Close
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The effects of abortion on psychological distress and intimate relationships in a community sample
(USC Thesis Other)
The effects of abortion on psychological distress and intimate relationships in a community sample
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6* x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. Bell & Howell Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. THE EFFECTS OF ABORTION ON PSYCHOLOGICAL DISTRESS AND INTIM ATE RELATIONSHIPS IN A COMMUNITY SAMPLE BY SHARON HELENE O’ NEIL A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVERSITY OF SOUTHERN CALIFORNIA IN PARTIAL FULFILLMENT O FTHE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN EDUCATION-COUNSELING PSYCHOLOGY DECEMBER 1999 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number 9987588 U M I* UMI Microform9987588 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 90007 This dissertation, written by Q!.ISJ.& ( under the direction of h.£.C..... Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY graduate Studies ■ D a te O .c t ol> e r.. .1 " 1 ., ... 19.9 9, DISSERTATION COMMITTEE Chairperson Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS LIST OF TABLES AND FIGURES 11! A BSTR A C T IV CHAPTER I. IN TR O D U C TIO N Importance of the Study History of Abortion Research Stress and Coping Theoretical Perspective Role of Social Support in Abortion Response Psychological Effects of abortion The Effects of Abortion on Intimate Relationships Social Conformity and Abortion Multicultural Research on Abortion Purpose of the Study Research Questions and Hypotheses Assumptions, Delimitations, and Limitations of the Study II. M E T H O D ......................................................................... Design Subjects Attrition Analyses Measures Analyses Nonlinear Effects Correlations Between Number of Abortions, Number of Children, and the 19 Other Variables Confirmatory Factor Analysis Path Models Specific or Nonstandard Effects Multiple-group Analyses IV . D ISCUSSIO N................................................................................................. 52 Discussion Conclusion Recommendations for Future Research III. RESULTS 35 R EFERENCES 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OFTABLES AND FIGURES TABLES Description of the Sam ple................................................................................................ 24 Reliabilities and Sources for the Measures...................................................................... 32 Correlations between Number of Abortions, Number of Children, and the Other 19 Variables..............................................................................................40 Summary of Fit Indexes and Nested Tests.......................................................................43 Intercorrelations Among the Latent Factors from the Initial CFA Model..........................44 Specific or Nonstandard Effects...................................................................................... 49 FIGURES CFA Model Depicting First Order Latent Constructs.......................................................42 Final Path M odel............................................................................................................... 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT In 1990, An American Psychological Association panel reviewed the literature on abortion and reported that severe effects following abortion are unlikely (Adler, David, Major, Roth, Russo, & Wyatt, 1990). Others (e.g., Speckhard & Rue, 1992) propose the existence of a “Post Abortion Syndrome” similar to Post Traumatic Stress Disorder. Neither side in the debate has offered substantial evidence for their position because both sides have based their conclusions on significantly Hawed research. This study examines abortion’s impact on psychological functioning and intimate relationships while addressing some of the sampling and measurement criticisms of prior research (i.e., volunteer bias, subject attrition, and limited follow-up of subjects). Structural Equation Modeling (Bender, 1995) is used to examine both predictors and consequences of abortion. A stress and coping framework is utilized and social support is considered as a moderator in the abortion response. In this study, abortion research volunteer bias is resolved by utilizing a community sample of women enlisted for longitudinal research on personal development and growth, not for research specific to abortion. Extensive attrition analyses are performed in this study to evaluate the possibility that those most affected by abortion may have dropped out of the study. Additionally, the measurement time frame is extended to four years. The results of this study indicate that abortion is not a public mental health concern. No psychological or relational effects for abortion were found on the variables examined. Social Conformity, with the variable of Religiosity in particular, predicted abortion occurrence. Number of Abortions predicted Number of Children. There was some suggestion, though not statistically significant, that abortion might be related to lower agitation. Further research would need to examine this possibility. Future research might also examine how abortion interacts with pre-existing psychiatric conditions. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 CHAPTER I INTRODUCTION The continuing debate over the psychological impact of abortion indicates abortion is an important area of psychological research. The two sides of the psychological debate closely parallel the two sides of the ongoing political debate. One side appears to downplay any psychological effects and the other side claims abortion results in dramatic effects. In 1990, an American Psychological Association (APA) panel reported that severe effects following abortion are unlikely (Adler, David, Major, Roth, Russo, & Wyatt, 1992). Others (e.g., Speckhard & Rue, 1992) propose the existence of a “Post Abortion Syndrome” similar to Post Traumatic Stress Disorder. Neither side in the debate h a s offered substantial evidence for their position because both sides have b ased their conclusions on significantly Hawed research. The purpose of this study is to examine abortion’s impact on psychological functioning and intimate relationships while addressing some of the methodological problems in prior research. Speckhard and Rue (1992) argue that definitive conclusions about the psychological sequelae of abortion cannot be made due to the methodological Haws in the research. These Haws relate to sampling and measurement issues. For example, those who volunteer for abortion research may have more positive attitudes about abortion. In the present study, abortion research volunteer bias will be resolved by utilizing a community sample of women enlisted for research on personal development and growth, not for research specific to abortion. It can also be argued that the women who are most affected by abortion might decline to continue in abortion studies, thereby eliminating any negative findings. For example, Adler (1976) found that women for whom abortion was more stressful were less likely to be represented at follow-up. Attrition analyses will be performed to determine whether or not subjects in this study selectively dropped out based on the variable of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. abortion. Another important criticism of abortion research is the limited long term follow- up of subjects. Generally, the psychological effects of abortion have been measured in the short term, ranging from one hour to about two years. Little is known about possible long term effects. This problem will be partially corrected by extending the time frame to four years. The present studv will revisit the issue of abortion and address some of the methodological problems from previous research. Both predictors and consequences of abortion will be considered. Utilizing a stress and coping framework, social support will be examined as a potential moderator in the adjustment to abortion. In addition to the controversy surrounding abortion, the abortion rates suggest this to b e an important area of study. On a worldwide basis, it has been estimated that 30 to 40 million induced abortions are performed legally each year, with a roughly equivalent number of illegal abortions performed (Masters, Johnson, & Kolodny, 1992). Half of the 6.4 million pregnancies occurring in the United States each year are unintended, and close to half of these end in induced abortion (Forrest & Henshaw, 1993). Elective abortion is the most common surgical procedure performed in the United States (Speckhard & Rue, 1992). Almost one half of American women have had at least one abortion (Forrest & Henshaw, 1993). U.S. abortion patients are comprised of all segments of the population. In 1987, almost 60% of abortion patients were under 25 years of age (Adler et al„ 1990). Most abortion patients (82%) were not married, and half had no prior births (Adler et al., 1990). 