The Official Journal
The North Carolina
Sociological Association: A
Refereed Web-Based Publication
Volume 1, Number 2
Power in North Carolina Parents:
Is There a Relationship Between
Family Structure and Adolescent Self-Efficacy?
The purpose of this research is to study the effects of family form on the self-efficacy of high school adolescents. Teens were chosen because they can cognitively comprehend the forces leading to a parental marital breakup or choices for parental singlehood. Additionally, they can potentially operate as a household partner and can self-report without parental intrusion. Yet, at the same time, such teens are greatly dependent upon their parents for economic and self-identity support. This means parental choices affecting divorce, cohabitation or remarriage could have pivotal consequences on these teens' self- efficacies when they are in the midst of making academic, vocational, and partner choices.
Our research agenda is to explore what factors impact an adolescent's psychosocial well-being. Bandura (1977) introduces self-efficacy as a valid measure of one's psychosocial well-being. He suggests an individuals belief regarding their ability to perform an efficacious action comes from a variety of socially constructed elements: performance outcomes, vicarious learning, emotional arousal, and verbal persuasion. Therefore, the social world of an individual is vitally connected to the perceived level of self-efficacy. Consequently, we will explore whether or not the family form in which a teen lives uniformly affects an adolescent's self-efficacy.
Family Structure and Adolescent Self-Efficacy
Self-efficacy theory suggests one's level of self-efficacy indicates one's willingness to approach the world as a change agent (Bandura, 1977, 1981, 1995, 1997). In effect, one is willing to try and accomplish more, expects to accomplish more, and is, indeed, motivated to achieve as one's level of self-efficacy increases. If family structure affects global self-efficacy uniformly, then the children within certain family forms are put at an automatic advantage interpersonally while others may likely be low achievers in a variety of arenas. However, if certain variables, such as the economic and emotional conditions of the family, the length of time since a separation or circumstances of custodial parent remarriage, mediate or increase self-efficacy, then the circumstances of a particular family setting would not have a direct impact on efficacy.
People live their lives within the context of external and internal barriers that impact their levels of self-efficacy (Bandura, 1982, 1986). External barriers arise from the material and social constructions of life. Internal barriers arise from the socialized and the experienced conditions of psychosocial perceptions. Individuals, therefore, face both real constrictions to action and internalized values and norms that affect efficacious action. A number of structural conditions impact self-efficacy. Social class, race, level of education, urban or rural background and gender all affect an individual's global self-efficacy (Birch, 1987). In effect, those lower on the hierarchy of the social world or distressed by life events are less likely to perceive themselves as capable of altering their social world. Those at the top of, or those who are supported by, their social networks are more likely to see themselves as capable of accomplishing a great deal.
Living in a single-parent home or dividing economic
resources following a separation can create both real and perceived economic
hardships on families. Divorce typically produces significant declines
in the standard of living in mother-headed homes, with life below the poverty
line often the result (Duncan and Hoffman, 1985). A child's comparison
of resources prior to and after the separation clearly displays a fall
in material possessions and social status. Negative outcomes, both psychological
and behavioral, are linked to poverty or perceived economic deprivations
(Amato and Booth, 1991).
The Family Structure Model
Family structure is expected to have a direct relationship to adolescent self-efficacy (Bandura, 1995; Schneewind, 1995). If non-traditional family forms hold uniform negative outcomes for children, then teens within such households should expect to have lower levels of self-efficacy. Thus, we hypothesize:
H1: Levels of adolescent self-efficacy
should be negatively correlated with living in a divorced, blended, never
married or cohabiting household compared to teens living in a first-marriage
The Economic Model
Economic status for teens in single-parent families and families experiencing a divorce are significantly lower than teens in intact or reconstituted families (Duncan and Hoffman, 1985). In such situations teens are not only susceptible to the lowering of self-efficacy related to income (Bandura, 1995; Birch, 1987), but also to a perceived loss to the self related to social comparisons (Rosenberg and Simmons, 1972). We propose:
H2: Levels of adolescent self-efficacy should be positively related to family income, economic resources and financial well-being.
H3: Levels of adolescent self-efficacy should be positively
related to positive expectations for an increase in family financial well-being.
The data gathered for this analysis is from The Youth and Careers Study of 1993 conducted by Dr. Luther Otto. Collected from 163 high school juniors paired with their mothers or guardian mothers from six high schools in two county school systems within North Carolina. The sample is a purposive cross-sectional design of 73 males and 90 females. Several national data sets are available that consider various aspects of adolescents' lives including self-esteem; however, to our knowledge none of these data sets contained measures of adolescent self-efficacy. Our desire to consider this variable in particular necessitated the use of this very specialized sample.
