Behaviour Research and Therapy 41 (2003) 795807 www.elsevier.

com/locate/brat

Masculinityfemininity as a national characteristic and its relationship with national agoraphobic fear levels: Fodors sex role hypothesis revitalized
W.A. Arrindell a,, Martin Eisemann b, Jo rg Richter c, Tian P.S. Oei d, Vicente s h, Lya Feldman h, E. Caballo e, Jan van der Ende f, Ezio Sanavio g, Nuri Bage rbara Torres i, Claudio Sica g, Saburo Iwawaki j, Chryse Hatzichristou k Ba Cultural Clinical Psychology Study Group1
a

Department of Clinical Psychology, University of Groningen, Heymans Institute, Grote Kruisstraat 2/I, 9712 TS Groningen, The Netherlands b Umea University, Umea , Sweden c Universita t Rostock, Rostock, Germany d The University of Queensland, Brisbane, Queensland, Australia e Universidad de Granada, Granada, Spain f Erasmus Universiteit, Rotterdam, The Netherlands g ` degli Studi di Padova, Padova, Italy Universita h Universidad Simo n Bol var, Caracas, Venezuela i Universidad del Pa s Vasco, San Sebastia n, Spain j Showa Womens University, Tokyo, Japan k University of Athens, Athens, Greece Accepted 30 November 2001

Abstract Hofstedes dimension of national culture termed MasculinityFemininity [Hofstede (1991). Cultures and organizations: software of the mind. London: McGraw-Hill] is proposed to be of relevance for understanding national-level differences in self-assessed agoraphobic fears. This prediction is based on the classical work of Fodor [Fodor (1974). In: V. Franks & V. Burtle (Eds.), Women in therapy: new psychotherapies
Corresponding author. Tel.: +31-50-3637607; fax: +31-50-3637602. E-mail address: w.arrindell@ppsw.rug.nl (W.A. Arrindell). 1 The authors listed represent the Cultural Clinical Psychology Study Group, which presently comprises the following members: s, R. Bentall, K.R. Bridges, A. Buchanan, V.E. Caballo, M.G. Calvo, G. Canalda, J. Castro, R. G. Aguilar, W.A. Arrindell, N. Bage Crozier, M. Davis, R.J. Edelmann, M. Eisemann, R.J. Farrer, L. Feldman, W. Frindte, A. Furnham, T. Ga rling, P. Gaszner, R. llai, E. Kasielke, Gillholm, M. Gustafsson, S.B. Hansson, P. Harris, C. Hatzichristou, B.L. Hudson, S. Iwawaki, M. Johnston, J. Ka J. Kenardy, C.C. Leong, A. Liddell, I. Montgomery, T.P.S. Oei, D.L. Palenzuela, D. Pennington, M. Peter, M.J. Pickersgill, L.A. n, E. Sanavio, C. Sica, M.A. Simo n, M. Surman, B. Torres, J. van der Ende, F. Zald var. Recinos, J.C. Richards, J. Richter, O. Ryde

0005-7967/03/$ - see front matter 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0005-7967(02)00188-2

796

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

for a changing society. New York: Brunner/Mazel]. A unique data set comprising 11 countries (total N = 5491 students) provided the opportunity of scrutinizing this issue. It was hypothesized and found that national Masculinity (the degree to which cultures delineate sex roles, with masculine or tough societies making clearer differentiations between the sexes than feminine or modest societies do) would correlate positively with national agoraphobic fear levels (as assessed with the Fear Survey ScheduleIII). Following the correction for sex and age differences across national samples, a signicant and large effect-sized national-level (ecological) r = +0.67 (P = 0.01) was found. A highly feminine society such as Sweden had the lowest, whereas the champion among the masculine societies, Japan, had the highest national Agoraphobic fear score. 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Sex roles; Agoraphobia; Cross-cultural assessment; National masculinityfemininity; Hofstede; Fodor

