The Social Science Journal 46 (2009) 493–505

Available online at www.sciencedirect.com

Gender role attitudes and attitudes to abortion:Are there gender differences?

Cynthia J. Patel ∗, Lucinda Johns

University of KwaZulu-Natal, South Africa

Received 13 May 2008; received in revised form 9 February 2009; accepted 13 February 2009

Abstract

This study examines gender differences in gender role attitudes and attitudes to abortion in a sampleof 141 undergraduate students. Religion and religiosity were treated as covariates. Using the short ver-sion of Spence and Helmreich’s (1978) Attitudes to Women Scale (AWS) taken from Spence and Hahn[Spence, J. T., & Hahn, E. D. (1997). The Attitudes Toward Women Scale and attitude to change incollege students. Psychology of Women Quarterly, 21, 17–34] and the abortion attitudes scale developedby Stets and Leik [Stets, J. E., & Leik, R. K. (1993). Attitudes about abortion and varying attitudestructures. Social Science Research, 22, 265–282], we found that females have more liberal attitudesto women and show stronger approval for women’s autonomy in abortion decision-making. Given thelow internal consistency of this three-item subscale, responses to the individual items were analyzed.More males than females feel that the father should have the right to prevent the mother from hav-ing an abortion. Most of the students indicate that the woman should have to tell the father beforehaving an abortion and most disagree that the abortion is entirely a woman’s decision. There wereno gender differences on the moral acceptability and availability subscales. The overall results reflectsomewhat conservative attitudes to women, and seemingly contradictory findings on the abortion issuein this sample of university students. Although they endorse the availability of abortion, at the sametime they find it morally unacceptable and show weak support for women’s autonomy in this deci-sion. The implications of these findings for the advancement of women’s rights in this country arediscussed.© 2009 Western Social Science Association. Published by Elsevier Inc. All rights reserved.

∗ Corresponding author at: School of Psychology, University of KwaZulu-Natal, King George V Avenue,Glenwood, Durban 4041, South Africa. Tel.: +27 312607619; fax: +27 312602618.

E-mail addresses: sebjo1zn@iafrica.com, patelc@ukzn.ac.za (C.J. Patel).

0362-3319/$ – see front matter © 2009 Western Social Science Association. Published by Elsevier Inc. All rights reserved.doi:10.1016/j.soscij.2009.02.006

494 C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505

Large scale surveys (Rule, 2004), smaller investigations (Patel & Kooverjee, in press; Patel& Myeni, 2008) and reports by Harrison, Montgomery, Lurie, and Wilkinson (2000) andVarga (2002) indicate that South Africans, in general, have negative attitudes to abortion. Atthe same time, South Africa is one of the most progressive countries in the world in termsof the legal recognition of sexual and reproductive rights (Cooper et al., 2004). Althoughwomen’s rights were accorded high priority in the country’s transformation from apartheid todemocracy, very little research has been done on the public’s attitude towards women and theirreproductive rights. The present study examines gender differences in gender role attitudesand attitudes to abortion in a sample of young South African students taken from a historicallydisadvantaged university campus. Religiosity and religion were statistically controlled in thestudy.

1. Gender role attitudes

Although considerable progress has been made with regard to women’s rights in SouthAfrica, Albertyn (2003) observes that these developments are mainly at the level of legal reform,while attention to cultural attitudes and values are not given priority. It means, therefore, thatuntil women’s organizations as well as the government engage in attempts at social reform,patriarchal attitudes about reproductive freedom, sexuality and gender will continue to prevailin this country. The study of gender roles is important since it aids our understanding of society’sbeliefs about a range of issues that affect women (Berkel, 2004).

That females in general have more egalitarian attitudes than males is a consistent findingin the literature on gender role attitudes (Berkel, 2004; Jones & McNamara, 1991; Knio &Elasmar, 1998; Kremer & Curry, 1987; Lottes & Kuriloff, 1992; Öngen, 2006; Toller, Suter,& Trautman, 2004). In their comparison of student cohorts of 1972, 1976, 1980 and 1992,Spence and Hahn (1997) found that there was a general increase in egalitarian attitudes overthe years, However, in all groups the males were always more traditional than the females.Similar findings were recorded by Loo and Thorpe (1998) in their 20-year replication study.Twenge (1997) provides further support for the gender gap in their meta-analysis of studiesconducted from 1970 to 1995. Although there has been limited research on gender and genderroles in South Africa, a consistent pattern of conservative attitudes towards women seems tobe emerging: Fifty-five percent of the police officers used in the Stephanus (2006) study holdnegative attitudes to women; male and female patients at a STI clinic in Cape Town endorsetraditional attitudes to gender roles and relations (Kalichman et al., 2005); male communitymembers were found to hold hostile attitudes to women (Kalichman et al., 2007). In Harper’s(2001) South African student sample female respondents reported more liberal attitudes thanmales.

