Attitudes of women toward intimate partner violence: a study of rural women in Nigeria

full article:

Introduction

The World Health Organization (WHO) defines intimate partner violence (IPV) against women as 'the range of sexually, psychologically, and physically coercive acts used against adult and adolescent women by current or former male partners'¹. Intimate partner violence is the third highest cause of death among people 15-44 years of age², and the most common form of violence against women. Its negative effects on women's health are serious enough to be recognized as a public health crisis with extensive effects on society^3-5. Lifetime worldwide prevalence of IPV has been suggested to be between 10 and 70% of women in marriage or current partnerships^6,7, and the lifetime prevalence of IPV in sub-Saharan Africa is reported at 20-71% in marriage or current partnerships^8,9. The prevalence is, however, suspected to be under-estimated due to under-reporting and a lack of standardized methodology².

Significant proportions of men and women in sub-Saharan Africa accept IPV as justifiable punishment for a woman's transgression of her normative roles in society^9-11, as well as for disobedience, adultery and disrespecting her husband's relatives¹². Empirical studies on IPV in rural areas in sub-Saharan African are scanty. This is regrettable, given that the majority of the African population resides in rural areas. However, available studies confirm high rates of IPV among rural women^9,10,13-16. Residency in rural areas has been associated with increased acceptance of IPV^7,10,14-19.

Relatively scanty empirical studies have been done on IPV among rural women in Nigeria; however, rates of violence against women in rural settings are believed to be difficult to estimate accurately for several reasons, including: (i) under-reporting; (ii) physical isolation associated with the rural milieu, which provides aggressors opportunities to engage in abusive behaviour; (iii) the patriarchal attitudes of rural law enforcement officers, which impede timely and effective responses to domestic violence reports; and (iv) acute difficulties encountered by rural women in using potentially supportive domestic violence services, if available.

In Africa, rural women have been reported to be conservative and are described as the bedrock of the socio-cultural values of traditional societies. These socio-cultural values define the gender norms of women and men (eg power, gender roles, responsibilities and obligations), and typically promote an imbalance of power between subordinate women and 'superior' men^16,20. This imbalance of power contributes to greater IPV among rural women in general^1,13.

Intimate partner violence among rural women in Nigeria

Although there is a paucity of studies on rural women's attitudes towards and experiences of violence, perceptions of male dominance over the subordinate female has been observed in Nigeria^21-23. Studies in Nigeria have shown that the deep-seated and rigid culture of patriarchy in rural communities makes reporting incidences of violence almost impossible, because doing so is viewed as causing indignity to the husband and being disrespectful of family members and elders whose roles include arbitrating in such matters¹⁵. An interest in the attitudes of rural women towards IPV arises in part from a need for information that might aid intervention programs necessary to reduce and prevent IPV in rural settings. The attitude of victims of violence is crucial to the success of violence intervention programs. For if the victim perceives IPV to be an integral part of 'male supremacy', culturally acceptable and a normal part of the marriage experience, she is unlikely to report such incidences of violence to appropriate health and law enforcement authorities, or to leave the marriage. Furthermore, researchers have found a direct relationship between positive attitudes toward violence against women and the actual occurrence of violence against women²⁴.

This study investigates the determinants of attitudes toward IPV among rural women in Nigeria. It offers insight into the social environment and norms surrounding domestic violence by specifically examining the relationship between attitudes towards IPV and demographic variables, socio-economic status (ie educational level, occupational status, and household wealth), and empowerment indicators (ie autonomy in household decisions, access to information, and literacy).

Methods

Study

Data from the 2003 Nigeria Demographic and Health Survey (NDHS) conducted between March and August 2003 in Nigeria were used for this study²⁵. The survey is a nationally representative, stratified, self-weighting probability sample collected in face-to-face interviews. The principal objective of the 2003 NDHS was to provide current and reliable data on fertility and family planning behaviour, child mortality, children's nutritional status, the utilization of maternal and child health services, and knowledge and attitudes towards HIV/AIDS. A related objective was to provide as many of these key indicators as possible for urban and rural areas separately, as well as for each of Nigeria's six geopolitical zones.

