INTERNAL POVERTY AND TEEN PREGNANCY.

Young, Tamera M. ; Martin, Sue S. ; Young, Michael E. 等

ABSTRACT

The subjects for the present study were drawn from the female students who participated in the National Education Longitudinal Study (NELS) initial eighth-grade data collection. Adolescent females who later became pregnant were matched on race, birth month, and birth year with adolescent females who did not report a pregnancy. The study examined selected predictor variables from the baseline 1988 wave of data in relation to the outcome variable of pregnancy status. Results indicated a statistically significant difference in locus of control between those females who later became pregnant and those who later did not experience a pregnancy during adolescence. Those who later became pregnant were much more likely to have an external locus of control (p = .0001). Females who later became pregnant were also more likely to have a poorer sense of personal efficacy (p = .0001). Finally, females who later experienced a teen pregnancy had more traditional occupational expectations (p = .006) and lower educational expec tations (p = .001) than did those who did not later report a teen pregnancy.

Adolescent childbearing has negative effects on the adolescent, the offspring, and society in general (Roosa, Fitzgerald, & Carlson, 1982; Elster, Lamb, Peters, Kahn, & Tavare, 1987; Donovan & Jessor, 1985; Mott & Haurin, 1988; U.S. Department of Health and Human Services, 1995). Research examining predisposing factors in adolescent childbearing has frequently been restricted to correlational studies, often even in the context of confounding enrichment programs. Thus, there is a great need to (1) identify the underlying factors that precede teen pregnancy and (2) address these factors in programs designed to reduce adolescent childbearing. The present study sought to examine the roots of adolescent childbearing using data acquired prior to pregnancy. Specifically, it explored the relationship between internal poverty (low educational and occupational aspirations, poor personal efficacy, and external locus of control) and adolescent pregnancy.

THEORETICAL BACKGROUND

Perceived Self-Efficacy and Locus of Control

This study is heavily rooted is Bandura's (1994) self-efficacy theory. Self-efficacy is defined as a person's beliefs about his or her ability to attain particular goals. It has been found to impact the coping ability and behaviors of the individual (Bandura, 1994).

This theory centers around three processes and three types of motivation. Efficacy-activating processes include cognitive, motivational, and affective processes (Bandura, 1994). Self-efficacy is believed to influence motivation--including causal attributions, outcome expectancies, and cognized goals. Self-efficacy influences causal attributions in that those who see themselves as efficacious attribute failure to lack of effort, while those who see themselves as inefficacious attribute failure to low ability. Efficacy influences outcome expectancy in that behavior is influenced by beliefs concerning personal capabilities and not just the expected outcomes of a behavior. Goal setting, another tool for enhanced motivation, is also influenced by perceived self-efficacy in several ways. Bandura asserts that efficacy determines the goals people set for themselves, how much effort they expend, how long they persevere in the face of barriers, and their resilience to failure. Efficacy also influences affective process es--those regulating emotional states and reactions such as stress, anxiety arousal, and depression (Bandura, 1994).

Another theoretical construct that is relevant for this study is locus of control (Rotter, 1975)--the degree to which individuals believe they have control over events in their lives versus the degree to which they believe they are victims of fate or external circumstances. Individuals with an external locus of control fail to see a connection between personal behavioral choices, well-being, and quality of life. These individuals often view life as being determined by (1) the lot into which they were born and (2) powerful others who restrict their upward mobility.

Life Options Model

According to this model, disadvantaged youths do not perceive themselves as having positive life options and therefore do not plan for their future and do not avoid barriers to that future such as early out-of-wedlock childbearing (Dryfoos, 1984). If disadvantaged youths do not perceive that doors are open to them, it is perhaps difficult for them to see teen pregnancy as closing any doors.

