Adolescent Responsibility 1 Running head: ADOLESCENT RESPONSIBILITY AND DEVELOPMENT PROBLEMS Adolescent Responsibility in Divorced Families: Internalizing Problems and Self-Esteem Shawna M. Andersen The University of Puget Sound Adolescent Responsibility 2 Introduction With a high rate of divorce in the United States, as well as in other countries, recent psychological research has focused on the multiple differences and similarities between divorced and intact families. The societal views on divorce seem to be changing from that of a negative occurrence to that of a more normative occurrence. One of the major underlying questions of psychologists when studying divorce is how such an occurrence may either positively or negatively impact a child or family. Research is shifting from the search for how a specific family structure may impact a child to the search for the underlying processes occurring in relation to diverse families. In recent research on the changing societal views of divorce, Amato (2001) found that even when taking into account the improvement of society‟s views on divorce, children of divorce continue to show numerous negative outcomes. Amato discusses the changes in outcomes of children from divorced families as compared to those from intact families. After taking into account the more normative views of divorce during recent times, the study shows a persistence of more negative outcomes among children of divorced parents in the 1990s than those of earlier decades. Lower levels of success at school, poorer behavior, more emotional problems, lower self-esteem, and more difficulties with interpersonal relationships are some of the negative outcomes found in children of divorce from recent times. If societal views are not responsible for differences between divorced and intact families, this brings to question what specific factors may be contributing to the differences seen. Researchers of divorce have not only assessed the negative outcomes of children in divorced families, but there has been research looking at the possible beneficial outcomes in children of divorce as well. Weiss (1979) was one of the first researchers to assess how responsibility levels of children from single-parent households may differ than those of dual- Adolescent Responsibility 3 parent households. He proposed the idea that those in divorced, single-parent households tend to take on a higher level of responsibility than the children of two parents. Weiss found some positive outcomes in children from divorced families, including higher self-esteem, greater independence, and increased feelings of competence, in relation to a child‟s increased levels of responsibility and power within the family. This model illustrates that when children are held at a higher level of accountability they may tend to “grow up faster” than those who are held at lower levels. The positive outcomes found in this study were limited, and some children showed them in combination with negative outcomes, including emotional stress and behavioral problems. However, this idea of growing up faster led to many further studies on responsibility and children of divorce. Additional evidence shows support for a relationship between children of divorce who are given more responsibility and a heightened sense of self-esteem (Barber & Eccles, 1992). Along with a number of other outcomes, findings demonstrate that increased household responsibility may be related to positive outcomes in children of divorce, such as rating themselves as self-sufficient, skilled, and responsible. This study is limited in its implications for generalizing because it did not directly examine a child‟s level of responsibility in relation to particular outcomes. However, inconsistency in the findings between whether divorce impacts a child positively or negatively points to the specific areas of study that require a deeper assessment. With such a large number of factors playing into divorced families, it is difficult to determine whether divorce per se is negative or positive for a child‟s outcome. The investigation of specific factors, such as responsibility levels, which are related to divorced families, will lead to more knowledge about the circumstances that may be beneficial or detrimental to a child‟s development. Adolescent Responsibility 4 Hetherington (1989) discusses numerous discrepant findings about the effects of divorce on children‟s development. Interestingly, she also found evidence to support a relationship between high levels of responsibility and positive outcomes in children of divorce. Within a longitudinal study of the effects of divorce on child outcomes, Hetherington found a small sample of girls from divorced families showing positive outcomes. This sample was composed of girls around the age of ten who played an especially responsible role within a divorced household and showed higher levels of independence. These girls displayed low externalizing and high prosocial behaviors. However, it is important to take into account the longitudinal nature of this study began with four-year-olds and followed up at the age of ten. Thus, the idea of higher responsibility as related to positive outcomes is limited in this study to girls in early and middle childhood. Hetherington and Stanley-Hagan (1999) discuss important follow-up research with the same sample of girls who showed positive outcomes in the context of heightened