Sample design

The 2003 NDHS was carried out using the list of standard enumeration areas from the 1991 Nigerian Population Census frame for data sampling. Based on this list, a stratified two-stage sampling procedure was used to select the sample of women. Most of the Nigerian population lives in rural areas, hence the number of clusters allocated to the urban areas in five out of the six zones was increased in order to obtain reasonable urban estimates. Overall, 365 clusters were selected, of which 165 were in urban areas, and 200 were in rural areas. Following allocation of the number of households to each state by urban and rural areas, the number of clusters was calculated based on an average sample take of 20 completed women's interviews (in 19 selected households) in urban areas, and 25 completed interviews (in 24 selected households) in rural areas. In each urban or rural area in a given state, clusters were selected systematically with equal probability. These procedures resulted in a probability sample of 8250 households.

Subjects

All women aged 15-49 years resident or visiting in the sampled households at the time of the survey were eligible for inclusion into the survey. A total sample of women (n = 7620) were included in the survey, of which 3911 were rural women. The overall response rate for women was 94%. This study is restricted to the sub-sample of 3911 rural female respondents.

Questionnaire

A comprehensive questionnaire covering issues ranging from demographic, socio-economic to health issues, as well as child health and welfare, women empowerment and social status, and husband's status was used. For the purpose of the current study, respondents were questioned on the justification for IPV, as well as other demographic and social issues.

Ethical considerations

The survey procedure and instruments for the 2003 Nigeria DHS are ethically approved by the Ethics Committee of the Opinion Research Corporation (ORC) Macro International Inc, Calverton, USA, and by the National Ethics Committee in the Federal Ministry of Health, Nigeria. Informed consent was obtained from all participants prior to participation in the survey, and collection of information was confidential. This study is based on analysis of secondary data with all participant identifiers removed. Ethical permission for use of the data in the present study was obtained from ORC Macro Inc.

Specification and measurement of variables

Dependent variables: Rural women's attitudes towards IPV were assessed by asking respondents if they would justify partner abuse of a woman for one or several reasons, such as: (i) if she goes out without telling him; (ii) if she neglects the children; (iii) if she argues with him; (iv) if she refuses to have sex with him; and (v) if she burns the food. Responses to these questions were transformed into a single dichotomous 'yes' or 'no' variable. Rural women who responded 'yes' to one or several of the attitude questions formed one group of the dichotomy, were considered to be the risk group, and were coded 1. However, women who responded 'no' to all the attitude questions (ie a firm negative response) formed the other group of the dichotomy, and were coded 0. This distinction was created to allow for meaningful interpretation of results in estimating the risk factors for patriarchal attitudes towards intimate partner violence.

Independent variables: The independent variables used in the logistic regression model included demographic characteristics, assessed using the following indicators: age (grouped as 15-18, 19-23, 24-28, 29-33, or >34 years); marital status (grouped as 'never married', 'currently married', 'formerly married'); region of residence (grouped as North Central, North East, North West, South East, South South, South West); ethnic affiliation assessed as a merger of Fulani/Hausa/Kanuri ethnic groups (categorization was based on the criteria of ethnic groups speaking a common language or dialect; sharing a sense of identity, cohesion and history; having a single set of customs and behavioural rules as in marriage, clothing, diet, taboos etc), Igbo, Yoruba, and others (a merger of other minor ethnic groups); and religious affiliation (classified as Christian, Muslim, Traditional and others). Socio-economic status was assessed using the following variables: highest level of education, classified as no education, primary, secondary or higher; and occupation was assessed as: professional/technical/managerial, clerical/sales/services/skilled manual, agricultural self-employed/agricultural employee/household and domestic/unskilled manual occupations; and not working; and wealth index, an indicator of the economic status of households that is consistent with expenditure and income measures. The wealth index was constructed using information about ownership of a range of household assets; each asset being assigned a weight (factor score) generated through principal component analysis, and the resulting asset scores were standardized in relation to a standard normal distribution with a mean of zero and standard deviation of one. Each household was then assigned a score for each asset, and the scores were summed for each household. Individuals were ranked according to the total score of the household in which they resided, and the sample was then divided into three quintiles: poor, middle, and rich²⁶.

