Boy/Girl Friend and Virginity Values, and Stigma Related to Condom Among Jimma University Students

Background Individual factors such as the value given to virginity and boy/ girl friend, and stigma attached to condom can strongly affect success in preventing early sexual initiation and in using condom consistently. To the best knowledge of the authors, no literature was available in Ethiopia on these issues until the time of this study. The objective of this study was to assess the value given to virginity, boy/ girl friend, and stigma related to condom access and use. Methods A cross-sectional survey was conducted on a random sample of 1986 students in May 2009 in Jimma University using both qualitative and quantitative techniques. Data were collected using piloted and pre-coded questionnaire and from six focus group discussions. Quantitative data were analyzed using SPSS for windows version 13 where descriptive statistics, ANOVA, and t-test were computed. P-value less than 0.05 was considered statistically significant and effect size was measured in Eta squared. The qualitative data findings were triangulated with the quantitative ones. Results Of 1986 respondents, 1612 (81.2%) were males, 365 (18.4%) females and 9 (0.4%) their gender not indicated. The age of respondents ranged from 17– 45 years with median of 20 years. Virginity value-scores were significantly lower among females (p< 0.001, Eta squared= 0.023). In contrast to many males, most females appeared to be not concerned about virginity during the focus group discussions. Many respondents of both genders reported that boy/girl friend is very important in campus life. Although the stigma to condom was slightly higher among females (p< 0.001, Eta squared= 0.009), most respondents of both genders had a stigmatizing attitude. Conclusion Lower virginity value among females with high value given to boy/girl friend appeared to indicate the liberalization of sex. Liberalization of sex and stigma to condom were occurring together. Sex educators targeting university students should focus on problems of liberalized unprotected sex in relation to success in life.


INTRODUCTION
Virgin ity is the practice of voluntarily refrain ing fro m some or all aspects of sexual activity for reasons such as chastity, prevention of unwanted pregnancy and sexually transmitted infections (STIs), or psychosociological reasons such as negative past experiences (1).
Early in itiation of sexual activ ity is linked to higher numbers of non-marital sex partners, minimal condom use, increased rates of STIs, increased rates of out-ofwedlock pregnancy and birth, increased single parenthood, decreased marital stability, increased maternal and child poverty, increased abortion, increased depression, and reduced happiness (2)(3)(4)(5)(6).
These consequences incur a high cost to society (7).
The available literatures in the area strongly recommend that sexual behavior among in-school youth and adolescents should be based on the ability to delay sexual init iation and to use condom consistently when abstinence is not possible. Success in these two strategies can be strongly affected by individual factors such as the value given to virginity and boy/ girlfriend, and stig ma attached to condom (2,3). However, to the best knowledge of the authors, no literature was found in Ethiop ia on the value given to virgin ity, boy/ girl friend and stigma to condom up until this study has been undertaken.
The purpose of this study was to address the aforementioned gap in the area. Findings of this study would enable reproductive health educators to select contents of sex education that do not conflict with values of their target , and guide health educators if the need arises to develop values that enhance the overall develop ment of sexual values among in-school adolescents and youth.

SUBJ ECTS AND METHODS
This cross-sectional survey was conducted in Jimma University, main campus in May 2009 using both qualitative and quantitative techniques on 1986 randomly selected regular undergraduate students. The same study subjects as Ambaw et al. (8) were used in this study. To determine the samp le size, a single population proportion formu lae was used taking 95% confidence interval and 3% margin of error. The variable considered for samp le size determination was the estimated prevalence of sexual intercourse among the source population, 32%. This variable was selected because it maximized the sample size. A contingency of 10% and a design effect of 2 were used. This resulted in a total sample size of 2044. Single stage cluster sampling technique was used to select 31 lecture classes of students using lottery method among a total of 206 clusters and all the members in the selected clusters were considered for the study. The qualitative part of the study included six focus groups of 7-10 purposively selected self-exp ressive participants who were not included in the quantitative section. Students fro m both gender and year level were included.