1994-95 data suggests African-Americans; Latinas; women aged 18-24; women with low incomes; and women who live with a partner outside of marriage, have never married, or are separated, have a higher percentage of abortion (Henshaw & Kost, 1996). Henshaw and Silverman (1988) report that in 1987,42 percent of women who had legal abortions were Protestant, 32 percent were Catholic, 1 percent were Jewish, 3 percent were of other Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 religions, and 22 percent reported they had no religious affiliation (cited in Masters et al., 1992). Prior to the legalization of abortion in the United States, permission for an abortion often required psychiatric determination of individual psychopathology (Speckhard & Rue, 1992). Early research tended to focus on psychiatric or medical complications of abortion, as some conception of disease was often required for abortion to be performed legally (Dagg, 1991). From 1935 to 1964, research conclusions confirmed expectations derived from Freudian psychological theory, which hypothesized that a denial of pregnancy would be unnatural for a woman, resulting in serious emotional trauma (Simon & Senturia, 1966). More recently, post legalization research may also have been influenced by psychoanalytic or sociocultural views about pregnancy and women. For example, a Menninger clinic study (i.e., Cavenar, Maltbie, & Sullivan, 1978), which was based on only two subjects, reported psychological distress occurring due to post abortion pregnancy fantasies based on Freudian concepts. The research from 1934 to 1964 on the psychological effects of abortion is replete with methodological problems including sampling bias, failing to make a distinction between spontaneous, illegal or therapeutic abortion, often using subjects who had criminal abortions, not using pretest measures of psychological functioning, and having an absence of uniformity in defining and evaluating emotional states (Simon & Senturia, 1966; Turell, Armsworth, & Gaa, 1990). In 1990, an A PA panel reviewed the research conducted between 1972 and 1989 on the psychological effects of abortion. This panel reported that severe psychological responses after abortion are unlikely (Adler et al., 1992). Adler et al.’ s (1992) conclusion and a further assumption that there is no scientific basis for a “Post Abortion Syndrome” (i.e., severe psychopathology similar to Post Traumatic Stress Disorder) for the great majority of abortion recipients is supported in other reviews of the research (e.g., Osofsky & Osofsky, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 1972; Smith, 1973; Bracken, Phil, Hachamovitch, & Grossman, 1974; Romans-Clarkson, 1989; Dagg, 1991; David, 1994). However, the current research on abortion is not free from many of the significant flaws in the earlier research. As Adler et al. (1990) and David (1994) point out, the abortion research reviewed by them has methodological Haws including lack of controls, sampling inadequacies, high subject attrition rates, lack of baseline data on subjects, and the lack of long term follow-up. A related concern is that small case studies reporting dramatic effects of abortion (e.g., Cavcnar et al., 1978) are often cited without reporting sample size or other important factors. For example, one case study (i.e., McCraw, 1989) which reported severe obsessive-compulsive disorder following an abortion, has been cited without mention of the subject’ s significant prior psychiatric history . In the present study, abortion is conceptualized as a stressor, which may lead to responses ranging from positive personal growth to psychopathology. The theoretical perspective in this study stems from Lazarus and Folkman’ s (1984) model of stress and coping. According to Lazarus and Folkman (1984), stress involves an interaction of the person and the environment in situations that the person appraises as straining his or her resources. In this model, a person brings stable resources to a situation. These stable resources include social support and personality variables such as self-esteem and locus of control. These resources combine with aspects of the situation, leading to how an event is appraised. These appraisals underly the strategies used to cope with the event. In this study, the involvement of a stable resource, social support, will be considered in the abortion response. In their review of social support studies, Cohen and Wills (1985) found support for two models describing social support’s association with well-being. In one model, it is proposed that social resources have an overall beneficial effect whether or not a person is under stress. The second model proposes that social support is related to well-being Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 primarily for a person under stress. Social support then, is viewed by Cohen and Wills (1985) as having both general, overall benefits as well as specific “buffering” effects. In this study, the buffering model of social support will be utilized conceptually. Social support will be considered as a moderator in the abortion response, buffering one from negative effects. In this view, abortion is a stressor. If abortion leads to psychological or relational distress, then subjects high on social support will be lower on distress following abortion than those having low social support. Support from one’s parents and one’ s family, as well as support for emotional problems from a variety of sources will be examined. Data pertaining to the partner in conception was not available for this study. A stress and coping framework will be utilized in this study, with social support examined a s a potential moderator in the adjustment to abortion. The Role of Social Support in Abortion Response Previous research has considered the role social support plays in the response to abortion. Perceived social support amongst 62 women obtaining abortions in an urban southern area was studied by Moseley, Follingstad, Harley, and Heckel (1981). They found that an important determinant of w omen’s psychological response to abortion was the (perceived) amount of support from the partner in conception, parents, and friends (Moseley et al., 1981). Opposition from the partner in conception, parents, or friends against the decision to have an abortion resulted in statistically higher levels of anxiety, depression, and hostility (Moseley et al., 1981). Interestingly, support from any one of these three social support groups offset opposition from another (Moseley et al., 1981). Similarly, in examining perceived social support amongst 283 women. Major et al. (1990) found that women who perceived high support from their family, friends, and partners had higher expectations for self-efficacy in coping with the abortion. Higher self-efficacy expectations predicted better psychological adjustment to abortion (Major et al., 1990). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 Bracken, Phil, Hacamovitch, and Grossman (1974) examined the importance of level of support from significant others in the decision to abort among a sample of 489 women. Bracken et al. (1974) reported that perceived partner support was significantly more important in predicting a favorable response to abortion among older women whereas perceived parental support was a more powerful predictor of favorable abortion response in younger women. Robbins and DeLamater (1985) found that 25% of their sample of 228 abortion recipients reported feelings of loneliness one week after abortion. In their sample, those who had support before, during, and after the abortion from the male partner, experienced less feelings of loneliness (Robbins & DeLamater, 1985). However, both Major et al. (1985) and Cozzarelli, Karrasch, Sumer, and Major (1994) reported that male partner accompaniment to an abortion clinic, increased perceptions of social support but did not predict more positive adjustment to an abortion. Potential explanations given for this result included that accompaniment might signify higher pre-abortion levels of distress and anxiety and that women in committed relationships may find the abortion decision more difficult (Cozzarelli etal., 1994). An issue related to social support is pressure or coercion to have an abortion. (This issue seems particularly relevant to counselors, who might inadvertently pressure a woman to have or not have an abortion. For example, a counselor might strongly encourage an unmarried, unemployed teenager to have an abortion without thoroughly considering her feelings, options, and right to autonomy.) In a review of the literature published since 1965, Romans-Clarkson (1989) reported that the women most likely to have problems following abortion are those who are pressured into having an abortion, either by relatives or because their pregnancy had medical or fetal contraindications. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 Depression Following Abortion Depression is one of the most researched abortion response v ariables. The present study incorporates depression, suicidality, and powerlessness in a dysphoria construct. In reviewing the research, it appears that short term depression may follow abortion for many women. However, it also appears that abortion alone is not enough to account for the short term depression. Personality and contextual variables may predispose some women to experience distress following abortion. In this study, variables such as self-efficacy and self-acceptance will be included in the analyses. Freeman (1977) reported from a sample of 108, that women w ho stated they had wanted their pregnancies were less likely to experience relief and more likely to experience depression al ter an abortion. However, at a four-month follow-up, the difference was no longer statistically significant (Freeman, 1977). Similarly, in a study of 360 women who underwent abortion, Greer, Lai, Lewis, Belsey, and Beard (1976) found a significant improvement in degree of depression three months after abortion. Freeman (1977) found that women who had wanted to get pregnant were more likely to describe the fetus as a child and to feel that abortion meant giving up a child. These women had the highest rate of short-term depression, though the majority were no longer depressed at the four-month follow-up (Freeman, 1977). Similarly, in a sample of 145 women receiving chemically induced medical (i.e. mifepristone) abortions, wantedness of the child was the factor that had the greatest effect on stress and regret six to eight months al ter abortion (Miller, Pasta, & Dean, 1998). In an examination of the effects of abortion on self-esteem, positive and negative affect, satisfaction with life, and beliefs about the fetus, Conklin and O’ Connor (1995) administered questionnaires to 817 subjects in physician’ s waiting rooms. They found that women who have had an abortion but who believe that fetuses are human scored lower on well-being variables than women who