REVIEW OF LITERATURE

Efficacy and Locus of Control

There is a dearth of research on general perceived self-efficacy relative to teen pregnancy. One study that did examine this relationship, conducted by Zimmerman, Sprecher, Langer, and Holloway (1995), found that, for females, a generalized sense of self-efficacy was a predictor of ability to say "no" to unwanted sex. St. Lawrence, Brasfield, Jefferson, Allyene, and Shirley (1994) found low self-efficacy to be associated with low levels of social support among teens. They also found that these teens were more likely to engage in casual sex, to have more nonmonogamous partners, to be victims of coercive sex, and to have higher rates of sexually transmitted disease than were adolescents with higher levels of social support (and perhaps higher levels of self-efficacy). Goh, Primavera, and Bartalini (1996) found self-efficacy to be associated with AIDS-preventive intentions. These studies suggest a need to examine self-efficacy relative to adolescent childbearing.

In regard to locus of control, the research reveals conflicting evidence. Holden, Nelson, Velasquez, and Ritchie (1993) found no significant difference between pregnant and nonpregnant teens. However, these researchers used only a small sample of 69 pregnant and 58 comparison adolescents, and their measure of locus of control had very low internal consistency (alpha = .23). Plotnick (1992), using data from the National Longitudinal Survey of Youth, found that high scores on internal locus of control were negatively related to the likelihood of premarital pregnancy among non-Hispanic white females. Morgan, Chapar, and Fisher (1995) found that adolescents with a history of pregnancy scored higher than nulliparous females on the Powerful Other (Health Locus of Control) subscale, which measures the belief that others control one's fate. Hanson, Myers, and Ginsburg (1987) found that adolescents who had a strong sense of control over their lives were less likely to experience teenage childbirth. Lewis, Ross, and Mi rowsky (1999) found that a low sense of control increased the risk of a subsequent nonmarital pregnancy.

Educational and Life Expectations

Educational expectations were found to correlate with teen pregnancy in some but not all studies. Hanson, Myers and Ginsburg (1987) found that female adolescents who had high educational expectations, and who had parents with high educational expectations for their daughters, were less likely to experience teenage childbirth. Plotnick (1992) found high educational expectations to be negatively related to the likelihood of premarital pregnancy. Thus, Plotnick recommended a policy that encourages the pursuit of higher education in order to reap the indirect advantages of reducing premarital childbearing and associated societal costs.

On the other hand, McCullough and Scherman (1991) found pregnant teens in a Cleveland encouragement program to have high educational aspirations. Rodriquez and Moore (1995) also found high educational aspirations among those enrolled in pregnant/parenting programs. One might argue that these expectations are attributable to the intervention programs in which the teens were involved, rather than to attitudes that existed at the time of conception. Regardless, these results indicate inconsistencies in the literature as to the role of educational expectations and aspirations in teen pregnancy.

Closely related to educational expectations are the future opportunities education affords. Perceived life options may also play a role in sexual decision making. Dryfoos (1984), Plotnick (1993), Sullivan (1993), and Bickel, Weaver, Williams, and Lange (1997) found that a limited sense of life options contributed to teen pregnancy. However, Holden, Nelson, Velasquez, and Ritchie (1993) found no significant difference between pregnant and nonpregnant teens in regard to future orientation. Mead (1986) asserted that those raised in poverty tend to value short-term gratification over long-term planning. Gordon (1996) found that her subjects had wanted their pregnancies, and that poor academic performance and poverty led to pregnancy rather than early pregnancy leading to dropping out of school and a life of poverty.

Life expectations may be influenced by internal resources. Church, Teresa, Rosebrook, and Szendre (1992) explored the relationship between rural general equivalency diploma (GED) students' willingness to consider certain occupations and self-efficacy, interests, and perceived incentives. They found self-efficacy to be related to specific occupations. The range of occupations considered by these students was predicted by general self-efficacy, range of interests, and range of incentives that the students associated with a career. They did not find a relationship between aptitude and self-efficacy. That is, perceived ability did not accurately reflect real ability. This suggests that aptitude means little in regard to life options if self-efficacy is low. Ralph, Lochman, and Thomas (1984) found poorer vocational-educational adjustment among pregnant teens as compared to a nulliparous comparison group. They recommended interventions that expose at-risk young women to future opportunities beyond early motherhood. Thus, career decision making should be addressed in programs designed to reduce out-of-wedlock births.