1. Introduction The essential feature of Agoraphobia is anxiety about being in places or situations from which escape might be difcult (or embarrassing) or in which help may not be available in the event of having panic-like symptoms (e.g. fear of having a sudden attack of dizziness or a sudden attack of diarrhea) or a panic attack. A panic attack is dened as a discrete period of intense fear or discomfort, in which a number of anxiety symptoms (e.g. palpitations, pounding heart, or accelerated heart rate, sweating, trembling or shaking, fear of dying, sensations of shortness of breath, chest pain or discomfort) developed abruptly and reached a peak within 10 min (American Psychiatric Association [APA], 1995, p. 405). The anxiety typically leads to a pervasive avoidance of a variety of situations that may include being alone outside the home or being alone at home; being in a crowd of people; travelling in an automobile, bus, or airplane; or being on a bridge or in an elevator (APA, 1995, p. 406). Some individuals are able to expose themselves to the feared situations but endure these with considerable dread; often an individual is better able to confront a feared situation when accompanied by a companion. In addition, individuals avoidance of situations may impair their abilities to travel to work or to carry out homemaking responsibilities (e.g. grocery shopping, taking children to the doctor) (APA, 1995, p. 406). The Diagnostic and statistical manual of mental disordersfourth edition (DSM-IV; APA, 1995, p. 406) distinguishes between Panic Disorder With Agoraphobia and Agoraphobia Without a History of Panic Disorder. Fodor (1974), among many others (e.g. Chambless & Goldstein, 1980; Chambless, 1982, 1989; Brehony, 1983; Wolfe, 1984), have advanced theoretical speculations for explaining the higher frequencies in adulthood of (agora)phobic conditions in females as compared to males by pointing to the dependence and superhelplessness that are inherent to the stereotypically feminine sex role. Thus, these authors have argued that as young girls grow to become women, traditional societal mores promote dependency, shaping their behaviour and aspirations by deemphasizing autonomy and mastery while fostering expectations of protection and guidance. By contrast, for boys socialization entails reinforcement for quite different behaviour: males are expected to be independent, confronting, and instrumental in their approach towards the world (cf. Woody & Chambless, 1989). In the classical view of Fodor (1974) in which, incidentally, high masculinity and high femininity are assumed to be mutually exclusive traits, phobic behaviour of any kind, albeit

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

797

agoraphobic fears in particular, is consistent with traditionally dened feminine roles (i.e. emotional, submissive, excitable, passive, house-oriented, not at all adventurous, avoidant of mastery experiences and of competition, displaying lack of assertiveness and a strong need for security and dependency) and at odds with the traditional male role (e.g. autonomous, achievement-oriented, initiative taking, assertive). Fodor (1974) notes that when the realistic stresses of adult life and marriage become overwhelmingly great, stereotypically emotional, passive, helpless women become anxious, wish to ee, dream of becoming more independent or of rescue or escape, and ultimately develop a phobia. In doing so they sink even further into an exaggerated version of the stereotypic feminine role, becoming dependent on those around them and avoiding autonomy, initiative or assertiveness. Fodor (1974) points out that this route is available for women since their socialization experiences have not prepared them to become mature adults. In reviewing the empirical evidence, Arrindell, Kolk, Pickersgill, and Hageman (1993) found virtually no support for Fodors suggestion of a positive association between feminine traits and (agora)phobic fears. Rather, in line with the body of research on sex roles and mental health in general (e.g. Davidson-Katz, 1991), they observed that lack of masculine traits coincided with high (agora)phobic fear. Arrindell et al. (1993, p. 136) emphasized that they do not interpret this outcome in conjunction with the null nding of no relationship between feminine traits and phobic fears as offering (partial) support for the Fodor (1974) hypothesis, as was done by others (Chambless & Mason, 1986, p. 234; Hafner & Minge, 1989, p. 711) on the basis of observed signicantly negative correlations between masculine traits and phobic behaviour. The argument advanced by Arrindell et al. (1993, p. 136) was that to do so would suggest that masculinity and femininity (measured as traits) are the opposite ends of one continuum, an assumption that has been challenged by empirical research (e.g. Heilbrun, 1976; Bem, 1981; Marsh, 1985). Importantly, however, Fodor (1974) and Chambless and Mason (1986) also clearly pointed out that a society that does not teach women to be instrumental, competent and assertive, rather than just house-oriented or nurturing and expressive, is one that breeds (agora)phobic women. Fodor (1974, pp. 141152) cites several sources (clinical observations, research ndings, reports in the media, childrens readers) from which she concludes that: [These sources] reect the beliefs of the larger society where different expectations exist for males and females and where from birth onwards a social sex role curriculum is prescribed (p. 141, italics added). Thus, Fodor (1974) actually also suggested that at the sociological or cultural level one should obtain a signicant correlation between social sex roles and (agora)phobic fears, a possibility which hitherto has not been empirically examined. Three arguments (outlined below) led the present authors to re-examine Fodors hypothesis from this national- or cultural-level approach, an angle which differs from the individual-level approach applied to date. Both approaches may yield entirely different results (e.g. Hofstede, 1980, p. 169). At the national level of analysis, one can gain an understanding of how broad (environmental, demographic, epidemiological and unalterable, constitutional) factors may inuence the health of large populations (e.g. Lynn, 1971; Lynn & Martin, 1995; Eysenck, 1977, 1982; Eysenck & Eysenck, 1983). Studies of this kind fall within the so-called holocultural tradition (e.g. Rohner, 1986, p. 38) where cultures (nations) are treated as units and culture (national) scores are correlated with one another.

