Reproductive and Contraceptive Knowledge, Contraceptive Self-Efficacy, and Contraceptive Behavior among Teenage Women
Levinson, Ruth Andrea, Adolescence
The relationship between reproductive and contraceptive knowledge (RCK) and contraceptive behavior has been studied extensively in an effort to address coitally active teenagers' ineffective patterns of contraceptive use. Educators have explored the relationship between RCK and contraceptive behavior because it is direct and efficient in terms of instruction and assessment. Unfortunately, providing information to adolescents on sexual and contraceptive matters does not necessarily assure that they will become better contraceptors or alter their sexual behavior. Many studies in the early assessment of the impact of sex education showed that educational programs that focus on facts about reproduction and contraception could increase knowledge, but demonstrated inconsistent and inconclusive effects on sexual behavior and increased contraceptive usage (Hansson, Jones, & Chernovetz, 1979; Kirby, 1984; Marsiglio & Mott, 1986; Whitley & Schofield, 1986; Zelnik & Kim, 1982).
Despite these ambiguous past findings, some educational programs have had an impact on teenagers' contraceptive and sexual behavior (Eisen, Zellman, & McAlister, 1990; Howard & McCabe, 1990). These exceptions suggest that exploring the relationship between sex education, RCK, and contraceptive behavior should not be abandoned. Rather, we may need to illuminate the factors that distinguish successful sex education programs from those which did not affect behavior. On examining those educational programs which have had an impact on behavior, we find they have focused on motivational factors which may affect teenagers' initiation into sexual intercourse and contraceptive use. Components of effective programs have been the development of a saliency of the probability and consequences of pregnancy as well as the provision of incentives to avoid it. Other components that have affected behavior change have been skills training in: (1) communication about the use of contraceptives and utilization of contraceptives; (2) linkages and utilization of clinics for health care services; and (3) sexual decision making (Bilodeau, Forget, & Tetreault, 1991; Carrera, Baker, & McCombs, 1978; Eisen, Zellman, & McAlister, 1990; Polit & Kahn, 1985; Schinke, 1984; Schinke & Gilchrist, 1977; Zabin, Hirsch, Smith, Streett, & Hardy, 1986). A commonality among these programs is that they have all utilized a RCK base, but have also focused on making salient the consequences of unprotected sexual relations (emotional factors) and on practicing domain-specific skills training (behavioral factors). Information was delivered within a social-psychological and behavioral context.
The success of these programs seems to indicate that when information is delivered in a social-psychological context rather than in a cognitive context alone, it is more valued and more likely to be used to activate effective contraceptive or prophylactic behavior. By contrast, the traditional sex education models prevalent in public education have used a didactic approach to delivery. The assumption underlying this information-processing approach is that behavior will change if based on cognitive input. Since information-alone models have not resulted in behavior change, the social-psychological approaches which utilize contraceptive-specific information may prove more successful.
In a previous study, self-efficacy theory was applied to analyze teenage women's sexual and contraceptive behavior. A domain-specific theory of behavior change (Bandura, 1977), it was operationalized in an 18-item scale as Contraceptive Self-Efficacy (CSE). CSE scores distinguished patterns of contraceptive use among teenage women when the effects of influential demographic, sexual experience, and psychosocial variables were controlled (Levinson, 1986). In two subsequent studies with samples that were demographically diverse, the relationship between CSE and contraceptive behavior was confirmed (Bilodeau et al., 1991; Levinson & Jaccard, 1991). In the current research, the relationships among teenage women's CSE, reproductive and contraceptive knowledge (RCK), and contraceptive behavior was investigated in two samples in order to determine whether CSE is related to RCK for both samples and whether CSE might influence the way knowledge translates into behavior.
The CSE scale assesses motivational barriers to contraceptive use among sexually active teenage women across 18 situational statements [ILLUSTRATION FOR FIGURE 1 OMITTED]. According to the self-efficacy construct, a woman's expectations as to whether she can and should exercise a component behavior will determine initiation and persistence in achieving a desired goal. The CSE scale measures the strength of a sexually active teenager's conviction that she can and should control sexual and contraceptive situations.