HYPOTHESES

The literature suggested the following hypotheses.

1. There would be a difference between adolescent females who later became pregnant and those who did not in regard to locus of control. Females who later became pregnant would have a more external locus of control than would the comparison group.

2. There would be a difference between adolescent females who later became pregnant and those who did not in regard to personal efficacy. Those females who became pregnant would have poorer personal efficacy.

3. There would be a difference between adolescent females who later became pregnant and those who did not in regard to future expectations. Adolescent females who later became pregnant would perceive their life options to be more limited (indicated by poor educational and occupational expectations) as compared to the comparison group.

METHOD

Data

The data analyzed for this study were drawn from the National Education Longitudinal Study (NELS). Data on various topics were obtained, via questionnaires and achievement tests, from a national probability sample of over 25,000 U.S. eighth graders in 1,000 schools (800 public and 200 private). Thus, the sample was representative of the nation's eighth-grade population (U.S. Dept. of Education, National Center for Education Statistics, 1990).

These eighth graders were reexamined in 1990 and 1992. In 1990, first follow-up data were collected from tenth-grade students and from those who had dropped out prior to the tenth grade (U.S. Dept. of Education, National Center for Education Statistics, 1993a, 1993b). Second follow-up data were collected in 1992 from twelfth-grade students and from those who had dropped out after the tenth grade (U.S. Dept. of Education, National Center for Education Statistics, 1995a, 1995b).

This national data set allowed for a longitudinal examination of the variables of interest. Specifically, the data allowed for comparisons between those who later became pregnant and those who did not. It permitted an exploration of possible predictor variables as they existed prior to the teen pregnancy. This longitudinal design facilitated a somewhat stronger case for causation than has been possible in the many studies that have examined correlates of teen pregnancy only after the pregnancy had occurred (sometimes even in the context of confounding enrichment programs for teen mothers).

Sample

The subjects for the present study were drawn from the female students who participated in the NELS initial eighth-grade data collection and subsequently in the tenth- or twelfth-grade survey or the first or second dropout survey. The subjects of specific interest were those females who later reported having a child, expecting a child, or dropping out due to a pregnancy, along with a comparison group of females. The study examined selected predictor variables from the baseline (eighth grade) 1988 wave of data in relation to the outcome variable of pregnancy status. The first follow-up dropout and first follow-up student data, as well as the second follow-up dropout and second follow-up student data were used only for assessing the outcome variable of later pregnancy status. For this study, adolescent females who became pregnant were matched with adolescent females who did not indicate having been pregnant. The matching variables were race, birth month, and birth year.

Variables

Internal poverty. Internal poverty was assessed in three areas: locus of control, personal efficacy, and future expectations (i.e., predisposing factors). Locus of control was measured using a six-item scale. Kaufman and Rasinski (1991) calculated Cronbach's alpha to determine reliability (internal consistency) using responses from the entire baseline data (males and females). They obtained an alpha of .678. For the present study, the researchers calculated this statistic for the females included in the analysis and obtained an alpha level of .674.

According to Kaufman and Rasinski (1991), correlations, a varimax rotated factor matrix, and factor statistics revealed a "personal efficacy" factor. The questions comprising this personal efficacy scale (a subset of the locus of control scale used in the NELS:88 database) are as follows: "My plans hardly work out," "I don't have enough control over my life," "Every time I get ahead something stops me," and "When I make plans I can make them work" (reverse scored). These were measured on a four-point Likert type scale, with 1 strongly agree and 4 = strongly disagree. Using the data from the females, Cronbach's alpha was calculated and a reliability estimate of .531 was obtained.