798

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

The three arguments were as follows. First, as pointed out above, within-society (i.e. individual level) correlational studies have as yet not provided much support for Fodors hypothesis in relation to dimensional measures of phobic fears, including agoraphobic fears (Arrindell et al., 1993). Second, both females and males within the same society can and have actually been found to score high or low on feminine and masculine traits (e.g. Bem, 1981). This is related to the empirical observation previously noted that masculine and feminine traits are not located at the opposite ends of one dimensional continuum, but can be measured as separate dimensions (e.g. Bem, 1981; Marsh, 1985). Third, in spite of the fact that Fodor (1974) and others (e.g. Foa, Steketee, & Young, 1984) have referred to agoraphobia as a womens syndrome, males too have been described in the clinical literature as agoraphobic sufferers (e.g. Hafner, 1986, ch. 7). Taking a national-level approach, the present study examined whether societies that provide females and males with equal opportunities for attaining mastery over their lives as opposed to those that offer these possibilities to a lesser extent, and which also differ on other characteristics relevant to the Fodor (1974) and Chambless and Mason (1986) hypothesis, would report lower levels of agoraphobic fears. It will be demonstrated below that there are masculine and feminine societies (not to be confused with masculine and feminine traits as conceptualized by for example Bem, 1981) that provide the sex role imperatives to which Fodor (1974) referred. 1.1. Hofstedes National MasculinityFemininity dimension Hofstede (1991, p. 5) denes culture (or mental software) as the collective programming of the mind which distinguishes the members of one group or category of people from another; culture determines how common basic problems are dealt with worldwide. In the anthropological sense, Hofstede (1998, p. 5) sees culture as broad patterns of thinking, feeling, and acting that have important consequences for the functioning of societies, of groups within those societies, and of individual members of such groups. Hofstede (1980) collected empirical evidence of differences in culture among paper-and-pencil responses on work-related value items from large samples of employees of subsidiaries of the same multinational business corporation (IBM) in 40 countries (N 116,000). (The total number of countries was later expanded to 50 and three regions.) The national samples were very well matched by occupational status, biological sex and age. Taking the country (n = 40), not the individual respondent, as the unit of analysis, a country-level or ecological factor analysis revealed four factors which accounted for about 50% of the total variance. The factor of interest to the present study was labelled MasculinityFemininity. [For the sake of saving space only this dimension will be thoroughly further discussed (see Hofstede, 1980, 1983, 1991, 2001 for a detailed description of the other three factors).]2 In Hofstedes view, Masculinity and Femininity differ in the social roles that are associated
2 Briey, the other three dimensions are power distance, uncertainty avoidance, and individualismcollectivism. Hofstede (1983) denes power distance as the extent to which the members of a society accept that power in institutions and organizations is distributed unequally. Uncertainty avoidance has to do with the degree to which the members of a society feel uncomfortable with uncertainty and ambiguity, leading them to support beliefs promising certainty and to maintain institutions protecting conformity. Individualism stands for a preference for a loosely knit social framework in which individuals are supposed to take care of themselves and their immediate families only, as opposed to collectivism, which stands for a preference for a tightly knit social framework in which individuals can expect their relatives, clan, or other in-group to look after them, in exchange for unquestioning loyalty.

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

799

with the biological fact of the existence of the two sexes, and in particular in the social roles that are attributed to men. Masculinity and Femininity refer to the dominant sex role pattern in the vast majority of both traditional and modern societies: that of male assertiveness and female nurturance (Hofstede, 1980, p. 277). Cultures labelled as masculine strive for maximal distinction between how men and women are expected to behave and to full their lives: [masculine cultures] expect men to be assertive, ambitious and competitive, to strive for material success, and to respect whatever is big, strong, and fast. [Masculine cultures] expect women to serve and care for the non-material quality of life, for children and for the weak. Feminine cultures, on the other hand, dene relatively overlapping social roles for the sexes, in which, in particular, men need not be ambitious or competitive but may go for a different quality of life than material success; men may respect whatever is small, weak, and slow (Hofstede, 1986, p. 308; Hofstede, 2000). More specically, survey ndings and correlational data with country-level indices have shown that the general societal norms behind low Masculinity countries and high Masculinity countries are among others the following ones (Hofstede, 1980, 1991). For low Masculinity (i.e. high Femininity) countries: sex roles in society should be uid: both men and women are allowed to be tender and to be concerned with interpersonal relationships; in the family, both fathers and mothers deal with facts and feelings; both boys and girls are allowed to cry, but neither boys nor girls should ght (unisex and androgyny as ideals); dominant values in society are caring for others and preservation; differences in sex roles should not imply differences in power between the sexes; both father and mother are used as models by boys and girls; men need not be assertive but can also take caring roles; everyone is supposed to be modest; stronger position of the mother in the family; people and warm relationships are important; womens liberation means that men and women should take equal shares both at home and at work. For high Masculinity countries: sex roles in society should be clearly differentiated: women are supposed to be tender and to take care of interpersonal relationships; in the family, fathers deal with facts and mothers with feelings; girls cry, boys do not, boys should ght back when attacked, girls shouldnt ght (machismo [ostentatious manliness] as model); dominant values in society are material success and progress; men should dominate in all settings; father is used as model by boys, mother by girls; men should behave assertively, ambitious and tough; women should care for (nurture); weaker position of the mother in the family; money and material objects are important; womens liberation means that women will be admitted to positions hitherto only occupied by men.