In the initial CSE study (Levinson, 1986), no relationship was found between RCK and contraceptive behavior. However, there was a relationship between knowledge and CSE, and between CSE and contraceptive behavior. Respondents with high CSE were conceptually compared to those who anticipated a risk of disease or pregnancy and were highly motivated to use contraceptives. It was posited that high CSE might be correlated with high knowledge scores because individuals who feel they can and should use contraceptives may be more receptive to information. Rosenthal, Moore, and Flynn (1991) have postulated along with other researchers that "A lack of a sense of competence in the sexual domain may be one explanation of the worrying gap between adolescents' knowledge about or intention to engage in safe sex and their actual behavior" (Turtle, Ford, Habgood, Grant, Bekiaris, Constantiou, Maeck, & Polyzoidis, 1989; Kashima, Gallois, Hills, & McCamish, 1991). In the current research this notion was tested to determine if interaction effects were operating between CSE, knowledge (RCK), and contraceptive behavior.
A secondary line of analysis looked at the individual responses on the 7-item Index of Reproductive and Contraceptive Knowledge (IRCK) which had a multiple choice format. Both correct and incorrect responses were examined in order to ascertain common knowledge and myths among teenage women. It was assumed that both correct and incorrect information had the potential to strongly affect behavior.
Research has shown that teenagers tend to utilize peers as their primary source of sex information (Banks & Wilson, 1989; Haas, 1979; Kallen, 1982; Spanier, 1977; Thornburg, 1981). Further, it has been documented that common myths prevail among teenagers concerning the conditions which affect whether a woman can become pregnant and how to use over-the-counter contraceptive methods (Coles & Stokes, 1985; Haas, 1979; Hayes, 1987; Zelnik & Kantner, 1977, 1980). It is important to be aware of these myths since they may seriously compromise teenagers' effective use of contraceptive methods (e.g., withdrawal, rhythm, condoms, and foam). In addition, since as noted, teenagers typically perceive friends to be more reliable sources of sex information than are teachers and other adults, they may feel they "know it all." Thus, they may attend sex education classes with only "half an ear."
Adolescent thought has often been categorized as egocentric, characterized by personal beliefs of invulnerability (Buis & Thompson, 1989; Elkind, 1967; 1978; Elkind & Bowen, 1979; Gruber & Chambers, 1987). One way of getting students' attention in sex education programs may be to challenge their myths and thus dispel their illusion of invulnerability and omnipotence.
Data were gathered on the relationship between a teenage woman's CSE, her use of contraceptives, and her RCK by means of a self-administered questionnaire completed by family planning clinic clients in the waiting room prior to their medical or counseling appointments. Receptionists and other clinic personnel were trained, and distribution and collection procedures were standardized. If a client was 20 years or younger and did not wish to participate in the studies, she returned the uncompleted form to the receptionist who then recorded on a refusal sheet the client's number, the date, and the time of her visit.
Participants in the original sample were from Sunnyvale, California (N = 258) and in the second sample from Chicago, Illinois (N = 263). The age of both samples ranged from 11-20 years, averaging 17.6 years. All respondents attended a family planning clinic for similar reasons. The reason for the clinic visit for half the Sunnyvale sample and 56% of the Chicago sample was birth-control related. The reason for clinic visit for approximately a third of both samples was pregnancy or abortion related.
Sixty-three percent of the Sunnyvale sample was European-American, and 21% was Latin-American. By contrast, 93% of the Chicago sample was African-American. The Sunnyvale sample lived in a suburban, lower-middle to middle-class environment, whereas the Chicago sample came from poverty to lower-middle-class communities in the inner-city. Half of the Sunnyvale sample was Catholic and 23% was Protestant, while 49% of the Chicago sample was Baptist, and 29% indicated they were not affiliated with any religion or left the question "blank."
In terms of educational status and aspirations, 90% of the Sunnyvale respondents intended to complete a degree beyond high school or to attend college; 71% of the Chicago sample were currently students, while 28% were not in school; 7% of the Chicago sample were high school dropouts with no intention of receiving a high school diploma or G.E.D; 32% of the Chicago sample aspired to finish high school or get a G.E.D.; 19% planned to get some training beyond high school or take some college courses; and 40% thought that they would finish college. Some of these later aspirations may be inflated, given that many of the responses came from teenage women who were currently not attending mainstream school programs.
Both samples were at high risk of pregnancy due to high frequency of unprotected coitus and irregular use of contraceptives. Approximately 20% of both samples said that they used no birth control method. A majority of participants said that they had engaged in unprotected coitus; 70% of the Sunnyvale sample said that there was at least one incidence within the last six months. Thirty-five percent of the Sunnyvale women had been pregnant at least once, but 91% said that they did not desire a pregnancy within the next six months. Seventeen percent of this sample had ever had an abortion, and 2% had a live birth (n = 7).