Finally, fixture expectations were assessed using several questions: whether the subject thought she would graduate from high school, how far she thought she would get in school, and what occupation she thought she would have at age 30. First, the expected occupations were categorized into three levels-high, medium, and low. The careers listed under the two professional headings were grouped together to form the high category. The middle category consisted of occupations under the headings of manager/administrator, proprietor or owner, protective services, school teacher, and technical. The low category consisted of clerical, craftsperson, farmer/farm manager, homemaker, laborer, military, sales, operative, don't know, and not planning to work. Second, the occupations were also categorized as traditional or nontraditional. The careers listed as traditional were clerical, homemaker or housewife only, professional-09 (e.g., nurse, librarian), sales (salesperson, advertising or insurance agent, real estate brok er), school teacher, service (beautician, private household worker), and not planning to work. The careers listed as being nontraditional for females were craftsperson, farmer, farm manager, laborer, manager/administrator, military, operative, professional-10 (e.g., clergy, dentist), proprietor or owner, protective services, and technical.

Teen pregnancy. The first follow-up student data and first follow-up dropout data were used to identify females who had experienced an early pregnancy. Later pregnancies were determined from the second follow-up student data and second follow-up dropout data. The first and second follow-up student and dropout surveys included the following question: "Do you have any children of your own?" Responses included "Yes, I do," "No," and "No, but I am expecting one." The first follow-up dropout data included responses from those who left school prior to the end of tenth grade (a list of reasons for their dropping out was provided, including pregnancy). The second follow-up dropout data were used to determine which teens dropped out after tenth grade due to a pregnancy.

Procedure

Females who reported that they had a child, or were expecting one, were identified from the 1990 and 1992 dropout and student data. These females were matched, on the basis of birth month, birth year, and race, with females who did not report having or expecting a child (1990 and 1992 dropout and student data). Such matches were found for 914 of the 937 females who had reported being pregnant or having a child. Thus, the data set used for this analysis came from a total of 1,851 females. Among these were 22 nonpregnant and 22 pregnant Asian/Pacific Islanders, 188 nonpregnant and 191 pregnant Hispanics, 210 nonpregnant and 211 pregnant Black non-Hispanics, 435 nonpregnant and 439 pregnant White non-Hispanics, and 47 nonpregnant and 55 pregnant American Indians. Also included were 2 nonpregnant and 2 pregnant females who gave multiple responses to the question on race, 1 pregnant female who refused to answer, and 10 nonpregnant and 16 pregnant females with missing responses.

Data were analyzed using appropriate SAS programs, including analysis of variance for interval-level variables and chi-square for categorical variables. Comparisons were made between those females who later became pregnant and those who did not. All of the predictor variables were taken from the baseline 1988 data.

Limitations

There were several limitations to this study. A separate data set was made consisting of those who reported having had a child and/or having been pregnant. It also contained subjects who claimed to have dropped out of school due to a teen pregnancy. The data did not allow for the identification of those students who were pregnant at a time before the survey, and yet did not drop out due to the pregnancy or give birth. Therefore, this data set may not be inclusive of all those who actually had been pregnant.

Likewise, the efficacy factor determined by a NELS technical report (Kaufman & Rasinski, 1991) was not a standard, or very good, measure of self-efficacy. To have examined the true predictive nature of self-efficacy as defined by Bandura (1994), a standard scale with questions that truly assess self-efficacy would have been needed. No such "real" self-efficacy scale was included in the NELS survey.

RESULTS

Locus of Control

An analysis of variance was performed to determine if there was a difference in locus of control between adolescent females who later became pregnant and those who did not. ANOVA results, shown in Table 1, indicated that there was a statistically significant difference in locus of control between the two groups (F = 64.42, p = .0001). Females who later became pregnant scored lower (M = 15.637) on this scale than did those who did not become pregnant (M = 16.670), with a lower score indicating a more external locus of control.

Personal Efficacy

ANOVA results (see Table 2) revealed a statistically significant difference in personal efficacy between the two groups (F = 42.07, p = .0001). Personal efficacy was lower in females who later became pregnant (M = 10.272) as compared to those who did not become pregnant in their teens (M = 10.836).