800

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

In addition to these norms, Hofstede (1980, pp. 296297) has pointed to some consequences for society at large. For feminine societies: belief in equality of the sexes; less occupational segregation (e.g. male nurses); men and women can both be breadwinners; men and women follow the same types of higher education. By contrast, for masculine societies: belief in inequality of the sexes; some occupations are considered typically male occupations, whereas other occupations are considered typically female; men are breadwinners, women are cakemakers; men and women follow different types of higher-level education. 1.2. Inferences and hypothesis Three inferences can be drawn from these key differences between masculine and feminine societies. First, based on Arrindell (1998), compared to masculine countries, the more feminine ones would offer both sexes, especially women, greater opportunities for the fullment of multiple social roles (employment, marriage, parenthood) that are associated with good self-rated health status, low morbidity, little restricted activity, infrequent use of medical care, low drug use (for both women and men), and that also have additive positive health effects for both sexes (see Barnett & Baruch, 1987; Cleary, 1987; Barnett & Rivers, 1996). In fact, many subjective wellbeing theories propose (and the empirical literature supports) that fullment of needs, goals, and desires is related to happiness (e.g. Emmons, 1986). Second, power inequalities (husbands dominance) in the family and the marital relationship are more frequently associated with satisfaction for husbands than for wives, whereas the highest levels of satisfaction are found in egalitarian couples (Gray-Little & Burks, 1983), especially in wives (Steil, 1997, p. 39). Sex role inequalities between spouses too have been found to be associated with unhappiness (Antill, 1983). In view of the Hofstede (1980, 1991, 1998) ndings, one could then argue that sex role complementarities are more likely to be found in masculine than in feminine societies. By contrast, sex role similarities in terms of feminine traits are more likely to be found in feminine than in masculine countries. Antill (1983) demonstrated that couples in which both partners were high on feminine traits (androgynous and feminine) were far happier than couples in which one or both were low on this dimension (undifferentiated and masculine). This empirical fact is not inconsistent with the emerging trend of feminine traits to be signicantly positively (rather than negatively) associated with measures of positive affect (see Arrindell, Vergara et al., 1997). Third, if Fodors hypothesis were correct, masculine societies would more likely than feminine societies be able to produce (agora)phobic individuals as the masculine ones would more likely be inclined to de-emphasize assertiveness, independence and other masculine traits in females

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

801

and to de-emphasize feminine behaviour in males than feminine societies would be inclined to do. By contrast, feminine societies are more likely than masculine societies to emphasize especially feminine traits in both females and males, without at the same time emphasizing or de-emphasizing masculine traits in either of the sexes. Hence, masculine societies are more likely than feminine ones to produce psychological ill-health. Hence, it was hypothesized that on a dimensional measure of agoraphobic fears, masculine nations would score signicantly higher than feminine ones, i.e. that Hofstedes national Masculinity scores would correlate positively with national Agoraphobic fear scores. The theoretical framework guiding the present study sees culture as a pathogenic agent (contributor to) and also as a pathoplastic modulator (e.g. the perception of illness is intimately inuenced by culturally determined beliefs) of (sub)clinical symptoms (e.g. Hallowell, 1970; Dra n, Westermeyer, Foulks, & Ruiz, 1999; Matsumoto, 2000, ch. 9 and 10). With guns, 1990; Alarco the present investigation an attempt is made to provide a further contribution to discovering whether differences (or similarities) across nations on dimensional measures of mental health can be predicted from Hofstedes dimensions of national culture (Arrindell, 1998). The Cultural Clinical Psychology Study Group aims at formulating a functional relationship between cultural characteristics and dimensions of psychopathology. If such a relationship can be established, it should be possible to predict behavioural manifestations from cultural characteristics and to infer cultural characteristics from manifestations of normal or abnormal behaviour (Draguns, 1990). 2. Method 2.1. National samples and inhabitants Eleven countries participated in the present study. Their inhabitants were students (volunteers) from the following countries: Australia (N = 701; 253 males, 448 females), East Germany (N = 357; 74 males, 283 females), Great Britain (N = 249; 98 males, 151 females), Greece (N = 446; 169 males, 276 females), Guatemala (N = 469; 208 males, 261 females), Hungary (N = 401; 143 males, 258 females), Italy (N = 1057; 475 males, 581 females), Japan (N = 275; 115 males, 160 females), Spain (N = 689; 303 males, 386 females), Sweden (N = 385; 184 males, 200 females), and Venezuela (N = 462; 164 males, 298 females). The mean ages for the national samples were as follows. For Australia, 23 years (SD = 8 years, range: 1770); East Germany, 24 years (5 years, 1854); Great Britain, 22 years (6 years, 1758); Greece, 22 years (2 years, 1848); Guatemala, 21 years (4 years, 1737); Hungary, 22 years (2 years, 1835); Italy, 22 years (3 years, 1851); Japan, 21 years (2 years, 1832); Spain, 22 years (3 years, 18 42); Sweden, 25 years (5 years, 1868); and for Venezuela, 21 years (3 years, 1637). (Variable Ns and dfs are due to missing data.) Foreign born and visa students were excluded from the original national subsamples. Following their exclusion, total N across the 11 nations was 5491. The data were collected in the mid 1990s. To maximize diversity on a relatively large number of measured and unmeasured variables, students within each country were recruited from different universities scattered across the country. In addition, each national sample was heterogeneous in terms of major course of study or