Twenty-seven percent of the Chicago women were defined as nonusers of contraceptives, while 28% were inconsistent users, and 45% were considered to be effective contraceptive users. Fifty-five percent of this sample had at least one baby, and 13% had ever had an abortion. Fifty-nine percent thought that having a baby would be "a big problem," while 27% thought that having a baby would be "some" or "not much" of a problem; another 12% thought a baby would be "no problem."
The pill had ever been used as the method of birth control (albeit sporadically) by 58% of the Sunnyvale sample and 60% of the Chicago sample. Thirty-six percent of the Chicago and 49% of the Sunnyvale sample had ever used the condom while 23% of the Sunnyvale sample and 8% of the Chicago sample had ever used condom and foam together. Thirty-four women in the Sunnyvale and five women in the Chicago sample had ever used the diaphragm. Sixty-six percent of the Sunnyvale sample and 15% of the Chicago sample had ever used less reliable methods of birth control such as withdrawal and rhythm.
The major instrument used to collect data about teenage women and their contraceptive behavior was the self-administered questionnaire. Respondents provided information on contraceptive use, demographics, sexual experience, an index of reproductive and contraceptive knowledge (IRCK), psychosocial factors, and CSE.
In the Sunnyvale and Chicago studies, a teenage woman's level of effective contraceptive use was evaluated from questions which asked (1) the current method(s) of birth control that she or her partner used, and (2) the consistency with which that method was used. Method effectiveness is often compromised by actual use patterns, especially by adolescents. In these two samples, an index of effective contraceptive use was created, taking into account the theoretical effectiveness of the method and how the individual actually used it. Hatcher and his colleagues have used the term "theoretical use effectiveness" to describe in percentiles what the effectiveness of a given method should be in preventing pregnancies if used correctly and consistently. They present "actual use effectiveness" percentiles which reflect how a given method actually worked in preventing pregnancies when women who did not want any more children used these methods over the course of a year (Hatcher, Stewart, Stewart, Guest, Stratton, & Wright, 1978, 1988).
In the present study, Hatcher's percentlie rates for each contraceptive method was scaled to construct a 0-20-point rating system for method effectiveness, with a score of 20 being equivalent to 100% effectiveness. If a respondent "always" used her listed method, she was assigned that method's theoretical use effectiveness score. For a woman who "mostly" used her method, Hatcher's average actual use effectiveness score was given. For a woman who "sometimes" used a method, the average actual use effectiveness score was multiplied by .25. This percentage was assigned to approximate the effectiveness rating of such occasional use. Respondents who used no form of contraceptives were scored 0. This weighting score is somewhat psychometrically arbitrary, but not unrealistic, and conveys a set of weights that make theoretical sense.
Thus, women who "always" used a theoretically effective method like the Pill received scores of 20, along with those women who "always" used condoms and foam together. For women who "always" used somewhat less theoretically effective methods, like foam alone or condoms alone, the score computed to 19. When a highly effective method like the Pill was used "mostly," the score dropped to 18, and "sometimes" usage of the Pill rated a score of 5. Foam alone, a less theoretically effective method, received a score of 15 when used "mostly," and a score of 4 when used "sometimes." The resultant efficacy scores were quadramodal and thus were grouped into four categories of effective contraceptive use where scores ranged from 1 (inefficacious contraceptive use) through 4 (highly efficacious contraceptive use). The mean contraceptive use score for Sunnyvale was 2.3, SD = 1.3, and for Chicago, the mean was 2.6, SD = 1.4.
CSE was assessed with the 18-item instrument described in Figure 1. In previous research we found that four factors adequately account for the variance in the scale and conceptually represent CSE (Levinson & Jaccard, 1991). The four factors are labeled Assertive Communication, Physicality of Sex, Taking Control, and Prevention of Unprotected Coitus. The factors and the items that constitute each factor are shown in Figure 2. The four factors were used as measurements of CSE in our analyses.
The demographic variables assessed on the questionnaire included age, ethnicity, and religion, as well as educational status and aspirations.
The Index of Reproductive Knowledge and Contraceptive Use (IRCK) was scored as the number of correct answers to seven questions which have been used frequently in nationwide surveys with clinic populations and in interview studies. These questions ask about the length of time a sperm can remain fertile, the probability of pregnancy with subsequent acts of unprotected coitus, the reliability of various forms of birth control, the time of the month when a woman is most fertile, and the effective use of condoms and withdrawal as methods of birth control. Knowledge scores ranged from 1-7. The mean correct score for Sunnyvale was 4.2, SD = 2.3. In the Chicago sample, the mean correct score was 3.1, SD = 1.8.