Demographically, the state is primarily Caucasian, with the African-American population comprising about one-fifth of the total. According to the 1990 Census, minorities comprised about 24.4 percent of the states population. The median state income for households in 1990 was $26,647 and 57.3% of the states households represent married couples with 14.9% single parent households (US Bureau of the Census, 1992).
Two counties were selected, one representing the traditional regionalism and the other the rapid growth and development region of the state. The traditional county is located in the western foothills. The rapid change county is located in the southeastern coastal plain. While the counties are rural in definition, a major interstate highway intersects both of them. Both have active distribution and service industries but no cities larger than 25,000 in population. Therefore, while the samples may represent the diversity of the state, they do not reach into the larger metropolitan centers nor the more isolated mountain communities. Both counties report a lower median household income than the state as a whole. However, in terms of proportion of divorced/separated households and households with children, they fall in line with the total population (US Bureau of the Census, 1992).
Both county school boards approved the survey research and each school was given a small grant commensurate with the number of matched surveys returned. All instruments were pre-tested, results coded, data analyzed and the instruments edited to assure the quality of the projects. Only juniors enrolled in required courses were eligible to participate. Project associates surveyed all students from each school on the same day within scheduled classroom settings.
The dependent variable in this study is a measure of adolescent global self-efficacy utilizing a standard Likert-type five item additive scale (Sherer & Adams, 1983). Subjects were asked how well they felt the following statements described themselves: When trying to learn something new, I soon give up if I am not initially successful; I give up easily; I give up on things before completing them; I avoid trying new things when they look too difficult to me; If something looks too complicated, I will not even bother to try it. These items were scored 1 = strongly disagree, 2 = disagree, 3 = agree, 4 = strongly agree. They were recoded to show higher scores meaning a higher level of self-efficacy. A reliability analysis of the five items indicated a mean inter-item correlation of .5105 and a standardized item alpha of .8391 indicating this scale is multidimentionally coherent and the items used provide the best fit for the scale (Agresti and Finlay, 1986).
Independent variables used in this study include six control variables: sex (coded as dummy variable of 0 = female and 1 = male), race (coded as a dummy variable of 0= nonwhite and 1= white), mother and fathers educational attainment (measured in levels of education completed), students grade point average (GPA from 0 to 4) obtained directly from school records and providing a cumulative average at the time of the study. Parents educational status, taken from the mothers survey, was utilized as a class measurement and, as school is a significant arena of life for the vast majority of teens, the use of grade point average is chosen to control for the expected effects of academic performance upon adolescent self-efficacy.
A series of dummy variables measuring the teens family structure were constructed from both the mothers and students surveys looking at first marriage, blended, divorced, never married and cohabiting families. Married or non-disrupted households were always coded 0.
In the economic model, students and mothers were asked questions about their family financial situation. First, mothers were asked global family income (coded in thousands of dollars in household income per year). Additionally, teens were to compare themselves with other families they knew. How would you describe your family financial situation over the past five years? Responses were: 1= we've lived much worse than most families, 2 = eve lived not as well as most families, 3 = eve lived about the same as most families, 4 = eve lived a little better than most families, and 5 = eve lived much better than most families. A third question asked their view of the family's economic well-being in the future. Over the next five years, would you say that your family finances will be: 1 = getting worse; 2 = staying about the same, 3 = getting better.
The data were analyzed using a series of multiple
regression models. By using a comparative appraisal of the two theoretical
models, we are able to delineate which of the two models provides the
greatest descriptive and analytical power.
The sample for this research was 55% female and 16% non-white. Twenty percent of the teens reported living in a family setting other than a first marriage household, with 8% of the mothers reporting being currently divorced or separated. Two percent of the mothers report they were never married. Matching this are teen reports of 10% of the sample group living in stepfamily households. Only three households (2%) reported a marital disruption within the past year. 93percent of the mothers reported the household make up had not changed within the last three years. The median grade point average of the sample was 2.824.
Median household income was well above the state average of $26,000 at just over $45,000. Households averaged under two children with a maternal age of 42. The vast majority of both parents were employed, with only 9% of the fathers and 2% of the mothers without a current paycheck. Only 7% of the teens perceived their family finances were below the community average, while almost half felt they lived above that standard. This sample was over represented by teens from financially secure households with working mothers, creating a potential class bias in the data.