Future Expectations

Chi-square tests were used to determine if there were differences in future expectations between adolescent females who became pregnant and those who did not. Variables of interest were the females' occupational expectations for age 30 (based on a low, middle, and high scale and a traditional versus nontraditional scale), how far they believed they would get educationally, and how sure they were that they would finish high school. These analyses produced mixed results (see Table 3). Although teens who later became pregnant had lower expectations in regard to educational attainment and finishing high school than did those who did not become pregnant, they did not have lower expectations in regard to level of career. Their expectations for careers that typically require education beyond high school were not statistically different from those of females who did not become pregnant. When the occupations were grouped into traditional and nontraditional, a statistically significant difference was found between fema les who later became pregnant and those who did not, [[chi].sup.2](1) = 7.689, p = .006. At the time of the baseline data collection, eighth-grade females who later became pregnant had more traditional occupational expectations.

DISCUSSION

The purpose of this study was to examine the relationship of various indices of internal poverty to later pregnancy status. The study showed that internal poverty was more likely among females who later became pregnant during their teens as compared with those who did not become pregnant. The results make a contribution to the literature in several ways. The subjects were drawn from a large probability sample, which enhanced external validity. A large number of later-pregnant females were matched with never-pregnant females on race, birth year, and birth month. Predisposing factors were examined longitudinally, which allowed for the assessment of predictor variables prior to conception. These qualities make the internal validity of this study greater than that of most studies in this area. Most studies could not make assumptions about causation because they were correlational, and there was also the greater likelihood of confounding variables (including enrichment programs for parenting teens). Finally, these results help to resolve contradictions in previous research.

Locus of Control

Results from ANOVA revealed that females who later became pregnant had a locus of control score that was more external relative to the score for the comparison group. This finding supports the studies by Plotnick (1992), Hanson et al. (1987), and Lewis et al. (1999), as well as the study by Morgan et al. (1995), which used the Powerful Other (Health Locus of Control) subscale. However, it contradicts the correlational study by Holden et al. (1993).

Personal Efficacy

The present study also found poorer personal efficacy in the eighth-grade females who later became pregnant as compared to their peers who did not report a pregnancy later. This finding provides some additional support for the notion that efficacy influences sexual decision making. It is consistent with the study by Zimmerman et al. (1995), which found self-efficacy to be a predictor of the ability to say "no" to sex, and with the study by Goh et al. (1996), which found self-efficacy to be associated with AIDS-preventive intentions. Further, it partially supports the study by St. Lawrence et al. (1994), which found low self-efficacy to be associated with low levels of social support among teens. These teens were also more likely to engage in casual sex, to have more nonmonogamous partners, to be victims of coercive sex, and to have higher rates of sexually transmitted disease than were those with higher levels of social support and perhaps higher levels of associated self-efficacy.

Future Expectations

In regard to future expectations, the results were mixed. It was found that educational expectations differentiated between the group of females who later became pregnant and the comparison group. This supports some previous research (Devaney & Hubley, 1981; Hanson et al., 1987; Plotnick, 1992), but contradicts the correlational studies by McCullough and Sherman (1991) and Rodriquez and Moore (1995). However, the latter two studies examined females who were already pregnant and in special programs. Therefore, they may have been measuring the effects of the programs, rather than antecedents to pregnancy.

Although expectations regarding education were lower for females who later became pregnant, the two groups did not significantly differ on expected career level. The results suggest that the occupational expectations of females who did not become pregnant were more realistic in terms of their educational expectations, unlike those for females who later became pregnant. It could be that it was beyond the ability of females who later became pregnant (and who, as determined by this study, had a more external locus of control) to comprehend the link between personal behaviors (such as abstaining from sex or obtaining an education) and outcomes (such as pregnancy or having a high-level career).

There were statistically significant differences between the two groups when careers were categorized into traditional versus nontraditional. Females who later became pregnant were more likely to expect to have a traditional occupation when they were 30 years old. These findings are in line with those of previous studies. For example, Ralph et al. (1984) noted poorer vocational-educational adjustment among pregnant teens as compared to a nulliparous group.

