Sexual Activity Among Zambian Female Teenagers: The Role Of Interpersonal Skills

Sexual Activity Among Zambian Female Teenagers: The Role Of Interpersonal Skills

Vijayan K. Pillai

One of the characteristics of African fertility is that the transition to motherhood occurs at an early age. African adolescents are more likely to become mothers than are Asian or Latin American adolescents. Most African countries have fertility rates higher than 10% per year for the 15- to 19-year-old female population. In Nigeria, for example, nearly two thirds of all females 20 years old and below have given birth at least once (United Nations, 1989).

In Zambia, social control over sexual relationships varies across ethnic groups. Among the Bemba, a large ethnic group, boys and girls are often separated at the age of eight or nine. Early sexual socialization takes place within same-gender peer groups under the supervision of elders. Traditionally, the Bemba have encouraged marriage before girls reach puberty. However, Zambian customs have been changing (United Nations Economic Commission for Africa, 1984). In particular, the age at marriage has steadily increased.

Two factors are associated with the recent changes in marriage patterns in Africa. First, increases in the number of years of schooling among African teenagers (Achola & Kaluba, 1989) and in the proportion of females attending school have contributed to the postponement of marriage. Second, mass schooling has had an effect on traditional family values and attitudes toward sex (Caldwell, 1982; Schuster, 1979). Thus, changes in education and family values have considerably increased marital age in many African countries, such as Zambia.

Further, because of greater participation in educational and recreational activities, African adolescents are increasingly away from the direct supervision of parents and other elders. In addition, the age at menarche is slowly decreasing, mainly due to improved nutrition. These developments, along with later age at marriage, have given adolescents greater opportunity for premarital sexual activity and have increased the chances of pregnancy.

During adolescence, a large number of factors may influence the initiation of sexual activity. One is adolescents’ ability to cope with changing body size and shape (Hetherington, Lerner, & Perlmutter, 1988; Lerner & Lerner, 1987). In particular, females have to develop interpersonal skills (particularly assertiveness) that will allow them to avoid unwanted sexual involvement (Schinke & Gilchrist, 1977). However, the importance of teaching adolescents skills related to coping and assertiveness is not well recognized by African policy makers.

Several theoretical approaches have focused on the role of interpersonal skills in preventing problem behaviors. For example, social learning theory (Bandura, 1986; DiBlasio & Benda, 1990) suggests that adolescents who can effectively implement their decisions are less likely to engage in problem behaviors. Studies have shown that primary prevention strategies, such as sex education, are more likely to be successful when complemented by decision-making and assertiveness skills (Barth, 1989; Kirby, 1985).

The present study examined the relationship between interpersonal skills (assertiveness) and sexual activity among female adolescents in Zambia. Three important factors associated with courtship behavior were controlled: emotional involvement, participation in traditional initiation ceremonies, and boyfriend’s age (Furstenberg, 1976; Zelnik, Kantner, & Ford, 1981; Lema, 1990; Thornton, 1990; Miller, McCoy, & Olson, 1986; Spanier, 1977; Oppong, 1983). The dependent variable, sexual activity, was divided into coital and noncoital behaviors.


The sample consisted of unmarried females between the ages of 13 and 19. Several studies on biosocial changes have suggested this age range for adolescence (see Flack, 1971). The females were drawn from seven randomly selected secondary schools (forms 3 through 7) in two large urban areas of Zambia (Copperbelt and Lusaka Central Provinces). Approximately 20% of the female students in each form were randomly selected.

A self-administered questionnaire was used to collect data. The questionnaire was pretested in an urban secondary school in Kabompo District. A few items were subsequently modified in order to enhance clarity.

Information on sexual activity during the previous two months was obtained. Frequency of coital and noncoital activities was measured on a 6-point scale, with responses ranging from very often to did not engage in the activity. Three groups were distinguished: those who engaged in noncoital activities only, those whose activities included intercourse, and those who did not have any sexual contact at all (the reference group). In the first model, the coital group was compared with the reference group. In the second model, the noncoital group was compared with the reference group.