2. Attitudes to abortion

Studies on gender differences in abortion attitudes yield diverse results. Although Bailey’s(1993) study did not reveal differences between feminists and non-feminists on abortion

C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505 495

attitudes, he did find that female college students are more approving of abortion than malestudents. Esposito and Basow (1995) and Stevens, Caron, and Pratt (2003) did not find genderdifferences in college students’ attitudes to abortion. Bryan and Freed (1993) and Carlton,Nelson, and Coleman (2000) found that while male and females have similar attitudes to abor-tion, on specific issues males are more in favour of abortion. In the former study men tend toapprove for soft reasons like choice of sex, while in the latter more females disagree that abor-tion after the first trimester is acceptable. Finlay’s (1996) study of gender differences amongseminary students demonstrate that men tend to hold more traditional views about abortionthan the general public while women are found to be more liberal. Significantly, however,both men and women approve of abortion for conditions like “rape” or when the woman’s lifeis threatened by the pregnancy. Jelen, Damore, and Lamatsch’s (2002) longitudinal analysisrevealed that men and employed women share a similar pro-choice profile compared to home-maker women who tend to be more pro-life. On the issue of gender differences in attitudes tomale involvement in the abortion decision, the findings have generally been unclear. In their1997 study, Nelson and Coleman found that females express lower support for male involve-ment in the decision than males. In 1999 however, Coleman and Nelson found that in generalthe sample endorsed higher than moderate support for male involvement. Jones (2006) alsofound higher than average support in his sample with no gender differences in attitudes to maleinvolvement.

These results point to the complex nature of gender differences and the need to examinedifferences on specific issues rather than overall approval or disapproval of abortion.

Past research on gender role attitudes and attitudes to abortion appears to have mixed results.Wall et al. (1999) point out that while it is expected that the more progressive the attitudes towomen’s societal roles the greater the approval of abortion, research conducted mainly in theU.S. show rather weak correlations between these variables. In their attempt to investigate thistrend across cultures using samples from the U.S., Croatia, the Czech Republic and Slovenia,they found that although variations across countries do exist, overall gender role attitudesand religiosity are found to be reasonable predictors of abortion attitudes with the strongestfindings in the Slovenia sample. Bailey (1993) mentions several studies that report a positiverelationship between these variables but no differences between feminists and non-feminists onabortion attitudes. Jelen and Wilcox (2003) point out that the expected shift to more pro-choicetendencies in American society has not taken place despite changing gender role attitudes andwomen’s increasing participation in the paid labour market. The findings of a recent studyexamining abortion attitudes and attitudes to contraception (Patel & Kooverjee, in press) seemto suggest that South African students have somewhat conservative attitudes to women.

3. Religiosity

Over the years religion has emerged as a significant predictor of both abortion attitudes(studies reported by Bailey, 1993; Bryan & Freed, 1993; Esposito & Basow, 1995; Krishnan,1991; Modi, 2002; Sahar & Karasawa, 2005; Stets & Leik, 1993; Stevens et al., 2003) andgender role attitudes (studies reported in Read, 2003; Wilcox & Jelen, 1991). A few of thesestudies were conducted with Christian samples, using “Catholic” and “Protestant” as religions

496 C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505

rather than as Christian affiliations or denominations. Other studies have used Christian samplesonly (Hollis & Morris, 1992) or Catholic versus non-Catholic (Wall et al., 1999). In instanceswhere other religions are included (Hess & Rueb, 2005), the numbers are too small to makeany meaningful comparisons.

Rather than use “religion” per se as a predictor of attitudes, several researchers chooseto assess the impact of religiosity on abortion attitudes (Sahar & Karasawa, 2005). Afterattempting to find religious differences between Christian and Muslim women on the sourcesof gender role attitudes, Read (2003) conclude that religiosity (and ethnicity) is more criticalin determining women’s gender role attitudes than religious affiliation. Several studies usedchurch attendance as an indication of religiosity, while others extend the concept of religiosityby including one or more of the following issues: beliefs and rituals, self-rating of religiousness,as well as attendance at services and meetings (Hess & Rueb, 2005; Krishnan, 1991; Wallet al., 1999). Since church attendance alone demonstrates a bias toward Christian practices,more inclusive measures attempt to embrace other religious groupings as well. Given SouthAfrica’s multi-cultural diversity, an appropriate measure of religiosity is required. Idehen’s(2001) Religious Orientation Test (ROT), which claims to be “. . .mindful of and meaningfulto all religious faiths” (p. 59) was developed on a Nigerian sample and has generated somepromising psychometric data. In addition, the increase in charismatic religious groups withtheir range of conservative ideologies may have implications for women’s reproductive health(Mhlanga, 2003) and gender role attitudes.