Women's empowerment was assessed using three indicators. The first was autonomy in domestic decisions, assessed by asking the women if they had final say regarding 'large household purchases', 'daily household purchases', 'visits to family or friends, 'own health', and 'food to be cooked each day'. Possible response options 'respondent alone', 'respondent and wife/partner', 'respondent and other person in the household', formed one group of the dichotomy, while the options 'husband/partner alone', and 'someone else' formed the other group of the dichotomy. The second indicator was access to media which was assessed using questions on frequency of listening to the radio, reading newspapers/magazines, and watching television. Responses were dichotomized into 'not at all' in one group; and 'less than once a week', 'at least once a week', and 'almost every day' in the other group. The third indicator literacy level, was considered a factor influencing access to information. It was assessed as the ability to read (being 'able to read whole sentences' formed one group of the dichotomy; while those 'able to read part of a sentence' and 'unable to read' were considered illiterate, and formed the other group of the dichotomy.

Analysis

Percentage distributions were made of the demographic and other relevant characteristics of the respondents. Many of the independent variables were transformed to reduce the number of categories wherever certain categories lacked enough subjects to enable meaningful statistical analysis. The transformations, however, remained logical. Only the predictor variables that were statistically significant in the bivariate analyses (p<.05) were all entered into the logistic regression model in a single block to control for possible confounding between these variables. The magnitude and direction of the relationship between the variables were expressed as odds ratios (OR) and significant levels expressed as p-values, and assumed at p<.05. Missing data were excluded from the analysis.

Results

Rural women who believe IPV is justified

The percentage of women who believed that IPV is justified is presented (Fig1). In total, 42% of the rural women justified IPV with at least one of the given reasons.



Figure 1: Percentage distribution of rural women's justifications for IPV.


Proportion of rural women with tolerant attitudes towards IPV by predictor variables

The frequency distribution of predictor variables associated with the rural women having tolerant attitudes towards violence is presented (Table 1). Significantly higher proportions of rural women who justified IPV were found among those who were currently married, resident in the North East region, of Hausa/Fulani/Kanuri ethnic group, Muslim, without education, and lived in poor households. Similarly, rural women who had no autonomy in household decisions pertaining to own health, large household purchases, household purchases for daily needs, visits to family or relatives, and on food to be cooked were found in significantly higher proportions to justify IPV compared with women who had full or partial autonomy. Finally, significantly higher proportions of rural women with tolerant attitudes towards violence were found among the women without access to newspaper, and television, as well as among the illiterate.

Table 1: Proportion of rural women with tolerant attitudes toward violence, by predictor indicators



Finding IPV justifiable: predictors of attitudes of rural women towards IPV

The results of the logistic regression analysis of attitudes of rural women towards IPV, and predictor variables are presented (Table 2). Rural women in the northern region (North Central, OR = 2.05, p<0.049, North East, OR = 11.76, p<0.000, and North West, OR = 2.48, p<0.015) as well as in the South South region (OR = 2.44, p <0.020) were at higher risk of justifying IPV, compared with women in the South West region. In addition, rural Muslim women (OR = 1.52, p<0.007) were at higher risk of justifying IPV compared with rural Christian women. Rural women with no education (OR = 1.39, p<0.219) and with primary education (OR = 1.03, p<0.038) were at higher risk of justifying IPV compared with women with secondary or higher education. Similarly, rural women in the poor (OR = 1.53, p<0.006) and middle wealth (OR = 1.50, p<0.006) quintiles were at higher risk of justifying IPV compared with their peers within the rich wealth quintile.

Table 2: Logistic regression analysis of rural women's attitudes towards IPV by predictor variables, with odds ratios (OR) and confidence intervals (CI)



In contrast, rural women belonging to the Yoruba ethnic group (OR = 0.27, p<0.016) were at lower risk of justifying IPV compared with rural women in the 'other' ethnic group (ie the reference category). Finally, rural women who had no access to radio (OR = 0.68, p<0.000) were at lower risk of justifying IPV compared with women who had access. However, occupation, literacy, and autonomy in household decisions were not significantly associated with the risk of justifying IPV.

Discussion

Intimate partner violence was shown to be tolerated under several circumstances; approximately 42% of the rural women in the present study would justify IPV for at least one of the given reasons. Comparable figures have been reported in several other studies in developing countries^27-29. However, the present results are lower than those reported in studies in similar developing countries^{9,10-12,29,30}. Although Nigeria is a signatory to most of the instruments aimed at eliminating gender-based violence, such as the International Conference of Population and Development (Cairo, 1994)³¹, the Beijing Declaration made at the Fourth World Conference on Women (Beijing, 1995)³², and the African Charter on Human and Peoples Rights, 1986³³, violence against women continues to be pervasive. This may be connected with the existence of discriminatory laws that condone and even legalize certain forms of violence against women (for instance The Penal Code Section 55 that applies in the northern states contains the compulsory requirement that a woman must appease her husband if he feels offended by her in the Igbo culture of in eastern Nigeria), and informal traditions^34,35.