Analysis of IRCK, Contraceptive Use, and CSE
The relationship between knowledge and contraceptive use was examined in several ways. First a zero order correlation was computed between the total knowledge scores, responses to each individual item (scored correct or incorrect), and the contraceptive behavior score. Second, each item was scored dichotomously in terms of the type of inaccurate knowledge that was evident. For example, IRCK item 6 asked, "When using a condom, a man should withdraw from the woman's vagina and take off the condom . . ." The response choices were (a) at any time after he ejaculates, (b) immediately after he ejaculates, (c) he should wait a while after he ejaculates, and (d) don't know. Since misinformation was fairly evenly distributed among the "wrong" answers, we wanted to test the effects of certain types of misinformation (see Table 2). These responses were also correlated with contraceptive behavior, but no significant correlations emerged. Third, we attempted to identify moderator variables that might interact with knowledge to predict contraceptive behavior.
This analysis used moderated multiple regression strategies with product terms (Jaccard, Turrisi, & Wan, 1990). The total knowledge score was multiplied by a given moderator variable, and the main effect terms and the product term were entered into a regression equation simultaneously in order to predict contraceptive behavior. A statistically significant regression coefficient for the product term is evidence of a moderated relationship. The following moderator variables were explored: the four factors of CSE, education level attained in school, age, pregnancy history, and abortion history. In addition to the total knowledge score, we explored the moderated analysis with each individual knowledge item as well as the individual "myth" items. None of the interaction terms were statistically significant. We also repeated the above analyses using pregnancy and abortion histories as the dependent variables to determine if knowledge was related to past experience with unintended pregnancy. No statistically significant relationships were observed.
A set of analyses were then undertaken to determine if CSE was related to knowledge. The four subscales were correlated with the total knowledge score using LISREL VII (Joreskog & Sorbom, 1988). This involved conceptualizing four latent variables corresponding to the sub-types of self-efficacy and an endogenous latent variable corresponding to knowledge [ILLUSTRATION FOR FIGURE 3 OMITTED]. Knowledge had a single indicator (the total IRCK score) and each latent variable had multiple indicators (the CSE items) as defined in our previous research (Levinson & Jaccard, 1991). The fit of the model described in Figure 3 was evaluated for each sample using the covariance matrix as input. The overall chi square tests of the model were statistically significant for both samples, [[Chi].sup.2] (143) = 282.23, Sunnyvale; [[Chi].sup.2] (143) = 176.26, Chicago. Due to the large sample size, however, goodness-of fit-indices were also examined. Bentler's Comparative Fit Index (1990) was satisfactory for Chicago (CFI = .90, GFI = .92), but marginal for Sunnyvale (CFI = .75, GFI = .88). For the Sunnyvale sample, correlated errors were introduced into the measurement model to improve model fit to satisfactory levels. These parameter changes did not fundamentally alter the correlational structure between the latent self-efficacy variables and the latent knowledge variable, [[Chi].sup.2] (136) = 203.62, GFI = .92, CFI = .88. The standardized residuals for the predicted versus the observed covariances were generally small and nonsignificant.
CSE was significantly related to knowledge (p [less than] 0.05) for three of the four subscales in both samples; Table 1 shows how the correlational pattern was duplicated in both samples. None of the standardized path coefficients were significant in either sample. Nevertheless, when considered as a group, the four CSE factors explained 12% and 28% of the variance in knowledge in the Chicago and the Sunnyvale samples, respectively. Thus, it appears that none of the four CSE factors contributes uniquely to explaining the variance in knowledge, but as a totality, the CSE factors provide reasonably good prediction of reproductive and contraceptive knowledge.
Table 1 Selected Statistics for Correlational Analyses between Knowledge and CSE Estimated Zero Order Correlation Sunnyvale Chicago CSE Factors Communication .30(*) .24(*) Physical .22(*) .35(*( Acceptance Control .16 .19 Prevention .28(*) .44(*) * p [less than] .01
Table 2 shows the results for both samples on the individual IRCK items. On the whole, the Chicago sample was less knowledgeable than the Sunnyvale about reproduction and contraception as evidenced by the higher percentages in the "don't know" and incorrect response categories. Although the implications of the results on the individual items are discussed later, the patterns on the individual items are highlighted here. The first item probed the respondent about when during the menstrual cycle the risk of pregnancy is greatest. It is notable that there was a great deal of confusion about the correct answer to this question in both samples; however, over half of the Sunnyvale sample did know that a woman is most vulnerable to pregnancy midway in her cycle, while less than 20% of the Chicago sample knew this. A slighter greater number of Chicago women thought that a woman was most vulnerable at the very end of her cycle. Nearly one quarter of both samples indicated that they did not know when a woman was most vulnerable to pregnancy. Item 3 asked for related information about how conception and reproduction occurs. The results on item 3 reveal that both samples were relatively unsure about the viability of sperm in a woman's body after coitus; over half of both samples answered incorrectly, and over one-third admitted that they did not know the answer.