have not had an abortion (Conklin & O’Connor, 1995). In contrast, women who have had an Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 abortion and who believe that fetuses are not human were adjusted as well as women who have not had an abortion (Conklin & O’ Connor, 1995). Cohen and Roth (1984) examined the relationship between two types of coping styles and immediate postabortion depression and anxiety (as measured by the depression and anxiety subscales of the Symptom Checklist-90) among 55 female subjects. The sample at the first testing, several hours prior to the abortion procedure, was slightly more anxious and slightly less depressed than psychiatric outpatients (Cohen & Roth, 1984). “Avoiders,” who tried to not talk or think about the abortion reported higher levels of depression and anxiety five hours postabortion than “Approachers," who talked about and expressed their feelings about the abortion and thought of ways to prevent a future unplanned pregnancy (Cohen & Roth, 1984). How the two groups might differ on the measures of depression and anxiety over time was not examined. Major, Mueller, and Hildebrandt (1985) examined attributions and coping expectancies in the post abortion adjustment of 247 women. They reported that women who blamed their pregnancy on their character coped less well with abortion than low self-character blamers and that women with high coping expectancies coped better than those with low coping expectancies (Major et al., 1985). Mueller and Major (1989) also studied the impact of attributional styles and coping expectations on 283 female subjects’ adjustment to abortion. They found that a poor attributional style (i.e. internal, stable, and global attributions made for negative events) combined with negative coping expectancies was related to higher levels of depression (Mueller & Major, 1989). Women who blamed their pregnancy on their character (a stable factor) rather than their behavior (a mutable factor) both expected to experience and experienced more negative consequences (i.e., depression) immediately following an abortion (Mueller & Major, 1989). In a small study (n=80) of grief response following first and second trimester elective abortions, Peppers (1987) used a scale measuring thirteen symptoms of grief: sadness, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 loss of appetite, irritability, sleep problems, difficulty concentrating, preoccupation of thoughts, repetitive dreams, depression, anger, guilt, problems returning to usual activities, exhaustion, and lack of strength. Peppers (1987) found that grief scores were high prior to the procedure, and significantly lower six weeks al ter the procedure, suggesting a relatively rapid resolution of the grief response. The wide variability in the scores suggests that some women have little difficulty with abortion, while some experience significant emotional effects (Peppers, 1987). An analysis of the findings indicated that the intensity of the grief response in this sample was dependent on the length of pregnancy (Peppers, 1987). In a study of personality variables involved in coping with abortion, Cozzarelli (1993) measured post abortion psychological distress (i.e., depression and mood) in a sample of 291 women. Pre-abortion levels of depression, as measured about one hour prior to the surgery, were somewhat higher than the means typically found in nondepressed populations (Cozzarelli, 1993). Level of depression approximately half an hour after the abortion procedure was closer to the mean for nondepressed populations (Cozzarelli, 1993). 3-week postabortion adjustment was better in those with high self-esteem, perceived control, and optimism through feelings of increased self-efficacy (Cozzarelli, 1993). Cozzarelli, Karrasch, Sumer, and Major (1994) explored the effects of partner accompaniment to an abortion clinic on immediate (i.e., 30 minute) and 3-week post abortion depression in Cozzarelli’s (1993) sample of 291 women. Depression was measured by Cozzarelli et al. (1994) using the 13 item Symptoms Checklist-90 Depression subscale (Derogatis, 1983). There were no overall differences in depression between the accompanied and unaccompanied women (Cozzarelli et al., 1994). Cozzarelli et al. (1994) reported that the average post abortion level of depression in their sample (M = .85) was closer to the mean for “normal” individuals (M = .36) than for psychiatric patients (M = 1.75) (Cozzarelli et al., 1994). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10 Overall, the research suggests short term depression may occur for some women following abortion, depending on various personality and contextual variables. Anxiety Anxiety levels are likely to be highest just prior to abortion, reflect the stress of an unwanted pregnancy, and decrease immediately following the procedure (Adler et al., 1990). A study by Rosen (1988) supported a transitory', situational nature of abortion related anxiety. Rosen (1988) studied state anxiety in 113 Swedish women prior to and following either surgical or medical abortion. She found a moderately elevated level of state anxiety immediately before the abortion, which then dropped to a normal level at the two week follow-up (Rosen, 1988). Bracken (1978) explored the psychological and procedure related origins of physical pain experienced during abortion and the relationship of physical pam with reported psychological reactions immediately following abortion. In his sample of 215 women, 85% found the abortion decision easy, felt fairly relaxed prior to the procedure, and felt calm following the abortion (Bracken, 1978). 15% of the women experienced a difficult abortion decision, were very anxious before and anxious and depressed immediately after the abortion, and experienced a very physically painful abortion (Bracken, 1978). Anxiety prior to the abortion was a function of the difficulty of making the abortion decision and married women found the decision more difficult than unmarried women (Bracken, 1978). Women reporting a less favorable psychological reaction immediately following the abortion, were also more likely to have experienced physical pain during the procedure (Bracken, 1978). The stability of these post abortion responses cannot be determined as assessments were not made beyond those immediately following abortion. There is some limited evidence that women who choose to terminate an unintended pregnancy may have better psychological functioning in some areas than those who choose other options. Zabin, Hirsch, Emerson, and Raymond (1989) studied 334 African- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11 American minors seeking pregnancy tests. Zabin et al. (1989) followed the three groups (i.e., not pregnant, carried to term, aborted) for two years. Educationally and economically, the abortion group was at an advantage over the childbearing group (Zabin et al., 1989). Also, alter two years, the abortion group was significantly lower on trait anxiety than the other two groups and had higher self-esteem and a more internal locus of control than the nonpregnant group (Zabin et al., 1989). In the current study, an agitation construct includes the variables of anxiety and hostility. Suicide Ideation Generally, it appears that suicide following elective abortion is rare and that when pre pregnancy mental health is controlled, abortion has no effect on completed suicide (Greer, Lai, Lewis, Belscy, and Beard (1976). In countries were legal abortion is available, suicide rates in pregnancy are lower than in non-pregnant women (Clare & Tyrell, 1994). Threats of suicide have been presented in legal cases following refused abortion in Ireland (Clare & Tyrell, 1994). Similarly, prior to the legalization of abortion in the United States, threats of suicide may have been used by some women to present psychiatric justification for permission for abortion. It appears from the research that abortion has no impact on suicide, except perhaps in cases of denied abortion. Suicide following elective abortion appears to be rare and not directly attributable to abortion. Six women out of the 360 studied by Greer et al. (1976) attempted suicide following abortion. However, each of these six had received outpatient psychiatric treatment prior to the termination of their pregnancies (Greer et al., 1976). In this study, the proportion of patients requiring outpatient treatment for psychiatric symptoms fell significantly from 29% before abortion to 19% following abortion (Greer et al., 1976). It appears that when pre-pregnancy mental health is controlled, abortion per se has no affect on completed suicide. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 Severe Mental Distress Following Abortion Major psychiatric disturbances following abortion usually occur in women with pre existing psychiatric problems, who are therefore predisposed to relapse (Zolese & Blacker, 1992). Furthermore, these rates are generally lower than those found in women following childbirth (Clare & Tyrell, 1994). In a prospective study in Britain, Brewer (1977) found rates of post abortion psychosis of 3 per 10,000 as compared with rates of psychosis following childbirth of 17 per 10,000. Conversely, David (1985) found evidence of increased psychiatric hospitalization within three months following abortion when compared with both postpartum rates and non- pregnant base rates. David (1985) studied the first time psychiatric admission rates of Danish women and found that the rates for women h av ing had abortions were consistently higher than the rates for women who delivered as well as for non-pregnant women. The presenting problems of the women in the three groups arc not defined. However, postpartum psychosis is generally defined as any mental disorder occurring within three months of delivery w hich is severe enough to require admission to a psychiatric facility (David, 1985). From this, one can surmise that women in each of the groups experienced a wide range of mental disorders including major affective disorders, brief reactive psychoses, posttraumatic stress, etc. Information on the comparative diagnoses and lengths of hospitalization was not reported. David (1985) reported a postabortion first psychiatric admission rate of 18.4 per 10,000 and a postdelivery rate of 12.0 per 10,000, compared with a rate of 7.5 per 10,000 for all Danish women aged 15 to 49. Important differences in the abortion group and the delivering group suggest that abortion per se may not be the cause of increased psychiatric admissions. Ninety-three percent of the women who delivered were married or in stable relationships, whereas only 18% of the women aborting were in stable relationships (David, 1985). Relationship status and the social support obtained from a partner are likely Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13 to be important in managing the possible stress following abortion. Further, women who were separated, widowed, or divorced were found to be especially at risk for psychiatric hospital admission following abortion (David, 1985). This finding suggests the possibility of higher psychiatric admission rates for women terminating pregnancies that may have been originally intended or wanted, as well as conceived in the context of an intact relationship (David, 1985). Most post legalization research indicates that abortion is less likely to cause psychiatric sequelae than childbirth (Freeman, 1977). However, the research suggests that psychological distress does occur following abortion. The distress (e.g., depression, regret), though above the level found in “normal” populations, docs not often reach levels found in clinical populations (Cozzarelli, 1993). Further, the affectivity (e.g., regret) reported by some women following abortion may have little impact on their overall psychological functioning (Robbins, 1979). For example, Robbins (1979) found no differences on MMPI scores across levels of regret and unwillingness to repeat an abortion procedure amongst a sample of 4 1 women. “Post Abortion Syndrome” Opinion papers and small case studies have reported posttraumatic stress occurring after abortion. Unfortunately, these reports are difficult to evaluate as they usually do not include pre-pregnancy baseline psychological data and other relevant information. Furthermore, it would clearly be improper to generalize from clinical case reports to the larger population of women obtaining abortions. However, this material is presented in order to present the full range of possible individual responses to abortion. Speckhard & Rue (1992) propose the existence of a “Post Abortion Syndrome” (PAS) diagnostically similar to posttraumatic stress disorder. They suggest that PAS occurs in a small percentage of women and is directly the result of abortion (Speckhard & Rue, 1992). While only providing anecdotal evidence of PAS, Speckhard & Rue (1992) argue that Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14 definitive conclusions about the psychological sequelae of abortion cannot be made due to the methodological flaws (e.g., limited long term follow-up of subjects, volunteer bias, and the high attrition rate of subjects who might be those most affected by abortion) in the research. Clinical case studies have reported posttraumatic responses following abortion but haven’ t fully discussed pre-existing social and psychological factors involved. For example, Erikson (1993) presented clinical case material on three women experiencing postabortion trauma conforming to the diagnostic criteria for posttraumatic stress disorder. Erikson (1993) theorizes that the existence of the developing fetus requires an adaptation in the woman’s internal psychological world and that destruction of the fetus assaults the stability of that internal structure, thereby causing traumatic conflict. Similarly, Butler (1996) describes the case of a 38 year old woman continuing to re-cxpcricnce the abortion procedure after 19 years. In this case study, it appears that the woman wanted the child and was coerced into an abortion by her parents (Butler, 1996). It appears that when pre abortion mental status and environmental variables are evaluated, abortion, per sc, no longer has a direct causal impact on severe psychological disorders, including the proposed “post abortion syndrome.” Intimate Relations The issue of abortion’s impact on intimate relations will be tested in this study in a variety of domains. Intercourse satisfaction, happiness within the relationship, dependence, and dating competence will be considered. Intercourse frequency or satisfaction and happiness within the relationship (as determined by separation) have been the most studied relational variables in abortion research. Barnett, Freudenberg, and Wille (1992) examined the social and psychological consequences of induced abortion on the relationship between women and their partners. Prior to abortion, the relationships in the aborting group were less harmonious and showed Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15 more conflicts than in the control group (Barnett et al., 1992). However, at the one year follow-up, the number of separations in the aborting group was not significantly higher than in the control group, nor were there any remaining qualitative differences in the relationships in either group (Barnett et al., 1992). Likewise, Robbins (1984) compared the relationships of 139 unmarried women who had abortions with those of 109 unmarried women who chose to have their babies. The women who had abortions, while more likely to be casually involved with their partners before becoming pregnant, were no more (or less) likely to end their relationships six weeks or one year al ter the abortion than were women who delivered!Robbins, 1979). Sexuality was another variable examined by Barnett et al. (1992). No significant differences between the aborting group and the control group were found as far as either coitus frequency or sexual satisfaction (Barnett et al., 1992). However, a reduction of libido lasting between 2 and 3 months was reported by 1 out of every 5 abortion patients (Barnett et al., 1992). Similar relationship adjustment results were reported by Greer et al. (1976). Among other variables, Greer et al. (1976) examined marital adjustment, other interpersonal relationship adjustment, and sexual adjustment following abortion in 360 women. They found that termination of pregnancy was not a significant factor influencing marital harmony three months post abortion (Greer et al., 1976). However, a significant improvement in other interpersonal relationships was found three months post abortion (Greer et al., 1976). In this study, sexual adjustment was defined in terms of the frequency when intercourse was enjoyable and the frequency of orgasm. Results indicated a significant improvement in sexual adjustment from (59% satisfactory) prior to abortion to both three months and two years after (74% satisfactory) abortion (Greer et al., 1976). Social Conformity The social conformity construct has been examined as a predictor of abortion occurrence in previous research and is included in the present study with the variables of religiosity, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 liberalism, and law abidance. Stein, Newcomb, & Bentler (1994) found that social conformity was not a significant predictor of abortion. However, they found a modest positive association between personality correlates of those who had abortions and those who engaged in drug use and risky AIDS behaviors (Stein et al., 1994). Licit drug use (i.e. cigarettes and alcohol) is examined in the present study. Conversely, Costa, Jessor. & Donovan (1987) found greater psychosocial unconventionality, greater political liberalism, and less religiosity to be correlated with women who have had an abortion. Religiosity is an aspect of social conformity which has been considered a s a factor in previous research on postabortion adjustment. Abortion rates indicate that similar percentages of Catholic women and Protestant women have had an abortion (Costa et al., 1987) and may have more negative postabortion responses (Russo & Dabul, 1997) than other women. For some women, practical necessity (e.g., financial concerns, relationship status) may conflict with personal, religious, or cultural values about abortion. Such a conflict might lead to higher post abortion distress. A factor that has been examined in post-abortion adjustment is the extent to which a woman has difficulty arriving at the decision to have an abortion (Major & Cozzarelli, 1992). Adler (1975) reported that women who found their abortion decision difficult were more likely to experience internally based negative emotions (i.e. regret, anxiety, depression, doubt, and anger) and were somewhat more likely to experience socially based negative emotions (i.e. guilt, shame, and fear of disapproval) after the abortion. Increased difficulty with the abortion decision has been associated with being married, being Catholic, the length of time a woman has been pregnant, negative attitudes toward abortion, and the lack of perceived social support from significant others (Major & Cozzarelli, 1992). The literature suggests a variety of predisposing factors relating to a positive or negative abortion experience. In the present study, religiosity, liberalism, and law abidance are included in the analyses. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17 Multicultural Research on Abortion Due to sparse multicultural research on abortion, it is difficult to hypothesize about differences in attitudes toward abortion and how they might relate to the psychological adjustment to an abortion. Presently, the few extant cross-cultural abortion studies only compare African-American and Caucasian women. Little is known about other groups, including those with high abortion rates such a s Latinas. The present study is ethnically diverse but has a modest representation of about 30% minorities. Abortion ratios indicate that African-American women have legal abortions at twice the rate of Caucasian women (Lynxwiler & Gay, 1994). Married African-American women are more likely to end a pregnancy through abortion than are married Caucasian women, a s are African-American women who have had a previous live birth, regardless of marital status (Dugger, 1998). Afriean-Amencans are more likely than Caucasians to view the inability to afford another child as a reason for a woman to seek an abortion (Hall & Ferree, 1986). Any restrictions on reproductive freedom most severely impact women with low incomes and women of color (Dugger, 1998). Prior to Roe v. Wade, women of color constituted 64% of deaths caused by illegal abortions, and since Roe v. Wade, 80% of all women who have died from self-induced abortions have been either African-American or Puerto Rican-American (Dugger, 1998). Using data from the General Social Survey (GSS) from 1972 to 1980, Combs and Welch (1982) found that the same demographic variables that are related to abortion support in the larger population-education, income, rural and southern residence, and religiosity-also predicted African-American attitudes (cited in Dugger, 1998). Hall and Ferree (1986) used the 1972 to 1984 GSS data to examine race differences (African- American and Caucasian only) in abortion attitudes. Controlling for 1 1 variables (e.g., SES, religious practice, and southern culture), they reported that African-American women Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18 were significantly more likely to have negative attitudes toward abortion than Caucasian women (Hall & Ferree, 1986). However, using the same GSS data but extending it from 1972 to 1988, other results were reported by Lynxwiler and Gay (1994). When dividing women by childbearing years, African-American and Caucasian women did not differ significantly in their abortion attitudes, although race differences appeared among older women and men (Lynxwiler & Gay, 1994). It appears then that racial differences in abortion attitudes can be accounted for by the greater opposition to abortion by African- American men and older African-American women (Dugger, 1998). Counselors’ beliefs or prejudices might also have an impact on a woman’s psychological adjustment to an abortion. Research has not directly addressed cross-cultural issues in pre- and post abortion counseling. However, Christensen (1988) points out a potential issue in counseling ethnic and minority women: stereotypes about the sexuality of some minority women (e.g., promiscuity, having children to get more welfare) may lead helpers to ignore, de-emphasi/.c, or exaggerate aspects of their sexuality. Consequently, those who are victims may be perceived as having invited the problem, or as not having been traumatized by certain experiences (Christensen, 1988). These reflections might relate to counseling women before or after an abortion. For example, a counselor might blame a woman for her pregnancy and respond to her with a lack of empathy or a lack of consideration of her individual response to the abortion. It is difficult to draw any conclusions about the differential impact of abortion on women from various cultures. Future multicultural research is needed to determine cultural variables involved in the abortion experience. Purpose of the Study As research may influence public policy, accurate information about the psychological and relational effects of abortion would be of great value. As previously stated, the purpose of this study is to respond to some of the methodological criticisms of previous Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19 abortion research and to determine the psychological and relational effects of abortion. Predictors and consequences of abortion will be examined. A stress and coping theoretical framework will be used and social support will be considered as a moderator in the abortion response. Research Questions 1 . Does abortion have an impact on psychological functioning and intimate relations? Psychological functioning will be examined in the areas of dysphoria, agitation, and agency. Intimate relations will be considered in the areas of happiness within a relationship, sexual satisfaction, dependence, and dating competence. 2. Do social conformity and licit drug use predict abortion occurrence? 3. Are there psychological and relational effects of abortion which are different due to high and low social support groups? For example, will those higher on social support have less negative psychological functioning and intimate relations following abortion than those lower on social support? Hypotheses Research is typically designed to reject a null hypothesis. An exception has been made in this study. This study is designed to accept, rather than reject, a null hypothesis regarding abortion and its effects. There have been mixed findings regarding the consequences of abortion. Much of this has been due to methodological problems. Due to the significant proportion of women having abortions and the potential influence of research on public policy, further investigation into the topic was warranted. The exception in research design appears justified because of the equivocal nature of previous abortion research, as well as important criticisms of abortion research methodology. For conceptual purposes, one could view the null hypotheses to be rejected in the affirmative. 1. Abortion is expected to have no effect on Dysphoria. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20 2. Abortion is expected to have no impact on Agitation. 3. Abortion is expected to have no effect on Agency. 4. Abortion is expected to have no effect on Intimate Relations. 5. Social Conformity is expected to be a significant predictor of abortion occurrence. 6. Licit Drug Use is not expected to be a significant predictor of abortion. 7. Subjects are not expected to differ on psychological or relational variables following abortion due to membership in the high or the low social support group. Assumptions of the Study 1 . The research design and data analysis procedures selected for this study were appropriate. 2. Subjects who participated in the study were not significantly different than those who chose not to participate. Delimitations of the Study 1 . The study is limited to subjects in Los Angeles County. 2. The study is limited to subjects who voluntarily agree to participate. Limitations of the Study 1 . The findings may not be generalizablc beyond Los Angeles County. 2. The sample is ethnically diverse yet has a modest representation of minorities, which may limit generalizability. 3. The findings are based on a community sample, most of whom had not sought treatment for psychological problems. The present results must be considered with caution when generalized to those who present for psychotherapy or other types of psychological help. 4. Differential experimental mortality of subjects may limit the findings. 5. It is possible that the constructs may reflect sample-specific covariances. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21 6. Validity is limited by the subjects’ honesty in the self-report measures. 7. Internal and external validity is limited to the instruments used in the study. 8. Under-reporting of abortion may occur in general and may have occurred in the sample. If this occurred, the variance may have been narrowed and the distribution changed, giving less statistical power to the study. 9. The responses to abortion of single versus married women were not examined. 10. The analyses were not limited to women who got pregnant. Women who had never been pregnant were included in the study. 11. A stress and coping theoretical framework was used for this study. However, many important aspects of this framework were not incorporated into this study. For example, data on cognitive appraisals specific to the abortion experience were not available. Further, information about the subjects’ reasons for abortion and data on social support specific to abortion were not available. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER II METHOD Design For the prospective analysis, the data was collected from 390 female subjects from the year 1984 and the year 1988. The analysis of social support as a moderator is cross- sectional and utilizes the 1984 data. The data used for all analyses was collected as part of a prospective study of adolescent, young adult, and adult development (Newcomb, 1992, 1994, 1997; Newcomb & Bentler, 1988a, 1988b). Subjects Data were obtained from an ethnically mixed community sample for an ongoing longitudinal study of personal growth and development that began in 1976 when subjects were contacted from 11 Los Angeles County schools in five districts (Newcomb, 1992; Newcomb & Bentler, 1988a). Sites were chosen to oversample lower socioeconomic status and minority areas in order to compensate for an expected lower participation rate amongst subjects from lower socioeconomic status schools. The original sample included 1,634 7th through 9th grade students. Women comprise 70% of the total sample. The remaining 30% are men. The portion of the sample used for this study contains 390 subjects and is entirely comprised of women. The current average age in the sample is 35 years old. African-American, Hispanic-Americans, and Asian-Americans comprise 34% of the sample. About one half of the sample reported having at least one abortion, with some reporting up to four. The individuals in this community sample have been surveyed every four years since they were in junior high school. The data used in the current study is from Year 9 (1984) and Year 13 (1988) of the development study. At each survey point, the subjects gave informed consent for their participation and were given certificates of confidentiality stating Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that their responses would be confidential. A demographic description of the sample i found in Table I. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24 Table 1 Description o f the Women in the Study Characteristic n % Age (Year 9) 21 145 37 22 130 33 23 106 27 24 8 Ethnicity (Year9) African-American 56 14 Latina 42 1 1 Caucasian 243 62 Asian-A merican 38 10 Other 10 3 Education (Year 13) <H.S. graduate 17 4 H.S. graduate 89 23 1-3 years college 170 44 College graduate 102 26 Adv anced degree 12 3 Income in past year (Year 9) None 46 12 <$2,000 57 15 $2,001 to $10,000 171 44 $10,001 to $22,500 111 29 >$22,501 5 1 Living situation (Year 9) Live alone 18 5 Live with parents 16 4 College housing 173 44 Roommates 46 12 Spouse 59 15 Cohabit 40 10 Other 38 10 Months employed or in school full-time (Year 9) None 84 22 One or two 30 ~8 Three or four 35 9 Five or six 16 4 Seven or Eight 27 7 > Eight 198 51 Number of children (Year 9) 0 304 78 1 78 20 2 7 2 3 1 .3 Note. N = 390 , H.S. = high school Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 Attrition Analyses A criticism of abortion research is that the subjects most affected by abortion might drop out of studies. This criticism was addressed with a series of attrition analyses. Data for this study were taken from two data gathering time periods, covering a span of four years. The two waves of data collection used in this study occurred in 1984 and 1988, Years 9 and 13 of the study. Intensive examinations of the attrition effects suggest that loss of subjects from Year 1 to Year 13 was minimally affected by any selective or systematic selection process over the twelve years of the study (Newcomb, 1992; Stein, Newcomb, & Bentler, 1993). Furthermore, most of the loss of subjects was due to the inability to obtain a current address for them and not due to voluntary' withdrawal from the study (Newcomb, 1997). Additional attrition analyses were performed for this study to determine w hether or not those who participated in the study in both Year 9 and Year 13 were different from those who dropped out of the study al ter Year 9. Correlations between each of the 21 variables in the study, including Number of Abortions, and drop out status were examined. None were found to be significant. A stepwise multiple regression analysis was then performed. It indicated that even the best predictor (i.e. Relationship with Family) of leaving the study was not significant. Next, a simultaneous multiple regression was performed. Each of the 21 variables in the study was analyzed and no systematic effect for attrition was found. If women who experienced psychological effects following abortion had dropped out of this study, then it would be expected that Number of Abortions would have been a significant predictor of attrition. However, even variables such as Depression and Anxiety were not significant predictors. There were no significant predictors of attrition from Year 9 to Year 13. A series of attrition analyses indicated that abortion did not predict dropping out of this study. The results of previous abortion research have been questioned due to the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26 possibility that women who do experience psychological effects differentially drop out of studies. In the present study, the question of attrition was examined and re-examined. Abortion had no effect on attrition. Measures Number of Abortions For Time 1 (Year 9), women were asked how many times they had abortions. This number represented a lifetime total. For Time 2 (Year 13), women were asked how many abortions they had in the past four years. Time 1 and Time 2 were mutually exclusive categories, as the amounts reported would otherwise overlap. About one half of the sample reported at least one abortion (Stein, Newcomb, & Bentler, 1994). The range of total abortions per subject was from 0 to 4. Number of Children Number of Children and Number of Abortions were measured in an identical manner. For Time 1 (Year 9), women were asked how many children they have. This number was used as an indicator of lifetime number of children. For Time 2 (Year 13), women were asked how many children they had in the past four years. Time 1 and Time 2 were mutually exclusive categories to prevent overlap. The range of total children per subject was from 0 to 3. Dysphoria The Dysphoria construct was represented by three measured variables: CES- Depression, Suicide Ideation and Powerlessness. Depression (CES-D). This scale contains the 20 self-report items that constitute the Center for Epidemiologic Studies— Depression Scale (CES-D) constructed by Radloff (1977). In this study, the CES-D scale was used as a single measure of Depression. The CES-D scale measures (lack of) positive affect, negative affect, impaired motivation, and impaired relationships. The reported reliabilities for the sections contained within the CES- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27 D scale are as follows: .79 for Positive Affect, .82 for Negative Affect, .73 for Impaired Motivation, and .59 for Impaired Relations (Radloff, 1977). Sample items for the CES-D scale include “I was happy,” “I had crying spells,” “I could not get along,” and “People were unfriendly.” Suicide Ideation. Suicide Ideation was represented by one scale containing four single-items: think about killing self, told someone kill self, end life with suicide, and tried to kill self. A five-point anchored scale that ranged from never (0) to always (5) was used to assess each item. Two of these items were taken from Petrie and Chamberlain’s (1983) Suicide Behavior Subscale: “I have been thinking about ways to kill myself” and “I have told someone I want to kill myself.” Two other items were developed by Newcomb & Bentler (1988): “I imagine my life will end with suicide” and “I tried to kill myself.” A greater amount of suicide ideation was indicated by higher scores. A reliability of .76 was reported for Suicide Ideation (Petrie & Chamberlain, 1983; Newcomb & Bentler, 1988). Powerlessness. Three single-item variables represented the Powerlcssness v ariable. Subjects were asked to rate their level of agreement with three statements: (a) “I feel I am not in control of my life, “ (b) “I feel that whether or not I am successful is just a matter of luck and chance, rather than my own doing,” and (c) “I feel that others are running my life for me.” Four levels of agreement were offered: not at all (0), a little (1), somewhat (2), very much (3). Cronbach’s alpha for these items was .65. The Powerlessness variable assesses a general lack of control over life events and has been validated in other samples and studies (Newcomb, 1986; Newcomb & Harlow, 1986). Agitation The Agitation latent construct was reflected in two scales: Anxiety and Hostility. Anxiety. Anxiety was represented by a single scale containing nine items. A scale ranging from (1) to (5) was used for each item. “Feeling nervous, fidgety, or tense” is an item sample. The Anxiety reliability is .75 (Mellingeretal., 1983). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28 Hostility. Hostility was reflected in nine items. A rating scale ranging from (1) to (5) was used. An item sample is “Easily upset, irritated, annoyed.” A reliability of .80 is reported for Hostility (Mellinger et al., 1983). Agency An Agency latent construct was represented by two scales: Self-acceptancc and Self- efficacy. Self-Acceptance. The measure for Self-Acceptance contained four items rated on a scale of ( I) to (5). “Like myself for what I am” is an item sample for Self-Acceptance. The reliability for the Self-Acceptance variable is .76 (Stein, Newcomb, & Bentler, 1986). Self-Efficacy. Five items from the Self-Efficacy Scale (Blatt, Quinlan, Chevron, McDonald, & Zuroff, 1982) were used to measure self-effficacy. A five-point anchored rating scale was used. Responses ranged from strongly disagree (1) to strongly agree (5). The five items were factor analyzed and found to reflect a unitary construct (McCarthy & Newcomb, 1992). Intimate Relationships The latent construct of Intimate Relationships was reflected in five measured scales: Intercourse Satisfaction (Newcomb & Rickards, 1995), Happy with Relationship (Newcomb & Bentler, 1988), the five-item dependence scale from the Depressive Experiences Questionnaire (DEQ; Blatt etal., 1982), the Dating Competence Scale created from a subscale of the Dating and Assertion Questionnaire (Levenson & Gottman, 1978), and an abbreviated version (Newcomb & Rickards, 1995) of the Dyadic Adjustment Scale (Spanier, 1976). Intercourse Satisfaction. Three items were used in a single scale to measure intercourse satisfaction: happy with sex life, problems with sex, and intercourse satisfaction. Happy with sex life was rated on a 7-point scale anchored at the low end as terrible (1) and at the high end as delighted (7). Intercourse satisfaction was rated on a 7- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29 point scale ranging from very dissatisfied (1) to very satisfied (7). A 5-point frequency scale ranging from never (1) to always (5) was used to assess the frequency of sexual problems. Happy with Relationship. A single-item reflected the Happy with Relationship variable. The item asked the subject to indicate how she felt during the past six months about the relationship with her date, mate, or spouse on a seven-point anchored rating scale, ranging from terrible (1) to delighted (7). Dependence. The five item DEQ scale assesses feelings about the self within an interpersonal relationship. The items are rated on Likert scales and include “1 often think about the danger of losing someone close to me” and “After an argument, I usually feel lonely.” The Dependency score of the DEQ has reported reliability of .81 (Blatt et al., 1982) and .61 (Newcomb & Rickards, 1995). Dating Competence. A shortened version of the nine-item dating scale from the 18- item Dating Competence Scale was used for this study. Five items were assessed on a four-point scale ranging from never (1) to always (4). Sample items include “have an intimate relationship with a member of the opposite sex” and “get a second date with someone you dated once.” The five items had a reported reliability of .72 (Newcomb & Rickards, 1995). Dyadic