a... . _ .. .. ~ .2. s... . hmwflm, . . gar .u . _ a. ,v u wwaflwgfi. , .31.. unfimfiwmmwmv _ . s, :1. V «Eli... Kiwi ‘5. 1- .u .1: I. . um. ”a Wannawfli ... w r“. . aw. ...... av .flulnlv n. 2:. t . tum ..Tu..~..r.~fllz I," W JunmwnuNaWM-w S1 .. . . 1...!) 3V}. “W 'WHWHHWMU. ulxa‘ fin... . .s .ifln... 1a . unflilv it .I. , u Hmwflhfiflflmfin ” ,‘E‘I . 3...... 9495.... «Lift. . Li! I . 9.1%.... »-. . . {3‘51 Or]..- 13» 11. a...;......... .4... Km. .klhiWfiv .r. 5.....fimn (uh . , . l .bl..l| .(a- . . . Luau... - h-.. . . ., :‘n‘ at): .r. n. 0“ . 5 1:... 5. 1* . “1‘9?“ 7. 15.30'” . t .0}! a D (i 0. I .V. .‘cl.l\gy .UO‘.InthIh“L ..\ . v | $03!. ‘0' . .nv .' . ‘Ivg viru‘.‘ I.” Er‘lll «fit... u..$ni..1»ot.nn.l‘..¥x~wxfihfihnl . ERG». L .51...- :...!.l.l.fld.r|t..t {III (9.... hr? .. .IJWlIthJLV“ .|\(o. in“? ”51%.- . utv.n.....snu1.b.1nn 4“ um.“ 5...... -1-.. . ' It‘ll-:‘kuI-hn‘ln" bzl‘uhxri .111..." ([1 5‘10‘s‘Kflbc9uPa . 7 t ..... owl?! Hl‘l‘v. . v2.5.4»... . I. i “1!... 313.51..“ . . 3%. nu..3nflbuw».ubyr\.. .Y Hrfuubiuuvri mun”?! .huvfihfiu... r. UR... 3%.- L y. “o '5 '1 h. i; ILA... it . PK. .I-ptr 3.- I Iv ":l..l|lnll..n til it: (otl‘clflo... I .15 . .. uni“: .i. .wnli4.?Il§v3.a.v..Ll 61...? - 11.59943». “HWY. 1.... 4......UvNIX‘1)... -H . i . . ¢ ..Ip...,.ru\3unv( .1 L I. I. b\ bl‘! : lb v| .I .11.... SMmhxll .021... .t»~5.1 I... ; 1.1.).ItLPvHI A. .ln’l‘lPOo‘A. Jfifihflukflv t w... .IN‘LIXI.?.I.JAJ.IU. 1‘ [U Q )O‘ltéhrl.kirn§ l».(v.l|.|u ‘u .. 1 .. .llvlvav 3 . i\\' .l IV... 11!. ‘0»01?VI1UJ.V..I..\O\, . v ‘5 t1... . ['11. “and..."nu. 1| «A4,.I‘l15rdnt’u, t! -f.‘ .00. a mi, l 2% . .v!“orv tkv.h..’ l , \ .lt)!u\lo.nF .l .u- 2‘... til I? vl‘l. pl .0... it! t\ . lib: it?" .I xii“ . Auntlrvtlvvt' lit}.l.l' . EP‘! . tCIII.‘ 5. l! f WO..MI«PIHI.IHH.LIJN$H14\.IUI\A s . . I-‘.PII4Q\‘\IO ‘ L . . Hk?if.l..... $2 .4... .8, -.>.J..\\n V W, l 1. H b . v F 9.... n. .. 5 12% 551‘, a -.‘1'1 ,‘\, Adlonud? 10.9.3.1?! .l 0,.I. :(‘il‘lifi . .lo . 1... ‘JV ‘l .. .l..l Os-.. .vlr...'t I! 3 . i , . . {-33,}.-.an ....|.ttl|vbl ”II. ...|.. l-l\.t|..‘uv.uhlb|. \ .oit. .‘I. If! .10’... wygéhfi....., . , 7 . ., : kiwi... ..§M.«wsw%sfik§§§fi 3.1. a \v 52! . ms... llllllllllllllllHllllHllllllIIHIllllllllllihlllllllllll 131293 01568 3646 LIBRARY .lichlgan State | university This is to certify that the thesis entitled PARTNER EMOTIONAL SUPPORT AND NETWORK SOCIAL SUPPORT: THE DIFFERENTIAL EFFECTS OF TWO SOURCES OF FUNCTIONAL SOCIAL SUPPORT ON ADOLESCENTS' ADJUSTMENT TO PARENTHOOD presented by Lisa B. Galasso has been accepted towards fulfillment of the requirements for M. A. Psychology degree in M (iajor professor 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution f 21*: PLACE u RETURN BOX tafmoam-‘cm‘ckom from your more. TO AVOID FINES rdum‘ouor before date' duo. E3 DATE DUE DATE DUE II MSU I. An Affirmative Action/Equal Opportunity Intuition Wain-9.1 PARTNER EMOTIONAL SUPPORT AND NETWORK SOCIAL SUPPORT: THE DIFFERENTIAL EFFECTS OF TWO SOURCES OF FUNCTIONAL SOCIAL SUPPORT ON ADOLESCENTS’ ADJUSTMENT To PARENTHOOD By Lisa B. Galasso A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1996 ABSTRACT PARTNER EMOTIONAL SUPPORT AND NETWORK SOCIAL SUPPORT: THE DIFFERENTIAL EFFECTS OF Two SOURCES OF FUNCTIONAL SOCIAL SUPPORT ON ADOLESCENTS’ ADJUSTMENT To PARENTHOOD By Lisa B. Galasso - This study was a step toward better understanding the mechanisms through which social support efi'ects positive adjustment to parenting among adolescents. This analytical model posited that partner emotional support and network social support would differentially influence adolescent parenting outcomes (N=l34; average age 15 years). The former support source was hypothesized to have a direct and reparative efi‘ect on poor self-esteem resulting from negative family environment variables while the latter support source was proposed to influence adolescents’ parenting ability both directly and as a moderator of stress. The adolescent’s adjustment to parenthood was measured by knowledge of infant development and positive parenting attitudes, while the predictor variables included measures of the adolescent’s family environment, self-esteem, partner emotional support, general network social support, and parenting stress. A series of regression analyses revealed that partner emotional support was unrelated to self-esteem, nor was self-esteem able to predict parenting outcome. General network support acted as a buffer against parenting stress, although this support source had no direct efi‘ect on adjustment to parenting. The results are discussed in terms of their implications for social support interventions with this population. ACKNOWLEDGMENTS I wish to thank my thesis committee, Anne Bogat, Chair, Robert Caldwell, and William Davidson for their help in all phases of this project fi'om its conception to its completion. Dr. Bogat’s comments on each draft provoked a great deal of thought; the transformations that this thesis went through were the direct result of her ability to stimulate intellectual growth and sophistication. The high standards that she set facilitated my entry into a new level of critical and professional thinking and writing. I also wish to thank her for her support and patience as I went through this process, resisting growth at many points along the way. I wish, also, to express my gratitiude to Dr. Caldwell for his thoughtful comments throughoutthis process; his ideas and advice were always valuable. I am especially indebted to him for the time he spent helping me with my statistical analyses, both conceptually and methodologically. Lastly, I would like to thank Dr. Davidson for his thought provoking comments and questions in the beginning phases of this project. Finally, I am indebted to the young mothers from YPED who participated in this study, and who courageously opened up their lives to us. TABLE OF CONTENTS List of Tables List of Figures Introduction The Current Research Stress Bufl‘ering and Main Efi‘ect Theories of Social Support Part One of the Theoretical Model Family Environment and its Influence on Adolescent Parenthood Teen’s Family Environment and Self-Esteem Partner Support and Self-Esteem Adolescent Adjustment to Parenthood Self-Esteem and Adjustment to Parenthood Part Two of the Theoretical Model Stress Stress and Adolescent Adjustment to Parenthood The Efl‘ects of Social Support on Parenting Stress Rationale for the Present Study Hypotheses ii iv 12 l4 l7 19 20 20 22 23 26 30 Methods Results Discussion References 31 4O 48 78 10 LIST OF TABLES Factor analysis of family environment variables: Two factor structure - adaptive and maladaptive parenting practices Hierarchical multiple regression analysis: Maladaptive family environment predicting adolescent parenting outcome (T3KIDIRIGHT) Maladaptive family environment predicting adolescent parenting outcome (T3KIDIRIGHT) Hierarchical multiple regression analysis: Maladaptive family environment predicting adolescent parenting outcome (FEQPPA) Maladaptive family environment predicting adolescent parenting outcome (FEQPPA) Hierarchical multiple regression analysis: Adaptive family environment predicting adolescent parenting outcome (T3KIDIRIGHT) Adaptive family environment predicting adolescent parenting outcome (T3KIDIRIGHT) Hierarchical multiple regression analysis: Adaptive family environment predicting adolescent parenting outcome (FEQPPA) Adaptive family environment predicting adolescent parenting outcome (FEQPPA) Descriptive statistics for dependent and independent variables iv °E ~ 62 63 65 67 68 69 70 figure LIST OF FIGURES Theoretical model Division of theoretical model into two distinct sources of social support Measurement model Regression 1 Regression 2 Regression 3 Regression 4 71 72 73 74 75 76 77 HVT RODUCT ION Over the past several decades, a diverse body of literature has accumulated which substantiates the beneficial effects of social support on psychological well-being and . physical health (Caplan, 1974; Cassel, 1976; Cobb, 1976; Cohen & Wills, 1985; Dean & Lin, 1977; Gottlieb, 1981; Henderson, 1980; House, Landis, & Umberson, 1988; Mueller, 1980; Schafer, Coyne, & Lazanis, 1981; Turner, 1981; Weiss, 1974). Investigations have demonstrated that a) the association between social support and well-being is attributable to an overall beneficial efl‘ect of social support on behavior and attitudes, and b) social support protects individuals from the potentially pathogenic efi‘ects of stressful events. These two processes by which social support has been shown to afi‘ect well-being are termed the main efl‘ect and stress buffering models, respectively, and the conditions under which each theory of support is said to operate have been the subject of lively debate in the literature. The present investigation is a further step in the ongoing process of more fiilly understanding the mechanisms through which social relationships enhance psychological health and adjustment to life events. Specifically, the transition to parenthood among pregnant adolescents, and the factors that afi‘ect this adjustment, will be examined. Both the main efi‘ect and stress buffering models of social support will be tested in this investigation of the role of supportive interactions related to adolescent parenting. Problem at Adolescent Pregnangg Adolescent pregnancy is associated with substantial risk for adverse educational and economic outcomes. Pregnancy is one of the most common reasons why adolescent girls leave school. One study reports that four out of five girls who become pregnant in high school drop out, compared to less than 10% of those who are childless (Drummond & Hansford, 1991). Lack of education and practical training leads to continued economic hardship for these young women. Drummond and Hansford (1991) report that ”a conservative estimate of the United States' public outlays attributable to teenage childbearing was $15 billion...[and] if present childbearing trends continue, two thirds of the nation’s teenage mothers will receive welfare assistance of an estimated $47 billion” (p 66). They argue that this trend in teenage childbearing has been a primary contributing factor to the feminization of poverty. In the psycho-social domain, adolescent pregnancy and parenting is associated with negative outcomes for both the mother and child that are far-reaching in their efi’ects. The parenting practices of teen mothers have been characterized as impatient, insensitive, and irritable; teen mothers are also prone to use more physical punishment with their children (Fulton, Murphy, & Anderson, 1991). Many studies have shown that adolescents know less about child development than do older parents (Azar, Robinson, Hekimian, & Twentyman, 1984; Fulton et al., 1991; Samuels, Stockdale, & Crase, 1994; Stoiber & Houghton, 1993). This lack of knowledge can result in inappropriate interactions and unrealistic expectations of their children’s behavior and abilities, and, consequently, these 3 faulty expectations can increase the incidence of child maltreatment and abuse. Inadequate mothering was found by Egeland (1980) to be related to misunderstanding the child and the nature of child-rearing. In this same vein, Reis (1993) found a significant positive relationship between appropriate knowledge of child development and favorable organization of the home environment. In yet another study of risk factors associated with adolescent parenting, Passino, Whitman, Borkowski, Schellenbach, Maxwell, Keogh, and Rellinger (1993) found that adolescent mothers experienced higher overall stress, and they reported feeling more socially isolated and restricted by their parenting roles and responsibilities than did comparable pregnant adults. As a result, these teen mothers were less interactive with their infants, and they scored lower on maternal afl‘ection, rate 'of- stimulation, flerdbility, positiveness, motivation, and overall quality of mothering. These inadequate maternal behaviors can have deleterious effects on the child's social, emotional, and cognitive development. Hence, the consequences of adolescent pregnancy have‘a negative impact on society as a whole and on women and children specifically. In light of this evidence of the negative consequences of adolescent childbearing, a growing body of literature has accumulated which has attempted to identify some of the variables associated with more adequate outcomes among these young women and their children. Social support has received the most attention as a variable that may efiectively moderate the harmful efl‘ects of the stresses related to adolescent parenting, as well as directly influence positive adjustment to parenthood. 4 The Current Research The current study is a step toward better understanding the mechanisms through which social support may efl‘ect positive adjustment to parenting among adolescents. The model proposed by the author (See Figure l) posits that adolescent parenting outcomes result fi'om several factors. First, the teen’s family environment and the emotional support provided by her current partner will afl‘ect her self-esteem, which, in turn, will afl‘ect the teen’s adjustment to parenting. A second independent influence on the teen’s parenting ability is the efi‘ect that her general social network support has, both directly and as a moderator of stress. At the heart of the present theoretical model is the postulation that two distinct modes of social support will influence the young mother? s ability to provide adequate care to her child (See Figure 2). These support sources are: a) emotional support from the teen’s partner and b) social support from the teen’s greater support network -‘ The former support source is assumed to be responsiveto general feelings of helplessness and loss of self-esteem; it is hypothesized to have a reparative efl‘ect on her self concept. The latter support source is assumed to be effective when the resources this type of tangible support provides are specific to the need elicited by a stressful event, such as needing transportation to a doctor's visit, needing money for diapers, or needing child care. The socially supportive relationships in this theoretical model are assumed to afl‘ect the teen’s parenting behavior both directly, as a main efl‘ect, and indirectly, as a moderator. These two sources of support and the mechanisms by which they are hypothesized to have their effects will be discussed in turn. 5 m Buflering and Main Effect Theories at Social Support Each conceptualization of social support in the present theoretical model, partner emotional support and network social support, is expected to exert a positive influence on the adolescent’s adjustment to motherhood. Whether this positive influence is due to a main efl‘ect of social support, whether social support mediates the harmful efl‘ects of stress related to parenting, or whether each source of support has both a direct and moderating efl‘ect are central questions to be addressed in the present investigation. One model that explains the process of social support is that of support as a main efi'ect. Cohen and Wills (1985) explain that a large social support network provides people with regular positive experiences and a set of stable, socially rewarding roles in the community. Thisasocial support may be related to overall well-being because it provides positive afi‘ect, a sense of predictability and stability, as well as recognition of self-worth; These authors further hypothesize that the general’benefit of social network support' could be related to physical health outcomes through emotionally induced efi‘ects on neuroendocrine or immune system functioning, or through influence on health-related behavior such as medical help-seeking (Cohen & Wills, 1985). Conversely, the lack of positive social relationships may lead to anxiety and depression, and these states may eventually influence physical health through a direct efi‘ect on physiological processes that influence susceptibility to disease, or through behavior that increases risk for disease (Cohen & Wills, 1985). A second model of social support is as a bufl‘er of stress, which, through an interactive process, affects health outcomes. According to Cohen and Wills (1985), support may buffer stress in two ways. [First], support may intervene between the stressful event and a stress reaction by attenuating or preventing a stress appraisal response. For example, the perception that others can and will provide necessary resources may redefine the potential for harm posed by a situation and/or bolster one's ability to cope with imposed demands and hence prevent a situation fiom being appraised as highly stressfirl. [Second], support may intervene between the experience of stress and the onset of the pathological outcome by reducing or eliminating the stress reaction or by directly influencing physiological processes. . .support may alleviate the impact of stress appraisal by providing a solution to the problem or by facilitating healthful behaviors (p.312). In their review, Cohen and Wills (1985) specify four types of social resources that operate as stress bufl‘ers. The first, esteem support, is information that a person is - - esteemed and accepted. The second is informational support which is support in the form of help in defining understanding, and coping with problematic events. The third type of support is social companionship, which involves spending time with others in leisure or recreational activities. It may reduce stress by fulfilling a need for afiliation and contact with others. The final support source is instrumental support, which is the provision of financial aid, material resources, and needed services. This type of support may help reduce stress by direct resolution of instrumental problems, or by providing the recipient with increased time for leisure activities. These four support resources function difi‘erently to mitigate the efl‘ects of stressful events. Stressful events oflen evoke feelings of 7 helplessness or threat to self-esteem. Esteem support may counterbalance threats to self- esteem, while informational support may help the individual to reappraise a stressor as benign or suggest appropriate coping responses which counter a perceived lack of control. Thus, esteem support and informational support appear to be responsive to a wide range of stressors. Conversely, the latter two types of support resources (instrumental support and social companionship) are assumed to be limited in their efl’ects, because the resources they provide are specific to the need elicited by a stressful event (Cohen .