levels of responsibility. They found evidence supporting a relationship between long-term responsibility levels and the tendency to have internalizing problems and low self-worth in late adolescence. The girls who displayed these negative outcomes were not only overburdened with household responsibility, but also the emotional support of their divorced, single-parent mothers during adolescence. Many studies may not have recognized this pattern of outcomes due to their largely internal nature (e.g., internalized emotions and levels of self-worth). This study, along with the earlier studies by Hetherington (1989) and Barber and Eccles (1992), bring up an interesting question about the impact of both household responsibility levels and emotional responsibility levels of adolescents from divorced families. Levels of responsibility for children of divorce have been studied in terms of household Adolescent Responsibility 5 chores (e.g., Barber & Eccles,1992) but also as the emotional support of a parent (e.g., Hetherington & Stanley-Hagan, 1999). It is important to assess both the emotional as well as the physical responsibilities of a child from a divorced family. There have been more recent studies regarding the effects of increased maternal disclosure to children after a divorce. Silverberg, Wallace, Lehman, and Raymond (2002) investigated some of the risk factors in child outcomes in relation to divorced, single-mothers who showed an increased level of disclosure to their daughters about sensitive topics. The results of this study provided evidence to support that a heightened level of mother-to-daughter disclosure about difficult topics is related to an increased level of the daughter‟s worrying about these issues and psychological distress. Another study about maternal disclosure examined both male and female adolescents from divorced, single-mother households (Silverberg, Wallace, & Lee, 2004). Silverberg et al. suggest that there may be initial short-term positive outcomes in children who experience increased maternal disclosure, such as feelings of importance. Disclosure may also help with development of coping strategies post-divorce. However, the researchers describe many long- term negative outcomes of especially high levels of disclosure to adolescents, such as mother- child role-reversal and adolescents showing psychological distress or anxiety. These findings, in concordance with the findings about increased household responsibility in divorce, point to the same underlying question regarding the specific long-term effects of overall responsibility on children from divorced families. Specifically, there is a lack of research that looks directly at the effects of different levels of both emotional and physical responsibilities of children from divorced families. An overburden of responsibilities of children in divorced families may indicate that these children are being treated as adults. To add to the previous research, the proposed study will evaluate the level of Adolescent Responsibility 6 responsibility as a specific factor that may influence the outcome of a child from a divorced family. However, the proposed study will analyze responsibility under different terms than much of the previous research. The proposed study suggests the necessity of assessing overall responsibility, in both physical and emotional terms. Increased maternal disclosure about sensitive topics has been described as an adolescent‟s increased emotional support of their mother, in studies by Silverberg et al. (2002, 2004). Therefore, emotional responsibility will be examined through parental disclosure to an adolescent. Furthermore, general household responsibility will be included in order to assess the more physical aspects of responsibility. In combining both household responsibility and parental disclosure, the proposed study will be able to tap into the same underlying construct. These aspects of responsibility are conceptually similar and will be combined to assess a type of global responsibility for children of divorce. The proposed study will assess the relations of long-term levels of overall responsibility, including household responsibility and parental disclosure, to internalizing problems and self- esteem levels in adolescents of divorced, single parents. The levels of each measure will be taken in terms of the adolescent‟s perceptions. Results are expected to reveal a curvilinear relationship, such that heightened levels of both household responsibility and parental disclosure to adolescents about sensitive topics will be related to more internalizing problems and lower self-esteem, moderate levels will be related to fewer internalizing problems and higher self- esteem, and very low levels of responsibility and disclosure will be related to more internalizing problems and lower self-esteem. Of particular interest are the effects of very high and very low levels of responsibility and disclosure over a period of time. These curvilinear results are expected for a number of reasons. As previously discussed studies by Weiss (1979) and Barber and Eccles (1992) show, Adolescent Responsibility 7 some level of household responsibility may be beneficial in fostering a positive image of one‟s self and feelings of self-worth within a family structure. It may be important to give children a certain level of responsibility in order to help them to learn about being responsible