IMPLICATIONS

The results of the study have a number of implications for teen pregnancy prevention. If an external locus of control and poor self-efficacy are antecedents to pregnancy, then these psychosocial factors need to be addressed. Empowering children and families can be instrumental in reducing teen pregnancy. Once those in positions of power understand the impoverished sense of personal efficacy perceived by pregnant teens, they can better implement programs that foster an internal locus of control and a deep sense of "I can."

The findings relating to educational aspirations also have implications. As part of an attempt to prevent adolescent childbearing, programs should promote an awareness of the importance of higher education, starting with young children.

Interestingly, expected occupational level was independent of later pregnancy status. Although teens who later became pregnant had lower expectations in regard to educational attainment and finishing high school than did those who did not later become pregnant, they did not have lower expectations in regard to level of career. There are several possible explanations for this finding. It could be due to the poor perceived ability to produce effects (poor self-efficacy) of teens who later became pregnant, or their poor ability to associate career outcomes with such behaviors as going to college (external locus of control). It also could be that a perception of limited life options (an understanding of the discrepancy between their educational and occupational expectations) came after the eighth grade. Perhaps the later realization that the level of education for which they felt destined did not allow for the actualization of their earlier career aspirations presented a future that may not have seemed so bright . Producing a child may thus have been perceived as a desirable alternative to a life that seemed to otherwise lack purpose and meaning. In other words, could purpose and meaning have been compromised by the reality of what it took to obtain a high-level occupation? What effect did this reality have on those who at eighth grade had an external locus of control? Further research must be conducted to answer these questions and explain this finding. Nevertheless, any program designed to reduce teen pregnancy would need to address self-efficacy and locus of control, help children to see education (as well as abstinence) as both instrumental and achievable, help children to make the connection between behaviors and outcomes, and provide them with the skills to develop realistic strategies for meeting life goals.

Directions for Future Research

This study indicates that teens are more vulnerable to pregnancy if they lack internal locus of control, personal efficacy, and educational aspirations. Future research should investigate the links between this internal poverty, the consequences that have been causally attributed (from correlational studies) to out-of-wedlock births (Sec. 101, Personal Responsibility and Work Opportunity Reconciliation Act, 1996), and those health risks connected to external poverty. Socioeconomic poverty has been associated with our most serious societal ills. Data from Healthy People 2000: National Health Promotion and Disease Prevention (U.S. Dept. of Health and Human Service, 1991) indicated that people with low income are at greater risk for heart disease, most types of cancer, infectious diseases such as HIV, obesity, high blood pressure, tobacco use, infant mortality, and traumatic injury and death as compared to those who are not poor. Thus, researchers need to examine whether well-being disparities can be exclusively attributed to socioeconomic status (or unwed births), or whether socioeconomic status, along with behaviors conducive to health and success, can be attributed to perceived self-efficacy, locus of control, and other mediating psychological resources. The relationship between internal and external poverty must be identified in order for society to treat its major social ills.

The accumulating knowledge on the role of internal poverty in teen pregnancy, drug abuse, crime, and other social problems must be applied to policies that truly impact families. The time has come for enrichment programs to replace impoverishment with empowerment. Therefore, future studies should investigate the application and long-term impact of prevention/intervention programs that attempt to go beyond symptoms and get at the root of the problem.

The authors are grateful for the comments by William C. Bailey, Steven A. Dennis, M. Jean Turner, George S. Denny, and Mary M. Whan on a previous draft of this paper.

Sue S. Martin, School of Human Environmental Science, University of Arkansas, Fayetteville.

Michael E. Young, Program in Health Science, University of Arkansas, Fayetteville.

Ling Ting, Computing Services, University of Arkansas, Fayetteville.

Reprint requests to Tamera M. Young, Program in Human Development and Family Science, HE 118, University of Arkansas, Fayetteville, Arkansas 72701. Electronic mail may be sent to heal00@aol.com.

REFERENCES

Bandura, A. (1994). Self-efficacy. In Encyclopedia of human behavior (Vol. 4, pp. 71-81). San Diego, CA: Academic Press.