Adiustment. A brief version of Spanier’s (1976) Dyadic Adjustment Scale was used only in Year 13 of this study. The complete Dyadic Adjustment Scale (Spanier, 1976) has a reliability of .96 and has shown strong content, criterion, and construct validity (Spanier, 1976). The 16-item shortened version used for this study has a reliability of .89 (Newcomb & Rickards, 1995). Four general areas reflected in this scale are consensus, affectional expression, cohesion, and satisfaction. Items for the shortened form were chosen on the basis of the largest factor loadings for each scale as found in previous research (Spanier, 1976). Participants were asked to evaluate their current or most recent Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 (if not currently involved) romantic involvement. Sample items include level of agreement in “making major decisions” and in “matters of recreation,” as well as how frequently the participant “regretted you were married (or lived together)” and “think things between you and your partner are going well.” Social Support Three multi-item scales were used to develop a latent factor of Social Support: good relationship with parents, good relationship with family, and support for emotional problems. The scales were measured using four bipolar items each. The scales measure the amount of acceptance, understanding, support, and respect experienced with parents, family and adults (Newcomb & Bentler, 1982). The good relationship with parents scale has a reliability of .82 (Newcomb & Bentler, 1986). The good relationship with family scale has a reliability of .84 (Newcomb & Bentler, 1986). A sample item for the good relationship with parents scale is “Parents show interest in me as a person” versus “Parents don’t seem to care about me.” “Family is very close to each other” versus “Family is not close at all” is a sample item from the good relationship with family scale. The Support for Emotional Problems scale includes support for psychological or emotional problems from three sources. These sources are family, friends, and organizations such as churches or counseling centers. Subjects indicated quantity of available support on a scale ranging from 0 (no units of support) to 9 (9 or more units of support). Social Conformity The social conformity latent construct was represented by three 4-item manifest measures of religiosity, (less) liberalism, and law abidance from the Bentler Psychological Inventory (BPI; Stein et al., 1986). Each of these is a four item bi-polar scale. A sample item for the religiosity scale is: “believe in religion or the Bible” versus “believe in science.” The period-free reliability for religiosity is .77 (Stein et al., 1986). A Liberalism scale sample item is: “support women’s liberation” versus “don’t feel women want or need Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31 it” and the period-free reliability is .54 (Stein et al., 1986). A sample item for the Law Abidance scale is: “return incorrect change” versus “keep incorrect change." The period- free reliability for law abidance is .85 (Stein et al., 1986). This construct has been used in numerous previous studies (e.g., Maddahian, Newcomb, & Bentler, 1988; McGee & Newcomb, 1992; Newcomb. Maddahian. Skager, & Bentler, 1987) Licit Drue Use Two measured-variable scales assessed how frequently two types of drugs (i.e. cigarettes and alcohol) had been used in the past six months. One item was used to measure cigarette use in Year 9. Two items were used to measure cigarette use in Year 13, as smokeless tobacco was added. Three items were averaged for alcohol (beer, wine, liquor). A 7-point scales ranging from never (1) to more than once a day (7) was utilized to measure both cigarette and alcohol use. A description of the reliabilities and sources for the measures can be found in Table 2. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 2 Reliabilities and Sources for Measures Used in the Analyses Measure Sample item Dysphoria CES-D Positive Affect Negative Affect Impaired Motivation Impaired Relations Suicide Ideation Think about killing self Told someone kill self End life with suicide Tried to kill self Powerlessness Feel not in control of life Success is matter of luck Others are running life Agitation Anxiety Hostility Agency Self-acceptance Self-efficacy Continued I was happy. I had crying spells. I could not get along. People were unfriendly. I have been thinking about ways to kill myself. I have told someone I want to kill myself. I imagine my life will end with suicide. I tried to kill myself. I feel I am not in control of my life. ...successful is just a matter of luck.... I feel that others are running my life for me. Feeling nervous, fidgety, or tense Easily upset, irritated, annoyed Like myself for what I am I set my personal goals as high as possible Item Reliability ranges Principal Sources Radloff (1 9 7 7 ) .79 0-3 .82 0-3 .73 0-3 .59 0-3 .76 0-5 Petrie & Chamberlain, 1983 Newcomb & Bentler, 1988 .65 0-3 Newcomb (1986); Newcomb & Harlow, 1986 .75 1-5 Mellinger et al., 1983 .80 1-5 Mellinger et al., 1983 .76 1-5 Stein, Newcomb, & Bentler, 1-5 1986; Blatt et al., 1982 McCarthy & Newcomb, 1992 CO IN ) 7 3 CD ■ o - 5 o Q . C o CD Q . $ l - H ■O CD — i 3 Table 2 continued— Reliabilities and Sources for Measures C/5 C / ) o ’ Intimate Relations 3 O Intercourse Satisfaction range of very dissatisfied to very satisfied 1-7 Newcomb & Rickards, 1995 l - H Happy with Relationship range of terrible to delighted 1-7 Newcomb & Bentler, 1988 CD O Dependence After an argument, 1 feel very lonely .81;.61 1-7 Blatt et al., 1982; o ■O V < Newcomb & Rickards, 1995 -5 c q ' Dating Competence get a second date with someone you dated or .72 1-4 Levenson & Gottman, 1978 l - H o Dyadic Adjustment Scale extent of agreement making major decisions .89 Newcomb & Rickards, 1995; o Spanier (1976) CD —s Social Support Newcomb & Bentler, 1986b T1 C Good Relation w / Parents Parents show interest vs. don't seem to care .82 0-1 Newcomb & Bentler, 1986b Good Relation w / Family Family is very close vs. not close at all .84 0-1 Newcomb & Bentler, 1986b CD Support for Emot Problem quantity of available support for emot prob 0-9 Newcomb & Chou, 1989 CD ■ a Social Conformity BPI; Stein et al., 1986 o Q . C Religiosity believe in religion vs. believe in science .77 a 0-1 a o Liberalism support women's lib vs. women don't want... .54 a 0-1 o -Q Law Abidance return incorrect change vs. keep it .85 a 0-1 -5 o Licit Drug Use Newcomb & Bentler, 1988b 5; Cigarette Use how often in the past six months.... 1-7 CD Q . l - H Alcohol Use 1-7 3 " o c Note. Unless otherwise indicated, reliabilities were computed using Cronbach's alpha. l - H ■O CD 3 c/ j c/ j o' o a: These alphas are computed as period-free with the formula of Heise (1969). CO CO 34 Analyses The EQS Structural Equation Program (Bentler, 1995) was used for the analyses of all confirmatory factor analysis (CFA) and path models, including the moderator multiple- group analyses. Goodness of fit was measured with the ratios between maximum likelihood chi-square estimates and the degrees of freedom (x 2 / df). It was also assessed by two fit indexes, the Bentler-Bonett nonnormed fit index (1989), which is identical to the Tucker-Lewis fit index, and the Comparative Fit Index (Bentler, 1989, 1990). Chi- square/degrees of freedom ratios of less than 2:1 and goodness of fit index values greater than .90 indicate well-fitting models (Newcomb, 1994). CFAs were run to determine the adequacy of the hyptithesized factor loadings, the degree of model fit, and the intercorrelations of the latent constructs. Initial path models were then tested on the basis of the CFA. Paths among latent constructs were tested using a subtractive approach, similar to stepdow n stepwise multiple regression. This approach starts with a saturated model that includes all hypothesized as well as non-hypothesized paths and all correlations among the independent variables and among disturbance (i.e. error) terms of the dependent constructs (Newcomb, 1994). All nonsignificant paths and correlations were then removed. The final model was evaluated by its resemblance to the hypothesized model. The presence of unexpected paths as well as the presence or absence of hypothesized paths provided the basis forevaluation. The multivariate Langrangian Multiplier test was used to add specific effects to the path model, and the Wald test was used to delete nonsignificant paths from the model (Chou & Bentler, 1990). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 CHAPTER III RESULTS Nonlinear Effects A preliminary analysis was executed to determine whether or not nonlinear effects could be found for this data set. This analysis was performed to rule out the possibility of nonlinear effects for any of the 20 variables in relation to abortion. Conceptually, it appeared possible that nonlinear effects might be found for this data set. For example, perhaps a woman who has had one abortion might be better adjusted than cither a woman who has none or a w oman who has had two or more abortions. There might be reasons a woman who never had an abortion might be higher on depression than a woman who had one. Likewise, a woman who has had two or more abortions might be higher on depression than a woman who has had one abortion. For example, a woman with an unwanted pregnancy might be opposed to abortion due to religious or other reasons. Or, this woman might be coerced into having the child by the partner or her family. She might have an unwanted child under difficult financial, relational, and emotional circumstances. A different woman might find herself pregnant, have an abortion, learn from the experience, and take a more active approach to birth control in the future. This woman might actually become empowered through her experience. A third woman might feel her life is out of her own control and might repeatedly have abortions. The experience of repeatedly becoming pregnant and having abortions might lower her self-esteem. As these examples