and developing feelings of independence. The proposed study suggests that the ideal level of household responsibility that is most beneficial for children of divorce is one that is moderate. Too much responsibility may overburden a child and show negative outcomes of more internalizing problems and lower self-esteem. Since the proposed study suggests that household responsibility and parental disclosure tap into the same concept of overall responsibility, there may also be an ideal moderate level of parental disclosure. As research by Silverberg et al. (2002, 2004) indicates, too much maternal disclosure can be detrimental. However, these researchers also discuss the importance of some recognition of the existence of sensitive divorce issues and the necessity to discuss these issues. If there is a moderate level of discussion about divorce and sensitive issues, the child may not show negative outcomes of low self-esteem and more internalizing problems. Of particular interest is how adolescents respond to excessive levels of responsibility, in terms of household responsibility and parental disclosure about these issues over a period of time. The expectation is to find that extremely low or high levels of overall responsibility will show a negative impact on the adolescent. According to the previous discussion of research, including Silverberg et al. (2002, 2004) and Weiss (1979), very low levels are expected to possibly prohibit an adolescent from developing normally. Under these circumstances, a divorced parent may be shielding an adolescent and treating him or her as a child. The proposed study anticipates these adolescents to show lower self-esteem and more internalizing problems. Extremely high levels of overall responsibility are expected to be related to similar outcomes. Adolescent Responsibility 8 An adolescent with very high levels of responsibility may develop low self-esteem and more internalizing problems through being held accountable for a burdensome load of responsibility. The proposed study suggests that when divorced parents treat an adolescent as an adult, through expecting extremely high levels of responsibility, there may be negative outcomes in the adolescent. These issues are important to assess for a number of reasons. Unlike much of the previous research, the proposed study will look at both male and female adolescents in maternal and paternal single-parent households. With our changing society, there is reason to believe that there may be an increasing number of single fathers. This is why it is important to assess both fathers and mothers in single-parent households. As Silverberg et al. (2004) illustrates, boys may also be affected by parental disclosure, therefore, indicating why the proposed study includes both male and female adolescents. Since studies by Hetherington (1989, 1999) show that responsibility may have different effects on the outcomes of children than on late adolescents, the proposed study will assess late adolescents. The proposed study will attempt to find not merely an initial impact of divorce, but the longer-term effects. In order to assess overall responsibility as it relates to the long-term development of adolescents, the proposed study will only take into account adolescents who have been living with their divorced, single-parent for at least two years. If research concerning adolescents‟ responsibility were to be taken too soon after divorce, there would be a possibility for the results indicating initial levels of self-esteem and internalizing problems. The participants will be asked to fill out surveys in order to assess their levels of household responsibility, their parent‟s disclosure to them, their internalizing problems, and their self-esteem. It is expected that the evidence from this study will support a statistically significant curvilinear relationship between Adolescent Responsibility 9 overall responsibility and internalizing problems, and between overall responsibility and self- esteem. This evidence will help to identify another important factor as to why there are so many differences in outcomes of adolescents from divorced families. It will add to the research on divorce and show how the structure of a family may not affect a child as much as the underlying factors of divorced families. Method Participants This study will include data from a sample of 200 adolescents in divorced, single-parent households. The participants will range in age from 16 through 18 years old. The participants will be used in the study only if they have been living in a divorced, single-parent household for at least two years. The study will be between-family in nature, by only assessing one child from a given family. However, descriptive data will be collected on the total number of children in the family. The number of male and female children will be approximately equal. The number of father-child versus mother-child dyads will be as equivalent as possible, given that the number of divorced, single-mother-child dyads in the United States is much greater than that of father- child dyads. The sample of adolescents will be a random, nationally representative sample of divorced, single-parent households in terms of parental education, ethnicity, household income, and family size. Procedure The participants will be contacted by telephone and asked to participant in the proposed study. After agreeing to participate in this