There were 85 female teenagers in the reference group, 263 in the coital group, and 82 in the noncoital group. These were cases where no relevant data were missing. Listwise deletion was used to drop cases.

Degree of involvement with a boy consisted of three variables (FEEL, LIKEDOING, and ONEFORME), with three categories for each: agree, neutral, and disagree. The age of the boy (BOYAGE) was divided into four categories: very young, young, old, and older. Intention to have an initiation ceremony (CERPRESS) was divided into three categories: those who intended to have a ceremony in the near future (yes), those who did not intend to have an initiation ceremony (no), and those who already had an initiation ceremony (conducted).

Assertiveness was measured using two items. Adolescents were asked how difficult it is for them to say no to an unexpected sexual advance (SAYNO), with responses grouped into two categories: difficult and not difficult. They were also asked how true it is that they find it difficult to say no to friends because they do not like to hurt their friends’ feelings (NOHURT): very true, true, somewhat true, or not true. It was hypothesized that females who are unable to assert themselves are more likely to engage in sexual activity than are those who have the ability to assert themselves.


A general linear model was appropriate for testing the hypothesis. Multiple classification analysis (MCA) was chosen over ordinary least squares techniques (Andrews et al., 1974). The MCA framework presents the coefficients of the categorical variables in terms of gross and net deviations from the grand mean. The category mean of each variable is the sum of the grand mean and gross deviation of each category. The net category means are the sum of the grand mean and the deviation of each category mean after controlling for other variables in the model.

In addition, the MCA technique yields eta and beta coefficients. The eta coefficient is identical to the multiple correlation coefficient when the dependent variable is regressed on all the categories of an independent variable. The beta coefficient is identical to the multiple partial correlation coefficient obtained when controlling for other variables in the model. (When the dependent variable is highly skewed, MCA results may yield values outside those permitted by probability. This problem is often encountered when any one of the dependent variable categories has less than 20% of the cases [Maddala, 1977]. In this study, none of the dependent variable categories had less than 20%).

Table 1 presents the results of the MCA analysis. It provides the parameter estimates from two separate models: coital and noncoital activities. SAYNO was significant at the .01 level across the two models, while NOHURT was significant at the .01 level only for the coital model.

Regarding the effect of SAYNO on the likelihood of coital activity, the difference between gross and net deviation of the “difficult” category was small, about 2%. The difference between eta and beta was also small. Thus, the effect did not change much when control variables were introduced.

In the first model, the percentage of females who engaged in coitus, were emotionally attached to their boyfriends, and found it difficult to say no was very high, about 90% (80 + 2 + 8; grand mean + net deviation of the “agree” category for FEEL + net deviation of the “difficult” category for SAYNO). The percentage of females who engaged in coitus, were only somewhat attached emotionally (neutral), and found it difficult to say no was lower, about 77% (80 – 11 + 8). Thus, greater emotional involvement slightly enhanced the odds of intercourse for adolescents who had trouble saying no to an unexpected sexual advance.

In the second model, the effect of SAYNO increased when other variables were controlled. In addition, the effect of FEEL on the likelihood of sexual activity was slightly stronger in the noncoital model. Approximately 66% of the females who had engaged in noncoital activities and were emotionally attached to their boyfriends found it difficult to say no (44 – 1 + 23). About 46% of the females who had engaged in noncoital activities and were somewhat attached emotionally (neutral) found it difficult to say no (44 – 21 + 23).

In sum, the two indicators of assertiveness, SAYNO and NOHURT, were associated with coital activity, and SAYNO was associated with [TABULAR DATA FOR TABLE 1 OMITTED] noncoital activity. The results suggest that adolescent females who lack assertiveness tend to be more vulnerable to sexual activity, and the likelihood of intercourse is considerably enhanced. Further, assertiveness not only is necessary for enforcing the decision to abstain from sexual activities, it also may be crucial should the adolescent decide to have sex. Contraceptive decisions often follow, and if the adolescent does not choose hormonal methods, the use of other means of birth control demands cooperation from her partner. Thus, assertiveness would be necessary for effective contraception.