Bickel, R., Weaver, S., Williams, T., & Lange, L. (1997). Opportunity, community, and teen pregnancy in an Appalachian state. Journal of Educational Research, 90(3), 175-181.

Church, A. T., Teresa, J. S., Rosebrook, R., & Szendre, D. (1992). Self-efficacy for careers and occupational consideration in minority high school equivalency students. Journal of Counseling Psychology, 39(4), 498-508.

Devaney, B., & Hubley, K. (1981). The determinants of adolescent pregnancy and childbearing. Final Report to the National Institute of Child Health and Human Development. Washington, DC: Mathematica Policy Research.

Donovan, J. E., & Jessor, R. (1985). Structure of problem behavior in adolescence and young adulthood. Journal of Consulting and Clinical Psychology, 53, 890-904.

Dryfoos, J. (1984). A new strategy for preventing unintended teenage childbearing. Family Planning Perspectives, 16, 193-195.

Elster, A., Lamb, M., Peters, L., Kahn, J., & Tavare, J. (1987). Judicial involvement and conduct problems of fathers of infants born to adolescent mothers. Pediatrics, 79, 230-234.

Goh, D. S., Primavera, C., & Bartalini, G. (1996). Risk behaviors, self-efficacy, and AIDS prevention among adolescents. Journal of Psychology, 130, 537-546.

Gordon, C. P. (1996). Adolescent decision making: A broadly based theory and its application to the prevention of early pregnancy. Adolescence 31(123), 561-584.

Hanson, S., Myers, D., & Ginsburg, A. (1987). The role of responsibility and knowledge in reducing teenage out-of-wedlock childbearing. Journal of Marriage and the Family, 49, 241-256.

Holden, G. W., Nelson, P. B., Velasquez, J., & Ritchie, K. L. (1993). Cognitive, psychosocial, and reported sexual behavior differences between pregnant and nonpregnant adolescents. Adolescence, 28(111), 557-572.

Kaufman, P., & Rasinski, K. A. (1991). Quality of the responses of eighth-grade students in NELS:88 (NCES No. 91-487). Washington, DC: U.S. Department of Education, Office of Educational Research and Improvement.

Lewis, S. K., Ross, C. E., & Mirowski, J. (1999). Establishing a sense of personal control in the transition to adulthood. Social Forces, 77(4), 1573-1599.

Marcus, T. L., & Corsini, D. A. (1978). Parental expectations of preschool children as related to child gender and socioeconomic status. Child Development, 49, 243-246.

McCullough, M., & Scherman, A. (1991). Adolescent pregnancy: Contributing factors and strategies for prevention. Adolescence, 26, 809-816.

Mead, L. (1986). Beyond entitlement: The social obligations of citizenship. New York: Free Press.

Morgan, C., Chapar, G. N., & Fisher, M. (1995). Psychosocial variables associated with teenage pregnancy. Adolescence, 30, 277-289.

Mott, F. L., & Haurin, R. J. (1988). Linkages between sexual activity and alcohol and drug use among American adolescents. Family Planning Perspectives, 20, 129-136.

Personal Responsibility and Work Opportunity Act of 1996, Pub. L. No. 104-193, 110 Stat. 2105 (1996).

Plotnick, R. (1992). The effects of attitudes on teenage premarital pregnancy and its resolution. American Sociological Review, 57, 800-811.

Ralph, N., Lochman, J., & Thomas, T. (1984). Psychosocial characteristics of pregnant and nulliparous adolescents. Adolescence, 19, 283-294.

Rodriquez, C., & Moore, N. (1995). Perceptions of pregnant/parenting teens: Reframing issues for an integrated approach to pregnancy problems. Adolescence, 30, 685-706.

Roosa, M. W., Fitzgerald, H., & Carison, N. A. (1982). Teenage parenting and child development: A literature review. Infant Mental Health Journal, 3, 4-18.

Rotter, J. B. (1975). Some problems and misconceptions related to the construct of internal versus external control of reinforcement. Journal of Consulting and Clinical Psychology, 43, 56-67.