study, each participant will be mailed a packet of all four surveys. The participants will be asked to complete each survey thoughtfully and fully. With the parent‟s consent, each adolescent will be asked to have no help or outside influence on Adolescent Responsibility 10 completing the surveys. An envelope with postage will be included for the participants to return the completed surveys by mail. Measures Participants will first be asked a set of descriptive questions. These will include questions about the participant‟s gender, age, amount of time living in a single-parent household, gender of single-parent, and number of siblings. Four constructs are of primary interest in the proposed study: (a) the level of parental disclosure as perceived by the adolescent; (b) the adolescent‟s level of perceived household responsibility; (c) the adolescent‟s level of internalizing problems; and (d) the adolescent‟s level of self-esteem. Parental disclosure. The instrument that will be used to measure the level of parental disclosure to the adolescent will assess various sensitive issues surrounding divorce. This instrument is a revised version of the scale used to measure mother-to-adolescent disclosure in the study by Silverberg et al. (2004). This scale was originally taken from the Parents‟ Disclosure Questionnaire and Parent Disclosure Survey, and adapted to focus only on items that would relate to sensitive issues discussed in divorced families. The topic areas of parental disclosure in this survey are finances, ex-husband, job ups-and-downs, parenting challenges, and personal concerns. The scale lists 26 items and response choices are as follows: 0= “My mother never says anything to me about this topic”; 1= “This topic has accidentally „popped out‟ a couple of times, when my mother didn‟t really mean it to”; 2= “My mother has talked with me about this topic, but without giving any details”; 3= “My mother has talked with me about this topic, giving a few details”; 4= “My mother has talked with me about this topic, going into quite a bit of detail”; and a choice that the topic does not apply to the participant. This scale will be revised for the proposed study, in order to address father-disclosure as well. Since the topics are Adolescent Responsibility 11 relatively gender neutral, the revisions made will be minimal, in exchanging “mother/father” for “mother” and “ex-husband/ex-wife” for “ex-husband”. After completion, the items that do not apply to the participant will first be removed. The ratings from the rest of the 26 items will be averaged in order to have a single score for parental disclosure. Higher scores indicate a higher level of parental disclosure about these topics. Household responsibility. The level of an adolescent‟s perceived household responsibility will be measured with a modified version of the Family Responsibility Index (FRI); (Bjorkquist, 2000). This is a 53-item instrument that is used to measure behaviors in family role responsibilities. The specific behaviors include yard work, laundry, house care and upkeep, kitchen clean-up, family business, housecleaning, car care, heavy housecleaning, family care, and preparing meals. The FRI is typically used to measure a spouse‟s level of responsibility in these tasks. Statements such as, “Mow lawn”, “Wash clothes”, “Put dishes away”, and “Care for family pets” are rated on a scale from 1 to 5, including 0= “Does not apply”. The instrument will be revised in order to assess an adolescent‟s level of responsibility in the same tasks that would be expected of an adult. The participant will be given options from 1= “Never participate in this task” to 5= “Always participate in this task” for each statement. Upon completion, the items that are responded with “Does not apply” will first be removed. The ratings from the rest of the 53 items will be averaged in order to have one score for household responsibility. Higher numbers represent a greater amount of household responsibility. Internalizing problems. In order to measure the level of an adolescent‟s internalizing problems, participants will complete the emotional tone subscale from the Self-Image Questionnaire (Galambos, Barker, & Almeida, 2003). This subscale is an 11-item instrument that was used by Galambos et al. to assess internalizing problems in adolescents over a period of Adolescent Responsibility 12 time. The questions pertain to depressive and anxious affect and include items such as, “I frequently feel sad” and “I am so very nervous”. Answers range from: 1= “Does not describe me at all” to 6= “Describes me very well”. The ratings from all of the 11 items will be averaged in order to have one score for internalizing problems. Higher scores are meant to indicate higher levels of internalizing problems. Self-esteem. Finally, the State-Self Esteem Scale will be used to measure the participants‟ levels of self-esteem (Polivy & Heatherton, 1991). This is a 20-item instrument used to assess self-esteem in terms of performance, social relations, and appearance. The scale is validated to measure the level of self-esteem as a state of being, instead of as an unchanging characteristic. Participants rate themselves on a scale of: 1= “Not at all” to 5= “Extremely”, on statements such as, “I am pleased with my appearance right now”, “I feel confident about my abilities”, and “I am worried what other people think of me”. The ratings from all 20 items will be