Surveys show that religion plays a significant role in the lives of most South Africans (SouthAfrican youth are optimistic about our future, 2005; Zantsi, Pettifor, Madikizela-Hlongwa,MacPhail, & Rees, 2004). In addition, the results of smaller studies indicate that university stu-dents (at a historically disadvantaged South African university) report high levels of religiosity(Patel & Myeni, 2008). More than half of the South African sample in Roothman, Kirsten,and Wissing’s (2003) study was made up of university students. They revealed high levels ofreligious well-being.

Much of the literature on gender differences report higher religious well-being amongwomen than among men (Cecero, Bedrosian, Fuentes, & Bornstein, 2006; Dorahy et al., 1998;Gauthier, Christopher, Walter, Mourad, & Marek, 2006; Hammermeister, Flint, El-Alayli,Ridnour, & Peterson, 2005; Miller & Hoffman, 1995; Roothman et al., 2003). Given the sig-nificant role that religion and religiosity appear to play in gender role attitudes and abortionattitudes we decided to remove the influence of these variables by treating them as covariates.

4. Education

The positive association between levels of education and abortion attitudes is a regularfinding (Bahr & Marcos, 2003; Bolks, Evans, Polinard, & Wrinkle, 2000; Lee, Kleinbach,Hu, Peng, & Chen, 1996—for their Chinese sample and not their American sample; studiesreported in Walzer, 1994). Education has also been positively linked to more liberal attitudestoward women (Kremer & Curry, 1987). In their study of the predictors of hostile and benevolentsexism, Glick, Lameiras, and Castro (2002) found educational level to be the strongest predictorof less sexist attitudes. Tu and Liao (2005) point out that while education is consistently linked

C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505 497

with the probability of having more egalitarian gender role attitudes, the positive effects aremore noteworthy for females than for men.

In view of these findings, it is expected that students at tertiary level would exhibit positiveattitudes to abortion and progressive gender role attitudes. By using a sample of universitystudents this study attempts to control for at least one background characteristic that is knownto affect both attitudes to abortion and gender role attitudes. In contrast to international researchwhich indicates that students generally possess favourable attitudes to abortion, a study of asample of South African Black female students shows overall negative attitudes to abortion aswell as a negative relationship between attitudes to abortion and self-rated religiosity (Patel &Myeni, 2008).

5. Aims

Drawing from a pool of students with a similar profile as the ones used in previous studies(Patel & Kooverjee, in press; Patel & Myeni, 2008), this study was planned to examine whether,after controlling for religion and religiosity, students’ attitudes to abortion and attitudes towomen vary by gender.

6. Method

6.1. The study setting

The study was conducted at the University of KwaZulu-Natal, South Africa. This tertiaryinstitution is in KwaZulu-Natal, the province with the largest population and a rich diversityof cultures. More than 80% of the province’s population is Black African, about 8.5% isIndian, 5% is White and 2% Coloured (or mixed ancestry). Christianity is the religion of themajority, with Hinduism and Islam being the largest practised religions of the non-Christiangroups (Rule, 2002). Although it is the third richest province in South Africa there are highlevels of poverty and unemployment. The highest rates of poverty exist among African people(64.4%), while there is almost no poverty among White people (PROVIDE, 2005). With regardto educational level of the adult population, more than 9% have no education, just over 20%have some or have completed primary school education, 36% have some secondary education,23.6% have Grade 12 and 9.8% have tertiary education.

The institution now known as the University of KwaZulu-Natal is the result of a merger, in2004, of two universities: the University of Durban-Westville (UDW) (initially established forthe Indian population but opened its doors to all races in 1984), a historically disadvantaged,black university during the apartheid era and the University of Natal (UN), a historicallyadvantaged institution. The study was conducted in 2006 when the last group of “UDW”psychology students completed their qualifications. As a historically disadvantaged institutionUDW attracted mainly Black and Indian students from lower socio-economic backgrounds. In2006 there were 44% Black, 49% Indian, 5% White and 1% Coloured students on the Westvillecampus of UKZN.

498 C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505

6.2. Sample

The sample was made up of 63 male and 78 female students (N = 141), from the Westvillecampus of the University of KwaZulu-Natal (previously known as the University of Durban-Westville). The ages of the students ranged from 19 years to 32 years with a mean of 22.11years.

Sixty eight percent of the sample (n = 96) was Black, 28% (n = 39) was Indian and therest (n = 6) were either White or Coloured. Of the total, 74% (n = 104) was Christian while16% (n = 23) was Hindu, 5% (n = 7) Muslim, and the rest were classified in the “Other” or“No religion” categories. Significantly higher levels of religiosity were recorded in the femalesample compared to the male sample (t = 3.31, p < .05).

6.3. Instruments

A 20-item scale measuring attitudes to abortion (Stets & Leik, 1993) was adapted for thepurpose of the present study. The authors subjected the items to factor analysis and foundthree factors, namely, availability (10 items), moral acceptability (7 items) and a women’sautonomy (3 items) in an abortion decision. They report internal consistency scores of .96,.95 and .73, respectively for these three factors. The factors are used as subscales in thisstudy.