Table 2 Examination of Individual Knowledge Questions Fact 1 A woman is most likely to get pregnant if she has intercourse: Sunnyvale Chicago a) a day or so before her period 4% 20% b) during her period 1% 6% c) halfway between periods 57% 18% d) risk is the same throughout 18% 29% e) don't know 22% 27% Fact 2 The least reliable method of birth control is: a) condom 3% 7% b) withdrawal 71% 60% c) rhythm 13% 4% d) diaphragm plus jelly or cream [less than] 1% 2% e) don't know 14% 25% Fact 3 A sperm can stay alive and able to fertilize an egg in the woman's body for as long as: Sunnyvale Chicago a) two hours 8% 11% b) one day 12% 17% c) 2-3 days 45% 32% d) don't know 36% 39% Fact 4 If a woman has intercourse several times without getting pregnant, it means: a) she probably can't get pregnant 2% 7% at this time in her life b) she probably can't get pregnant 13% 24% as easily as most women c) she probably is normal, but 83% 58% by chance she didn't get pregr those times d) she probably won't ever get 1% 10% pregnant Fact 5 A man's sperm is released: Sunnyvale Chicago a) only at the time of ejaculation 38% 25% b) only after an orgasm is complete 5% 3% c) in different amounts throughout 55% 56% intercourse d) don't know 1% 14% Fact 6 When using a condom, a man should withdraw from the woman's vagina and take off the condom: a) at any time after he ejaculates 21% 14% b) immediately after ejaculation 44% 45% c) he should wait a while after he 31% 16% ejaculates d) don't know 4% 24% Fact 7 The most reliable method of birth control is: Sunnyvale Chicago a) condom 16% 18% b) foam 4% 4% c) diaphragm plus jelly 72% 38% d) rhythm 3% 5% e) don't know 5% 30%
Items 2 and 7 asked respondents about the reliability of various methods of birth control: the condom, foam, the diaphragm plus jelly or cream, withdrawal, and rhythm. Again, the Chicago sample was less well informed than the Sunnyvale sample, with about one quarter stating that they did not know which methods were either "most" or "least" reliable.
The results of item 4 are interesting because we can see a pattern in the differences between the samples in assessment of the meaning of not becoming pregnant from several acts of unprotected intercourse. The Chicago sample tended to interpret the lack of a pregnancy as more of an indicator of infertility than did the Sunnyvale sample. The Sunnyvale sample was more likely to accurately ascribe the nonoccurrence of a pregnancy to chance.
Both items 5 and 6 asked about information that is critical to the effective use of condoms. The amount of misinformation on these items was impressively high. Only slightly more than half of both samples knew that sperm is released in different amounts throughout intercourse. One-quarter of the Chicago sample and more than one-third of the Sunnyvale sample held the erroneous belief that sperm is released only at the time of ejaculation. Less than half of both samples did not know how to use a condom effectively, as indicated by the responses on item 6. A substantial percentage of both samples incorrectly believed that a man should or can wait a while after ejaculation to withdraw and to remove a condom.
The results of this study did not support the hypothesis regarding the possible interaction effects between know]edge, CSE, and contraceptive use. Further, no significant relationships between several measures of knowledge and behavior were found even when the effects of influential study variables such as education level and previous pregnancy experiences were taken into account. These results were surprising initially even though they mirror those of a number of other studies (Bilodeau et al., 1991; Fishman, Collier, Stewart, & Swarts, 1974; Kirby, 1984; Pope, Westerfield, & Walker, 1985; Zelnik & Kantner, 1977). It was particularly perplexing that the social-psychological theoretical framework and analytic approach utilizing CSE as a moderator variable did not reveal any new understanding about the relationship between adolescent women's reproductive and contraceptive knowledge and their contraceptive behavior. These results were especially striking since CSE was highly correlated with and explained a significant proportion of the variance in knowledge. However, after considering (1) the strong relationships between three of the four CSE factors and knowledge, as well as (2) the content of the IRCK items, and (3) the measurement of contraceptive use, a new understanding of the data was reached which has already influenced our ongoing research.