averaged in order to have one score for self-esteem. Higher numbers indicate a lower level of self-esteem. These numbers will be inverted so that higher numbers indicated a higher level of self-esteem. Data Analysis The analyses that will be conducted in the proposed study will test the following hypotheses: Adolescents from divorced, single-parent households will display a curvilinear relationship between their levels of overall responsibility and internalizing problems, and between their levels of overall responsibility and self-esteem. Analyses will begin by obtaining an averaged rating for each instrument. A primary set of data analyses will be conducted in order to assess the descriptive statistics of the participant population and their responses to the surveys. The means and standard deviations will be calculated for the following items: (a) age Adolescent Responsibility 13 of participants, (b) amount of time living in a divorced, single-parent household, (c) number of siblings, (d) level of parental disclosure, (e) level of household responsibility, (f) level of internalizing problems, and (g) level of self-esteem. The frequency of male and female participants, as well as of paternal and maternal single-parent households will be calculated. All of these descriptive statistics will be conducted in order to assess whether the data set resembles that of a normal distribution. Before beginning the next set of analyses, the predictor variable of overall responsibility will be created. The averaged ratings for household responsibility and parental disclosure will be transformed to create a composite score that indicates the level of overall responsibility for each participant. Since the measures for household responsibility and parental disclosure are based off of different scales, each averaged rating will be first standardized. After the z-scores are computed they will be averaged together to create an overall responsibility score. This new variable, of overall responsibility, will be used in all of the subsequent analyses. A set of multiple regression analyses will be conducted in order to determine whether a curvilinear relationship exists between the level of one‟s overall responsibility and internalizing problems, and between the level of one‟s overall responsibility and self-esteem. To determine whether a statistically significant curvilinear relationship exists, the analyses will follow two steps. First, the variable of overall responsibility will be entered into the analysis as a linear component in the prediction of the outcomes. Second, a new variable will be constructed by squaring the level of overall responsibility, in order to create a quadratic predictor. This statistic will be entered as the curvilinear component. The squared and non-squared predictor variables will be used in a multiple regression analysis to determine whether they predict the level of one‟s internalizing problems. If the hypothesis is supported, the quadratic predictor will significantly Adolescent Responsibility 14 predict internalizing problems; such that the results will form a U-shaped trend (see Figure 1). If the results do not show a pure U-shape with the multiple regression analysis, a scatterplot of the data will also be assessed. This will help in determining whether a curvilinear trend exists. The same sequence of analyses will be used in determining whether a curvilinear relationship exists between the level of one‟s overall responsibility and self-esteem. The squared and non-squared predictor variables of overall responsibility will be used in a multiple regression analysis to determine whether they predict the level of one‟s self-esteem. If the hypothesis is supported, the results will form a U-shaped trend (see Figure 2). A scatterplot of the data will also be assessed with overall responsibility and self-esteem. Significant results of a curvilinear relationship are expected for both sets of analyses. Conclusions Since there has been discrepancy in previous research about the effects of responsibility on children from divorced families, the proposed study attempts to connect some of the gaps. In doing so, it is expected that a curvilinear relationship exists between overall responsibility and internalizing problems, as well as between overall responsibility and self-esteem. Specifically, the hypothesis is that extremely high and low levels of overall responsibility will be related to high levels of internalizing problems and low levels of self-esteem, and that moderate responsibility will be ideally related to lower internalizing problems and higher self-esteem. This research will contribute to the growing research on divorce. The most recent research is shifting to show that the underlying factors that are related to divorced families may have more of an impact on children than the actual structure of a divorced family. This research could be useful in advising parents of divorce about inappropriate versus appropriate behaviors and expectations of their children. It is expected that the results will Adolescent Responsibility 15 illustrate another circumstance where divorce is not necessarily detrimental, if parents know where to draw the line in levels of responsibility and disclosure. If this research is made accessible to parents, they may realize that it is beneficial to discuss a moderate level of issues surrounding divorce with children, as well as hold a