& Wills, 1985). » These conceptions of support resources are central to the conceptual model of adolescent parenting under investigation here. In the present model, the partner of the adolescent is expected to provide emotional support, referred to as esteem support by Cohen and Wills, while social support from the adolescent’s greater support network constitutes the tangible assistance outlined by these researchers. Although Cohen and Wills predict that: the types of support in their four-fold division of social support will exclusively show bufl‘ering efi‘ects for stressful events, the evidence in the literature for the stress-bufi'ering hypothesis is at best equivocal. Hence, in the present investigation, both the moderating and main efl‘ects of these support sources will be tested in relation to parenting outcomes among adolescents. Specifically, partner support is hypothesized to afl‘ect the teen’s self-esteem directly as well as to moderate the relationship between her family environment and self.esteem. Similarly, the role of network social support is proposed to afl‘ect the teen’s parenting attitudes directly as well as to bufl‘er the effects of stress related to her role as a parent. The conceptual links that tie this theoretical model together will be discussed in order to provide the contextual background against which 8 these supportive interactions are hypothesized to occur. Part One Qt The Theoretical Model The focus of Part One (See Figure 2, Part 1) is on the effects of family environment and partner emotional support on self-esteem and adjustment to parenthood. First, the adolescent’s family environment is expected to afl‘ect her adjustment to parenthood directly; the underpinnings of this link in the model are borrowed fi'om attachment theory which can help explain the transmission of parenting behaviors intergenerationally. In addition, the family environment of the teen is expected to effect her self-concept directly, which in turn is expected to influence her ability to provide adequate care to her child. Second, the emotional support provided by the partner of the adolescent is expected to afi‘ect the teen’s self-esteem directly. This support source is also expected to moderate the relationship between the teen’s family environment and her self- esteem; again, attachment theory may, help explain the mechanisms by which: this interaction may occur. Each of these conceptual links in the present theoretical model will be discussed in turn in this section. Fan—1112' Environment and its Influence on Adolescent Parenthood Research indicates there are specific social and psychological reasons why adolescents become pregnant (cf. Barnett, Papini, & Gbur, 1991; MacLanahan, 1988; Sullivan, 1993). Although it is beyond the scope of the current investigation to explore this rich literature, a general discussion of some of the variables leading to adolescent pregnancy is important for understanding the backdrop of familial characteristics that influence the parenting attitudes and behavior of these young women. 9 One variable that has commonly been linked to the incidence of teenage pregnancy is family disruption (McLanahan, 1988). Family disruptions have a profound effect on children and adolescents. Growing up in a single parent family has important consequences for adult family behavior of daughters. For instance, adolescent girls fi'om single parent households are more likely to be sexually active and have premarital births than are adolescents fi'om two-parent families (Hogan & Kitagawa, 1985). This may be due to the fact that single parent families and family disruptions are linked to less efl‘ective discipline and lack of parental supervision. McLanahan (1988) examined variation among single parent households to try to explain the trend that women who, spend part of their childhood in these families are more likely to give birth before marriage and to have their own marriages break up. A reasonable explanation for this is ofi‘ered by a socialization hypothesis which says that daughters who live with single mothers see this behavior as an acceptable and viable alternative. Also, single parents have'more dificulty maintaining authority and control over their daughter’s dating, another factor that is related to early pregnancy and child rearing. In fact, both very permissive parenting styles and very strict parenting styles put adolescents at risk for early pregnancy (Miller, McCoy, Olson, & Wallace, 1986; Hogan & Kitawaga, 1985; Peterson, Rollins, & Thomas, 1985). Among disrupted families, it has been suggested that when the teen is missing familial or maternal support, she may feel lonely and isolated and seek an intimate relationship in order to fulfill her need to be loved and cared for (Townsend & Worobey, 1987). Moreover, pregnant adolescents report lower levels of perceived love, attention, and interdependence fi'om their mothers than nonpregnant teens, and they also perceive 10 less affection, fewer demands, and more rejection from their mothers (Olson & Worobey, 1984). Although family disruptions, and other familial constellation variables seem to be primarily discussed in the literature as causal factors of teenage pregnancy, these are nonetheless important factors that also affect the adolescent's adjustment to parenting through their efi‘ects on her developing concept of self, and through the process of intergenerational transmission of parenting behaviors. The intergenerational transmission of parenting behavior speaks to the direct link in the present model between the teen’s family environment and her subsequent parenting behavior. Attachment theory may provide a means for understanding the mechanisms by which this link in the model operates. Attachment theory posits that the mother, or primary care-giver, provides a secure base for the infant who in turn feels that she is a stable, steady, dependable force in his or her life and is emotionally. available. This positive attachment allows the child to explore his or her world confidently and to develop autonomy and self-reliance. Bowlby (1969) claimed that the quality of attachment, not merely its presence, was central to healthy developmental outcomes. Bowlby further believed that the primary attachment relationship was a prototype for later social relationships. The importance of attachment to the present model centers around the intergenerational effects of attachment; the way parents organize their own childhood experiences is a powerful predictor of how they will parent their own children The efl‘ects of attachment history on subsequent attachment relationships has been investigated in several studies which have shown that parental attachment history, as inferred fiom 11 interviews, is related to the quality of attachment in the next generation (Main, Kaplan, & Cassidy, 1985; Morris, 1980; Ricks, 1985). Hence, in relation to the current model, a teen parent who had a poor attachment relationship with her mother may use this same poor attachment representation as her working model of how parents should relate to children, and, in turn, may exhibit similar poor attachment behavior with her own children Although actual attachment patterns are not being measured in the current study, this theoretical construct can be used to understand similar parenting practices among mothers and daughters. The literature on child abuse and maltreatment, and on the continuity and discontimrity of patterns of abuse transmitted intergenerationally (Belsky, 1993; Ciccetti, Toth, & Lynch, 1995; Ciccetti & Garmezy, 1993; Egeland, Jacobvitz, & Sroufe, 1988; Egeland & Sroufe, 1981) also provides a vehicle for exploring continuity and change in parenting practices relevant to the current investigation. .F or example, Egeland,- Jacobvitz, and Sroufe (1988) followed the child care practices (for three years) of women who had been abused themselves as children. Although their findings support the notion that maltreatment is continuous across generations, not all parents who had been abused themselves abused their own children. These researchers interpret their findings fi'om an attachment theory perspective. They suggest that working models of the self others, and relationships are ongoing constructs that are susceptible to change, especially in cases where a supportive relationship was available to the neglected child, thereby enhancing her feelings of self-worth. A discussion of the dynamic role of supportive relationships as determinants of change in levels of self-esteem will be covered subsequently. First, a more 12 detailed discussion of adolescent self-esteem, and how it is affected by family constellation variables will lay the groundwork for that discussion. Teen ’s F amil Environment and el -Esteem The family constellation variables in the present model that are hypothesized to precede the adolescent’s concept of the self include maternal parenting style (whether it was permissive, authoritarian, or authoritative), mother's attitudes towards being a parent (i.e., positive parenting attitude versus negative parenting attitude), and mother's psychological control or acceptance of the teen. The present investigation is founded upon the premise that these family environment variables have a direct efi‘ect on the adolescent’s developing self-esteem, which will in turn influence her adjustment to her new role as a parent. As attachment theory predicts, adolescents whose family environments were lacking positive care-giving will have lower notions of self worth, and will resultingly have more dificulty providing adequate care to their own children. Evidence for this link in the model is plentifirl in the literature. Many studies have examined the relationship of self-esteem to the quality of the relationship that the adolescent had with her own parent. Adolescent's perceptions of their attachment relationships with their mother, as well as dimensions of parental acceptance, have been positively associated with self-esteem (McCormick, 1994; Paterson, 1995). One study, by Abernathy, Robbins, Abernathy, Grechbaum, and Weiss (1975), found that family experiences in adolescence appeared to be critical in the development of attitudes, personality traits, and self-concept. They specifically suggest that low self-esteem was derived fiom an unsatisfactory identification with a mother who was neither warm nor 13 respectful to the adolescent. Babikian and Goldman (1971) similarly report that “a basic deficiency in the early formative years contributes to a poor structuralization of the ego” (p. 759). Olson and Worobey (1984) similarly found that low self-esteem among a group of adolescent mothers was associated with early maternal rejection. Hence, parental nurturance has consistently been found to be a strong predictor of self-esteem during adolescence and early adulthood; in fact, this parenting quality was found to be a stabilizing influence during these transitional years (Buri, 1992). It has also been shown that parental authoritativeness is directly related to self-esteem, whereas parental authoritarianism is inversely related to self concept (Buri, 1989; Johnson, 1991). In terms of parental psychological control and acceptance of the adolescent, the dimension of control has been shown to be negatively related to self-esteem. Conversely, individuals with high self-esteem perceive their parents as using less psychological control, and as not being overly firm in making and enforcing rules and regulating their behavior (Litovsky, 1985). Optimal self-concept, then, develops in an atmosphere of acceptance that allows the adolescent autonomy and that fosters her growing self-competence. In substandard familial environments where competent, sensitive parenting was not readily available, and anxious or avoidant attachment patterns have developed, adolescents may suffer negative consequences such as low self-esteem and poor self-concept. In fact, the child’s primary attachment with the mother serves as the ongoing context for determining both the quality of self-concept as well as its vulnerability to outside influences. Hence, adolescents with strong maternal bonds are likely to have positive self images that are resilient to outside influences, such as peer groups. Those with less 14 positive patterns of attachment, however, should have less positive self images that are more open to outside influence (Hofi‘man, Ushpiz, & Levy-Shifl‘, 1988; Kalish & Knudson, 1976). For example, an adolescent whose self-esteem is low and whose relationship with her mother is strained, will be more likely to turn to a peer group for support and validation. A primary assumption of the present model is that adolescents with low self-esteem, resulting from less than optimal parental caregiving, will be more open to the influence of partner social support. This assumption is central to the moderating hypothesis of social support in this model: among teen’s with low self-esteem, partner social support is expected to have a reparative efi‘ect. Teens with higher self- esteem and who are more self-accepting are believed to have internalized a positive image of the self which is not in need of repair. Patner smart and Self-Esteem A reason for examining partner support is that numerous studies have documented the more positive efi‘ects of partner or spouse support, as opposed to other support sources, on maternal adjustment to parenting (Chen, Telleen, & Chen, 1995; Crnic, Greenberg, Robinson, & Ragozin, 1984; Schamess, 1993; Thompson, 1986; Thompson & Peebles-erldns, 1992; Tilden, 1983; Wandersman, Wandersman, & Kahn, 1980; Vaz, Smolen, & Miller, 1983). For instance, O’Hara (1983) found that the social support provided by spouses, as opposed to other confidants, was especially important for reducing postpartum depression among adult mothers. Some researchers suggest that emotional support fi‘om a partner may help provide mothers with the emotional resources necessary to give adequate care to their children (Egelend et al., 1988). Other studies 15 (Brown, Bholchain, & Harris, 1975; Paykel et al., 1980) have found that confiding husbands and boyfiiends served stress-protective functions for women while other confidants, such as friends or relatives, did not. This may occur because such intimate relationships are conducive to the adequate provision of information and esteem support. Perhaps the confiding nature of these relationships counteracts stressors by increasing feelings of self-esteem and personal eflicacy. The current model proposes that partner support influences self-esteem in two ways. First, a main efi‘ect of partner support is proposed. Namely, a supportive environment fostered by partner support leads to greater self-esteem which leads to better adjustment to parenting. Illustratively, Sarnuels and her colleagues (1994) found that the most robust predictors of maternal adjustment to parenthood, among adolescents, were older maternal age, marital status, social support, self-esteem, and contact with baby's father. In this study, however, contact with the baby's father was the single most robust predictor of the mother‘s adjustment to parenting. Crnic et a1. (1984) attempted to study the breadth of social support influence on maternal stress and the mother-infant relationship to eighteen months of age, among mothers aged 16 to 38 (mean age was 24 years). Their findings, not surprisingly, revealed that mother's perceived satisfaction with intimate and community support had a positive influence on life satisfaction, satisfaction with parenting, and quality of interaction with her child. Interestingly, and most relevant to the current thesis, was that the mother‘s intimate relationship support had the most positive relationship