The participants were required to state the extent to which they agree or disagree withthe statements based on a 5-point Likert scale which ranged from strongly disagree tostrongly agree (coded 1–5). Factor analysis of the attitude items using a South Africansample (Patel & Myeni, 2008) resulted in a similar, but slightly more differentiated factorstructure than that of Stets and Leik (1993). The reliability analysis of the former studyand a study using a similar sample (Patel & Kooverjee, in press) revealed very satisfac-tory internal consistency values (Cronbach’s alpha) of .85 and .87, respectively. The higherthe scores on the three subscales, the stronger the belief that abortion should be avail-able, that abortion is morally acceptable and that women should have autonomy in decisionmaking.

The Religious Orientation Scale (ROT) developed by Idehen (2001) measures levels ofreligiosity using 6 items and a 1–5 response format option. High scores on this scale reflecta more superficial religious orientation while low scores are said to reflect a deeper religiousorientation. The author reports the following information on the scale’s reliability: test–retestreliability of .75, internal consistency of .80, and split-half reliability of .82. The validity figuresare as follows: Using a sample of 736 respondents, the correlation between the ROT and thePurpose-in Life scale was .143 (p < .001). The author’s one-factor solution accounting for82% of the variance provides evidence for the construct validity of the scale. A study using asimilar sample of South African students revealed Cronbach’s alpha of .88 on the ROT (Patel,Ramgoon, & Paruk, in press).

The short version of the Attitudes to Women Scale (AWS) developed by Spence and Helm-reich (1978) cited in Spence and Hahn (1997) contains 15 items which measure gender rolebeliefs. Spence and Hahn report internal consistency scores of .84 for their male sample and.81 for their female sample. Sherman and Spence (1997) cite research which showed that this

C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505 499

version is highly correlated with the original version and yielded a one-factor structure forboth men and women. The 4-point response options range from “agree strongly” to “disagreestrongly”. The higher the score on this instrument the more egalitarian the attitude of therespondent.

6.4. Procedure

The students completed the questionnaires during their research methodology lectures aspart of a class exercise aimed at demonstrating scaling techniques and descriptive statistics.

It was pointed out that participation was voluntary and the possibility of publication of thefindings was also discussed.

7. Results

The inter-item consistency values (Cronbach’s alpha) indicated mostly satisfactory levelsof reliability for the scales: .87 for the availability subscale, .83 for the moral acceptabilitysubscale, .61 for the women’s autonomy subscale of Stets and Leik’s (1993) attitudes to abortionscale, .87 for the Religious Orientation Test (ROT) and .77 for the Attitudes to Women Scale(AWS).

A one-way MANCOVA, was undertaken with gender as the independent variable, attitudesto women, the three abortion attitudes subscales (moral acceptability, availability and women’sautonomy) as dependent variables and religion and religiosity as covariates. Box’s M test wasperformed to evaluate the homogeneity of variances and covariances. The results indicated thatthe assumptions were not violated (Box’s M = 15.44, F = 1.49, p = .13). Using Wilk’s criterion,the gender main effect was found to be significant for the combined set of dependent variables.(Wilk’s λ = .76, F(4, 134) = 10.62, p < .001).

Controlling for the influence of religion and religiosity, the separate ANCOVAs revealedsignificant gender differences on attitudes to women and women’s autonomy in the abortiondecision. Scores on the availability subscale and moral acceptability subscale did not differ bygender (Table 1).

Table 1Summary of the univariate ANCOVAs on the availability, moral acceptability, and women’s autonomy subscales,and the Attitudes to Women Scale.

Measures Group estimated marginal mean (S.E.) F

Males (n = 63) Females (n = 78)

Availability 29.69(1.04) 31.55(.93) 1.71Moral acceptability 15.16(.72) 15.42(.64) .07Women’s autonomy 5.69(.36) 7.19(.32) 9.51*

AWS 28.14(.72) 34.25(.65) 38.49**

Note. AWS = Attitudes towards Women Scale.∗ p < .01.∗∗ p < .001.

500 C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505

8. Discussion

After controlling for religiosity and religion the results indicated that female students weremore liberal in their attitudes to women than their male counterparts. This finding adds to theconsistent pattern of research which indicates that females tend to endorse more egalitariangender roles than males do. This is not surprising since the scale measures issues that are directlypertinent to women’s roles at work and in society at large. Given their history of limited social,political and economic power it is argued that women have more to gain from a more egalitariansociety than men do (Beere, King, Beere, & King, 1984). However, the level of support asreflected in the mean scores of this sample is worthy of comment. It is interesting to note thatthe mean AWS score for males in our study (Mean = 28.08) is only slightly higher than that ofthe American male cohort of more than 27 years ago (Mean = 26.04), while our female average(Mean = 34.29) is lower than the female cohort scores of 1992 (Mean = 36.34) as reported inSpence and Hahn (1997). While the AWS does not encompass issues of reproductive freedom,these findings confirm that although there was a gender difference on this scale, the sample ingeneral hold rather conservative attitudes to women. That our female student sample’s averageis lower than that of the American sample of more than a decade-and-a-half ago is also a causefor concern, given the progress this country has made in terms of raising awareness of women’srights.