The results of the present study suggest social, religious and cultural influences in attitudes towards violence among the rural women. Women in the three northern regions, women who were Muslims, and women who had primary or no education were more likely to report tolerant attitudes towards violence compared with their counterparts from the southern region, Christians, and those with secondary or higher education. Regarding the findings of tolerant attitudes of rural women in the northern regions and Muslim women towards violence, this is highly indicative of the socio-cultural and religious restrictions imposed on them, and is thus in agreement with the findings from recent work from Nigeria^29,36. This is an expected finding, given that in Nigeria the Sharia Penal Code (that portion of a state's laws that deal with defining the elements of particular crimes and specifying the punishment for each crime), applicable in northern states, permits husbands to 'correct' their wives as long as such correction does not result in grievous harm, which is defined as loss of sight, hearing, power of speech, facial disfigurement or life-endangering injuries³⁷. Also related is the fact that northern women are more socio-culturally subordinate and economically disadvantaged than their peers in the southern region, as well as the gender-restrictiveness of the predominant ethnic groups (ie Hausa/Fulani/Kanuri) in the northern regions, which predisposes them to IPV^29,36,38.

Tolerant attitudes to IPV of rural women with primary or no education, and lower household wealth have also been reported in previous studies^10,14,18,29. It has been posited that such women of low socio-economic status are likely to experience violence due to their limited resources^18,35,39. Although the mechanism by which poverty increases the risks of violence is still unclear, low socio-economic status probably reflects a variety of conditions that, in combination, increase women's risk of victimization⁴⁰. This may also be related to their attitudes towards violence, and is an important finding for policymakers in their efforts to change societal attitudes towards IPV and minimize violence against women.

The finding of reduced risk of rural women from the Yoruba ethnic group tolerating IPV was corroborated in other studies²⁹. This ethnic difference may be explained in terms of the social institutions of gender and women's autonomy, in which ethnic groups that are more gender egalitarian are less likely to justify IPV²⁹. The more egalitarian status of Yoruba women in relation to Yoruba men is exemplified by a study on fertility desires, which shows that Yoruba women are better able to negotiate future pregnancies and completed family size after they have successfully borne several children for their husbands and husbands' lineages. In effect these women's value depends upon, and is confirmed by, their reproductive success⁴¹.

Finally, the finding that lack of access to media (radio) was associated with a lower risk of justifying violence among the rural women is worthy of note. Although this is in contrast to findings in a nation-wide study from Nigeria²⁹, the explanation for this finding is unclear. It does, however, underscore the importance of structural empowerment in forming women's attitudes toward violence, which may necessitate the introduction of structural changes (eg improved literacy and education for women, and improved media access), along with other interventions to alter societal norms and attitudes toward IPV and reduce exposure to violence, a point made in a previous study¹⁴.

Policy implications

The findings in this study have important policy implications. Violence against women as well as existing patriarchal attitudes towards IPV must be emphasized as public health problems in Nigeria, warranting the formulation and implementation of policies to counter the effects of IPV. In addition, law enforcement agencies must be empowered to intervene and, if necessary, prosecute perpetrators of violence against women. Changing social, cultural and religious norms would require long-term action, using formal (legislature, law enforcement) and informal (community, traditional and religious leader) means.

Study limitations

A number of caveats need to be taken into consideration when interpreting the results of this study. Focusing on women's attitudes toward IPV will not, however, fully enhance policy-makers ability to capture the entire scope of societal norms regarding domestic violence. In addition, it does not capture the issues that motivate partner abuse or attitudes towards partner abuse, such as a husband's drinking habits. Therefore, wider measures of attitudes toward IPV need to be made using qualitative research methods.