to the outcome variables at each measurement period. Young, unmarried mothers fared far worse, exhibiting less positive maternal attitudes and l6 behavior. Taken together, these findings support Belsky's (1984) theory that during the transition to parenthood, partner support is more important and beneficial than support fiom community members or friends, and it is a major support source that affects competent parenting and more positive parent-child interactions. The current thesis centers on this notion: mothers with greater emotional support fiom their partners and more partner-relationship satisfaction will benefit from increased self-esteem which will in turn allow them to respond more appropriately to their children. Second, it is also hypothesized in the present study, that partner support moderates the relationship between the teen’s family environment and her self-esteem. Hence, among emotionally-deprived teens, social support from partners is expected to have a reparative influence on self-esteem and, thus, indirectly afl‘ect adjustment to parenting. Inherent in this model is the notion that self-esteem is a malleable construct that is open to the reparative influence of positive socially supportive relationships. Egeland and his colleagues (1988) specify three types of relationship variables that seem to exert a reparative influence on poor self-concept resulting fiom negligent familial environments: 1) the availability of an emotionally supportive relationship during the individual’s childhood, 2) an ongoing therapeutic relationship during any period of the individuals life, and 3) formation of a stable, satisfying relationship with a partner during adulthood. These authors claim that such experiences provide the basis for developing alternative working models of relationships and, thereby, permit the individual to form a nurturing relationship with her child despite having experienced parental abuse or maltreatment. Whether this supportive relationship directly enhances self-esteem, which in 17 turn enables her to provide more positive care giving to her own children, or whether it repairs, or mediates her faulty working model of herself in relation to others resulting fi'om inadequate parenting, will be tested in the present theoretical model. Among adolescents whose self-esteem has been damaged from poor early relationships, the current model proposes that the availability of emotional support fiom partners will have a reparative efl‘ect upon teens’ self-esteem. In other words, partner emotional support will moderate the negative efl‘ects of the teen’s early home environment on her self-esteem, which in turn will lead to more positive adjustment to parenthood. Before considering the evidence for the proposed link between self-esteem and positive maternal adjustment, let us first consider the variables that constitute the adolescent’s adjustment to parenthood. Adolescent Adjustment to Parenthood Healthy child development depends critically on the childirearing behavior of the mother or primary care-giver. The mother’s knowledge, attitudes, and emotional state afi‘ect her behavior with the child, which in turn afi‘ects the child’s development (Bell, 1970). Children of adolescent mothers are at risk for developmental delays and other problems, so the nature of adolescent parenting attitudes, knowledge, and emotional state are of importance to researchers. Broussard and Hartrrer (1971) reported a relationship between unfavorable perceptions by the mother toward her infant and later emotional disorders in preschool children. Relatedly, a variety of parental mental disorders have been associated with emotional or cognitive problems in children (Gamer, Gallant, & Grunebaum, 1976). Passino et al. (1993) assert that a major determinant of parenting efi‘ectiveness is the mother’s general personal adjustment. Specifically, a mother who is 18 psychologically well-adjusted should be better able to deal with the numerous stressors associated with raising a child and be more efi‘ective in her parenting interactions. For example, Colletta and Gregg (1981) found that the level of emotional distress experienced by adolescent mothers was less for those with greater personal resources and more adequate coping styles. Investigations into the relationship between parental knowledge of developmental milestones and child rearing practices have shown that lack of knowledge can lead to inappropriate interactions and unrealistic expectations of children’s behavior and abilities (Azar et al., 1984; Epstein, 1980; Miller, 1988; Roosa, 1983). Consequently, faulty ermectations in the direction of over-estimation of infant abilities have been linked to incidences of child maltreatment and abuse (De Lissovoy,1975; Fulton et al., 1991; Reis & Hertz, 1987). Similarly, McAnarney and Thiede (1981) found that the younger the adolescent mother, the less she used typical maternal behavior such as touching, synchronous movement, higher-pitched voice, and physical closeness. Other studies have shown that young mothers do demonstrate warmth and physical interaction with their infants, but engage in relatively less verbal interaction (Field, “ridmayer, Stringer, & Ignatofl‘, 1980). This survey of the literature indicates that knowledge of infant development and maternal attitudes are appropriate measures of adolescent adjustment to parenting. Hence, the specific maternal characteristics under investigation, and which comprise the dependent variables in the present study, are a) the mother’s perception of her infant and herself as a parent, and b) her knowledge of infant development. 19 Sell-Esteem and Adjustment to Parenthood Higher levels of self-esteem have been positively associated with young mothers’ adjustment to the maternal role (Sroufe, Pianta, & Egeland, 1988). This positive link between young mothers’ notions about the self and subsequent maternal attitudes and behavior toward the infant have been reported by several researchers (Held, 1981; Kempe & Helfer, 1974; Koniak-Grifin, 1988; Mercer, 1977; Patten, 1981). Illustratively, one study (Williams, Joy, Travis, Gotowiec, Blum-Steel, Aiken, Painter, & Davidson, 1987) found that teen mothers who have higher levels of self-esteem report better relationships with their children, experience less role conflict, and tend to report a more positive emotional state. Unger and Wandersman (1985) similarly report that among adolescent parents, self-esteem and feelings of mastery significantly predicted parenting skill and satisfaction. Rogosch (1992) similarly found that emotional support fiom network members influenced self-esteem, which predicted more adaptive parenting attitudes. It is less clear, however, how self-esteem is afl‘ected by pregnancy,'and the many changes that occur during this transitional period. For instance, the literature on self- esteem among pregnant adolescents has been mixed. One study reports that unwed adolescent mothers tend to exhibit low self-esteem (Barnett, Papini, & Gbur, 1991), which may be due to difficulty in functioning academically, economically, psychologically, socially, and as a parent, as well as being distressed over body image. Another study found no difl‘erences in self-esteem between parenting and non-parenting adolescents (Streetman, 1987), while still other findings show that pregnant and parenting teens have significantly higher self-esteem than non-pregnant or non-parenting controls (Crase & 20 Stockdale, 1989), which might be due to the extra attention associated with the birth of a baby. A further study of the changes in self-concept of unmarried mothers, from the beginning of the third trimester to shortly alter birth, demonstrated that the self-concept of these teens underwent changes that were suggestive of better adjustment over the course of pregnancy and parturition (Patten, 1981). Thus, there is a clear lack of consensus in the literature of how self-esteem is affected by pregnancy and childbearing among adolescents. An important factor that is ignored in these investigations, but which will be addressed in the present investigation, is familial background variables, which we have already seen afl‘ects the adolescent’s developing self-concept. Although the literature clearly shows that self-esteem is associated with more positive maternal firnctioning, it is less clear whether this construct is influenced by pregnancy and parenting status m .22, or whether changes in self-esteem result from changes in the presence or absence of supportive interpersonal relationships. Part Two Qt The Theoretical Model The second part of the conceptual model under investigation comes from community psychology literature (See Figure 2, Part 2). The thrust of this section centers on the direct and stress-bufl‘ering efl‘ects of general network social support on adolescent adjustment to parenthood. Each of these links of this second part of the model will be discussed in turn. §Less_ In recent times, research on stressors has undergone a conceptual shifi. Earlier paradigms viewed stress as a phenomenon resulting fiom exposure to major life changes 21 or life events (Dohrenwend & Dohrenwend, 1974). The more recent view of stress has been from the perspective of minor, or everyday, events. Lazarus and his colleagues (1984) have developed a measure of stressors that focuses on everyday events or “hassles.” They conceptualize hassles as “experiences and conditions of daily living that have been appraised as salient and harmful or threatening to the endorser’s well-being” (Lazarus, 1984, p. 376). In fact, research sparked by this paradigm shift has successfiilly demonstrated that minor daily stressors, or hassles, provide a more powerfirl prediction of adaptational outcomes, such as psychological symptoms, than the more widely researched life events measures (Chamberlain & Zika, 1990). A major criticism of the life events approach to stress has been that major life stresses, such as death, divorce, or the loss of a job, are low-hquency occurrences for most individuals. Alternatively, the hassles approach, which measures the cumulative impact of relatively minor daily stressors, has greater adaptational significance. Illustratively, Chamberlain and Zika (1990) studied the difl‘erences between life events and hassles as predictors of well-being variables. They found a relatively low association between life events and hassles suggesting that these variables provide substantially different measures of stressors. Further, they found a high association between hassles and mental well-being, but only a low association between life events and the outcomes measures. This suggests that minor stressors are an important influence on psychological well-being. The hassles approach to stress measurement, however, is not without its critics; Dohrenwend and Shrout (1985) claim that the hassles scale is composed of items that could also represent psychological symptoms, which confounds this measure as a tool for studying relations between stress and psychological 22 outcomes. Hassles, conceptualized as the irritating, fi'ustrating, annoying, and distressing demands that characterize everyday interactions, are especially applicable to families with young children (Crnic & Greenberg, 1990). Minor parenting hassles appear to be an important source of stress in the familial context contributing to less parental satisfaction and less firnctional families. Crnic and Greenberg (1990) report that parenting hassles account for a significant portion of the variance in the prediction of parent andfamily outcome, even when mother’s ratings of children’s behavior problems are partialled out. Hence, the cumulative impact of hassles, or daily stressors, and their potentially adverse effect on parental functioning and the parent-child relationship, is especially salient for adolescent parents. M and Adolescent Adjustment to Parenthood Although the birth of a child can be an event thatis associated with considerable joy, it is also a life event that has the potential for decreasing parents’ psychological well- being (Brown & Harris, 1980; Stemp, Turner, & Noh, 1986). Wandersman, Wandersman, and Kahn (1980) point out that the transition to parenthood is marked by many potentially stressful changes. There is the new responsibility for the well-being of the infant, financial resources must be reallocated, and a reorientation of relationships within the social network most ofien occurs during this transition. Highly stressed parents may lose their ability to care for their children in a warm, sensitive, and competent manner. Parents, on the other hand, who experience fewer minor irritations in their daily routines, should have more energy and enthusiasm for child rearing, and should perceive parenting 23 to be less stressful than more burdened mothers do (Gelfand, Teti, & Radin-F ox, 1992). For teenage mothers, the stress associated with the transition to motherhood may be exacerbated by the normal developmental tasks of adolescence (Bacon, 1974). In fact, previous studies have shown that adolescent mothers experience higher levels of stress due to parenting than their adult counterparts (Brown, Adams, & Kellan, 1981; Colletta & Gregg, 1981; de Anda, 1992; Roosa, Fitzgerald, & Carlson, 1982; Ventura, 1980). For example, East, Matthews, & Felice (1993) found that adolescent mothers who reported high stress associated with parenting had low confidence in their mothering role, low acceptance of their children, and low empathy for their children’s needs. Teenage mothers- have been shown to be at higher risk for psychological distress (Brown et al., 1981), lower self-esteem (Marini, 1978), and decreased personal eflicacy (McLaughlin & Micklin, ‘ - - 1983) as compared to older mothers of equal educational level and socioeconomic status. Hence, negotiating the compounded stresses of new motherhood with all of its responsibilities, as well as the physical, cognitive, and psychosocial changes and crises associated with adolescence, can create a disequilibriating experience for the teen. As a result, the efl‘ects of socially supportive interactions are especially salient for adolescent pregnancy and parenting outcomes. The Eflects oj Social Support on Parenting Stress The literature provides ample evidence that mothers of young children, especially adolescent mothers, are at elevated risk for psychological distress. There is a growing literature that identifies social support as an important factor in influencing these distress levels (Paykel, Emms, Fletcher, & Rassaby, 1980). Among adolescent mothers, social 24 support has been associated with increased acceptance of the infant (Colletta, 1981) and greater parental satisfaction (Schilmoeller, Baranowski, & Higgins, 1991). One way that social support positively impacts the adolescent’s adjustment to parenthood is through a direct effect. The underlying assumption of this social support model is that embeddedness or social connectedness provides access to social resources that increase the mother’s coping capabilities (Gore, 1978). This direct effect model assumes that social support will be beneficial regardless of the degree of stress the young mother is experiencing. Hence, the general beneficial efl‘ect of social support could be related to the young mother’s integration in a support network, which helps her to cope with negative experiences that otherwise might have increased. the experience of psychological distress (Moos & Mitchell, 1982). An empirical demonstration of this type of support is provided by Collins, Dunkel-Schetter, Lobel, and Scrimshaw (1993) who found that women (ages 18 to 42) who received‘more prenatal support experienced better progress in labor and delivered babies with higher Apgar scores. In addition, women with less support showed greater incidence of depressive symptomatology both prenatally and 6-8 weeks postpartum. In this same vein, a study by D’Arcy and Siddique (1984) found evidence of a strong direct efi‘ect of spousal and community social support upon mental health of adult mothers of preschool-age children. These findings, among others (Andrews, 1978; Camp, Holman, & Ridgway, 1993; Norbeck & Anderson, 1989), are consistent with the main effect theory of social support which suggests that embeddedness in supportive social networks are important for psychological well-being. The second model of social support involves the protective, or mediating effect of 25 social support upon the stressors related to the teen's transition to parenthood. This model is based on theory borrowed from the greater community mental health literature regarding the bufl‘ering efl‘ects of social support on stress. Psychosocial stress is posited to have negative efi‘ects on well-being among individuals with little or no social support, whereas these harmful efl‘ects will be decreased for individuals with stronger support systems (Cohen & McKay, 1983). This phenomena has been studied extensively, with confirmatory findings, in relation to adolescent pregnancy and postnatal adjustment (Barrera, 1981; Cutrona, 1984; Norbeck & Tilden, 1983; Paykel et al., 1980; Turner, Grindstafl‘, & Phillips, 