At a superficial level, the outcomes relating to abortion attitudes seem to be slightly morepromising. Female students appeared to be more positive on issues of women’s autonomyin the abortion decision-making process than the male sample. No differences were found onavailability and moral acceptability of abortion. A study by Patel and Kooverjee (in press) usinga similar sample, also found gender differences on women’s autonomy. Although females inthis study showed stronger support for women’s autonomy in the abortion decision than males,the means indicate that the overall level of support is rather weak. Moreover, the low internalreliability of this subscale suggests that this result needs to be treated with caution. The lowalpha coefficient (.61) for this three-item subscale could be a function of its heterogeneity,the inadequate number of items making up this dimension or a combination of these issues.Inspection of responses on individual items revealed the following: significantly more of themales (78%) compared to the females (59%) felt that the father should have the right to preventthe mother from having an abortion (χ2 = 5.98, p < .05); more than 80% of both males andfemales said that the woman should have to tell the father before having an abortion; and 22%of males and 32% of females supported the idea that abortion is entirely a woman’s decision.Chi-square analyses revealed no gender differences in the responses to the latter two statements.There were striking similarities in the extent of agreement between a previous study of femalestudents (Patel & Myeni, 2008) and this sample of females on the first two issues, with a notabledifference on the third issue “Abortion should be entirely the woman’s decision” (almost 55%agreement in the earlier study compared to 32% in this study). These high levels of overallsupport for the fathers’ involvement in the abortion decision for both males and females appearto be in line with more recent studies (Coleman & Nelson, 1999; Jones, 2006) compared toan earlier study by Nelson and Coleman (1997). As with Jones’s (2006) sample, our samplewas more pro-life than pro-choice: The present sample’s (overall) performance on the moralacceptability (Mean = 15.30) and women’s autonomy (Mean = 6.52) subscales resemble those

C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505 501

of the pro-life group (Means = 16.28 and 7.96, respectively) in Stets and Leik’s (1993) study.Current measures of the male role in the decision making process assess extent of agreementwith fathers right to object to termination of pregnancy as can be seen in the individual itemsformulated by Stets and Leik (1993) and Bryan and Freed (1993). Given evidence that in thiscountry, the opposite could be a reality with fathers’ insisting on termination (Varga, 2002),future assessments of fathers’ role in abortion decision making need to be more explicitlyformulated in order to provide an accurate assessment of what male involvement means in thiscontext.

On the availability subscale, however, our sample’s mean of 30.71 was even higher thanthat of the pro-choice group (Mean = 27.44) in Stets and Leik’s (1993) study. This suggeststhat while our group as a whole were inclined to find abortion morally unacceptable and dis-agreed with women’s autonomy in the decision, they tended to strongly favour its availability.These seemingly disparate findings may be a function of the fact that the availability sub-scale contains items tapping support for abortion availability in traumatic situations like rape,incest or danger to the mother’s health. Research has consistently shown high endorsementfor abortion in these instances (Carlton et al., 2000; Finlay, 1996; Hollis & Morris, 1992;Lindell & Olsson, 1993; Patel & Kooverjee, in press). Other items in this scale cover thegovernment’s involvement in abortion accessibility. Since South Africa has liberal abortionlaws, respondents may feel obligated to agree with current legislation. In their review on abor-tion research, Jelen and Wilcox (2003) report the idea that political party identification tendsto be more enduring than abortion attitudes and that partisanship determines issue attitudesrather than vice versa. In the case of the present sample, knowledge of party affiliation wouldhave afforded an opportunity to test this hypothesis. It is possible that, by virtue of the itemscovered, student responses on the availability subscale reflect a more public picture of their atti-tudinal structure while responses to the moral acceptability subscale represent more personalperspectives.

These findings have implications for the outcomes of South Africa’s initiatives in imple-menting strategies that are meant to create gender equality. If traditional attitudes are present, itmeans that to improve the situation of women in this country, fundamental changes in the out-looks of both men and women need to occur. That these conservative attitudes prevail regardlessof religious beliefs, religiosity and educational level suggests that there may be other culturaldynamics, worldviews and values that need to be considered. Stets and Leik (1993) proposethat abortion attitudes are more influenced by a monolithic, closed minded approach than byreligious orientation.