Conclusion

In conclusion, identifying factors associated with women's attitudes towards IPV not only pin-points priority groups who require intervention, but also identifies possible obstacles to such interventions. However, the situation in Nigeria is complex, partly as a result of religious doctrines, and partly due to traditional/ cultural beliefs that tend to have a powerful influence on women themselves. Ending IPV in Nigeria, as in many other patriarchal societies, requires long-term commitment and strategies involving all of society. This may require stronger commitments by governments to passing and enforcing laws that ensure women's legal rights and the punishment of abusers. In addition, community-based strategies can focus on empowering women, reaching out to men, and changing the beliefs and attitudes that permit abusive behaviour. Only when women are treated as equal members of society will violence against women change from being an invisible norm into a shocking aberration.

Acknowledgements

The authors thank Measure Demographic and Health Survey (ORC Macro) for the data used in this article, the three anonymous reviewers for their very useful comments and our advisor and mentor.

References

1. WHO/WHD. World Report on Violence (WHO/FRH/WHO/97.8#). Geneva: World Health Organization, 1997.

2. Krug EG, Dahlberg LL, Mercy JA, Zwi AB, Lozano R. World report on violence and health. Geneva: World Health Organization, 2002.

3. Grisso JA, Schwarz DF, Hirschinger N, Sammal M, Brensinger C, Santanna J et al. Violent injuries among women in an urban area. New England Journal of Medicine 1999; 341: 1899-1905.

4. Kyriacou DN, Anglin D, Taliaferro E, Stone S, Tubb T, Linden JA et al. Risk factors for injury to women from domestic violence against women. New England Journal of Medicine 1999; 341: 1842-1898.

5. Bacchus L, Mezey G, Bewley S. Domestic violence: prevalence in pregnant women and association with physical and psychological health. European Journal of Obstetrics and Gynecology and Reproductive Biology 2004; 113: 6-11.

6. Ellsberg MC, Rena R, Herrera A, Liljestrand J, Winkvist A. Wife abuse among women of childbearing age in Nicaragua. American Journal of Public Health 1999; 89: 241-244.

7. Heise L, Raikes A, Watts CH, Zwi AB. Violence against women: a neglected public health issue in developed countries. Social Science and Medicine 1994; 39: 1165-1179.

8. Jewkes RK, Levin J, Penn-Kekana L. Risk factors for domestic violence: Findings from a South African cross-sectional study. Social Science and Medicine 2002; 55: 1603-1617.

9. Koenig MA, Lutalo T, Zhao F, Nalugoda F, Wabwire-Mangen F, Kiwanuka N et al. Domestic violence in rural Uganda: evidence from a community-based study. Bulletin of the World Health Organization 2003; 81: 53-60.

10. Hindin MJ. Understanding women's attitudes towards wife abuse in a non-western society. Bulletin of the World Health Organization 2003; 81: 501-508.

11. Rani M, Bonu S, Diop-Sidibe N. An empirical investigation of attitudes towards wife-beating among men and women in seven sub-Saharan African countries. African Journal of Reproductive Health 2004; 8: 116-136.

12. Haj-Yahia MM. Belief about wife-beating among Arab men from Israel: the influence of patriarchal ideology. Journal of Family Violence 2003, 18: 193-206.

13. WHO. WHO multi-country study on women's health and domestic violence. (WHO/FCH/GWH/02.01). Geneva, World Health Organization, 1999.

14. Lawoko S. Factors associated with attitudes toward intimate partner violence: a study of women in Zambia. Violence and Victims 2006; 21: 645-655.

15. Ilika AL. Women's perceptions of partner violence in a rural Igbo community. Africa Journal of Reproductive Health 2005; 9: 77-88.

16. Karamagi CAS, Tumwine JK, Tylleskar T, Heggenhougen K. Intimate partner violence against women in eastern Uganda: implications for HIV prevention. BioMed Central Public Health 2006; 6: 284.

17. El-Zanaty F, Hussein EM, Shawky GA, Way AA, Kishor S. Egypt Demographic and Health Survey 1995. Calverton, MD: National Population Council (Egypt) & Macro International Inc, 1996.

18. Gonzaláles-Brenes M. Domestic violence and household decision-making: evidence from East Africa. (Online) 2004. Available: http://www.sscnet.ucla.edu/polisci/wgape/papers/7_Gonzalez.pdf (Accessed 9 September 2008).

19. Lawoko S. Predictors of attitudes toward intimate partner violence: A comparative study of men in Zambia and Kenya. Journal of Interpersonal Violence 2008; 23: 1056-1074.