In Zambia, there are few programs directed at the prevention of teenage pregnancy. Adolescents who become pregnant are expelled from school and provided with little support. Needed are pregnancy prevention approaches that involve the family, school, community, and peers, impacting adolescents’ behavior at the intrapersonal, interpersonal, and social levels. However, the influences of intrapersonal and social variables on adolescent pregnancy have received greater attention than have the effects of interpersonal factors. The findings of the present study suggest that, in Zambia, female adolescents’ inability to say no is significantly associated with the occurrence of sexual activity.

As the country modernizes, the interaction of young men and women, without family supervision, is likely to increase. Thus, they will need greater interpersonal skills. Unfortunately, the skills needed to negotiate personal relationships are not being taught. Comprehensive pregnancy prevention programs must therefore include efforts to increase the capacity of adolescents to assert themselves in situations involving sexual activity and contraception.


Achola, P., & Kaluba, H. (1989). School production units in Zambia: An evaluation of a decade of presidential experiment. Comparative Education, 25, 165-180.

Andres, F. M., Morgan, J., Sonquist, J. A., & Klem, L. (1974). Multiple classification analysis. Ann Arbor, MI: Institute of Social Research, University of Michigan.

Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs, NJ: Prentice-Hall.

Barth, R. P. (1989). Reducing the risk: Building skills to prevent pregnancy. Santa Cruz, CA: Network Publications.

Caldwell, J. (1982). Theory of fertility decline. New York: Academic Press.

DiBlasio, F. A., & Benda, B. B. (1990). Adolescent sexual behavior: Multivariate analysis of a social learning model. Journal of Adolescent Research, 5, 450-467.

Flack, R. (1971). Youth and social change. Chicago: Markham.

Furstenberg, F. (1976). Unplanned parenthood: The social consequences of teenage childbearing. New York: Free Press.

Hetherington, E. M., Lerner, R. M., & Perlmutter, M. (Eds.). (1988). Child development in life span perspective. Hillsdale, NJ: Lawrence Erlbaum Associates.

Kirby, D. (1985). The effect of selected sexuality education programs: Toward a more realistic view. Journal of Sex Education and Therapy, 11, 28-37.

Lema, V. (1990). Sexual behavior, contraceptive practice and knowledge of reproductive biology among adolescent school girls in Nairobi, Kenya. East African Medical Journal, 67, 87-94.

Lerner, R. M., & Lerner, J. V. (1987). Children in their contexts: A goodness of fit model. In J. B. Lancaster, J. Altmann, A. S. Rossi, & L. R. Sherrod (Eds.), Parenting across the life span: Biosocial dimensions (pp. 377-404). Hawthorne, NY: Aldine.

Maddala, G. S. (1977). Econometrics. New York: McGraw-Hill.

Miller, B., McCoy, J., & Olson, T. (1986). Dating age and stage as correlates of adolescent sexual attitudes and behavior. Journal of Adolescent Research, 1, 361-371.

Oppong, C. (Ed.). (1983). Male and female in West Africa. Boston: Allen and Unwin.

Schinke, S., & Gilchrist, L. (1977). Adolescent pregnancy: An interpersonal skill training and approach to prevention. Social Work in Health Care, 3, 159-167.

Schuster, I. G. (1979). New women of Lusaka. Palo Alto, CA: Mayfield.

Spanier, G. B. (1977). Sources of information and premarital sexual behavior. Journal of Sex Research, 13, 73-88.

Thornton, A. (1990). The courtship process and adolescent sexuality. Journal of Family Issues, 11, 239-273.

United Nations. (1989). Adolescent reproductive behavior: Evidence from developing countries (Vol. 2). New York: United Nations.

United Nations Economic Commission for Africa. (1984). Interrelationships among infant and childhood mortality: Socioeconomic factors and fertility in Zambia. Lusaka: Central Statistical Office.

Zelnik, M., Kantner, J., & Ford, K. (1981). Sex and pregnancy in adolescence. Beverly Hills, CA: Sage.

Thomas R. Barton, Ph.D., Associate Professor, Department of Social Work, University of North Texas. itudes, and b haviors betwe

COPYRIGHT 1999 Libra Publishers, Inc.

COPYRIGHT 2001 Gale Group