802

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

the faculty in which the students were enrolled. Thus, the samples did not include only psychology students (for further details see Arrindell, Eisemann et al., 2003). Despite being an atypical subsample relative to the general population of their own countries, several workers have argued that university students are still very much the products of their respective cultures. Thus, Williams, Satterwhile, and Saiz (1998, p. 57) have argued that while university students may have been inuenced by educational experiences which have brought them into contact with other cultures, this would seem likely to produce a conservative error in viewing them as cultural representatives, i.e., such experiences would tend to reduce differences between national samples. By making these national student samples comparable in terms of major variables that can be expected to covary with self-reports of fear (sex and age), one would obtain a narrow but very well comparable set of national samples which are functionally equivalent in many respects: education, intelligence (which enables the comprehension of abstract tasks), marital and work status, intellectual orientation (which enables students to appreciate the value of psychological research and be willing to cooperate in it). 2.2. Measures 2.2.1. Hofstedes national Masculinity index Hofstede (1991, p. 83) computed a Masculinity index score for each of the 50 countries and three regions involved in the IBM study. Scores were obtained in a range from about zero (0) for the most feminine (Sweden) to about 100 for the most masculine country (Japan) by multiplying factor scores by 20 and adding 50. Hofstede (1980, 1991, 1998, 2001) report an array of correlations supporting the construct validity of the dimension of interest. National Masculinity scores were retrieved from Hofstede (2001) and are displayed in Table 1.
Table 1 National Agoraphobic fear scores (as assessed with the Fear Survey ScheduleIII) and Hofstedes national MasculinityFemininity index scores for 11 nations Agoraphobic fear observed/unadjusted Spain Great Britain Italy Guatemala Australia Hungary Greece East Germany Japan Venezuela Sweden 6.06 7.11 7.57 8.35 7.09 9.11 7.47 6.28 11.03 8.12 5.67 Agoraphobic fear adjusted Hofstedes national Masculinity for sex and age Femininity index 6.13 7.12 7.67 8.45 7.02 9.02 7.42 5.89 11.20 8.02 5.85 42 66 70 37 61 88 57 66 95 73 5

Higher scores on the Fear Survey ScheduleIII and Hofstedes MasculinityFemininity index point to higher scores on Agoraphobic fears and Masculinity.

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

803

2.2.2. Fear Survey Schedule (FSS)III Agoraphobic fear subscale The FSS-III Agoraphobic fear subscale was derived by Arrindell (1980) on the basis of a factor analysis of the responses of a large sample of non-institutionalized phobic club members (N = 703). This dimension has been shown to be replicable, and invariant across a limited number of Western national samples, the Netherlands, Australia, Canada and the UK (Arrindell, Emmelkamp, & van der Ende, 1984; Arrindell, Oei, Evans, & van der Ende, 1991; Arrindell et al., 1990). Data supporting the cross-national factorial validity of the FSS-III dimensions across the 11 national samples involved in the present study are given in Arrindell et al. (2003a). The internal consistency reliabilities (males and females pooled) pertaining to the FSS-III Agoraphobic fear subscale for the 11 national samples were as follows: 0.74 (Australia), 0.70 (East Germany and Spain), 0.76 (Great Britain and Greece), 0.79 (Guatemala), 0.82 (Hungary), 0.75 (Italy), 0.72 (Japan), 0.71 (Sweden), and 0.78 (Venezuela). These levels of internal consistency are considered to be satisfactory (e.g. Cicchetti, 1994). To ensure translation equivalence of the FSS-III, the original US version (Wolpe & Lang, 1964) was translated into each of the languages and independently back-translated in order to check for discrepancies between original and new versions so as to make adaptations, where necessary. All questionnaire sets were completed anonymously. 3. Results 3.1. Foregoing analyses The major demographic variables on which the national samples differed and which on the basis of previous ndings (Arrindell et al., 1987, p. 220) were expected to correlate with selfassessed fears were biological sex (overall chi - square = 97.19, df = 10, P 0.001) and, to a lesser extent, age (overall F = 25.46, df = 10 and 5367, P 0.001). By using the approach of statistical control, these variables were treated as covariates in an analysis of covariance and controlled for in order to obtain corrected mean national FSS-III Agoraphobic fear scores for each national sample (e.g. Tatsuoka, 1971). The contributions of sex, age and their joint effects in reducing error in the statistical design were each statistically signicant. For sex: F = 169.63, P 0.0001; for age: F = 5.99, P 0.025; for sex and age conjointly: F = 89.20, P 0.0001. Table 1 gives for each national sample the observed and corrected means on the dimensional measure of interest. Next, the observed and corrected national means were correlated with the Hofstede national Masculinity index using Pearsons productmoment correlation coefcient. 3.2. Main ndings The observed national FSS-III Agoraphobic fear scores correlated 0.995 (P 0.0001) with their corrected counterparts. Japan had the highest national Agoraphobic fear scores (both uncorrected and corrected), whereas Sweden had the lowest scores (both uncorrected and corrected). The uncorrected national FSS-III Agoraphobic fear scores correlated +0.72 (P 0.01) with Hofstedes national Masculinity scores. As could be expected then, and in line with prediction, the