Correlational analyses revealed that attitudes to women were not related to levels of religios-ity or attitudes to abortion. Jelen and Wilcox (2003) argue that although abortion and genderpolitics tend to converge at times, it may be incorrect to treat the former as an extension of thelatter. Drawing on several studies with varied findings on gender attitudes as a factor in abortionattitudes, the authors comment on the apparent complexity of this relationship. In the presentstudy the failure to find a significant correlation may be traced to a more basic cause. It is pos-sible, as Bailey (1993) points out, that the AWS is dated. In addition the internal consistencyof the items, as indicated by Cronbach’s alpha of .77 suggests that the items may be slightlyheterogeneous. Moreover the items in the AWS relate to women’s equality in the workplaceand in the home. The scale does not tap into attitudes on more sensitive and controversial

502 C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505

issues of women’s reproductive freedom. While tremendous strides have been made in termsof the recognition of the productive role that women play in society, it is possible that soci-etal attitudes about their reproductive rights remain conservative. Women’s freedom to controltheir fertility and in particular the highly emotive and complex nature of abortion, places itin a realm which is further complicated by religious teachings and the cultural importance offertility.

The present study’s findings on gender role attitudes appear to correspond with that of thefew South African studies reviewed: Overall endorsement of egalitarian gender roles was low;compared to males, females were found to have more liberal attitudes to women. Differences inabortion attitudes are less clearly differentiated. While the students show strong overall supportfor the availability of abortion, their reported moral unacceptability of abortion is similar toStet’s and Leik’s pro-life sample and they defend the fathers’ role in the decision makingprocess. Further research would need to examine the possibility of public versus personaloutlooks in influencing these seemingly ambiguous findings.

While it cannot be claimed that the participants accurately represent the population ofKwaZulu-Natal, the convenience sample of research methodology students offers an opportu-nity for us examine patterns of relationships suggested from previous studies (Patel & Myeni,2008; Patel & Kooverjee, in press) using participants with similar characteristics. At best thecurrent sample represents that sector of the population with higher educational levels and fromlower socio-economic status levels.

8.1. Recommendations

In order to assist the government in its drive to promote women’s rights and in particularto ensure easier access to fertility control options, the negative attitudes of young educatedpeople of this country need to be addressed. As Loo and Thorpe (2005) point to the functionof educators in effecting change by opening up discussions around gender roles, challengingtraditional roles, increasing awareness and eventually behaviours of future generations, wealso need to recognize the critical roles of religious leaders and women’s organizations incollaborating on a shared agenda.

Since reproductive freedom, is seen as one of the central issues in women’s health inSouth Africa, all relevant stakeholders need to accept responsibility in ensuring that traditionalattitudes to women do not present obstacles to this goal.

References

Albertyn, C. (2003). Contesting democracy: HIV/AIDS and the achievement of gender equality in South Africa.Feminist Studies, 29, 595–615.

Bahr, S. J., & Marcos, A. C. (2003). Cross cultural attitudes toward abortion: Greek vs. Americans. Journal ofFamily Issues, 24, 402–424.

Bailey, W. T. (1993). College students’ attitudes toward abortion. Journal of Social Behavior and Personality, 8,749–756.

Beere, C., King, D., Beere, D., & King, L. (1984). The Sex-Role Egalitarianism Scale: A measure of attitudestoward equality between the sexes. Sex Roles, 10, 563–567.

C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505 503

Berkel, L. A. (2004). A psychometric evaluation of the Sex-role Egalitarianism Scale with African Americans. SexRoles, 50(9–10), 737–742.

Bolks, S. M., Evans, D., Polinard, J. L., & Wrinkle, R. D. (2000). Core beliefs and abortion attitudes: A look atLatinos. Social Science Quarterly, 81, 253–260.

Bryan, J. W., & Freed, F. W. (1993). Abortion research: Attitudes, sexual behavior, and problems in a communitycollege population. Journal of Youth and Adolescence, 22, 1–22.

Carlton, C. L., Nelson, E. S., & Coleman, P. K. (2000). College students’ attitudes toward abortion and commitmentto the issue. The Social Science Journal, 37, 619–625.

Cecero, J. J., Bedrosian, D. R., Fuentes, A., & Bornstein, R. F. (2006). Religiosity and health dependencyas predictors of spiritual well-being. The International Journal for the Psychology of Religion, 16(3),225–238.

Coleman, P., & Nelson, E. (1999). Abortion attitudes as determinants of perceptions regarding male involvementin abortion decision. Journal of American College Health, 47(4), 164–171.

Cooper, D., Chelsea, M., Orner, P., Moodley, J., Harries, J., Cullingworth, L., et al. (2004). Ten years of democracyin South Africa: Documenting transformation in reproductive health policy and status. Reproductive HealthMatters, 12(24), 70–85.

Dorahy, M. J., Lewis, C. A., Schumaker, J. F., Akuamoah-Boateng, R., Duze, M. C., & Sibiya, T. E. (1998). Across-cultural analysis of religion and life satisfaction. Mental Health, Religion & Culture, 1(1), 37–43.

Esposito, C. L., & Basow, S. A. (1995). College students’ attitudes toward abortion: The role of knowledge anddemographic variables. Journal of Applied Social Psychology, 25, 1996–2017.