1990; Wandersman et al.,l980). For example, Camp, Holman, and Ridgway (1993) found that teens who were more-anxious during pregnancy received more support, and that adolescent mothers with more negative attitudes toward the pregnancy had more, rather than less, social support, regardless of their age. They interpret their findings as indicating that those with more stress received more socialsupport, which is consistent with a stress-bufl‘ering model. Further, in a study of parenting stress, Crnic and Greenberg (1990) found that under high levels of stress, mothers with greater support satisfaction exhibited more positive maternal behavior than mothers with low support. Specifically, in this study, both fiiendship and community support consistently acted to moderate mothers’ experience of daily hassles. Interestingly, intimate support was not a significant moderator of stress in this study, suggesting that emotional support fi'om fiiends, rather than partners, bufl‘ered mothers fi'om the adverse effects of the daily hassles of parenting. Perhaps the support provided fiom other mothers (i.e., fiiendship) was protective because they were more likely to share the common daily experience of hassles 26 related to child care duties. Further, a longitudinal investigation, by Norbeck and Anderson (1989), found significant stress-bufi‘ering effects of social support on anxiety in pregnancy among low income women. Although these researchers operationalized stress as the negative event scores fi'om a life events questionnaire, as opposed to daily hassles, they found that Time 1 measures of stress and social support significantly predicted Time 2 anxiety. Lastly, in a study of the stress-bufi‘ering hypothesis among married couples, Steinglass (1988) reports that individual’s perceptions of social embeddedness mediated reactions to stressfirl life events. . ' t The current study is interested in determining whether general embeddedness in a social network directly efl‘ects positive care-giving, or whether social support provided underconditions of stress leads to more positive maternal attitudes and behavior. The litaature suggests that both of these models of social support, direct effect and-stress- bufl‘ering, can. adequately explain the beneficial efi‘ects of social support on maternal adjustment and well-being. One of the main contributions of the current study is to test both of these models of social support simultaneously. Rationale or the Present Stu There is a growing literature on the beneficial efl‘ects of social support on maternal functioning, especially among adolescent mothers who experience a more dificult transition to parenthood, and are at greater risk for adverse outcomes, than their adult counterparts. However, social support interventions for this population have had mixed results (Klerman, 1979; 1993). Hence, it appears that we still have not unequivocally identified the specific social support variables that have the most beneficial efi‘ects for 27 pregnant and parenting adolescents, nor have we determined the specific mechanisms at work in socially supportive interactions. The present investigation tested an analytical model of two distinct support sources and their differential efi‘ects on adolescent adjustment to parenting. Two difl‘erent measures of social support were used, which were intended to access two independent sources of social support, partner emotional support and network social support, and each was tested using a moderating and main efi‘ect model. A specific functional measure of social support was intended to be sensitive to the specific functions of partner support, as opposed to a global structural measure which only identifies the presence of such a relationship, since the intention is to assess the availability of a close, emotionally supportive relationship and to determine the function of this type of support. In terms of the dimension of network social support included in the present model, a compound measure :of firnctional support including a wide variety of support functions was assumed to be most effective in assessing the efl‘ects of this type of social support. The association between social support and well-being has been attributed to both a main efl‘ect model and a stress buffering model. The buffering model is most likely to be found when the social support measure assesses interpersonal resources that are responsive to the needs elicited by stressful events. The main efl‘ect model is more likely to be found when the support measure assesses a person's degree of integration in a large social network or community. In order to clarify the operative mechanisms at work in either of these models, investigators need to test the effectiveness of specific support 28 resources in response to specific stressors. Additionally, there is a need for clear theoretical models of the mediating processes between support and well-being in relationships. The construction of the present theoretical model was an effort to clarify two specific mechanisms through which social support may have an effect on adolescent's adjustment to parenting. Zk Present Theoretical Model of Adolescent Parenting Social support was hypothesized to afl‘ect parenting behaviors both directly and indirecfly in this theoretical model. The first theoretical assumption in this model of adolescent parenting was that the quality of partner support would influence the teen's worldng model of the selfin relation to others. This change was expected to provide her - with enhanced self-esteem and feelings of belonging in a mutual, supportive relationship, which, in turn, was expected to influence her ability to enter into a mnturant relationship with her child. Hence, this model was concerned with the quality of support provided ' through the teen’s relationship with her partner and the intermediate processes that affected the adolescent's feelings and attitudes towards motherhood. The second major theoretical link in this model concerned the effects of support on stress associated with the transition to parenthood. The type of support specific to this link in the model was functional social support provided by the teen's greater family and community network structure. This type of support was supposed to be responsive to specific stressors related to the teen's new role as a parent, and it was hypothesized to exert both a protective or bufi‘ering efl‘ect against stress, as well as a main effect on the teen's psychological adjustment and well-being. 29 In summary, the present investigation was an effort to specify the mechanisms through which social support may have a beneficial effect on adolescent adjustment to parenting. Although this is an area that has been extensively studied in the field of psychology, much of the research in this domain has been criticized as lacking clear theoretical underpinnings and lacking clear conceptualizations of how social support afl‘ects well-being. There have been numerous attempts to document the occurrence of social support, with little efl‘ort to test theories of how it actually operates. The present study attempted to clearly conceptualize the construct of social support, taking into account the type, amount, source, and nature of support to be studied; inherent in this model was the notion that specific types of support would be related to specific aspects of adjustment. This theoretical model was an efl‘ort to understand more fully the multidimensionaL transactional nature of firnctioning during the transition to parenthood, and the mechanisms through which socially supportive relationships may afl‘ect adaptation throughout this process. This model does not attempt to capture the full complexity of factors that contribute to positive adjustment to parenting. Instead, it was intended to be one of many building blocks of knowledge about the processes that afi‘ect maternal adjustment among adolescents. Hwotheses 1. Family environment variables will have a direct efl‘ect on the outcome variables. 2. Family environment variables will predict self-esteem (e. g., positive and negative family environments will lead to higher and lower levels of self-esteem, respectively). 3. Partner emotional support will predict self-esteem (e.g., adolescents whose partners 30 provide greater amounts of emotional support will have higher levels of self-esteem than those who receive little or no emotional support from their partner). 4. An interaction is predicted between partner emotional support and family environment. 5. Self-esteem will be positively related to the outcome variables. For instance, higher levels of self-esteem will be associated with greater knowledge of infant development, and more positive parenting attitudes. 6. Hassle fi'equency will have a direct negative efl‘ect on the outcome variables. For instance, as frequency of hassles increase, positive parenting attitudes and knowledge of infant development will decrease. 7. Network social support will have a direct positive effect upon the outcome variables. 8. Network social supportflS SB) will bufl‘er the stress related to parenting (hassle frequency), leading to more positive maternal adjustment; . 31 ME THODS Research Participants The participants (n=134) were recruited fi'om an alternative school for pregnant/parenting adolescents in a mid-sized Midwestern city. This program ofi‘ered a variety of services including two meals per day, child development classes, nutrition and traditional instruction, cab service to and from school, and child care. Teens could remain in the program for two semesters after they had given birth, then they had to return to their home school, unless they were under 16 in which case they could remain in the program until their sixteenth birthday. This program served predominantly low income families (school records indicated that about 60% of the students received AFDC). The majority of the sample (45 %) were Afiican American, 28 % were European American, 13 % were Latina, 4 % were Native American, and 10 % were ofother or mixed ethnic/racial backgrounds. The average age of the students was 15, with a range fi'om 12 to 20 years of age. The current study involved the analysis of a set of data collected as part of a larger, longitudinal study involving mentoring of pregnant/parenting teens. The participants at Time 3 made up the sample of subjects for the present investigation. Time 3 data was deemed most optimal for testing the present hypotheses because the dependent variable in this model was the adolescent’s adjustment to parenthood. At Time 3, the adolescent had been a parent for at least six months, as opposed Tl when she had not yet given birth, or T2 when her infant was a newborn, and she was undergoing many changes as she adjusted to her new role as a mother. Although data for the present study was based primarily on 32 T3 interviews, there are three instruments that measure the teen’s family environment which were administered at Time 1 and Time 2. These instruments were intended to measure constructs that remain fairly constant over time, namely, the adolescent’s perception of her mother’s parenting style and the dimensions of maternal acceptance and control. M_eas_u_rg Several existing measures were utilized for this research project. (See Figure 3: Measurement Model). AMent’s Fa_r_m_lz° Environment Adolemt ’s Parent ’5 Parenting Style. The Parenting Attitude Questionnaire (PAQ) (Buri, 1989) is a 30-item inventory that measures the parenting style of the teen’s mother. This inventory, which was only administered at T1, yields three main factors: authoritative parenting style, authoritarian parenting style, and permissive parenting style. The Cronbach alphas obtained for this sample for each factor are .80, .73, and .46, respectively. Participants responded to items on a 5-point Likert scale (1=strongly disagree, 5=strongly agree). An example of an item on the authoritative parenting style factor is “my mother gives her children direction and guidance in a reasonable and fair way;” an example of an item on the authoritarian parenting style factor is “my mother does not allow me to question any decision that she makes;” and an example of an item on the permissive parenting style factor is “my mother does not view herself as responsible for directing and guiding my behavior as I grow up.” Adolescent ’s PerceLtion of Her Pa_r_ent ’s BeLravior. The Children’s Report of 33 Parental Behavior Inventory (CRPBI) (Schaefer, 1965) is a 30-item inventory that measures the adolescent’s perception of her relationship with her mother. This measure was administered postnatally, at T2. Factor analysis of this measure yielded four factors with adequate Cronbach alphas for this sample: acceptance (.91), psychological control (.82), easy parenting (.79), and hard parenting (.66). A previous factor analytic study of this measure, on a sample of 350 5th, 6th, 9th, and 10th graders, yielded this same factor Wu"? structure consisting of two dimensions of control and two dimensions of discipline (Kawash, 1988). Subjects responded to items on a 3-point Likert scale (1=not like my mother, 3=a lot like my mother). An example of an item on the acceptance scale is “my mother isa person who gives me alot of care and attention;” an example of an item on the psychological control scale is “my mother is a person who, if I have hurt her feelings, stops talking to me until I please her again;” an example of an item on the easy parenting scale is “my mother isa person who gives me as much fieedom as I want;” and an example of an item on the hard parenting scale is “my mother is a person who believes in having a lot of rules and sticking to them.” Adolescent ’s Mother ’5 Perception at Her m Parenting Style. The Family Experience Questionnaire (FEQ) (Frank, Jacobson, & Hole, 1986) is a 62-itern inventory that measures the respondent’s attitudes toward being a parent. This inventory was administered to the W at Time 2. Factor analysis of this scale yielded two factors: positive parenting attitude with a Cronbach alpha for this sample of .83, and negative parenting attitude with a Cronbach alpha for this sample of .85. Responses were made on a 4-point Likert scale (1=strongly agree, 4= strongly 34 agree). Some examples of items making up the positive parenting attitude scale are “I live for my children,” and “as a parent, I never stop enjoying seeing the world through my children’s eyes.” The negative parenting attitude scale was made up of items such as “being a parent makes me feel drained and depleted,” and “I did not know how much anger I had inside of me until I became a parent.” Mm»! The Rosenberg Self-Esteem scale (Rosenberg, 1965) consists of 10 items measuring self-worth and self-acceptance on a Likert scale ranging fiom (1) strongly disagree to (4) strongly agree. Items include “All and all, I tend to'feel I am a failure” and “IfeelthatIamavaluableperson, atleastequalwithothers.” Thisscalehas demonstrated good test-retest reliability with a sample of teenage pregnant and parenting mothers (r=.96; Colletta, 1981). The Cronbach alpha obtained for this sample at Time 3 was .78. The sum score of all scale items was used in the analyses. M’ m Measures Partner @1211 ,Supmrt. A modified version of the Maternal Social Support Questionnaire (N orbeck, Lindsey, & Carrieri, 1981) was employed to measure partner social support. For the original scale, the authors report a test-retest reliability of .92. This instrument was designed to measure the structural aspects of the mother’s social network. These items include the number of supporters listed, the relationship of the supporters to the respondent, the length of time the respondent has known each supporter, the proximity of the supporter to the respondent in terms of where they live, and the fi'equency of contact between the respondent and the supporter. The response format for 35 each of these items ranges from 1 to 5 with the higher numbered responses indicating greater frequency or proximity. Participants were asked to list up to 20 supporters and indicate for each how much emotional support, advice and information, and practical assistance was provided to them specifically about parenting, and how much general, overall support about issues other 1 than parenting was provided. Each type of support provided was rated on a 5—point Likert scale, ranging fi'om none at all (1) to a great deal (5). The correlation between the number of supporters reported by the teens for T1 and T2 was r--.44, and for T2 and T3 was r= .53. For the present study, the only support variable fi'om this questionnaire that was used in the analyses was the emotional support provided by the adolescent’s partner. Scores range from 1 (none at all) to 5 (a great deal). Respondents who did not list a partner as a supporter were given a score of zero as their partner emotional support score. fletwork Sgial Spppgrt. The Inventory of Socially Supportive Behavior (ISSB), a unidimensional measure of social support (Barrera, Sandler, & Ramsey, 1981), was used in the present investigation to measure network social support. This scale had adequate test-retest