite is true, in that correlations with these factors while not significant were negative in direction, and the few significant correlations between item scores and age were also all in a negative direction., i.e. younger people more strongly held stereotypic beliefs.

In the future the scale will help assess the strength of gender stereotypes held by individuals and groups. The scale is easily understandable with no reverse items and is culturally appropriate. The scale can be used as an outcome indicator for interventions designed to encourage gender friendliness, including reducing incidents of gender-based violence and improving the recruitment or promotion policies of organizations. The scale is developed for the adult population, but its simplicity suggests that in time it may be possible to produce a child version to help understand and address the stereotyping issues evident in the educational system as set out in the introduction to this paper.

Conclusion

Gender stereotypes develop at a very young age and universally, in the West as well as in countries like Bangladesh, they have a wide ranging influence on many aspect of life including academic growth, the formation of personality traits, emotional expression, career progression and the shaping of behaviour in general (Ruble & Martin, 1998; Chaplin & Aldao, 2013; Hutson-Comeaux & Kelly, 2002; Plant et al., 2000; Bonebright et al., 1996). This study contributes to the goal of understanding and changing gender stereotypes in Bangladesh by providing the reliable, valid, and psychometrically sound measure that researchers and policy makers have sought (Mills et al., 2012). The Strength of Gender Stereotyping Scale provides a psychometrically sound, short in length, easy to administer, less time-consuming scale (it takes about four or five minutes to complete even in its oral form) that assesses the strengths of gender stereotypes in the adult population of Bangladesh. The results presented in this paper show that gender stereotypes in Bangladesh of the type that perpetuate inequality in favour of men are strong and above all pervasive, equally held by both males and females and by the young and old. It is a huge challenge for the educational system and for programmes aimed at encouraging social change.