The content of the CSE statements are behaviorally specific to the kinds of cognitive, emotional, and physical situations and demands that teenage women experience over time in being sexually active and in trying to use contraceptives. The data have been consistent with the theory indicating that CSE is related to contraceptive behavior. Unlike the measure of CSE, the measure of knowledge was not directly related to knowledge that was method specific for this sample; therefore, this assessment of knowledge was unlikely to reflect or have an impact on effective contraceptive practices. Although IRCK items 5 and 6 asked about information that is critical to effective condom use, very few young women used condoms as a primary method of birth control (see Table 2). Similarly, items i and 3 have behavioral implications for those who use methods such as withdrawal and the rhythm or the calendar/basal body temperature method, but only a very few women used these methods of birth control. Finally, the Pill was the primary method of contraception for both samples (followed by no method at all), and none of the IRCK items tapped critical knowledge related to Pill use. In sum, we believe it is critical in future research to assess reproductive and contraceptive information that is method and practice specific.
We also believe that if the measure of contraceptive use were method specific rather than being a single indicator of the overall effectiveness of contraceptive use, the chances of finding linkages between method-specific knowledge and method-specific behavior would be greatly increased. We were unable to conduct these types of analyses due to the very small number of women who used methods other than the Pill and no method at all. Currently we are finding, as are other researchers and clinicians, that young women are increasingly using the condom and the sponge as methods of birth control and prophylaxis. As these numbers increase and the threat of AIDS becomes more salient, assessment of method-specific contraceptive use should become more feasible as well as pertinent.
The results on the IRCK items 5 and 6 are particularly significant when considered in the context of STD and HIV prevention rather than in the context of birth control. There is a significantly large amount of misinformation in both samples that seriously compromises effective condom use. It will be imperative in future AIDS and sex education interventions to teach and to assess information related to effective condom and sponge or diaphragm usage. Multiple-item indicators of knowledge should be developed which are specific to these contraceptive and prophylactic methods. For example, it will not only be critical to know when and how to put on and to remove a condom (items 5 and 6), but it will be important to know how to select condoms for purchase. Condom users will need to know that not all lubricated condoms contain nonoxynol-9 and that not all condoms prevent the transmission of the HIV virus (e.g., lamb skin condoms do not provide protection). In order to use condoms effectively in STD prevention, one will need greater specificity of knowledge as well as the motivation to use condoms.
The last IRCK item for discussion is number 4. We think it is important to note the sample differences concerning a woman's subjective assessment of the likelihood of pregnancy given several experiences with unprotected intercourse. Over half of the Chicago women have had a baby, and a higher percentage have ever been pregnant. This sample's early fertility rate and their tendency to ascribe possible fertility problems in response to IRCK item 4 suggests that lower overall knowledge scores and high pregnancy rates may be linked to a lack of motivation to avoid pregnancy.
Further support for this interpretation is found in the somewhat weaker relationship evidenced between the knowledge score and the four CSE factors for the Chicago sample as compared to the Sunnyvale sample in the regression analyses. While a significant proportion of the variance in knowledge was explained by three of the four CSE factors for both samples, the strength of this relationship is more pronounced for the Sunnyvale women. Thus, for the Chicago women there are other factors operating on the knowledge variable in addition to CSE. Other researchers have demonstrated that social and environmental factors are strong influences on contraceptive use for black teenagers (Abrahamse, Morrison, & Waite, 1988; Hogan, Astone, & Kitagawa, 1985; Ladner, 1987). The patterns in the zero order correlations also indicate, particularly in the Sunnyvale sample, that young women who feel confident in their ability to prevent unprotected intercourse tend to have high IRCK scores. These young women may attend, absorb, and value this information because they are highly motivated to avoid an unwanted pregnancy, and find information related to reproduction and contraception salient.
For both samples, the zero order correlations also indicate that young women who feel able to be assertive in communicating their sexual and contraceptive needs, and who feel comfortable with their sexuality, tend to have higher knowledge scores. Other researchers have found positive sexual self-concept, sexual self-esteem, or erotophilia to be related to high knowledge (Byrne & Fisher, 1983; Fisher, Byrne, Edmunds, Miller, Kelley, & White, 1979; Hatcher, 1976; Winter, 1988). The one CSE factor that was not related to knowledge was the one which assessed a young woman's confidence that she could be in control of the different sexual and contraceptive situations that occur with a male partner. None of the knowledge questions tap the dynamics and power relations of male and female interactions and sexual negotiations; thus, it is not surprising to find that these two variables are unrelated.