The dependent variables of the study were virg inity value, boy/ girl friend value, and stigma related to condom. The dependent variables were measured using five-point Likert scale developed by the investigators. There were 5 items in the virginity value scale, 6 items in the boy/ girl friend value scale, and 4 items in the stigma related to condom acess and use scale. The items were totaled in each of the scales to get an interval level of measurement as recommended by Streiner and Norman for data that is not severely skewed (9). Normality of distribution was checked by observing expected normal probability plots rather than histograms and Kolmogorov-Smirnov tests as recommended by Tabachnick and Fidell for large sample size (10).
Professionals in the field of sociology and psychology were consulted to assure content validity. Internal consistency of the scales was checked using Cronbach's alpha test after administering on a randomly selected class of 40 students before the actual study (these were later excluded fro m the main study). On the 40 sample of students the virginity value scale had a reliability of 0.80, the boy/ girlfriend value scale had a reliability of 0.83, and the stigma related to condom scale had a reliab ility of 0.81. Independent variables of the s tudy included sociodemographic characteristics such as age, sex, ethnicity, religion, and church/mosque attending habit, education year level, faculty, and place of origin (rural or urban). Independent variables were measured using self reports on a questionnaire developed by the investigators.
The quantitative data were collected by trained Jimma University instructors using a self-administered, piloted, and precoded questionnaire prepared in simple English in classrooms arranged for that purpose. Supervision was made to assure independence of responses and the questionnaires were collected back immed iately after co mpletion by the respondents. The qualitative data were collected by two trained and experienced facilitators in each focus group who used tape recorders and took notes to document the discussion. Each focus group took 40-75 minutes depending on the range of ideas generated by the study participants.
Gu iding questions used for the focus group discussion were: Is virg inity important? How? Is boy/ girlfriend important in a campus life? How? What do you feel if you see a student in your campus carrying condom or if you are seen by others carrying condom? The quantitative data were analyzed using SPSS for windows version 13.0. The independent-samples ttest, one-way between groups ANOVA with post hoc Tukey tests, and descriptive statistics were computed. Normality of distribution and equality of variance were checked before t-test and ANOVA tests were computed. Effect sizes of t-test and ANOVA tests were measured in Eta squared values. A p-value of less than 0.05 was taken as statistically significant. The data from the qualitative part of the study was transcribed, coded and categorized, and finally triangulated with the quantitative findings.
The proposal was reviewed and ethically cleared by Jimma University and info rmed consent was obtained fro m every one of the participants before data collection. The following operational defin itions were used:  Virginity: A condition of remain ing totally abstinent from sexual intercourse (penis to vagina sex).  Virginity value : Belief about virginity that serves as a guideline to help individuals make decisions whether to remain virgin or to get a virgin partner. High score on the virgin ity value scale shows high value to virgin ity and vice versa.  Boy/ girlfriend: Opposite sex friend with whom one has sexual relat ionship.  Boy/girlfriend value : A belief about boy/girl friend that serve as a guideline to help individuals make decisions about whether to have a boy/girl friend or not. High score on the boy/girl friend value scale shows high value given to boy/ girl fiend in campus life and vice versa.  Stigma related to condom: A negative attitude towards people carrying condom with them. High score on the condom stigma scale shows high stigma to condom and vice versa.

RES ULTS
The detail background characteristics of study participants had been reported in another article (8).
Virgin ity value had normal d istribution with a med ian value of 19 with scores ranging from 5-25. The independent samples t-test analysis showed that the mean virgin ity value-scores were significantly h igher among males (p< 0.001, Eta squared= 0.023), those who had the habit of frequently attending church/ mosque (p< 0.001, Eta squared= 0.017), and those who had rural origin (p< 0.001, Eta squared= 0.029) compared to their counter parts.