20. Watts C, Keogh E, Ndlovu M, Kwaramba R. Withholding of sex and forced sex: dimensions of violence against Zimbabwean women. Reproductive Health Matters 1998; 6: 57-65.

21. Ejiofor L. Dynamics of Igbo democracy: behavioural analysis of Igbo politics in Agiuyi clan. Ibadan: University Press, 1981.

22. Uchem RN. Overcoming women's subordination, an Igbo African and Christian perspective. Enugu: Snapp Press, 2001.

23. Okeke GC. Christian marriage in Nigeria in the 21st century: problems and challenges. Bigard Theological Studies 2002; 2: 19-35.

24. Hanson RK, Cadsky O, Harris A, Lalonde C. Correlates of battering among 997 men: family history, adjustment and attitudinal differences. Violence and Victims 1997; 12: 191-208.

25. National Population Commission. Nigeria Demographic and Health Survey 2003. Calverton, MD: National Population Commission and ORC/Macro, 2004.

26. Filmer D, Pritchett LH. Estimating wealth effects without expenditure - or tears: an application to educational enrolments in States of India. Demography 2001; 38: 115-132.

27. Fawole OI, Aderonmu AL, Fawole AO. Intimate partner violence: wife beating among civil servants in Ibadan. African Journal of Reproductive Health 2005, 9: 54-64.

28. Ogunjuyigbe PO, Akinlo A, Ebigbola JA. Violence against women: an examination of men's attitudes and perceptions about wife beating and contraceptive use. Journal of Asian and African Studies 2005; 40: 219-229.

29. Oyediran KA, Isiugo-Abanihe UC. Perceptions of Nigerian women on domestic violence: evidence from 2003 Nigeria Demographic and Health Survey. African Journal of Reproductive Health 2005; 9: 38-53.

30. Khawaja M, Linos N, El-Roueiheb Z. Attitudes of men and women towards wife beating: findings from Palestine refugee camps in Jordan. Journal of family Violence 2008; 23: 211-218.

31. United Nations Population Information Network (POPIN). Report of the International Conference on population and Development, Cairo, 5-13 September 1994. Report no. A/CONF.171/13: Report of the ICPD (94/10/18). (Online) 1994. Available: http://www.un.org/popin/icpd/conference/offeng/poa.html (Accessed 9 September 2008).

32. United Nations. Fourth World Conference on Women Action for Equality, Development and Peace 4-15 September 1995 - Beijing, China. Report n. A/CONF.177/20. (Online) 1995. Available: http://www.un.org/esa/gopher-data/conf/fwcw/off/a--20.en (Accessed 9 September 2008).

33. African Commission on Human and People's rights. Protocol to the African Charter on Human and Peoples' Rights on the rights of women in Africa. (Online) 2003. Available: http://www.achpr.org/english/_info/women_en.html (Accessed 9 September 2008).

34. Odimegwu CO, Okemgbo CN. Gender role ideologies and prevalence of violence against women in Imo State, Nigeria. Anthropologist 2003; 5: 225-236.

35. Amnesty International. Nigeria Unheard Voices: Violence against women in the family. (Report no. AFR 44/004/2005). (Online) 2005. Available: http://asiapacific.amnesty.org/library/pdf/AFR440042005ENGLISH/$File/AFR4400405.pdf (Accessed 1 May 2007).

36. Robson E. Wife seclusion and the spatial praxis of gender ideology in Nigerian Hausaland. Gender, Place and Culture 2008; 7: 179-199.

37. Women's International Network. Family violence in Lagos, Nigeria. Lagos: Women International network News, 1998.

38. Kritz MM, Makinwa-Adebusoye P. Ethnicity, work and family as determinants of women's decision-making autonomy in Nigeria. Population and Development Program Working Paper Series No. 97.06. Ithaca, NY: Cornell University, Department of Rural Sociology, Population and Development Program, 1997.

39. Yllo K. Sexual equality and violence against wives in American states. Journal of Comparative Family Studies 1983; 14: 676-686.

40. Heise L. Violence against women: an integrated, ecological framework. Violence Against Women 1998; 4: 262-290.

41. Bankole A. Desired fertility and fertility behaviour among the Yoruba of Nigeria: A study of couple preferences and subsequent fertility. Population Studies 1995; 49: 317-328.