804

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

corrected national FSS-III Agoraphobic fear scores were also positively and signicantly correlated with Hofstedes national Masculinity scores: ecological r = +0.67 (P = 0.012, one-tailed test). In accordance with Cohen (1992), this association reected a large effect size.

4. General discussion On the basis of the large effect-sized, signicantly positive ecological correlation found in the present study between Hofstedes national Masculinity index and the national FSS-III Agoraphobia subscale scores, one could predict a pattern or trend reecting the countries with highest Masculinity scores to have the highest Agoraphobic fear scores and vice versa. To illustrate, Arrindell et al. (1987) compared 558 US, 498 British and 339 Dutch university students on the FSS-III subscale of interest. Arrindell et al. (1987, p. 223) reported Agoraphobic fear scores adjusted for the inuences of biological sex of 21.22 for US, 22.60 for British, and 18.88 for Dutch students. Individual-level analysis indicated that each of the between-country comparisons on the measure of relevance was statistically signicant (P 0.001). Indeed, paralleling this observation, Great Britain had the highest Masculinity score (66), followed by the US (62) and the Netherlands which clearly had the lowest Masculinity score (14). The point in time at which Arrindell et al. (1987, p. 236) suggested that the Hofstede dimensions might be useful in explaining national differences in general stress and (un)well-being, they did not have sufcient national data at their disposal in order to empirically support the above-mentioned contention. In relation to the more intensive agoraphobic fears in Great Britain observed by Arrindell et al. (1987), Draguns (1990, p. 265) notes that: [This outcome] curiously parallel[s] the distinguished contributions of British clinicians and researchers to the understanding and treatment of this disorder... The present ndings would however demonstrate that Hofstedes MasculinityFemininity dimension should not be overlooked as being a(nother) potential explanatory factor in relation to crossnational variations in agoraphobic (avoidance) behaviour. Indeed, as Draguns (1990, p. 252) suggested: the Hofstede dimensions, in spite of having originated in the domain of industrialorganizational psychology, are equally relevant to clinical psychologists. It has been demonstrated that the Hofstede dimensions are useful in explaining national differences in subjective well-being, which is an element of positive affect (Diener, Diener, & Diener, 1995; Arrindell, Hatzichristou et al., 1997). The present ndings thus expand this view in relation to agoraphobic fear, which is a subcomponent of the higher-order negative affect dimension. A forthcoming study extends these ndings even further in relation to other forms of phobic fears as well (Arrindell et al., 2003b). Further studies containing larger numbers of national samples than considered in the present investigation are needed to study how broad environmental, demographic, epidemiological and unalterable, constitutional factors independently of each other and in complex interactions with one another bring about national differences in phobic behaviour.

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

805

Acknowledgements The contribution of Dr W.A. Arrindell was made possible by a fellowship (19911996) of the Royal Netherlands Academy of Arts and Sciences (KNAW).