reliability (r=.82) and internal consistency (.92) in a sample of pregnant adolescents (Barrera, 1981). This 40-item scale measures how often participants perceive that they have received functional support fiom their support network on a 5-point Likert scale (1=not at all, to 5=about every day). Some examples of the items are “gave you under $25 to keep” and “agreed that what you wanted to do was right.” This measure yields an overall support score which is the average of the 40 items. Stokes and Wilson 36 (1984) indicate that this instrument is an appropriate measure of global social support, reporting an alpha of .92 for their study sample. The Cronbach alpha obtained for this sample was .93 at Time 3. M A modified version of the Hassles Scale (Kanner, Coyne, Schaefer, & Lazarus, 1981) was administered. The original version of this scale included 117 items and was validated on a sample of EuroAmerican men and women between the ages of 45 and 64; test-retest reliability was .79. In the 24-item version of this scale used for this study, redundant items and items and words that suggested psychological and somatic symptoms were eliminated (cf, Dohrenwend, Dohrenwend, Dodson, & Shrout, 1984) as well as items inappropriate to a teen population. The Cronbach alpha obtained for the current sample at Time 3 was .86. The domains tapped in this revised scale were work, health, family, fiiends, the environment, practical considerations, and chance occurrences. The participants were asked to indicate on a 4-point Likert scale, ranging from none (1) to a great deal (4), how much of a hassle each item had been for them during the previous month. Items such as sex, intimacy, your health, and your medical care were included on the scale. The scale yielded three separate stress scores: (1) hasslefiequency, a simple count of the number of items checked (scores range from 1 to 24); (2) hassle severity, the sum of the severity ratings (range: 24 - 96); and (3) hassle intensity, the average intensity of all the stressors perceived by the subject (range: 1- 4). In the current study, only the hassle frequency score was utilized in order to simplify the number of variables being entered into 37 the statistical analyses. Adolescent ’s Adjustment to Parenthood Knowledge at Intant Development. The Knowledge of Infant Development Inventory (KIDI) was developed by MacPhee (1981). The original 75-item scale includes several subsections. Only the 20-item “Milestones” subscale was used for this study. Questions such as “Most babies can sit on the floor without falling over by 7 months” and “A two-year-old is able to reason logically, much like an adult” are answered in an agree/older/younger/not sure format. That is, one indicates whether one agrees that the statement is age-appropriate, whether the statement is true of older or younger infants, or whether one is not sure about the statement. Only one score was used for the present . study: the percentage of times the respondent provided correct judgements of her infant’s abilities. The Cronbach alpha for this study sample was .54 at Time 3. Adolescent ’s Perception at Her QM Parenting Attitudes. The Family Experience Questionnaire (FEQ) (Frank, Jacobson, & Hole, 1986) is a 62-item inventory that measures the teen’s attitudes toward being a parent. This inventory was administered postnatally, at T2. Factor analysis of this scale yielded two factors: positive parenting attitude with a Cronbach alpha for this sample of .87, and negative parenting attitude with a Cronbach alpha for this sarnple of .82. Responses were made on a 4-point Likert scale (1=strongly agree, 4= strongly agree). Some examples of items making up the positive parenting attitude scale are “I live for my children,” and “as a parent, I never stop enjoying seeing the world through my children’s eyes.” The negative parenting attitude scale was made up of items such as “being a parent makes me feel drained and depleted,” and “I did 38 not know how much anger I had inside of me until I became a parent.” Procedure Participation in the research study was voluntary and available to all pregnant adolescents who were first-time mothers enrolled in a special school for pregnant and parenting adolescents during a four-year period. Adolescents and their parent(s) or legal guardian(s) were asked to participate in the study during their intake interview into the special school. All adolescents older than 18 gave their written consent; younger teens gave their verbal consent. Signed consent was also obtained from each teen's parent or legal guardian. ' Each adolescent was interviewed three times by trained undergraduate interviewers: prior to the assignment of experimental condition and before the birth of the child ('1‘ 1, prenatal) and one month (T2, postnatal) and six months (T3, follow-up) after the birth. Financial reimbursement was received for each interview in which the teen participated ($5 for T1; $10 for T2; and $15 for T3). Afier the T1 interview, teens were randomly assigned to a schooling as usual group or an experimental group. The 9-month intervention paired teens with a trained female mentor fiom the community who was to meet weekly with the teen for the purpose of enhancing her social support network, her parenting skills, and her educational achievement. Preliminary analyses indicated no difl‘erences between the mentored and non- mentored teens on any T1, T2 or T3 outcome variables collected as part of the larger study; therefore, the two groups were combined in the analyses presented in this paper. Complete data for the three time periods in the larger study was not available for every 39 teen. At Time 1, of the 175 teens who were eligible to participate in the study, 3 did not complete a T1 protocol. At T2, attrition reduced our sample to 155. At T3, there were 134 teens who had complete interviews. There were no difl‘erences detected between our final sample and those participants lost through attrition or incomplete data. 4O RES UL TS As stated previously, the current investigation was an attempt to test a theoretical model of the efl‘ects of two types of social support on adolescents’ adjustment to parenthood. Also, given the debate in the literature about the process by which social support mitigates well-being, both main efl‘ect and mediating theories of social support were examined. Statistical evaluation of this theoretical model, using SPSS-Wmdows (Statistical Package for the Social Sciences), was carried out in two stages. First, a factor analysis of the nine adolescent family environment variables was conducted to reduce . . these to two factors that represented positive and negative family environments. Second, a series of hierarchical multiple regression analyses were performed to test the hypotheses of the study and to determine whether the model was an appropriate fit for the data. A more detailed description of each stage of the analysis and the findings of this study are presented below. Factor A The independent variable labeled “Adolescent’s Family Environment” (See Figure 3: Measurement Model) consists of nine variables that are intended to measure dimensions of the adolescent’s relationship with her own mother. These include the teen’s perception of her mother’s parenting style and the adolescent’s mother’s report of her own parenting attitudes. The teen’s perception of her mothers’ parenting style were measured with the Parenting Attitude Questionnaire (PAQ), which consists of three factors, authoritarian, authoritative, and permissive parenting styles, and with the Children’s Report of Parental Behavior (CRPBI), which consists of four factors, psychological control, psychological 41 acceptance, strict parenting, and easy parenting. The mothers’ report of her own parenting attitudes were gathered using the Family Experience Questionnaire ( FEQ) which consists of two factors, positive parenting attitudes and negative parenting . attitudes. Reduction of these nine factors into two dimensions, positive and negative family environment, was attempted by performing a principle components factor analysis in which two unrotated factors were extracted. Due to double loading, permissive parenting style was discarded fi'om the original factor analysis. The resulting two-factor sohrtion for- the remaining variables is reported in Table l. The four variables that loaded onto Factor 1 were psychological acceptance (-.70), hard parenting (.83), psychological control (.78), and authoritarian parenting style (.66). This factor was conceptualized as representing a family environment in which unaccepting, strict, psychologically controlling authoritarian parenting practices were dominant. The variables that loaded onto Factor 2 were positive parenting attitude (.75), negative parenting attitude (.33), and authoritative parenting style (.63). This factor was conceptualized as reflecting a family environment in which authoritative parenting practices, and both favorable and unfavorable attitudes towards parenting, were dominant. Standardized scores for each of these two factors were generated for each subject. Factor analyses of the FEQ were conducted to determine whether the adolescents’ and their mothers’ responses resulted in the same factor structure (e. g., positive and negative scales). The responses of both groups did, in fact, result in identical positive and negative scales with reliabilities of .83 and .85 for the mothers, and .87 and .82 for the teens. It appears that the structure of the factors did not difl’er for these two groups; 42 however, the relationship between the positive and negative scales is different for adolescents and their mothers. Among the mothers of teens, a positive relationship exists between the positive and negative scales of the FEQ (r= .16), while among the teenage mothers an inverse relationship (r= -. 15) exists between these two scales. An r to Z transformation revealed that these correlations are significantly difi‘erent fiom each other (1; = -10.00; p < .001). In summary, factor analysis of the variables constituting the adolescents’ family environment resulted in two factors that reflected adaptive and maladaptive parenting practices. The first factor, maladaptive parenting practices, reflects an environment in which parents tended to be strict, controlling, and unaccepting of the teen; the second factor reflects an environment in which more mature, accepting parenting dominated. Further analyses showed that adolescents and their mothers do not share similar views regarding the relationship between positive and negative parenting attitudes. Hierarchical Mqu’ le Rm’ n Am The second stage of the data analysis involved a series of hierarchical multiple regressions to test the hypotheses of this study. Because the proposed model had two variables that constituted the adolescent’s family environment (adaptive and maladaptive parenting practices) and two outcome variables (adolescent positive parenting attitude and correct knowledge of infant development), four main regressions were computed (See Figures 4, 5, 6, and 7). In addition, each of these four regression analyses was computed in two parts due to the limitations of SPSS’s hierarchical multiple regression program for handling moderator variables (See Figures 4, 5, 6, and 7: Self-esteem moderates family 43 environment and adjustment to parenting). Hence, in the first part of each regression, self- esteem was entered as the dependent variable, predicted by family environment, partner emotional support, and the interaction term, family environment X partner emotional support. In the second part of each regression, the adolescent’s adjustment to parenthood was the dependent variable, predicted by family environment, self-esteem, network social support, parenting stress, and the interaction term, network social support X parenting stress. The criteria for statistical significance was set at p < .05 in each regression. R ° n ne: M tive F ' Environment Predica'n Know e Injant Development Results of the first hierarchical regression analysis are reported in Table 2. In part one, self-esteem is entered as the dependent variable, with maladaptive family environment (MFE), partner emotional support (PBS), and the interaction term, MFE X PES, entered as the independent variables. Only maladaptive 'family'environment significantlypredicted self-esteem (t_= -2.45) and accounted for a significant portion of variance in self-esteem (See Table 3). Specifically, maladaptive family environment accounted for 9% of the variance (F (2, 80) = 7.91), while partner emotional support and the interaction between maladaptive family environment and partner emotional support only increased the variance accounted for by 1%. In part two of this regression (See Table 2, Part Two), correct knowledge of infant development was entered as the dependent variable, and MFE, self-esteem, network social support, hassle frequency, and the interaction terrrr, hassle frequency X network social support, were entered as the independent variables. Two variables emerged as significant 44 predictors of the outcome variable: hassle frequency (1 = 1.96) and the interaction term, hassle frequency X network social support (1 = -1.97). However, none of these predictor variables accounted for a significant portion of the variance in correct knowledge of infant development (See Table 3). area Results of the second hierarchical regression analysis are reported in Table 4. In part one, self-esteem was entered as the dependent variable, with maladaptive family environment (MFE), partner emotional support (PBS), and the interaction term, MFE X PES, entered as the independent variables. Only maladaptive family environment significantly predicted self-esteem (t_= -2.45) and accounted for a significant portion of variance in self—esteem (See Table 5). Specifically, maladaptive family environment accounted for 9% ofthe variance (F (2, 80) = 7 .91), while partna' emotional support and the interaction between maladaptive family environment and partner emotional support only increased the variance accounted for by 1%. In part two of this regression (See Table 4, Part Two), adolescent positive parenting attitude was entered as the dependent variable, and MFE, self-esteem, network social support, hassle frequency, and the interaction term, hassle frequency X network social support, were entered as the independent variables. Two variables emerged as significant predictors of the outcome variable: hassle fi'equency (t = -2.25) and the interaction term, hassle fi'equency X network social support (1 = 2.19). Only network social support (R2= .09; F (5, 77) = 2.67) and the interaction between network social 45 support and parenting stress (R2 = .15; F (5, 77) = 2.68) accounted for a significant portion of the variance in adolescents’ positive parenting attitudes (See Table 5). ° n Three: A tive Famil Environment Predictin Correct Know e o Injant Development Results from the third hierarchical regression analysis are reported in Table 6. In part one, self-esteem was entered as the dependent variable, with adaptive family environment (AFE), partner emotional support (PBS), and the interaction term, AFE X PES, entered as the independent variables. None of these independent variables significantly predicted self-esteem, nor did they account for a significant proportion of the variance (See Table 7). In part two of this regression (See Table 6, Part Two), correct knowledge of infant development was entered as the dependent variable, and adaptive family environment, self- esteem, network social .support, hassle frequency, and the interaction term, hassle fiequency X network social support, were entered as the independent variables. Again, none of these variables emerged as significant predictors of the outcome variable. Additionally, none of these variables accounted for a significant amount of variance in the outcome measure (See Table 7). Attitudes Results fi'om the fourth hierarchical regression analysis are reported in Table 8. In part one, self-esteem was entered as the dependent variable, and adaptive family environment (AFB), partner emotional support (PBS) and the interaction term, AFE X 46 PBS, were entered as the independent variables. None of these independent variables significantly predicted self-esteem, nor did they account for a significant proportion of the variance (See Table 9). In part two of this regression (See Table 8, Part Two), adolescent positive parenting attitudes was entered as the dependent variable, and adaptive family environment, self-esteem, network social support, hassle fi'equency, and the interaction term, hassle frequency X network-social support, were entered as the independent variables. The variables which emerged as significant predictors of the outcome measure were hassle frequency (1 = -2.04) .and the interaction between hassle fi'equency and network social support (1 = 2.01). Additionally, analysis of variance (See Table 9) revealed that only the interaction term, network social support X hassle fi'equency, accounted for a significant proportion of the variance in adolescent positive parenting attitudes (Rz= .134; F (5, 77) = 2.40). Post Hoe Am The finding that partner emotional support was not a significant predictor of self- esteem, neither direcfly nor as a moderator variable, was inconsistent with the literature. Therefore, an attempt was made to firrther explore the relationship between partner emotional support and self-esteem among this sample by changing the way partner emotional support was scored. Because the original scoring method involved assigning a score of zero to all participants who did not list a partner as a supporter (N=55), the mean level of partner emotional support seems to have been under-represented for this sample. Perhaps the lack of significant findings was due to this negatively skewed distribution of 47 scores; in order to test this hypothesis, a series of analyses were performed using two different sets of scoring criteria for partner emotional support. The first post Leg analysis involved excluding fi'om the analyses all subjects who did not list a partner as a supporter, and then re-running all of the hierarchical multiple regression analyses with these new partner emotional support scores. Using this method of scoring, the number of subjects who listed a partner as a supporter was 79, but this number dropped to 54 when those subjects who did not meet the criteria of having complete data for the other measures were excluded. The findings of these analyses were that no significant relationships emerged. The second post hm analysis involved scoring partner emotional support dichotomously. Individuals who listed a partner as a supporter were given a score of 1, and those who did not list a partner as a supporter were assigned a score of O. This scoring method ignored the amount of emotional support provided by the partners, and only took into account presence or absence of partner emotional support. The series of hierarchical multiple regression analyses were then computed using flris revised scoring method. Results fiom these analyses did not reveal any significant findings. 