Finlay, B. (1996). Gender differences in attitudes toward abortion among Protestant seminarians. Review of ReligiousResearch, 37, 354–360.

Gauthier, K. J., Christopher, A. N., Walter, M. I., Mourad, R., & Marek, P. (2006). Religiosity, religious doubt,and the need for cognition: Their interactive relationship with life satisfaction. Journal of Happiness Studies, 7,139–154.

Glick, P., Lameiras, M., & Castro, Y. R. (2002). Education and Catholic religiosity as predictors of hostile andbenevolent sexism toward women and men. Sex Roles, 47(9–10), 433–441.

Hammermeister, J., Flint, M., El-Alayli, A., Ridnour, H., & Peterson, M. (2005). Gender differences in spiritualwell-being: Are females more spiritually-well than males? American Journal of Health Studies, 20(2), 80–84.

Harper, M. (2001). Individual differences in gendered person perception: A multifactorial study. Unpublishedmaster’s thesis. Johannesburg, South Africa: Rand Afrikaanse University.

Harrison, A., Montgomery, E. T., Lurie, M., & Wilkinson, D. (2000). Barriers to implementing South Africa’sTermination of Pregnancy Act in rural KwaZulu-Natal. Health Policy and Planning, 15, 424–431.

Hess, J. A., & Rueb, J. D. (2005). Attitudes toward abortion, religion, and party affiliation among college students.Current Psychology, 24(1), 24–42.

Hollis, H. M., & Morris, T. M. (1992). Attitudes toward abortion in female undergraduates. College Student Journal,26, 70–74.

Idehen, E. E. (2001). The development and evaluation of a religiosity scale. IFE Psychologia, 9(2), 58–69.Jelen, T. G., Damore, D. F., & Lamatsch, T. (2002). Gender, employment status, and abortion: A longitudinal

analysis. Sex Roles, 47, 321–330.Jelen, T. G., & Wilcox, C. (2003). Causes and consequences of public attitudes toward abortion: A review and

research agenda. Political Research Quarterly, 56(4), 489–500.Jones, B. H., & McNamara, K. (1991). Attitudes toward women and their work roles: Effects of intrinsic and

extrinsic religious orientations. Sex Roles, 24(1–2), 21–29.Jones, R. K. (2006). Male involvement in the abortion decision and college students’ attitudes on the subject. The

Social Science Journal, 43, 689–694.Kalichman, S. C., Simbayi, L. C., Kaufman, M., Cain, D., Cherry, C., Jooste, S., et al. (2005). Gender attitudes,

sexual violence, and HIV/AIDS risks among men and women in Cape Town, South Africa. The Journal of SexResearch, 42(4), 299–305.

Kalichman, S. C., Simbayi, L. C., Cain, D., Cherry, C., Henda, N., & Cloete, A. (2007). Sexual assault, sexual risksand gender attitudes in a community sample of South African men. AIDS care, 19(1), 20–27.

504 C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505

Knio, R., & Elasmar, M. G. (1998). Factors influencing gender role attitudes among Lebanese youth. RetrievedJanuary 28, 2009, from http://list.msu.edu/cgi-bin/wa?A2=ind9811d&L=aejmc&F=P&P=4276.

Kremer, J., & Curry, C. (1987). Attitudes toward women in Northern Ireland. The Journal of Social Psychology,127(5), 531–533.

Krishnan, V. (1991). Abortion in Canada: Religious and ideological dimensions of women’s attitudes. Social Biology,38, 249–257.

Lee, Y. T., Kleinbach, R., Hu, P. C., Peng, Z. Z., & Chen, X. Y. (1996). Cross cultural research on euthanasia andabortion. Journal of Social Issues, 52(2), 131–148.

Lindell, M. E., & Olsson, H. M. (1993). Swedish students’ attitudes toward abortion. Health Care for WomenInternational, 14, 281–291.

Loo, R., & Thorpe, K. (1998). Attitudes towards women’s roles in society: A replication after 20 years. Sex Roles,39(11–12), 903–912.

Loo, R., & Thorpe, K. (2005). Relationships between attitudes toward women’s roles in society, and work and lifevalues. The Social Science Journal, 42, 367–374.

Lottes, I. L., & Kuriloff, P. J. (1992). The effects of gender, race, religion, and political orientation on the sex roleattitudes of college freshmen. Adolescence, 27, 675–688.

Mhlanga, R. E. (2003). Abortion: Development and impact in South Africa. British Medical Bulletin, 67,115–126.

Miller, A. S., & Hoffman, J. P. (1995). Risk and religion: An explanation of gender differences in religiosity. Journalfor the Scientific Study of Religion, 34(1), 63–75.