References
n, R. D., Westermeyer, J., Foulks, E. F., & Ruiz, P. (1999). Clinical relevance of contemporary cultural psyAlarco chiatry. Journal of Nervous and Mental Disease, 187, 465471. American Psychiatric Association (1995). Diagnostic and statistical manual of mental disorders, international version (4th ed.). Washington, DC: American Psychiatric Association. Antill, J. K. (1983). Sex role complementarity versus similarity in married couples. Journal of Personality and Social Psychology, 45, 145155. Arrindell, W. A. (1980). Dimensional structure and psychopathology correlates of the Fear Survey Schedule (FSS-III) in a phobic population: a factorial denition of agoraphobia. Behaviour Research and Therapy, 18, 229242. Arrindell, W. A. (1998). Femininity and subjective well-being. In G. Hofstede et al. (Eds.), Masculinity and femininity: the taboo dimension of national cultures. Thousand Oaks, CA: Sage. Arrindell, W.A., Eisemann, M., Richter, J., Oei, T.P.S., et al. Cultural Clinical Psychology Study Group (2003a). Phobic anxiety in 11 nations. Part 1: Dimensional constancy of the ve-factor model. Behaviour Research and Therapy, in press. Arrindell, W. A., Eisemann, M. et al., Cultural Clinical Psychology Study Group (2003b). Phobic anxiety in 11 nations: II. Hofstedes dimensions of national culture predict national-level variations (in preparation). Arrindell, W. A., Emmelkamp, P. M. G., & van der Ende, J. (1984). Phobic dimensions: I. Reliability and generalizability across samples, gender and nations. Advances in Behaviour Research and Therapy, 6, 207245. Arrindell, W. A., Hatzichristou, C., Wensink, J., Rosenberg, E., van Twillert, B., Stedema, J., & Meijer, D. (1997). Dimensions of national culture as predictors of cross-national differences in subjective well-being. Personality and Individual Differences, 23, 3753. Arrindell, W. A., Kolk, A. M., Pickersgill, M. J., & Hageman, W. J. J. M. (1993). Biological sex, sex role orientation, masculine sex role stress, dissimulation and self-reported fears. Advances in Behaviour Research and Therapy, 15, 103146. Arrindell, W. A., Oei, T. P. S., Evans, L., & van der Ende, J. (1991). Agoraphobic, animal, death-injury-illness and social stimuli clusters as major elements in a four-dimensional taxonomy of self-rated fears: rst-order level conrmatory evidence from an Australian sample of anxiety disorder patients. Advances in Behaviour Research and Therapy, 13, 227249. Arrindell, W. A., Pickersgill, M. J., Bridges, K. R., Kartsounis, L. D., Mervyn-Smith, J., van der Ende, J., & Sanderman, R. (1987). Self-reported fears of American, British and Dutch university students: a cross-national comparative study. Advances in Behaviour Research and Therapy, 9, 207245. Arrindell, W. A., Solyom, C., Ledwidge, B., van der Ende, J., Hageman, W. J. J. M., Solyom, L., & Zaitman, A. (1990). Cross-national validity of the ve-components model of self-assessed fears: Canadian psychiatric outpatients data vs. Dutch target ratings on the Fear Survey ScheduleIII. Advances in Behaviour Research and Therapy, 12, 101122. Arrindell, W. A., Vergara, A. I., Torres, B., Caballo, V. E., Sanderman, R., Calvo, M. G., van der Ende, J., Oosterhof, n, M. A. (1997). Gender roles in relation to assertiveness and L., Castro, J., Palenzuela, D. L., Zald var, F., & Simo Eysenckian personality dimensions: replication with a Spanish population sample. Sex Roles, 36, 7992. Barnett, R. C., & Baruch, G. K. (1987). Social roles, gender, and psychological distress. In R. C. Barnett, L. Biener, & G. K. Baruch (Eds.), Gender and stress. New York: The Free Press. Barnett, R. C., & Rivers, C. (1996). She works/he works: how two-income families are happier, healthier, and betteroff. New York: HarperCollins. Bem, S. L. (1981). Bem Sex-Role Inventory: professional manual. Palo Alto, CA: Consulting Psychologists Press.