48 DISCUSSION The primary purpose of this study was to explore the differential effects of two specific sources of functional social support on adolescents’ adjustment to parenthood. A theoretical model was developed, based on the literature, which emphasized the importance of social support for well-being among women undergoing the challenges of pregnancy and parenthood. Among adolescent women, child rearing is especially stressfirl, and the benefits of social support for these youngsters and their children have been described in the literature (e.g., Camp et al., 1993; Thompson, 1986; Unger & Wandersman, 1985). However, extant research sufl‘ers from a lack'of clarity regarding ‘ who in the teenage mothers’ network provides the most efl‘ective support, and how this support actually firnctions to efl'ect positive adjustment to parenthood. For instance, much of the literature focuses on the types of stressors faced by adolescent mothers (e.g., Barth, Schinke,-& Maxwell, 1983; de Anda et al., 1992; Passino et al., 1993) or describes the quality of these parent-child relationships and the negative outcomes often associated with them (e.g., Crnic et al., 1984; McAnarney, Lawrence, & Aten, 1979; Ragozin et al., 1982). The construct of social support has, for the most part, been described in general terms in the literature, and has been loosely applied to the problem of adolescent parenthood. For example, the value of social support as a predictor of positive adjustment has been extolled by many researchers; however, closer scrutiny of the evidence reveals that there is considerable debate in terms of which specific facets of social support are responsible for which aspects of positive outcome. The types of questions that continue to provide grist for the mill in this domain of psychology include who in the teen’s 49 network provides the most effective support? Under what conditions are adolescents most receptive to help (docs social support exert a reparative efl‘ect)? How does support actually influence positive adjustment (what are the mechanisms through which it operates)? What types of support are most efl‘ective? Answers to questions like these are fundamental to the evolution of social support theory, as well as to the practical application of social support interventions with adolescent parents. Hence, the goal of the current investigation was to ask very specific questions about two difl‘erent sources of social support. To that end, this study focused on partner emotional support and general network social support, and specified an analytical model that attempted to explain how these support sources function to afl'ect positive adjustment to parenthood among adolescents. Another primary goal of the current study was to test whether the two types of social support specifiedin this model interact with certain environmental variables to efi‘ect positive adjustment, or whether positive outcomes are due to a direct efl‘ect of social support. The literature supports both direct and mediating models of social support, so it is still unclear under what conditions each mode of social support operates. ngor Finding The results of this study suggest that the proposed model of social support, as a whole, does not accurately fit the data for this sample of adolescent mothers. The first part of this model (See Figure 2: Part One) was concerned with the efi‘ects of partner emotional support on adjustment to parenthood. In this part of the model, the adolescent’s family environment was predicted to a) afl‘ect self-esteem, which in turn 50 would affect parenting ability, and b) directly afl‘ect adjustment to parenthood. Partner emotional support was presumed to a) directly afl‘ect self-esteem and b) act as a mediator between family environment and self-esteem. Self-esteem was then presumed to impact adjustment to parenthood. The only significant finding from this piece of the model was that maladaptive family environment had a direct inverse relationship with self-esteem. Adaptive family environment, however, did not emerge as a significant predictor of self- esteem, nor was self-esteem significantly related to either of the outcome variables. Lastly, and perhaps most'detrimental to the viability of this part of the model, was the finding that partner emotional support had no efl’ect on self-esteem, and was unrelated to positive adjustment to parenting. .- The second part of the model (See Figure 2: Part Two) dealt with the direct and stress-bufi‘ering efi‘ects of network social support on adjustment to parenting. This part of the. model was a slightly better fit for the data than the first part. 3A consistent-finding among these variables was that parenting stress had a significant negative impact on adjustment to parenting. Most relevant to the current thesis was the finding that network social support served a stress-bufl‘ering function in relation to adjustment to parenthood, although it did not exert a direct effect on outcome. In summary, it appears that certain links in this model were supported by the data, but as a whole this theoretical model does not accurately describe this sample of adolescents. In the following pages, these findings will be more fully discussed. Partner Emotional Sgpport There were two main frameworks utilized in this model for conceptualizing the 51 efl’ects of partner emotional support upon adolescents’ self-esteem. First, a direct efl‘ect of partner emotional support on self-esteem was proposed based on research that suggests that social embededness is responsible for the beneficial effects of social support (Cohen & Wills, 1985). The premise behind this finmework was that mere involvement in a supportive intimate relationship would foster feelings of self-esteem and personal efiicacy among parenting adolescents. Moreover, support fi'om a male partner has a special cultural meaning inconnection with the birth of a child (Thompson, 1986). However, among this sample of teenage mothers, involvement in an intimate relationship was not significantly related to feelings of self worth. An explanation for the current lack of findings might be that the partner relationships of these teens, although emotionally supportive, were also the source of conflict (Thompson, 1986) and stress - partner characteristics that were not analyzed in the present model. Also, most of the literature documenting the positive effects of partner and spouse support on maternal adjustment involves adult populations (e.g., Belsky, 1984; Crnic et al., 1984; O’Hara, 1983; Paykel et al., 1980). An exception, however, is a study by Samules et al. (1994) who found that support fiom the baby’s father enhanced adjustment to parenting among adolescents (mean age 17.5 years). However, their results showed that older mothers (ages ranged, fi'om 14 to 20) seemed to have a better situation for parenting; they were more likely to be married and have more contact with the baby’s father, and they had higher self—esteem. These authors suggest that the delay of childbearing, even at these younger ages, brings a more positive outcome. Similarly, Crnic et al. (1984) found that among parents between the ages of 16 to 38, the older mothers’ reports of partner support were positively related 52 to efi‘ective behavior toward their infants. The lack of findings regarding a direct efi‘ect of partner support in the current study emphasizes the important point that the circumstances surrounding adolescent parenthood (the mean age in the current study was 15 years) may be substantially difl‘erent fiom those faced by adults, or even older adolescents; therefore, adolescent groups should not be seen as homogeneous, and caution should be used in generalizing findings fiom adult populations to adolescent ones. A second framework used to conceptualize the proposed moderating role of partner emotional support was based on the work of Egeland et al. (1988) who specify difi‘erent types of relationships (one of which is an emotionally supportive relationship with an adult partner) that they believe exert a reparative influence on poor self-concept resulting fi'om negative familial environments. Thus, in the current study, partner emotional support was hypothesized to have a reparative efl‘ect on the self-esteem of adolescents who came from homes in which maladaptive parenting practices dominated. The data did not, however, support this hypothesis. Several important difi‘erences exist between the current sample and those studied by Egeland et al. which may help to explain the inconsistencies in findings. First, unlike the current sample, all of the women in the Egeland study had been abused as children, and they were being studied to see whether they would continue the cycle of abuse, and maltreat their own children. They were also older than the teens in the current study (average age was 20.5 years opposed to the current sample’s mean age of 15 years), and they had been parents for a longer period of time (mothers were observed longitudinally with their children from birth to three years). Another important difference is the quality of intimate relationships among adolescents 53 and their adult counterparts. It is generally the case among adolescents that intimate relationships are more sporadic, transient, and less stable than adult relationships because biological maturation ofien precedes social, cognitive, and emotional maturation during adolescence. Hence, early maturing adolescents may enter into sexual relationships without having the psychological and emotional sophistication needed to sustain them (Schonert-Reichl & Ofi‘er, unpublished manuscript). It may be the case that the teens and their partners in this study were not yet mature enough to engage in the types of stable, secure relationships, described by Egeland et al., that could affect qualitative changes in self-image. Evidence for this can also be gleaned from research on the consequences of adolescent marriages. Lamb and Elster (1990) note that adolescent marriages are highly unstable and may delay or prevent the successfirl completion of developmental tasks commonly associated with adolescence, such as identity formation, autonomy, and intimacy. Theyalsoreportthatthese marriagestend not tobe strong, aredistressed, and the partners do not generally identify each other as sources of emotional support. Lastly, the lack of variance accounted for by partner emotional support suggests that there are myriad other factors, not depicted in the current model, that more importantly influence self-esteem among the teenage mothers in this sample. A cursory survey of the literature highlights some of the areas that could possibly account for a greater proportion of the variance in the current model. Many researchers agree that parenting adolescents rely more fiequently on their mothers for support than any other support source, including the baby’s father (Colletta & Gregg, 1981; Unger & Wandersman, 198 5). Maternal support, however, is not always available, nor is it 54 necessarily given without conflict, which can negatively impact feelings of self-worth (Davis & Rhodes, 1994; Olson & Worobey, 1984). A related finding is that changes in levels of maternal support, whether positive or negative, have been shown to reduce self- esteem in parenting adolescents (Bogat, Caldwell, & Guzman, unpublished manuscript). Other factors that have been associated with low self-esteem among parenting adolescents are high density social support networks (Dunst, Vance, & Cooper, 1986) and the number of stressful life events facing the adolescent (Young, 1990). Tolan, Miller, and Thomas (1988) found that adolescents who experience more daily hassles have more negative self- images. Implications for the current study might be that the benefits of partner emotional support on self-esteem were far outweighed by burdensome stressfirl events such as poverty or violence. Finally, Renick (1990) cautions that the level of social support perceived by adolescents is inextricably linked to their feelings of self-worth. There is a difi‘erence between how much support is perceived by an individual and how much support she actually receives, because perceptions of social support are situationally determined and may vary fi'om day to day depending on factors such as mood. These findings strongly suggest that social support be viewed as a multidimensional rather than unidimensional concept, and that specific sources of social support be evaluated for their unique efl‘ects on feelings of self-worth and personal adjustment. ALaptive and MMtive Farrah Environments The only significant finding regarding family environment was that maladaptive parenting practices negatively impacted the adolescent’s self-esteem. This finding is consistent with previous research on the efi‘ects of parenting style on self-esteem 55 (Abernathy et al., 1975; Buri, 1989; McCormick, 1994; Olson & Worobey, 1984; Paterson, 1995). An unusual finding was that adaptive parenting practices had no impact on self-esteem; this is inconsistent with current literature (Buri, 1992) . This perhaps speaks to the ambiguous meaning behind the adaptive family environment variable that resulted when the nine variables making up the family constellation of the adolescent were factor- analyzed. Specifically, it was expected that the positive and negative parenting attitude scales of the FEQ would load onto positive and negative family environment factors, respectively. Instead, both of the parenting attitude scales loaded onto one factor which did not instantly fit an intuitive “positive parenting” schema (whereas the other factor consisted of negative parenting practices that intuitively fit a “negative parenting” schema). A modified conceptualization of this factor was developed that was labeled “Adaptive Family Environment” because it was assumed that respondent’s who have a more mature understanding of parenting might be more likely to endorse both positive and negative items. For example, well-seasoned parents may not view the items “I have the knowledge I need to be a good parent” and “My kids are always trying my patience” as contradictory statements, although the structure of the Family Experience Questionnaire (e. g., positive and negative subscales) categorizes the former as indicative of positive parenting attitudes and the latter as indicative of negative parenting attitudes. Alternatively, individuals who view parenting as either all good or all bad may be less flexible in their thought patterns, and unable to tolerate behavior fiom their children that is inconsistent with their preconceived notions of what constitutes appropriate behavior. Hence, this style of parenting may be less adaptive because it entails “either-or” thinking 56 and inflexible expectations. Analyses revealed that the “adaptive family environment” variable was unrelated to any of the other constructs in this model which lends further support to the explanation that this dependent variable may not have been a valid construct in this theoretical model. Lastly, family environment turned out to be unrelated to outcome among this sample of teenage