Findings from the qualitative part of the study were also more or less similar. For examp le a fourth year male student made the following remark wh ile expressing the concern he has for the virginity of his anticipated partner: (p= 0.003, Eta squared= 0.01), age groups (p< 0.001, Eta squared= 0.008), and religions (p< 0.001, Eta squared= 0.018). Post hoc comparisons using Tukey test showed that the mean virginity value score among Oro mos was significantly higher than Amharas, and the mean score of A mharas was in turn significantly higher than Gu rages. Students of public and med ical faculties had significantly higher mean v irgin ity value scores compared to students of humanities and social sciences faculty. Students in the age group 20-24 years had higher mean virginity value score than in the age group 17-19 years. Likewise, Muslims and Protestants had higher mean virginity value scores compared to Orthodox Christians. All other differences observed in the one-way between-groups analysis test were not statistically significant (table 1). The second category is the category that emphasized the disadvantages of a boy/ girl friend including wasting time, bro ken heart as a result of bro ken relationships (main ly the males were reported to break the relationship), loss of freedom (the boy/ girl fiend has to permit to go to parties, even to go outside the campus), and decreased social interaction with other students. A fourth year male student described the likelihood of breaking a relationship in the university because of the high possibility of getting new sexual partners as: Seblewongel and Bezabih in the book "Love unto Crypt" ("Fikir Eske Mekabir," the title of a well known Amharic Novel written by Hadis Alemayehu, Ethiopia) loved each other until their death because they had no choices. But in this university, there are too many Bezabihs for Seblewongel and too many Seblewongels for Bezabih. (Many smiles and noddings). Results from the one-way ANOVA also showed that level of boy/ girl friend value had a statistically significant difference among groups of different educational year levels (p< 0 .001, Eta squared= 0.018), relig ions (p< 0.001, Eta squared= 0.013), and age groups (p= 0.002, Eta squared= 0.007). Post hoc analysis using Tukey test showed that the mean score of boy/ girl friend value was significantly lower among students in the age group 17-19 years compared to 20-24 years or 25-45 years. But there was no statistically significant difference between age groups 20-24 years and 25-45 years. Orthodox Christians and Protestants had significantly h igher boy/ girl friend values compared to Muslims, and year-three and above students had higher boy/ girl friend value compared to year-one and year-two students. But there was no statistically significant difference between year-one and year-two or yearthree and year-four and above. Boy/ girl friend value scores did not show statistically significant difference among the faculties (table 2).
Stig ma related to condom had a normal distribution with med ian 9 and scores ranging from 4-20. The independent-samples t-test analysis showed that the mean scores of condom related stigma were significantly higher among females (p< 0.001, Eta squared=0.009), those who had the habit of frequently attending church/ mosque (p= 0.029, Eta squared= 0.002), and those who had urban origin (p= 0.005, Eta squared= 0.004) co mpared to their counter parts.
Generally, all females and most males expressed their negative outlook to wards condom in the focus group discussions. A second year female student expressed her strong disapproval of carrying condom as: If I see a female student carrying condom with her, just…it is like a priest carrying pistol with him instead of a cross. If I see my boy friend carrying condom, I suspect that he is looking for other girls. A third year male student also added: If you are seen with condom, you are caught-redhanded. If you have a girl friend, your relationship will be spoiled immediately. Some male participants had the view that the possession of condom is not as such a seriously stigmatized act. For examp le, a fourth year male student had described his view as: Although there are students who discriminate those who carry condom with them, there are also others who do not take it that much seriously except some light critics designed for making fun. People of urban origin are of this type. I have some friends who carry condom with them; I never feel bad about them. I know that sexual practice is much more rampant than is usually reported. The one-way between-groups ANOVA test showed that stigma related to condom did not have a statistically significant difference among students of different educational year level (p= 0.111), ethnicities (p= 0.138), and facult ies (p= 0.728) (table 3).

DISCUSS ION
The purpose of the study was to assess the value University students give for virginity and boy/ girl friend, and the stigma they have for condom. It has not investigated factors that influence values and attitudes other than socio-demographic characteristics of the respondents. Many of the socio-demographic factors are known to be unlikely or less likely to be changed with interventions.