St. Lawrence, J. S., Brasfield, T. L., Jefferson, K. W., Allyene, E., & Shirley, A. (1994). Social support as a factor in African-American adolescents' sexual risk behavior. Journal of Adolescent Research, 9, 292-310.

Sullivan, M. (1993). Culture and class as determinants of out-of-wedlock childbearing and poverty during late adolescence. Journal of Research on Adolescence, 3, 295-316.

U.S. Dept. of Education, National Center for Education Statistics. (1990). National Education Longitudinal Study, 1988 [computer file]. Chicago, IL: National Opinion Research Center [producer], 1989. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor].

U.S. Dept. of Education, National Center for Education Statistics. (1993a). National Education Longitudinal Study, 1988: First Follow-Up (1990) [Dropout Data] [computer file]. U.S. Dept. of Education, Office of Educational Research and Improvement [producer], 1992. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor].

U.S. Dept. of Education, National Center for Education Statistics. (1993b). National Education Longitudinal Study, 1988: First Follow-Up [Student Data, Vol. I & II] [computer file]. U.S. Dept. of Education, Office of Educational Research and Improvement [producer], 1992. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor].

U.S. Dept. of Education, National Center for Education Statistics. (1995a). National Education Longitudinal Study, 1988: Second Follow-Up (1992) [Dropout Data] [computer file]. U.S. Dept. of Education, Office of Educational Research and Improvement [producer], 1994. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor].

U.S. Dept. of Education, National Center for Education Statistics. (1995b). National Education Longitudinal Study, 1988: Second Follow-Up (1992) [Student Data] [computer file]. U.S. Dept. of Education, Office of Educational Research and Improvement [producer], 1994. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor].

U.S. Dept. of Health and Human Services. (1991). Healthy People 2000: National health promotion and disease objectives (DHHS Publication No. PHS 91-50213). Washington, DC: U.S. Government Printing Office.

U.S. Department of Health and Human Services. (1995). Report to Congress on out-of-wedlock childbearing (DHHS Publication No. PHS 95-1257-1). Hyattsville, MD: U.S. Government Printing Office.

Zimmerman, R. S., Sprecher, S., Langer, L. M., & Holloway, C. D. (1995). Adolescents' perceived ability to say "no" to unwanted sex. Journal of Adolescent Research, 10, 383-399.

Table 1
ANOVA for Locus of Control by Later
Pregnancy Status
Variable n M SD F p
Later Never Pregnant 878 16.670 2.766 64.42 .0001
Later Pregnant 897 15.637 2.657
Source df Sum of Squares Mean Square
Model 1 473.593 473.593
Error 1773 13033.735 7.351
Total 1774 13507.327
Table 2
ANOVA for Personal Efficacy by
Later Pregnancy Status
Variable n M SD F p
Later Never Pregnant 878 10.836 1.847 42.07 .0001
Later Pregnant 897 10.272 1.816
Source df Sum of Squares Mean Square
Model 1 141.125 141.125
Error 1773 5948.010 3.355
Total 1774 6089.136
Table 3
Chi-Square Test of Independence: Pregnancy and Expectations
Variable n df
Occupational expectation (career level) 1,791 2
Occupational expectation (traditional,
 nontraditional) [*] 1,791 1
Highest education expected [**] 1,833 5
How sure of high school graduation [**] 1,829 3
Variable [[chi].sup.2]
Occupational expectation (career level) 3.885
Occupational expectation (traditional,
 nontraditional) [*] 7.689
Highest education expected [**] 90.282
How sure of high school graduation [**] 49.301
Variable p
Occupational expectation (career level) .143
Occupational expectation (traditional,
 nontraditional) [*] .006
Highest education expected [**] .001
How sure of high school graduation [**] .001
(*)Females who later became pregnant had more traditional
occupational expectations than did those who did not become
pregnant.
(**)Females who later became pregnant had lower expectations
in regard to educational attainment and finishing high
school did those who did not become pregnant.