To examine the question on the relationship of adolescent self-efficacy and family structure, we performed a series of OLS regressions for each outcome hypothesis separately. Each model held a series of comparative regression analyses. By entering family structure variables toward the end of these equations, we show the effect of family structure against other theoretically important variables. Each of these is explored in order.
The results of the first model are in Table 2. Model 1 shows the effect of all control variables and family structure on adolescent self-efficacy. Race and the students grade point average are statistically significant, suggesting being white and achieving higher levels of academic performance are positively correlated to self-efficacy. Neither sex of the student, the presence of siblings nor the educational attainment of parents are significant factors in the equation.
None of the family structure variables (Model 2) are statistically significant. In effect, whether a teen resides in a first marriage, divorced, blended, never married or cohabitating household makes no difference in teen self-efficacy outcomes. This is surprising considering both the cultural expectation of pathology in "non-traditional" households and the literature noted earlier. These results strongly suggest, at least where efficacious issues are concerned, the variations in household structure are inconsequential.
The second theoretical perspective proposed the overall economic conditions under which a teen lives are directly linked to levels of self-efficacy. The results of the OLS regression equations to test the effects of economic stress on adolescent self-efficacy are found in Table 3.
Hypothesis 2 predicted higher levels of adolescent self-efficacy would be linked to higher levels of family income and economic well-being, supposing a teen would feel more powerful with access to adequate financial resources. Model 1 does not support this hypothesis. However, when adolescent's perception that her/his family's economic situation is better than those of her/his peers is included in the equation along with income, both are significant (See Model 2). Adolescent's perception is positively correlated to self-efficacy, while income is negatively correlated. Model 3 does not indicate a better fit, as neither the equation itself is as robust in either statistical significance nor R2.
Again, none of the family structure variables approaches significance in any model. Teens from single parent households would be expected to show lower levels of self-efficacy owing to the psychosocial and economic stresses associated with life in such a home. Following that, teens who live in a blended household would have such economic stresses mediated and hold levels of self-efficacy somewhere between youth in a single-parent and first marriage household. All three equations test this, yet none were statistically significant.
Discussions and Conclusions
The results of the study are eye catching. The main findings of this study provide no support for the general hypothesis that family structure holds any significant detrimental relationship to adolescent self-efficacy. Contrary to the proposed hypothesis, being raised in a non-normative family structure has no effect upon the measured levels of adolescent self-efficacy. Teens self-efficacy is either much more resilient to family disturbances and social stigma or, in the face of such elements, youth garner other sources to maintain high levels of perceived agency.
When testing Hypothesis 2, income is not, by itself, significant to adolescent self-efficacy. Income only becomes modestly significant when in the presence of a teens self-evaluation comparison of her or his family's economic situation within the community. Teens who perceive their family as having more economic resources in comparison to other families in their vicinity, whether it is true or not, see themselves and their significant others as holders of wealth and financial power, leading to increased self-efficacy.
Yet, surprising and contrary to prediction, while
the family's higher comparative financial status is positively correlated
to youth self-efficacy, income is negatively directed. This means
youth from lower income families hold higher levels of self-efficacy.
There are a number of possible explanations. First, youth in lower income families may live in a social environment that encourages agency. In effect, the belief is one must accomplish things on her or his own. Second, many lower income youth, in comparison to teens from wealthier households, may hold part-time jobs that elevate self-efficacy through both the work and their paycheck. Third, as the median income of the data set is well above the income averages for the counties, the effect of a heavy middle income bias may be seen. In this case, income may hold some curvilinear effect. Further analysis is needed to clarify this relationship.
Among the control variables we found across models that being white was positively correlated to adolescent global self-efficacy. While minority teens may glean self-esteem from family and peers on levels equal to that of whites (Rosenberg & Simmons, 1972), in the rural south, where endemic racism still has a significant hold and where minorities are disadvantaged in both the academic and vocational arenas, finding white teens feeling more powerful and efficacious in their world is not surprising. Had the sample been drawn from an urban area or a community of youth where ethnic mix is more even, the results may show a different pattern.
As expected, there was a consistent positive correlation across models between teen self-efficacy and the students Grade Point Average. Self-efficacy has been linked to intellectual development and cognitive ability as well as to school performance and career development (Bandura, 1995).