mothers. This is surprising, given the wealth of evidence in the literature regarding the influence of family constellation on later organization of the home environment (Belsky, 1993; Cicchetti & Garmezy, 1993; Main et al., 1985; Morris, 1980; Ricks, 1985). Perhaps an explanation for the diversion of the current findings fi'om mainstream literature is the lack of sensitivity of the outcome measures in the present investigation. In particular, many of the previous findings cited in this thesis focus on attachment behavior and the intergenerational effects of attachment. However, in the current investigation, the measures used to assess the efi‘ects of family environment on adolescent adjustment to parenthood focused on appropriate knowledge of infant development and positive maternal attitudes. These particular outcome variables seem to be reliable indices of the adolescent’s adjustment to parenting, but they appear not to measure the intergenerational effects of attachment and parenting behavior. Network Social Support and Parenting M A finding of this study that was consistent with previous research (cf. De Anda et al., 1992) was that stress had an adverse effect on the adolescent’s parenting outcome. More frequent daily parenting hassles were associated with decreased knowledge of infant development and less positive parenting attitudes. General network social support, 57 however, served as a buffer against the daily parenting hassles experienced by these teenagers. Specificaly, although the tangible support provided by the teen’s support network did not directly afl‘ect her knowledge and attitudes about parenting, under conditions of increased stress it served a protective firnction. It has been suggested that the buffering effect of social support occurs when functional support (e. g., financial aid, material resources, needed services, social companionship, information, etc.) either reduces stress by a direct resolution of instrumental problems, or provides the recipient with increased time for personal actvities such as relaxation (Cohen & Wills, 1985). Hence, under conditions of increased parenting stress, the adolescents in this sample were able to elicit from their social support network the help that they needed, which beneficially afi‘ected their parenting attitudes and behavior. Thoits (1982), however, emphasizes that the stress-bufi‘ering efl‘ects of social support on adjustment, that are found in cross-sectional designs such as the present investigation, must be interpreted with caution. She warns that social support and stress may have a reciprocal relationship. The adequacy of social support afl‘ects the severity of stressful events encountered, and the severity of stress afl‘ects the adequacy of social support. This, she argues, may account for the buffering efl'ect in investigations that do not establish a baseline level of stress and social support, before measuring the interaction of the two. It is of theoretical importance, then, to investigate, over time, the mutual efi‘ects of stress and social support upon each other (Thoits, 1982) before drawing any definitive conclusions about the protective nature of social support. Lastly, functional network support did not appear to directly influence adjustment 58 to parenting, although it did account for a significant proportion of variance in positive parenting attitudes. This is not surprising given the fact that a main efi‘ect is most likely to be found when the social support instrument assesses the individual’s degree of integration in a social network or community. Hence, in the current investigation, the instrument used (the Inventory of Socially Supportive Behavior-IS SB) may not have been sensitive to the adolescent’s degree of embededness in her network, although the type of support it measured (e. g., instrumental support) did account for a significant amount of variance in parenting attitudes. Another caveat of the IS SB, which is a compound measure of enacted firnctional support, is that it assesses actual support recieved in the recent past, as opposed to perceived availability of support. Cohen and Wills (1985) argue that “past use” measures, such as the 18 SB, may to some degree reflect psychological distress which leads to increased use of social support. @nclusions and Suggestions (or Future Research The purpose of the present study was to examine the specific functions of two sources of social support in relation to adolescent parenting attitudes and knowledge. The types of support accessed in the current investigation were perceived support fiom a partner and enacted support from one’s network. These sources of support were tested in a theoretical model of adjustment to parenting which did not, as a whole, fit the data for the current population of adolescents. Ifreplicated, these findings could have significant implications, especially in the areas of policy and prevention. There has been increasing interest in the fathers of babies born to adolescent mothers. This interest is based on the fathers’ potential economic contributions to the family, especially in light of the current 59 climate of social welfare downsizing. Social service agencies throughout the country are developing programs that provide incentives for young fathers’ continued involvement with their children and the mothers (cf. Rabinovitz, June 16,1996, New York Times). The main thrust behind these interventions is to foster personal responsibility in these young I men, while at the same time reducing the need for government assistance among adolescent mothers and their children. There is also considerable debate as to the potential influences of father presence on the mother’s well-being and the child’s socioemotional development, especially in light of evidence of the importance of the family system for healthy child development (Elster & Lamb, 1982). However, according to the cinrent findings, efforts to include adolescent fathers in interventive efi‘orts and policy decisions may be premature. The current findings suggest that male partners do not provide a significant amount of support, and further research is needed to assess what efl‘ects they have on the mother and child’s socioemotional development. The present investigation was unable to substantiate any benefits of partner support on adolescent mothers’ adjustment. An important question that presents itself then, is whether this relationship has any detrimental efl'ects on maternal firnctioning. Ecologically-valid investigations into the efi‘ects of partner support and conflict on maternal and child adjustment and well-being are suggested before conclusions are drawn about the value of inclusion of the fathers into these young families. Also, in light of the current finding that younger adolescents are not receptive to the reparative influence of partner social support compared to older adolescents (cf., Sarnuels et al., 1994), developmentally sensitive research needs to be conducted. The goal of such research may be to identify the 6O developmental tasks that must be completed before intimate relationships can become a significant source of emotional support. Finally, the theoretical contribution of the current study is that further support has been found for the stress-bufi‘ering effects of social support, which can be helpful in planning empirically based support interventions with this population. This study also emphasizes the need to conceptualize social support as a multidimensional phenomenon and the importance of developing more complex theoretical models of the efl‘ects of social support on adjustment and well-being. Future research should use a developmental fi'amework to identify the benefits and detriments of partner support to both the mother and child, as well as examine what other types and sources of social support within young mothers’ network exert a reparative influence. Efl‘orts should continue to be made toward understanding the facets of social support that may lead to the development of interventions that help shape young women into strong, empowered, and efl’ective mothers. TABLES 61 Table 1 Eggs); Anglysis 9f Family Environment Variables: 1m Epgtor gagga - Adaptive grid Maladaptive Parenting Practices Factor 1: Baggy}: Maladaptive Family Environment Adaptive Family Environment Family Environment Variable Factor Loadilg Factor Loading CRPBI Easy Parenting - .43“ .28 CRPBI Psych. Acceptance - .70 .35 CRPBI Hard Parenting .83 .32 CRPBI Psych. Control .78 .15 FEQ Pos. Parenting Attitude - .19 , .75 FEQ Neg. Parenting Attitude .18 .33 Authoritarian Parenting Style .66 .45 Authoritative Parenting Style - .37 .83 ' bolded numbers indicate variables that load onto a particular factor 62 Table 2 Agglggam Parenting Outcome (I3KIDIRIGHD I i R re ion Anal siS'Malada tive Famil Environ e P Predictor Variable beta 3 Part One: Predicting Self-Esteem Maladaptive Family Environment (MFE) - .43 - 2.45 .02 Partner Emotional Support - .04 - .33 .74 MFE X Partner Em. Supp. .17 .95 .35 Part Two: Predicting Correct Knowledge of Infant Development (T3KIDIRIGHT) Maladaptive Family Environment (MFE) - .07 - .58 .58 T3 Self-Esteem .05 .42 .67 T3 Network Social Support .64 1.8 .08 T3 Hassle Frequency .98 1.98 .05 Hassle Frequency X Network Social Support - 1.35 - 1.97 .05 63 Table 3 il vironment Predictin d lescent Parentin Predictor Variable N AR 2 F 2 Part One: Predicting Self-Esteem Maladaptive Family Environment (MFE) 82 .09 7.91 .01 Partner Emotional Support 82 .09 3.99 .02 MFE X Partner Em. Supp. 82 .10 2.95 .04 Part Two: Predicting Correct Knowledge of Infant Development Maladaptive Family Environment (MFE) 82 .01 .37 .54 T3 Self-Esteem 82 .01 .52 .60 T3 Network Social Support 82 .01 .34 .79 T3 Hassle Frequency 82 .01 .27 .90 Network Soc. Supp. X Hassle Freq. 82 .06 1.00 .43 Table 4 Hierarchigl Multime Regression Analysis: Maladaptive Family Environment Predictng Mglescem Parenting ngmg (FEQPPA) Predictor Variable beta t a Part One: Predicting Self-Esteem Maladaptive Family Environment (MFE) - .43 - 2.45 .02 Partner Emotional Support - .04 - .33 .74 MFE X Partner Em. Supp. .17 .95 .35 Part Two: Predicting Adolescent Positive Parenting Attitudes Maladaptive Family Environment (MFE) .18 1.55 .13 T3 Self-Esteem .17 1.39 .17 T3 Network Social Support - .46 - 1.36 .18 T3 Hassle Frequency - 1.07 - 2.25 .03 Hassle Frequency X Network Social Support 1.43 2.19 .03 65 Table 5 Maladagivg Family Environment Predicting Adolescent Parenting Outcome (EEQEEA): Predictor Variable N AR 2 F A Part One: Predicting Serf-Esteem Maladaptive Family Environment (MFE) 82 .09 7.91 .01 Partner Emotional Support 82 .09 3.99 .02 MFE X Partner Em. Supp. 82 .10 2.95 .04 Part Two: Predicting Adolescent Positive Parenting Attitudes Maladaptive Family Environment (MFE) 82 .02 1.84 .18 T3 Self-Esteem 82 .05 2.1 1 .13 T3 Network Social Support 82 .09 2.67 .05 T3 Hassle Frequency 82 .10 2.05 .10 Network Soc. Supp. X Hassle Freq. 82 .15 2.68 .03 66 Table 6 Adolescent Parenting Outcome (T3KIDIRIGHT) Predictor Variable beta t L Part One: Predicting Self-Esteem Adaptive Family Environment (APE) .04 .22 .83 Partner Emotional Support - .02 - .21 .84 AFE X Partner Em. Supp. .09 .56 .58 Part Two: Predicting Correct Knowledge of Infant Development Adaptive Family Environment (AFE) - .12 - 1.07 .29 T3 Self-Esteem .08 .69 .49 T3 Network Social Support .64 1.84 .07 T3 Hassle Frequency .94 1.91 .06 Hassle Frequency X Network Social Support - 1.31 - 1.93 .06 67 Table 7 Maggiyg Family Envimnment Predicting Adolescent Parenting Outcome (IsnglBlQED: Predictor Variable N AR2 F L Part One: Predicting Sen-Esteem Adaptive Family Environment (AFE) 82 .01 .87 .35 Partner Emotional Support 82 .01 .44 .65 AFE X Partner Em. Supp. 82 .02 .40 .76 Part Two: Predicting Correct Knowledge of Infant Development Adaptive Family Environment (AFE) 82 .01 .88 .35 T3 Self-Esteem 82 .03 1.01 .37 T3 Network Social Support 82 .03 .68 .57 T3 Hassle Frequency 82 .03 .51 .73 Network Soc. Supp. X Hassle Freq. 82 .07 1.17 .33 68 Table 8 Hierarchical Multiple Regregion Analysis: Adagtive Family Environment Pmictigg ole n ntin come FE PPA Predictor Variable beta t Part One: Predicting Self-Esteem Adaptive Family Environment (AFE) .04 .22 Partner Emotional Support - .02 .21 AFE X Partner Em. Supp. .09 .56 Part Two: Predicting Adolescent Positive Parenting Attitudes Adaptive Family Environment (AFE) .12 1.07 T3 Self-Esteem .11 .9 T3 Network Social Support - .4 - 1.19 T3 Hassle Frequency - .96 - 2.04 Hassle Frequency X Network Social Support 1.31 2.01 823 69 Table 9 iv I vi ent Predi in Adolescent Parenti Outcome E PPA ' Predictor Variable N AR2 F L Part One: Predicting Self-Esteem Adaptive Family Environment (AFE) 82 .01 .87 .35 Partner Emotional Support 82 .01 .44 .65 AFE X Partner Em. Supp. 82 .02 .40 .76 Part Two: Predicting Adolescent Positive Parenting Attitudes Adaptive Family Environment (AFE) 82 .04 3.12 .08 T3 Self-Esteem 82 .05 1.94 .15 T3 Network Social Support 82 .09 2.54 .06 T3 Hassle Frequency 82 .09 1.91 .12 Network Soc. Supp. X Hassle Freq. 82 .14 2.40 .05 70 Table 10 ' 'v Statisti r e ndent and lnde ndent Variables Mean SD Min Max N W Maladaptive Family Environment 0.00 1.00 -2.21 2.63 88 Adaptive Family Environment 0.00 1.00 -3.21 2.26 88 Partner Emotional Support 2.45 2.21 0.00 5.00 134 T3 Self-esteem 3.07 .45 1.90 4.00 125 T3 Network Social Support 2.84 .61 1.83 4.78 125 T3 Hassle Frequency 13.85 4.91 1.00 24.00 125 mm T3 Conect Knowledge of Infant Development 64% 17% 17% 92% 125 T2 Adolescent Positive Parenting Attitudes 78.68 8.74 55.00 102.00 112 FIGURES 71 .282 3:885. H _ 23mm 808:835— macaw 950,—. ton—96 Boom c.3382 tong—am acetoEm .85.—um Ema: £5 \ sea 72 2255 325 e .838 seam cap as e82 .3635 e 8%5 a 2:5 73 roan—am Boom c1950 Z mecca—E 3 Eon—33.34 9:3... Eco: 20883.82 H m .oSmE 852.33 .m> .838 .md .8525 388m a 58m $.20 u 5.20 6502 .8 2% 5.85 n GE 8355‘ mama—93m 92 a 8m u 38% 0mm Hoe—.835 bian— 950,—. €85 GE 59 2. gmrxom a. ton—3m >083on 23.5 2. 8.3: £5 \ seem Bouoam 855 fl 74 toqqsm Boom _ commuocmod ” v 2.55 2.5m mama—28 cam—Scents ”5533 9.3:an $5.838 .23 $5338: ”38$ .aoaaoicm 38am 96302 EOE EU— ma. $5683 2 80583..“2 3:8... Eooummflaom 2. mmmumh c.3332 toqasm mason—68m 23.5 mg. Esteem 658m ma. 87.8w Eda \ 805m 75 N 863.33— ” n 23$ 83:2 3.83.— 938; ”mcucocam 8 .5§m=_v< 980,—. mmmfih toaasm 38m c.3382 .6532..— oamum 2. 3:3: £5 \ 30% guru—om mp. or»: 9.3:qu Estate—Bu ”6323 $55.3.— mcfiobcoo .23 $53385 cocoa..— EeauohE—m £89m 03332 toga—am .8885 655 fl 76 m gag—mom ” 0 239m 562 EU— mu. msuaoam 8 Eon—3&3. ”.80.”. mmmmmh E896 38m 3.8332 Egagcm 03% E. 87.3: .55 \ 305m gfimraow B. 2.? mew—.22. 033E653 $5682. 25% 8.63% 9:32. 32 .3638.“ mew—5.8a 928a uouoamucoaaoagm bag...” 3525 toga—am . .2285 852 fl v 3.323% n 5 0.59m 3.9.. 9:823 gag—653 $52.28 Scan 395:? 25%? Be. £333.“ mac—.22. 358a 83:3 wannabe 328m ”mat—BB.— 8 «55:33. Macon. Sgfimrhom a. “covamucofiaocmgm amaum 02:25 77 mam—2. toga—am 38m .3502 tongsm fiaouoEm— .65—cm 2. 55:00.5 0.3:: mg. 8.3:: 339 \ macaw REFERENCES Abernathy, V., Robbins, D., Abernathy, G.L., Grechbaum, H., & Weiss, IL. (1975). Identification of women at risk for unwanted pregnancy. WM thiLtLyJ32, 1027-1031. Andrews, G. (1978). Life event stress, social support, coping style, and risk of psychological impairment. Journal of Nervous Mental Dimrdgrs, 166, 307-316. Azar, S.T., Robinson, D.R., Hekimian, E., & Twentyman, CT. (1984). Unrealistic expectations and problem-solving ability in maltreating and comparison mothers. 1911M of consulting and Clinical Psychology, 52, 687-691. Babikian, H.M., & Goldman, AA (1971). A study of teenage pregnancy. Amm'can lgurnfl ngsychiatg, 128, 755-760. Bacon, L. (1974). Early motherhood, accelerated role transition and social Pathologiw- W 33-41. Barnett, J.K., Papini, DR, & Gbur, E. (1991). Familial correlates of sexually active pregnant and nonpregnant adolescents. Adolmgg, 26, 457-472. Barrera, M. (1981). Social support in the adjustment of pregnant adolescents: Assessment issues. In B.H. Gottlieb (Ed), Social networks and m'gl mm; (pp. 69- 96). Beverly Hills: Sage Publications. Barrera, M., Sandler, I.N., & Ramsay, TE. (1981). Preliminary development of a scale of social support: Studies on college students. American 10ml of Qommm' P ch 1 9 435-447. Barth, RP., Schinke, S.P., & Maxwell, 1.8. (1983). Psychological correlates of teenage motherhood. Journal of Youth and Adolescence, 12, 471-487. Bell, SM. (1970). The development of the concept of object as related to infant- mother attachment. Child Development, 41, 291-311. 78 79 Belsky, J. (1984). The determinants of parenting: A process model. Child Development, 55, 83-96. Belsky, J. (1993). Etiology of child maltreatment: A developmental-ecological analysis. Psychological Bulletin. 114. 413-434. Bogat, GA, Caldwell, R., Davidson, W., Guzman, B., & Galasso, L. (unpublished manuscript). Maternal support among pregpant and parenp'pg flplmms, Bowlby, J. (1969). Attachment and Loss. New York: Basic Books. Broussard, E. R, & Hartner, M. (1971). Further considerations regarding matemal perceptions of the firstborn In J Hellmuth (Ed ) BMW M (p. 231-256). New York: Brunner/Mazel. Brown, G.W., & Harris, T. (1980). Causes and logic in epidemiological psychiatry: IV. Further comments on the vulnerability model. 