In previous research on CSE and contraceptive behavior (Levinson, 1984, 1986; Levinson & Jaccard, 1991), we have argued for the need to teach teenagers about sexual decision-making and contraception by relating to the specific kinds of needs, feelings, pressures, and situations that teenagers experience as sexual human beings. We have stated that teenagers will be empowered to be more responsible about their sexual behavior and contraceptive use if they are more prepared emotionally, cognitively, and behaviorally to deal with the actual feelings and situations they encounter. The research reported here extends that behavior-domain specific approach to the selection and assessment of knowledge items and to the assessment of contraceptive usage. Some critical knowledge areas where teenage women may tend to be misinformed have also been identified. It may be effective for sex educators to challenge these myths with their students, especially where they are critical to effective contraceptive and prophylactic practice.
The author wishes to thank the W. M. Keck Foundation which supported this research, and Debra Zatkowsky and Diana Pesotsky who provided assistance in the preparation of this manuscript.
Abrahamse, A., Morrison, P., Waite, L. (1988). Beyond stereotypes: Who becomes a single teenage mother? Santa Monica, CA: Rand Corporation.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84(2), 191-215.
Banks, I. W., & Wilson, P. I. (1989). Appropriate sex education for black teens. Adolescence, 24(93), 233-245.
Bentler, P. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238-246.
Bilodeau, A., Forget, G., & Tetreault, J. (1991). L'auto-efficacy relative a la contraception chez les adolescentes: La validation d'une echells de mesure [Contraceptive self-efficiency among female and male adolescents: The validation of the French version of Levinson's scale] Unpublished paper. Montreal: Department de Sante Communautaire.
Buis, J. M., & Thompson, D. N. (1989). Imaginary audience and personal fable: A brief review. Adolescence, 24(96), 773-781.
Byrne, D., & Fisher, W. A. (1983). Adolescents, sex, and contraception. Hillsdale, NJ: Lawrence Erlbaum Associates.
Carrera, M., Baker, E., & McCombs, K. (1978). A sex education program that worked. SIECUS Reports, 6(4), 6.
Coles, R., & Stokes, G. (1985). Sex and the American teenager. New York: Harper & Row.
Eisen, M., Zellman, G., & McAllister, A. (1990). Evaluating the impact of a theory-based sexuality and contraceptive program. Family Planning Perspectives, 22(6), 261-271.
Elkind, D. (1967). Egocentrism in adolescence. Child Development, 38(4), 1025-1034.
Elkind, D. (1978). The child's reality: Three developmental themes. Hillsdale, NJ: Lawrence Erlbaum Associates.
Elkind, D., & Bowen, R. (1979). Imaginary audience behavior in children and adolescents. Developmental Psychology, 15, 38-44.
Fisher, W., Byrne, D., Edmunds, M., Miller, C., Kelley, K., & White, L. (1979). Psychological and situation-specific correlates of contraceptive behavior among university women. The Journal of Sex Research, 15, 38-55.
Fishman, S., Collier, P., Stewart, J., & Swartz, D. (1974). The impact of family planning classes on contraceptive knowledge, acceptance, and use. Health Education Monographs, 2, 246-259.
Gruber, E., & Chambers, C. V. (1987). Cognitive development and adolescent contraception: Integrating theory and practice. Adolescence, 22, 661-670.
Haas, A. (1979). Teenage sexuality: A survey of teenage sexual behavior. New York: MacMillan.
Hansson, R. O., Jones, W. H., & Chernovetz, M. E. (1979). Contraceptive knowledge: Antecedents and implications. The Family Coordinator, 28(1), 29-34.
Hatcher, R., Stewart, G., Stewart, F., Guest, F., Stratton, P., & Wright, A. (1978). Contraceptive technology 1978-1979. New York: Wiley and Sons (reprinted, 14th edition, 1988-1989).
Hatcher, S. (1976). Understanding adolescent pregnancy and abortion. Primary Care, 3, 407-425.
Hayes, D. (Ed.). (1987). Risking the future: Adolescent sexuality, pregnancy and childbearing (Vol. 1). Washington, DC: National Academy Press.
Hogan, D. P., Astone, N.M., & Kitagawa, E. M. (1985). Social and environment factors influencing contraceptive use among black adolescents. Family Planning Perspectives, 17(4) 165-168.
Howard, M., & McCabe, J. B. (1990). Helping teenagers postpone sexual involvement. Family Planning Perspectives, 22(1), 21-26.