806

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

Brehony, K. A. (1983). Women and agoraphobia: a case for the etiological signicance of the feminine sex-role stereotype. In V. Franks, & E. D. Rothblum (Eds.), The stereotyping of women: its effects on mental health. New York: Springer. Chambless, D. L. (1982). Characteristics of agoraphobics. In D. L. Chambless, & A. J. Goldstein (Eds.), Agoraphobia: multiple perspectives on theory and treatment. New York: Wiley. Chambless, D. L. (1989). Gender and phobia. In P. M. G. Emmelkamp, W. T. A. M. Everaerd, F. W. Kraaimaat, & M. J. M. von Son (Eds.), Annual series of European research in behavior therapy, vol. 4: fresh perspectives on anxiety disorders. Amsterdam/Lisse, The Netherlands: Swets & Zeitlinger. Chambless, D. L., & Goldstein, A. J. (1980). Anxieties: agoraphobia and hysteria. In A. M. Brodsky, & R. T. HareMustin (Eds.), Women and psychotherapy: an assessment of research and practice. New York: The Guilford Press. Chambless, D. L., & Mason, J. (1986). Sex, sex-role stereotyping and agoraphobia. Behaviour Research and Therapy, 24, 231235. Cicchetti, D. V. (1994). Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychological Assessment, 6, 284290. Cleary, P. D. (1987). Gender differences in stress-related disorders. In R. C. Barnett, L. Biener, & G. K. Baruch (Eds.), Gender and stress. New York: The Free Press. Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155159. Davidson-Katz, K. (1991). Gender roles and health. In C. R. Snyder, & D. R. Forsyth (Eds.), Handbook of social and clinical psychology: the health perspective. New York: Pergamon. Diener, E., Diener, M., & Diener, C. (1995). Factors predicting the subjective well-being of nations. Journal of Personality and Social Psychology, 69, 851864. Draguns, J. G. (1990). Normal and abnormal behavior in cross-cultural perspective: specifying the nature of their relationship. In J. J. Berman (Ed.), Nebraska symposium on motivation, 1989, vol. 37: cross-cultural perspectives. Lincoln: University of Nebraska Press. Emmons, R. A. (1986). Personal strivings: an approach to personality and subjective well-being. Journal of Personality and Social Psychology, 51, 10581068. Eysenck, H. J. (1977). National differences in personality as related to ABO blood group polymorphism. Psychological Reports, 41, 12571258. Eysenck, H. J. (1982). The biological basis of cross-cultural differences in personality: blood group antigens. Psychological Reports, 51, 531540. Eysenck, H. J., & Eysenck, S. B. G. (1983). Recent advances in the cross-cultural study of personality. In J. N. Butcher, & C. D. Spielberger (Eds.), Advances in personality assessment, vol. 2. Hillsdale, NJ: LEA. Foa, E. B., Steketee, G., & Young, M. C. (1984). Agoraphobia: phenomenological aspects, associated characteristics, and theoretical considerations. Clinical Psychology Review, 4, 431457. Fodor, I. G. (1974). The phobic syndrome in women: implications for treatment. In V. Franks, & V. Burtle (Eds.), Women in therapy: new psychotherapies for a changing society. New York: Brunner/Mazel. Gray-Little, B., & Burks, N. (1983). Power and satisfaction in marriage: a review and critique. Psychological Bulletin, 93, 513538. Hafner, R. J. (1986). Marriage and mental illness: a sex-roles perspective. New York: The Guilford Press. Hafner, R. J., & Minge, P. J. (1989). Sex role stereotyping in women with agoraphobia and their husbands. Sex Roles, 20, 705711. Hallowell, A. I. (1970). Fear and anxiety as cultural and individual variables in a primitive society. In I. Al-Issa, & W. Dennis (Eds.), Cross-cultural studies of behavior. New York: Holt, Rinehart and Winston. Journal of Social Psychology, 9, 2547 (1938). Heilbrun, A. B. Jr (1976). Measurement of masculine and feminine sex role identities as independent dimensions. Journal of Consulting and Clinical Psychology, 44, 183190. Hofstede, G. (1980). Cultures consequences: international differences in work-related values. Beverly Hills, CA: Sage. Hofstede, G. (1983). Dimensions of national cultures in fty countries and three regions. In J. B. Deregowski, S. Dziurawiec, & R. C. Annis (Eds.), Expiscations in cross-cultural psychology. Lisse, The Netherlands: Swets & Zeitlinger. Hofstede, G. (1986). Cultural differences in teaching and learning. International Journal of Intercultural Relations, 10, 301320.

W.A. Arrindell et al. / Behaviour Research and Therapy 41 (2003) 795807

807

Hofstede, G. (1991). Cultures and organizations: software of the mind. London: McGraw-Hill. Hofstede, G. (1998). Masculinity/femininity as a dimension of culture. In G. Hofstede et al. (Eds.), Masculinity and femininity: the taboo dimension of national cultures. Thousand Oaks, CA: Sage. Hofstede, G. (2000). Masculine and feminine cultures. In A. E. Kazdin (Ed.), Encyclopedia of psychology, vol. 5. Washington, DC: American Psychological Association. Hofstede, G. (2001). Cultures consequences: comparing values, behaviors, institutions, and organizations across nations (2nd ed.). Thousand Oaks, CA: Sage. Lynn, R. (1971). Personality and national character. Oxford: Pergamon. Lynn, R., & Martin, T. (1995). National differences for thirty-seven nations in extraversion, neuroticism, psychoticism and economic, demographic and other correlates. Personality and Individual Differences, 19, 403406. Marsh, H. W. (1985). The structure of masculinity/femininity: an application of conrmatory factor analysis to higherorder factor structures and factorial invariance. Multivariate Behavioral Research, 20, 427449. Matsumoto, D. (2000). Culture and psychology: people around the world (2nd ed.). Belmont, CA: Wadsworth. Rohner, R. P. (1986). The warmth dimension: foundations of parental acceptancerejection theory. Beverly Hills, CA: Sage. Steil, J. M. (1997). Marital equality: its relationship to well-being of husbands and wives. Thousand Oaks, CA: Sage. Tatsuoka, M. M. (1971). Multivariate analysis: techniques for educational and psychological research. New York: Wiley. Williams, J. E., Satterwhite, R. C., & Saiz, J. L. (1998). The importance of psychological traits: a cross-cultural study. New York: Plenum. Wolfe, B. E. (1984). Gender ideology and phobias in women. In C. S. Widom (Ed.), Sex roles and psychopathology. New York: Plenum. Wolpe, J., & Lang, P. J. (1964). A Fear Survey Schedule for use in behaviour therapy. Behaviour Research and Therapy, 2, 2730. Woody, S., & Chambless, D. (1989). Social and personality factors predicting sex differences in fearful behavior. Poster at EPA, April, Boston.