None of the issues raised suggest an endemic
pathology to family structure. In fact, the research suggests a null effect
(Davison 2003) of family structure on adolescent self-efficacy, at least
in a relatively small sample on a localized level. Greenstein (1999) and
Hines (1997) indicate that the emotional support of parents makes the difference
in the lives of children. The strongest findings of this research support
that contention. If this is the case, the uniform disdain that has come from
conservatives related to family form is to be questioned.
Programs and policies that attempt to solve adolescent
problems by encouraging two adults to stay married are not likely to be
a panacea to solving all adolescent ills. Our research suggests the need
for policy considerations that provide various support systems to parents,
such as helping adults communicate with teens (e.g., teen led parent-teacher-student
conferences or flexible work schedules that encourage parents to spend
time with their children) and encouraging administrators to rethink the social
role of non-athletic events in the life of the school. Ultimately, the quality
of the relationship with parents, not the structure of the family, needs
to be our focus.
Table 1: Means and Standard Deviations for Self-Efficacy Sample (N=163)
Table 2: Family Structure Regression (N=163)
Dependent variable = adolescent self-efficacy
*= significant at the .05 level
**=significant at the .01 level
***=significant at the .001 level
Non-standardized betas in parentheses.
Table 3: Economic Deprivation Regression (N=163)
Dependent variable = adolescent self-efficacy
*= significant at the .05 level
**=significant at the .01 level
***=significant at the .001 level
Non-standardized betas in parentheses.
Agresti, Alan and Barbara Finlay. 1986. Statistical Methods
for the Social Sciences. San Francisco, CA: Dellen Publishing Company.
Amato, Paul and Alan Booth. 1991. Consequences of Parental Divorce
and Marital Unhappiness for Adult Well-Being. Social Forces,
69 (3): 895-914.
Bandura, Albert. 1977. "Self-Efficacy: Toward a Unified Theory of
Behavioral Change." Psychological Review, 84: 191-215.
Bandura, Albert. 1981. The Assessment and Predictive Generality
of Self-Precepts of Efficacy. Journal of Behavior Therapy and Experimental
Psychiatry, 13: 195-199.
Bandura, Albert. 1982. "Self-Efficacy Mechanism in Human Agency."
American Psychologist, 37:122-147.
Bandura, Albert. 1986. "The Explanatory and Predictive Scope of
Self-Efficacy Theory." Journal of Social and Clinical Psychology,
Bandura, Albert. 1995. Self-Efficacy in Changing Societies.
New York, NY: Cambridge University Press.
Bandura, Albert. 1997. Self-Efficacy: The Exercise of Control.
New York, NY: Freeman Publishing.
Birch, Dianne. 1987. Adolescents Whose Parents are Divorced:
An Interview Study and Ethnographic Analysis. Blacksburg,Virginia:
Unpublished Doctoral Dissertation.
Davison, Beth. 2003. "In Search of the Null." Presidential Address for the North Carolina Sociological
Association Meetings, Boone, NC.
Demo, David & Alan Acock. 1988. The Impact of Divorce on Children.
Journal of Marriage and the Family, 50: 619-648
Duncan, G.J., and S.D. Hoffman. 1985. "Economic consequences of
Marital Instability." In M. David and T. Smeeding (eds.), Horizontal
Equity, Uncertainty and Economic Well-Being, pp. 427-467. Chicago, IL:
University of Chicago Press.
Greenstein, Theodore. 1999. "Employment, Early Maternal." In C.A.
Smith (ed.), The Encyclopedia of Parenting Theory and Research, pp.154-155.
Westport, CT:Greenwood Press.
Hines, Alice. 1997. Divorce-Related Transitions, Adolescent Development
and the Role of the Parent-Child Relationship: A Review of the Literature.
Journal of Marriage and the Family, 59: 375-388.
Rosenberg, Morice, and R. Simmons. 1972. Black and White Self-Esteem:
The Urban School Child. Washington, D.C.:American Sociological Association.
Schneewind, Klaus. 1995. Impact of Family Processes on Control Beliefs.
In Albert Bandura (ed.), Self-Efficacy in Changing Societies,
pp. 114-148. New York, NY: Cambridge University Press.
Sherer, Mark and Carol Adams. 1983. "Construct Validation of the
Self-Efficacy Scale." Psychological Reports, 53(3): 899-902.
US Bureau of the Census. 1992. Current Population Reports,
Household, Families and Children: A 30-Year Perspective. Washington,
DC:US Government Printing Office.
© 2003 North Carolina Sociological Association