Jaccard, J., Turrisi, R., & Wan, C. (1990). Interaction effects in multiple regression. Beverly Hills, CA: Sage Publications.
Joreskog, K., & Sorbom, D. (1988). LISREL 7.16. Mooresville, Indiana: Scientific Software, Inc.
Kallen, D. J. (1982). Talking about sex revisited. Journal of Youth and Adolescence, II(1), 11-23.
Kashima, Y., Gallois, C., Hills, R., & McCamish, M. (1991). Intentions, past behavior, future behavior: When does intention to use a condom predict actual use? (Manuscript under review.)
Kirby, D. (1984). Sexuality education: An evaluation of programs and their effects. Santa Cruz, CA: Network Publications.
Ladner, J. (1987). Black teenage pregnancy: A challenge for educators. Journal of Negro Education, 56(1), 53-56.
Levinson, R. A. (1984). Contraceptive self-efficacy: A primary prevention strategy. Journal of Social Work and Human Sexuality, 3(1), 1-15.
Levinson, R. A. (1986). Contraceptive self-efficacy: A perspective on teenage girls' contraceptive behavior. The Journal of Sex Research, 22(3), 347-369.
Levinson, R. A., & Jaccard, J. (1991, August). Contraceptive self-efficacy in two samples: Implications for interventions. In L. Fisher (Chair), Theories, issues, and practice in adolescent health promotion. Symposium conducted at the 99th Annual Meeting of the American Psychological Association, San Francisco, CA.
Marsiglio, W., & Mott, F. (1986). The impact of sex education on sexual activity, contraceptive use, and premarital pregnancy among American teenagers. Family Planning Perspectives, 18(4), 151-162.
Polit, D. F., & Kahn, J. R. (1985). Project redirection: Evaluation of a comprehensive program for disadvantaged teenage mothers. Family Planning Perspectives, 17(4), 150-154.
Pope, A. J., Westerfield, C., & Walker, J. (1985). The effect of contraceptive knowledge source upon knowledge accuracy and contraceptive behavior. Health Education, 16(3), 41-44.
Rosenthal, D., Moore, S., & Flynn, I. (1991). Adolescent self-efficacy, self-esteem, and sexual risk-taking. Journal of Community & Applied Social Psychology, 1, 77-88.
Schinke, S. (1984). Preventing teenage pregnancy. Progress in Behavior Modification, 16, 31-64.
Schinke, S., & Gilchrist, L. (1977). Adolescent pregnancy: An interpersonal skill-training approach to prevention. Social Work in Health Care, 3(2), 159-167.
Spanier, G. (1977). Sources of sex education information and premarital sexual behavior. Journal of Sex Research, 13, 73-78.
Thornburg, H. D. (1981). Adolescent sources of information on sex. The Journal of School Health, 51, 274-277.
Turtle, A.M., Ford, B., Habgood, R., Grant, M., Bekiaris, J., Constantinou, C., Maeck, M., & Polyzoidis, H. (1989). AIDS-related beliefs and behaviors of Australian university students. Medical Journal of Australia, 150, 371-376.
Whitley, B. E., & Schofield, J. W. (1986). A meta-analysis of research on adolescent contraceptive use. Population and Environment, 8(3-4), 173-203.
Winter, L. (1988). The role of sexual self-concept in the use of contraceptives. Family Planning Perspectives, 20(3), 123-127.
Zabin, L. S., Hirsch, M. B., Smith, E. A., Streett, R., & Hardy, J. (1986). Evaluation of a pregnancy prevention program for urban teenagers. Family Planning Perspectives, 18(3), 119-126.
Zelnik, M., & Kantner, J. F. (1977). Sexual and contraceptive experience of young unmarried women in the United States, 1976 and 1971. Family Planning Perspectives, 9, 55-71.
Zelnik, M., & Kantner, J. F. (1980). Sexual activity, contraceptive use and pregnancy among metropolitan-area teenagers: 1971-1979. Family Planning Perspectives, 12(5), 230-237.
Zelnik, M., & Kim, Y. J. (1982). Sex education and its association with teenage sexual activity, pregnancy, and contraceptive use. Family Planning Perspectives, 14(3), 117-126.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Reproductive and Contraceptive Knowledge, Contraceptive Self-Efficacy, and Contraceptive Behavior among Teenage Women. Contributors: Levinson, Ruth Andrea - Author. Journal title: Adolescence. Volume: 30. Issue: 117 Publication date: Spring 1995. Page number: 65+. © 1999 Libra Publishers, Inc. COPYRIGHT 1995 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.