Modi, D. (2002). Factors influencing women’s attitudes towards abortion at the University of Pennsylvania.Perspectives in Psychology, 5, 42–50. Retrieved on 20 February 2008 from http://bespin.stwing.upenn.edu/∼upsych/Perspectives/2002/Modi.pdf.

Nelson, E., & Coleman, P. (1997). Attitudes towards the level of men’s involvement in abortion decisions. Journalof Humanistic Education & Development, 35(4), 217–224.

Öngen, D. (2006). Attitudes towards women: A study of gender and academic domain differences in a sample ofTurkish university students. Social Behavior and Personality, 34(5), 467–486.

Patel, C. J., & Myeni, M. C. (2008). Attitudes to abortion in a sample of South African female university students.Journal of Applied Social Psychology, 38, 736–750.

Patel, C. J., & Kooverjee, T. (in press). Abortion and contraception: Attitudes of South African university students.Health Care for Women International.

Patel, C. J., Ramgoon, S., & Paruk, Z. (in press). Exploring religion, race and gender as factors in life satisfactionand religiosity of young South African adults. South African Journal of Psychology.

PROVIDE (2005). A profile of KwaZulu-Natal: Demographics, poverty, inequality and unemployment. PRO-VIDE Background Paper Series, 1(5), 1–21. Retrieved on 8 December 2008 from http://www.elsenburg.com/provide/documents/BP2005 1 5%20Demographics%20KZ.pdf.

Read, J. G. (2003). The sources of gender role attitudes among Christian and Muslim Arab-American women.Sociology of Religion, 64, 207–222.

Roothman, B., Kirsten, D., & Wissing, M. (2003). Gender differences in aspects of psychological well being. SouthAfrican Journal of Psychology, 33(4), 212–218.

Rule, S. (2002). Spirituality in South Africa: Christian beliefs. In HSRC public attitudes in contemporary SouthAfrica: Insights from an HSRC survey. Cape Town: HSRC Press.

Rule, S. (2004). Rights or wrongs? Public attitudes toward moral values. Human Sciences Research Council Review,2(3), 4–5.

Sahar, G., & Karasawa, K. (2005). Is the personal always political? A cross-cultural analysis of abortion attitudes.Basic and Applied Social Psychology, 27, 285–296.

Sherman, P. J., & Spence, J. T. (1997). A comparison of two cohorts of college students in response to theMale–Female Relations Questionnaire. Psychology of Women Quarterly, 21, 265–278.

South African youth are optimistic about our future (2005). Retrieved on 25 October 2005 from www.sagoodnews.co.za/faith/451615.htm.

C.J. Patel, L. Johns / The Social Science Journal 46 (2009) 493–505 505

Spence, J. T., & Hahn, E. D. (1997). The Attitudes Toward Women Scale and attitude to change in college students.Psychology of Women Quarterly, 21, 17–34.

Stephanus, F. (2006). The relationship between sex role orientation and rape victim blame among police officers inthe Cape Peninsula. Unpublished master’s mini-thesis. Bellville, South Africa: University of the Western Cape.

Stets, J. E., & Leik, R. K. (1993). Attitudes about abortion and varying attitude structures. Social Science Research,22, 265–282.

Stevens, S. R., Caron, S. L., & Pratt, P. (2003). Decade in review: The importance of religion in shaping the sexualattitudes of college students in the 1990s. Journal of College and Character, 2. Retrieved on February 2, 2004,from http://www.CollegeValues.org/articles.cfm.

Toller, P. W., Suter, E. A., & Trautman, T. C. (2004). Gender role identity and attitudes toward feminism. Sex Roles,51(1–2), 85–90.

Tu, S., & Liao, P. (2005). Gender differences in gender-role attitudes: A comparative analysis of Taiwan and CoastalChina. Journal of Comparative Family Studies, 36(4), 545–566.

Twenge, J. M. (1997). Attitudes toward women, 1979–1995. Psychology of Women Quarterly, 21, 35–51.Varga, C. A. (2002). Pregnancy termination among South African adolescents. Studies in Family Planning, 33,

283–298.Wall, S. N., Frieze, I. H., Ferligoj, A., Jarosova, E., Pauknerova, D., Horvat, J., et al. (1999). Gender role and religion

as predictors of attitude toward abortion in Croatia, Slovenia, the Czech Republic, and the United States. Journalof Cross-Cultural Psychology, 30(4), 443–465.

Walzer, S. (1994). The role of gender in determining abortion attitudes. Social Science Quarterly, 75(3), 687–693.Wilcox, C., & Jelen, T. G. (1991). The effects of employment and religion on women’s feminist attitudes. Interna-

tional Journal for the Psychology of Religion, 1, 161–171.Zantsi, K., Pettifor, A., Madikizela-Hlongwa, L., MacPhail, C., & Rees, H. (2004). Religiousness and sex-

ual behaviour among South African youth. Retrieved on 1 October 2007 from http://gateway.nlm.nih,gov/MeetingAbstracts/102284317.html (Abstract).