11W Psychigpy, 137, 584-585. Brown, G.W., Bholchain, M., & Harris, T. (1975). Social class and psychiatric disturbance in urban populations. Sociology, 9, 225-254. BrownH: Adams T &Kellan,S (1981) Wm Greenwich,CT: JAI Press. Buri, J .R (1989). Self-esteem and appraisals of parental behavior. W WA. 33-49. Buri, J .R (1992). Stability of parental nurturance as a salient predictor of self- esteem. Psychplpgical Reports, 71, 535-543. Camp, B. W. ,Holman, S. ,& Ridgway, E. (1993). The relationship between social support and stress in adolescent mothers. Developmental m Beflvipgal Pfi'mgs, 14, 369-3 74. Caplan, G. (1974). Support ystems and communig mm hang. New York: Behavioral Publications. Cassell, J .C. (1976). The contribution of the social environment to host resistance. Amerim Journal of Epidemiology, 104, 107-123. 8O Chamberlain, K., & Zika, S. (1990). The minor events approach to stress: Support for the use of daily hassles. British J ournal of Psychology, 81, 469-481. Chen, S., Telleen, S., & Chen, E. (1995). Family and community support of urban pregnant students: Support person, function, and parity. 19pm pf Commm'ty Psychplpgy, 23, 28-33. Cicchetti, D., Toth, S.L., & Lynch, M. (1995). Bowlby’s dream comes full circle: The application of attachment theory to risk and psychopathology. Advances in Clinig hildP ch lo 17 1-75. Cicchetti, D., & Garmezy, N. (1993). Prospects and promises in the study of resilience. Special Issue: Milestones in the development of resilience. W P h 010 5 497-502. Cobb, S. (1976). Social support as a moderator of life stress. mm Mill, 300-314. Cohen, S., & McKay, G. (1984). Social support, stress, and the bufl‘ering hypothesis: A theoretical analysis. In A. Baum, J .E. Singer, & S.E. Taylor (Eds), k f h 1 health (pp. 253-267). Hillsdale, NJ: Erlbaum. Cohen, S., & Wills, TA. (1985). Stress, social support, and the buffering hypothesis. Psychplogy Bulletip, 98, (2), 310-357. Colletta, N.D., & Gregg, CH. (1981). Adolescent mothers’ vulnerability to stress. 19pm pf Nervous and Mental Dime, 169, 50-54. Collins, N.L., Dunkel-Schetter, C., Loel, M., & Scrimshaw, S.C.M. (1993). Social support in pregnancy: Psychosocial correlates of birth outcomes and postpartum depression. Journal of Personalig and Social Psychology, 65, 1243-1258. Crase, S.J., & Stockdale, DP. (1989, November). W prsgmt adolescents. Paper presented at the National Council on Family Relations Annual Conference, New Orlean, LA Crnic, K.A., & Greenberg, MT. (1990). Minor parenting stresses with young children. Child Dflelopmmt, 61, 1628-1637. Crnic, K.A, Greenberg, M.T., Robinson, N.M., & Ragozin, AS. (1984). Maternal stress and social support: Effects of the mother-infant relationship from birth to eighteen months. Amgjgg Journal of Orthppgchiatgz, 54, 224-235. 81 Cutrona, CE. (1984). Social support and stress in the transition to parenthood. Journal pf Abnormal Psychology, 93, (4), 378-390. D’Arcy, C., & Siddique, CM. (1984). Social support and mental health among mothers of preschool and school age children. Social Psychiatry, 19, 155-162. Davis, AA, & Rhodes, J .E. (1994). Afiican-American teenage mothers and their mothers: An analysis of supportive and problematic interactions. mm P c 01 22 12-20. - Dean, A, & Lin, N. (1977). The stress-buffering role of social support. loamy! pr§r_v_st m4 Menfl Disgse, 165, 403-417. de Anda, D., Darroch, B, Davidson, M., Gilly, J., Javidi, M., Jefl‘ord, S., Komorowski, R, & Morejon-Schrobsdorf, A (1992). Stress and coping among pregnant adolescents. 19211.31, of Adolescent Research, 7, 94-109. DeLissovoy, V. (1975). Child care by adolescent parents. Children Tm 2, 22- 25. Dohrenwend, B.S., & Dohrenwend, BR (1974). W W New York: Wiley- Dohrenwend, B.S., & Shrout, PE. (1985). Hassles in the conceptualization and measurement of life stress variables. American Psyphplpgist, 40, 780-785. Dohrenwend, B.S., Dohrenwend, B.P., Dodson, M., & Shrout, PE. (1984). Symptoms, hassles, social supports, and life events: Problem of confounded measures. f P h 10 93 222-230. Drummond, RJ. (1991). Dimensions of self-concept of pregnant unwed teens. qumal prsycholpgy, 125, 65-69. Drummond, RJ., & Hansford, S.G. (1991). Dimensions of self-concept of pregnant unwed teens. Joumsl of Psychology, 125, 65-69. Dunst, C.J., Vance, S.D., & Cooper, CS. (1986). A social systems perspective of adolescent pregnancy: Determinants of parent and parent-cth behavior. 3mm fle_al_th_10_um_§l.l. 34-48- East, P.L., Matthews, K.L., & Felice, ME. (1994). Qualities of adolescent mothers’ parenting. Joprnal of Adolescent Health 15, 163-168. 82 Egeland, B. (1980). Preliminary results of a prospective study of the antecedents of child abuse: An American sample. In C.H. Kempe, AW. Franklin, & C. Cooper (Eds), The abused child in the family and in the community. New York: Pergamon Press. Egeland, B., Jacobvitz, D., & Sroufe, LA. (1988). Breaking the cycle of abuse. Child Devslopment, 59, 1080-1088. Egeland, B., & Sroufe, LA. (1981). Attachment and early maltreatment. Child mvslppment, 52, 44-52. Epstein, AS. (1980). New insights into problems of adolescent parenthood. Bpllsm pf High Scope Educational Research No. 5. Field, T.M., Widmayer, S.M., Stringer, S., & Ignatofl‘, E. (1980). Teenage lower- class, black mothers, and their preterm infants: An intervention and development follow- up. Child myelppment, 51, 426-436. Frank. 8. Jacobson. & Hole. (1986) WWW Unpublished manucsript, Michigan State University, East Lansing, MI. Fulton, AM., Murphy, KR, & Anderson, S.L. (1991). Increasing adolescent mothers’ knowledge of child development: An intervention program. A l 26 73-81. Gamer, E., Gallant, D., & Grunebaum, H. (1976). Children of psychotic mothers: An evaluation of l-year-olds on a test of object permanence. Mum Psyghiatfl, 33, 311-317. Gelfand, D.M., Teti, D.M., & Radin-Fox, CE. (1992). Sources of parenting stress for depressed and nondepressed mothers of infants. J gm of Clinipd Child P ch 1 21 262-272. Goldberg, D. (1972). The detection of pgchiatric illness by gusstipm’ e. London: Oxford University Press. Gore, S. (1978). The efi‘ect of social support in moderating the health consequences of unemployment. Journal pf Hflth and Socig Belgvipr, 19, 157-165. Gottlieb, BR (1981). Social networks ahd gzcial sppppg, Beverly Hills, CA: Sage. Held, L. (1981). Self-esteem and social network of the young pregnant teenager. A l 16 905-912. 83 Henderson, S. (1980). A development in social psychiatry: The systemic study of social bonds. Journhl of Nervous and Mental DiseaseJ68. 63-69. Hofinan, M.A, Ushpiz, V., & Levy-Shifl‘, R. (1988). Social support and self- esteem in adolescence. Journhl of Youth aid Adolescence. 17. 307-316. Hogan, D.P., & Kitagawa, EM. (1985). The impact of social status, family structure, and neighborhood on the fertility of Black adolescents. American Jpg_rn__al pf Sg’plogy, 90, 825-855. House, J .S. (1981). Work, stress and social support. Reading, MA: Addison- Wesley. House, J .S., Landis, KR, & Umberson, D. (1988). Structures and processes of social support. Annpgl Reviews in Sociology, 14, 293-318. Johnson, BM. (1991). Systemic patterns of parenting as reported by adolescents: Developmental difi‘erences and implications for psychosocial outcomes. W W 235-252. Kalish, RA, & Knudtson, F.W. (1976). Attachment versus disengagement: A life-span conceptualization. Humah Develppment, 19, 171-181. Kanner, AD., Coyne, J.C., Schaefer, C., & Lazarus, RS. (1981). Comparison of two modes of stress measurement: Daily hassles and uplifts versus major life events. f Behavi Medicine. 4. 1-39. Kaplan, B.H., Cassell, J.C., & Gore, S. (1977). Social support and health. W14. 47-58. Kawash, GR (1988). A factor analysis of a short form of the CRPBI: Are children’s perceptions of control and discipline multidimensional? Mpflsyphoipgy, 12;, 57-67. Kempe, C., & Helfer, RE. (1974). Helping me battergd shilg us his my“ , Philadelphia: Lippincott. Klerman, L. (1979). Design problems in evaluating service programs for school- age parents. Evfistipn and the Health Professiohgs, 2, 55-70. Klerman, L. (1993). Adolescent pregnancy and parenting: Controversies of the past and lessons for the future. Journal of Adolescent Healm, 14, 553-561. 84 Koniak-Griflin, D. (1988). The relationship between social support, self-esteem, and maternal-fetal attachment in adolescents. Resegrch in Nursing and Health 11, 269- 278. Lamb, M.E., & Elster, AB. (1990). Adolescent parenthood. In G.H. Brody & I.E. Sigal (Eds), Methods of family research. Hillsdale, NJ: Lawrence Elbaum Associates. Lazarus, RS. (1984). Puzzles in the study of daily hassles. mm mm; 375-389. Litovsky, V.G. (1985). Perceptions of child rearing and self-concept development during the early adolescent years. Journal of Youth and Adplssgmcs, 14, 373-3 87. MacLanahan, S. (1988). Intergenerational consequences of family disruption. Ammm JddmsJ pf Sociology, 94, 130-152. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. Monpgpsphs for the; Sfl'gy fdr Rgmch in Child Development, 50, 66-104. Man'ni. MM- (1978). Commences of childbgn'ngmghildsnamsmttm £01m Seattle: Battelle Human Affairs Research Centers. McAnarney, ER, Lawrence, RA, & Aten, M]. (1979). Premature parenthood: A preliminary report of adolescent mother-child interaction. W 328- 338. McAnarney, E. R, & Theide, H. A. (1981). Adolescent pregnancy and child rearing: What we learned m a decade and what remains to be learned. mg mm; 91- 103 McCormick, CB. (1994). Parent-child attachment and self-esteem in adolescence. Jm of Youth and Adolescence, 23, 1-18. McLaughlin, S.D.E., & Micklin, M. (1983). The timing of the first birth and changes in personal eficacy. Journal of Marriage and M Family, 45, 47-55. McPhee, D. (1981). Manual: Knowledge of Infant Develppmgh Invdrhpgy. Unpublished manuscript, University of North Carolina, Chapel Hill, North Carolina. Mercer, RT. (1977). Nursing care for patients at risk. Thorofare, NJ: Slack. 85 Miller, SA (1988). Parents’ beliefs about children’s cognitive development. Child Development, 59, 259-285. Miller, B.C., McCoy, J K Olson, T.D., & Wallace, D. (1986). Dating age and stage as correlates of adolescent sexual attitudes and behavior. Journal of Adolescent march; 361-371. Moos, RH., & Mitchell, RE. (1982). Social network resources and adaptation: A conceptual framework. In T.A Wills (Ed), Basic processes in helping relstiphships (pp. 213-232). New York: Academic Press. Morris, NM. (1981). The biological advantages and disadvantages of teenage pregnancy. Amm'm dernal pruhlic Health, 14, 412-427. Mueller, D. P. (1980). Social networks: A promising direction for research on the relationships of the social environment to psychiatric disorder. W 1_,4 147-161. Norbeck, J.S., & Anderson, NJ. (1989). Life stress, social support, and anxiety in mid- and late- pregnancy among low income women. R h in N ' H 12 281-287. Norbeck, J .S., Lindsey, AM., & Carrieri, V.L. (1981). The development of an instrument to measure social support. Nursing Research 30, 264-269. Norbeck, J .S., & Tilden, V. (1983). Life stress, social support, and emotional disequilibrium in complications of pregnancy: A prospective, multivariate study. 193mg! WM. 30-46- O’Hara, M.W. (1983). Postpartum depression: A role for social network and life stress variables. Mof Nervous and Mental Disease 171 336-341. Olson, C.F., & Worobey, J. (1984). Perceived mother-daughter relations in a pregnant and non-pregnant adolescent sample. Adolescence 19 781-794. Passino, AW., Whitman, T.L., Borkowski, J.G., Schellenbach, C.J., Maxwell, S.E., Keogh, D., & Rellinger, E. (1993). Personal adjustment during pregnancy and adolescent parenting. Adolescence 28 97-122. Paterson, J. (1995). Adolescent attachment to parents and fi'iends in relation to aspects of self-esteem. Journhl of Youth dn_d Adolescence, 24, 365-376. 86 Patten, M. (1981). Self-concept and self-esteem: Factors in adolescent pregnancy. Adolescence. 16. 765-778. Paykel, E.S., Emms, E.M., Fletcher, J ., & Rassaby, ES. (1980). Life events and social support in pubertal depression. British Joum_al of Psychology, 136, 339-346. Peterson, G.W., Rollins, B.C., & Thomas, D.L. (1985). Parent and adolescent conformity: Compliance and internalization. Youth ahd Socim, 16, 397-420. Ragozin, AS., Basham, RB., Crnic, K.A, Greenberg, M.T., & Robinson, NM. (1982). Efl‘ects of maternal age on parenting role. MW 627- 634. . Reis, J. (1993). Black and white adolescent mothers’ child-rearing beliefs and behaviors. Infght Mental Health Journal, 14, 221-233. Reis, J., & Hertz, E.J. (1987). Correlates of adolescent parenting. Adplesm 22, 599-609. Renick, M. (1989, April). Perceived social support and its relgp'ghdhip tp glphgi Wbles womally-gchieving students. Paper presented at the biennial meeting of the Society for Research in Child Development, Kansas City, MO. Ricks, M.H. (1985). The social transmission of parental behavior: Attachment across generations. Mpnpggaphs pf the Socisty for Rmh in Chd' d Dsyelppmsm, 59, 21 1-227. Rogosch, F.A, Mowbray, C.T., & Bogat, GA (1992). Determinants of parenting attitudes in mothers with severe psychopathology. mm P h l 4 469-487. Roosa, M.W. (1983). A comparative study of pregnant teenagers’ parenting attitudes and knowledge of sexuality and child development. M A 1e 12 213-224. Roosa, M.W., Fitzgerald, H..,E & Carlson, NA (1982). Teenage and older mothers and their infants: A descriptive comparison. A l 17 1-17. Rosenberg, M. (1965). Socie_ty and the adolescent self-m s. Princeton, NJ: Princeton University Press. Samuels, V.J., Stockdale, D.F., & Crase, SJ. (1994). Adolescent mothers’ adjustment to parenting. qurnal of Adplescence, 17, 427-443. 87 Schafer, C., Coyne, J .C., & Lazarus, RS. (1981). The health-related firnctions of social support. Journal of Behavioral Medicine. 4. 381-406. Schaefer, ES. (1965). Children’s Report of Parental Behavior Inventory. Scharness, S. (1993). The search for love: Unmarried adolescent mothers’ views of, and relationships with, men. Adolescence. 28. 425-437. Schilrnoeller, G.L., Baranowski, M.D., & Higgins, BS. (1991). Long-term support and personal adjustment of adolescent and older mothers. Adolesceng, 26, 787- 797. Schonert-Reichl, K, & Offer, D. (unpublished manuscript). ngdpghelpspd sppifl mpppst in gdplssc_e_n_ce: The role of non-related Ms. Institute for Urban and Minority Education, Columbia University. Sroufe, L.A., Pianta, RC., & Egeland, B. (1988). Continuity and discontinuity in maternal sensitivity at 6, 24, and 42 months in a high-risk sample. Who, 481-487. Stemp, P. S, Turner, RJ., & Noh, S. (1986). Psychological distress 1n the postpartum period The significance of social support Weaning Family, 48, 271-277 Stoiber, K.C., & Houghton, T.G. (1993). The relationship of adolescent mothers’ expectations, knowledge, and beliefs to their young children’s coping behavior. _In_faht, magmas. 61-79. Stokes, J. P. & Wilson, D. G. (1984). The Inventory of Socially Supportive Behaviors: Dimensionality, prediction, and gender difl'erences.Am W W 53-69 Streetrnan, LG. (1987). Contrasts in the self-esteem of unwed teenage mothers. W22. (86), 459-464- Thoits, P. A (1982). Conceptual, methodological, and theoretical problems rn studying social snppert as a buffer against life stress mm W 145- 159 Thompson, MS. (1986). The influence of supportive relations on the psychological well-being of teenage mothers. Social Forgg, 64, 1006-1024. 88 Thompson, M., & Pebbles-Wilkins, W. (1992). The impact of formal, informal, and societal support networks on the psychological well-being of Black adolescent mothers. Social Work, 37, (4), 322-328. Tilden, VP. (1983). The relation of life stress and social support to emotional disequilibrium during pregnancy. Research in Nursing and Health 6, 167-174. Townsend, J.K., & Worobey, J. (1987). Mother and daughter perceptions of their relationship: The influence of adolescent pregnancy status. Adolesdsng, 22, 487-496. Turner, R]. (1981). Social support as a contingency in psychological well-being. f H h d SocrflBehsvior. 2_2, 357-367. Turner, RJ., Grindstafl‘, C.F., & Phillips, N. (1990). Social support and outcome in teenage pregnancy. Journal of Health and Social Behavior, 31, 43-57. Unger, D. G., & Wandersman, L. P. (1985). Social support and adolescent mothers: Action research contributions to theory and application. W Ismfl, (1) 29-45 Vaz, R, Smolen, P., & Miller, C. (1983). Adolescent pregnancy: Involvement of the male partner. Jmml pf Adolescent Health Cge, 4, 246-250. Ventura, 8.]. (1980). Trends and differentials in births to unmarried women: U.S., 1970-1976. 21191 Hsslth Statistics, Series 21, No, 36. Wandersman, L., Wandersman, A, & Kahn, S. (1980). Social support in the transition to parenthood. Jdrhml pf Communigy Pachplpgy, 8, 332-342. Weiss, R (1974). The provision of social relationships. In A Rubin (Ed), an' g M (pp. 17-26). New York: Prentice-Hall. Williams, T.M., Joy, L.A, Travis, L., Gotowiec, A, Blum-Steele, M., Aiken, L.S., Painter, S.L., & Davidson, SM. (1987). Transition to motherhood: A longitudinal study. MW; 251-265- "71111111llll'lllllll“