However, describ ing an attribute in terms of background characteristics of the study subjects is a necessary step for further understanding. This has been done using a large random sample and strong statistical techniques. Qualitative techniques were combined with focus group discussion to elucidate the perceptions underlying the values and attitudes.  Male gender, rural origin , and church/ mosque attending habit were found to be associated with higher virgin ity value. Tradition in many parts of Ethiopia and religious teachings discourage sexual intercourse before marriage. The majority of males participated in the focus groups stressed two main points concerning the virginity of their anticipated sexual partners: girls retain p leasant memories fro m their past relationships, and the possibility of continuing with previous relationships. According to Co x FD, it is very hard to predict what the effects of non-marital sexual relat ionships will have on marital sexuality (11). Concerning their own virginity, some males confessed that the strong sexual feeling makes them practice sex in contrary to their value. Following puberty the sexual desires of males is known to be strong to the extent that inaccessibility stops them (11,12). Males of rural origin believed that sexual practice weakens academic performance, and it is not acceptable fro m religious point of view. Non-marital sexual involvement is known to reduce intellectual, social, and other areas of involvement (11).
Some males during the focus group were in favor of sexual intercourse before marriage for they believed premarital sex is necessary to rule out sexual incompatib ility and females want experienced boys. These findings show the typical myths related to sexual intercourse (13). There is sufficient evidence that most sexual in-co mpetencies are psychological (11,12) and over all sexual satisfaction is determined by one's mental and emotional processes rather than the tactile maneuvers. It is not compatibility with sex that leads to love; it is the presence of love that makes sex co mpatible (11). The tendency of females to be less concerned for their virgin ity and the virginity of their anticipated partners indicates the liberalization of sex. The freedo m to move up socio-economically, to choose a partner, and to leave unsatisfactory relationships is intertwined with increased non-marital Male students with no church/ mosque attending habit, and urban origin had higher values for boy/ girl friend compared to their counter parts strengthening the evidence of sex liberalizat ion. Gender had very small effect size in the quantitative findings and it was not man ifested in the qualitative findings. The advantages of having a boy/ girl friend wh ich were manifested during the focus group discussions were its being source of pride, its being source of confidence, support and respect. The disadvantages were wastage of time, psychological trau ma when the relationship breaks; loss of freedom imposed by partner, and decreased social interactions. These are common problems of a relationship based on only sexual co mmit ment rather than commit ment to a total relationship (11).
Students below the age of 20 or below the third year educational level had lower value to boy/ girl friend. A study on other sexual behaviors in Jimma University had similar findings (8,13,14). The find ing that Muslims had lower values to boy/girl friend than Christians needs further study. A previous study had shown that Muslims were mo re absolutist than hedonistic in their sexual values (14). Boy/girl friend value did not vary among the different faculties showing that it was a common value.
Female gender, church/ mosque attending habit, and urban origin had shown a slightly higher level of stigma related to condom compared to their counter parts. Social expectations and religious teachings that emphasize chastity which indirectly crit icize condom possession and/or use could be responsible. All the focus group participants equated possessing condom with practicing sex whether they stigmatized it or not. For example, those participants who feel good when they possess condom feel so because they perceive that others would conclude they are practicing sex. The slightly higher stig ma related to condom among the urban students may be caused by a reporting bias in the direction of mass media teachings among the rural students. Sexual intercourse with out condom was 44% more likely among students of rural o rig in compared to urban students (8). In the focus group discussion, students of urban origin were less stigmatizing the possession of condom. The absence of difference in the level of stig ma with educational year level, ethnicit ies, and faculties may show that the factors affecting the stigma are similar among these different groups.
To conclude, in contrast to many males, most females do not value virginity. Many students of both genders feel that a boy/ girl friend is very important in the campus life. These are indicators of the liberalization of sex. Yet, most of the respondents stigmat ize condom indicating that making condom accessible to university students must be complemented with efforts to minimize the stigma related to it. We also recommend that sex educators targeting university students should focus on problems of liberalized, unprotected sex in relat ion to success in life.