child accountable for a moderate level of help around the house. However, results would also show that relying too much on a child to take on the tasks of an adult and emotionally support a single-parent may lead to detrimental effects. There are a few limitations of the proposed study. First of all, the design of the study uses measures that needed to be modified in order to assess the necessary constructs. The levels of one‟s household responsibility and parental disclosure are difficult constructs to measure, and may be more accurately assessed with observation and personal interviews. This type of design would take much more time and be very costly in conducting. Also, the study will be limited to adolescents from divorced, single-parent households. Therefore, the results cannot be generalized to children of other ages or types of households. Because of the nature of the proposed study, if a significant curvilinear relationship is found between overall responsibility and internalizing problems or between overall responsibility and self-esteem it would not imply that a cause-and-effect relationship exists. The design of this study, being that of an assessment of the adolescents‟ perceptions of each construct, limits the study to only indicate a relationship between the variables. Other possible factors could account for the internalizing problems and self-esteem of participants. For instance, a lack of time that an adolescent has could be related to overall responsibility and vary with the levels of self-esteem and internalizing problems of participants. For instance, too much or too little available time could be detrimental. There could also be an inverted time-order relationship, such that Adolescent Responsibility 16 adolescents with high self-esteem and low internalizing problems could be taking on a moderate level of overall responsibility. If the proposed study finds support of a curvilinear relationship between overall responsibility and adolescent outcomes, there are numerous other areas where a similar model could be examined. For instance, a similar pattern may be seen in non-divorced, single-parent households. This relationship could also be looked at in families where remarriages or multiple remarriages are impacting a child‟s level of responsibility. This curvilinear model is not only limited to the context of divorce and single-parents, but could be researched in families with parents that are more dependent on their children because of drug abuse or handicaps. In conclusion, it is expected that the proposed study will add to the growing knowledge of the underlying processes that may occur in families of divorce. It will connect some of the discrepant findings and provide a more complete picture of how responsibility levels may affect an adolescent from a divorced family. The proposed study will provide numerous practical applications for advising divorced, single-parents in how much responsibility to give to and expect from their children. Adolescent Responsibility 17 References Amato, P. (2001). Children of divorce in the 1990s: An update of the Amato and Keith (1991) meta-analysis. Journal of Family Psychology, 15, 355-370. Barber, B. L., & Eccles, J. S. (1992). Long-term influence of divorce and single parenting on adolescent family and work-related values, behaviors, and aspirations. Psychological Bulletin, 111, 108-126. Bjorkquist, P. M. (2000). Family Responsibility Index. In K. Corcoran & J. Fischer, Measures for clinical practice: a sourcebook (pp. 308-313). New York: Free Press. Galambos, N. L., Barker, E. T., & Almeida, D. M. (2003). Parents do matter: Trajectories of change in externalizing and internalizing problems in early adolescence. Child Development, 74, 578-594. Hetherington, E. M. (1989). Coping with family transitions: Winners, losers, and survivors. Child Development, 60, 1-14. Hetherington, E. M., & Stanley-Hagan, M. (1999). The adjustment of children with divorced parents: A risk and resiliency perspective. Journal of Child Psychology and Psychiatry, 40, 129-140. Polivy, J., & Heatherton, T. F. (1991). Development and validation of a scale for measuring state self-esteem. Journal of Personality and Social Psychology, 6, 895-910. Silverberg, S. B., Wallace, S., & Lee, S. A. (2004). Sensitive mother-to-adolescent disclosures after divorce: Is the experience of sons different from that of daughters? Journal of Family Psychology, 18, 46-57. Adolescent Responsibility 18 Silverberg, S. B., Wallace, S., Lehman, S. J., & Raymond, M. (2002). Mother-to-daughter disclosure after divorce: Are there costs and benefits? Journal of Child and Family Studies, 11, 469-483. Weiss, R. S. (1979). Growing up a little faster: The experience of growing up in a single-parent household. Journal of Social Issues, 35, 97-111. Adolescent Responsibility 19 Figure Caption Figure 1. Hypothesized Relations between Overall Responsibility and Internalizing Problems Adolescent Responsibility 20 6.0 5.0 Internalizing Problems 4.0 3.0 2.0 1.0 -1.0 0 1.0 Overall Responsibility Note. Higher numbers of internalizing problems indicate a higher level of internalizing problems. Adolescent Responsibility 21 Figure Caption Figure 2. Hypothesized Relations between Overall Responsibility and Self-Esteem Adolescent Responsibility 22 5.0 4.0 Self-Esteem 3.0 2.0 1.0 -1.0 0 1.0 Overall Responsibility Note. Higher numbers of self-esteem indicate higher levels of self-esteem.