demonstrate, it is conceivable that a woman who had one abortion might be less depressed (or anxious etc.) than a woman who had either no abortions or two or more. Therefore, an examination of nonlinear effects was performed. ANOVAs were run to look for main effects, one for each of the 20 variables being studied. The means for each of the variables in relation to Number of Abortions were examined. One possible nonlinear effect was found for the Relationship with Family Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36 variable, suggesting women who had two abortions might have better relationships with their families than women who had one abortion. However, it was expected that at least one finding would be made purely by chance when running 20 ANOVAs. Indeed, when using the Bonferrom (Dunn, 1961) procedure to correct for capitalizing on chance due to multiple comparisons, this effect was no longer significant (.05/20 = .003). The alpha level was adjusted to account for the 20 comparisons being made. The resulting alpha level was nonsignificant, indicating the finding was most likely a product of chance. As this omnibus test was not significant, further analyses of potential nonlinear effects were not indicated. Therefore, continuing with analyses of linear effects was deemed appropriate. Analyses were also performed to determine if including women who had never been pregnant affected the results of the study. All women in the sample were included in this study. To consider this possible confound, ANOVAs were run for the Year 9 and the Year 13 data. The means for three groups were examined. The three groups were a) never pregnant, b) pregnant, no abortions, and c) at least one abortion. Each variable was examined in a separate ANOVA. Significant differences were suggested between the three groups for the Year 9 variables Dating Competence, Religiosity, Law Abidance, Cigarette Use, and Alcohol Use. The Bonferroni procedure to correct for capitalizing on chance was used. The p values associated with individual tests were adjusted to account for the nineteen comparisons being made (.05/19 = .003). After that procedure, Dating Competence no longer showed a significant between group difference. The remaining significant results indicated differences on the variables Religiosity, Law Abidance, Cigarette Use, and Alcohol Use. Religiosity, Law Abidance, Cigarette Use, and Alcohol Use remained significant after adjusting the p values to account for comparisons of the three groups (.05/3 = .016). No other variables showed significant differences between the three groups for Year 9. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37 Women who had been pregnant and had not had an abortion were higher in Religiosity then women who had at least one abortion. There was no significant difference between women who had never been pregnant and the other two groups on Religiosity. Women who had never been pregnant and women who had been pregnant and not had any abortions scored significantly higher on the Law Abidance measure than women who had at least one abortion. Women who had at least one abortion also used significantly more cigarettes than women who had never been pregnant. Differences were also found for Alcohol Use. Women who had never been pregnant and women who had had at least one abortion had higher Alcohol Use than women who had been pregnant and not had any abortions. No other differences between the three groups w ere found on any other variables for Ycar 9. For the Year 13, significant differences were suggested for the variables Happy with Relationship, Dating Competence, Religiosity, Law Abidance, Cigarette Use, and Alcohol Use. After using the Bonferroni procedure to avoid capitalizing on chance, by adjusting the p values to account for nineteen comparisons (.05/19 = .003), the variables Happy with Relationship and Dating Competence were no longer significant. The remaining results were very' similar to the Year 9 results and remained significant al ter adjusting the p values to account for comparisons of the three groups (.05/3 = .016). The Year 13 results for Religiosity, Law Abidance and Cigarette Use were identical to the Year 9 results. For Year 13 Alcohol Use, there was no longer a significant difference between women who had never been pregnant and women who had been pregnant and not had any abortions. However, a significant difference remained suggesting women who had been pregnant and not had any abortions used less alcohol than women who had at least one abortion. No other differences were found for any of the Year 13 variables. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38 Correlations Between Number of Abortions, Number of Children and the 19 Other Variables Correlations were run between Number of Abortions, Number of Children, and the 19 other variables. These correlations do not imply any direction of effects or causality. All of the associations found for both Number of Abortions and Number of Children were small. For example, the amount of variance explained between Year 9 Number of Abortions and Year 9 Dating Competence was .023. Indicated were associations between the Year 9 Number of Abortions vanable and the Year 9 variables of Dating Competence (. 15, p < .001), (less) Religiosity (-12, p < .01), (less) Law Abidance (-27, p<.001). Cigarette Use (.23, p < .001), and Alcohol Use (.12, p < . 01). Women who had abortions appeared more competent in dating, less religious and law abiding, and higher in cigarette and alcohol use than women who had not had abortions. Year 9 Number of Abortions were correlated with the Year 13 variables of Number of Abortions (. 13, p < .001), Dating Competence (.11, p < .01), (less) Law Abidance (-.17, p < .001), Liberalism (.10, p < . 05), Cigarette Use (.23, p < .001), and Alcohol Use (.12, p < . 01). The associations significant in Year 9 decreased for each of the variables except Cigarette Use. Religiosity was no longer significantly correlated with Number of Abortions in Year 13. A small but significant association between Liberalism and abortion became apparent in Year 13. No other associations between the variables and Number of Abortions in Years 9 or 13 were found. Correlations between Number of Children and the other variables are also provided for comparison. These results indicated Year 9 Number of Children was associated with Year 9 Number of Abortions (.11, p < .01), CES-D (.15, p < .001), (less) Self-Acceptance (-. 15, p < .001), (less) Self-Efficacy (-. 13, p < .001), Dependence (. 10, p < .05), and Cigarette Use (. 17, p < .001). The correlations between Year 9 Number of Children and Year 13 variables were as follows: Number of Abortions (. 19, p < .001), (less) Self- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39 Acceptance (-. 11, p < .01), (less) Relationship with Parents (-. 12, p < .01), Religiosity (. 15, p < .001), and Cigarette Use (.17, p < .001). These correlations are presented in Table 3. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 Table 3 CORRELATIONS BETWEEN NUMBER OF ABORTIONS, NUMBER OF CHILDREN, AND THE 19 OTHER VARIABLES N=390 NUMCHLD9 NUMABF YEAR 9 NUMCHLD 1.00 NUMABRT .11° 1.00 CES-D .15° .01 SUIC IDEATION .01 .00 POWRLES .08 -.03 ANXIETY -.05 .00 HOSTILITY .04 .04 SELF-ACCEPT -.15C .04 SELF-EFFIC -.13C .04 INTERC SAT -.04 .04 HAPPY W REL -.01 -.04 DEPENDENCE .10* .01 DATING COMP -.02 .15' REL PARENTS -.04 -.03 REL FAMILY .04 .02 S EMOTIONAL .00 -.07 RELIGIOSITY .08 -.12° LAW ABID .03 -.27c LIBERALISM -.01 .03 CIG USE .17C .23c ALCOHOL USE -.18 .19' YEAR 1 3 NUMCHLD -.03 .05 NUMABRT .19c .13' CES-D .04 -.01 SUIC IDEATION -.09 -.04 POWRLES .06 .00 ANXIETY -.04 -.06 HOSTILITY -.03 -.03 SELFACCEPT -.11D .01 SELF-EFFIC -.05 -.03 INTERC SAT -.01 -.02 HAPPY W REL .06 -.04 DEPENDENCE .01 -.02 DATING COMP .04 .11° DYADIC ADJ -.09 -.01 REL PARENTS -.12° -.01 REL FAMILY -.03 .03 RELIGIOSITY .15C -.07 LAW ABID -.01 -.17' LIBERALISM -.03 .10* CIG USE .17c .23' ALCOHOL USE -.07 .12° Two-tailed level of significance: a=p <.05 b=p <.01 c=p <.001 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 41 Confirmatory Factor Analysis Prior to constructing the across-time path model, the correlations among the latent constructs were examined to verify that the chosen measured variables reflected the latent constructs in a statistically reliable manner. Confirmatory Factor Analyses (CFAs) were run to determine if the chosen measured variables reliably captured the latent constructs. As the latent constructs Agitation, Intimate Relations, Social Conformity, and Licit Drug Use had been used in previous research, they were expected to be reliable. The latent constructs Dysphoria, Agency, and Social Support were constructed for this study. For the CFA model, all factor loadings were freed, factor variances were fixed at 1.00 in order to identify constructs, and all factors were allowed to freely correlate without any imputation of causality among them. The initial fit of the measurement model was tested and found to be good, indicating the hypothesized model adequately describes the sample data. All of the hypothesized factor loadings were in the expected directions and significant (p c.OOl), as were all of the covariances among the latent variables. Therefore, the hypothesized factor loadings will be presented in the path models. The fit summaries are as follows: x 2 (946, N=390) = 8398.01, NNFI= .90, CFI=.92. The CFA model is presented in Figure 1 . A summary of the fit indexes and nested tests can be found in Table 4. Table 5 provides the intercorrelations among the latent factors from the initial CFA model. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. YEAR 9 S U ia O E PEATION M 1 Agitation SELF -ACCEPTANCE Agency \ NTER SAT Intimate \ Relations J HAPPY WITH R EL J DATING CO M PE TEN CE Social \ Support J _ SociaTX 4 0 ^ Conformity J CIGARET TE U S E Drug Use J ALCOHOL U S E » Of Abortion* • Of CMdren FIGURE 1 CFA MODEL DEPICTING FIRST-ORDER LATENT CONSTRUCTS Dysphoria^ YEAR 13 Dysphoria I SUICPE IDEATION * L PO W ERL ESS NE SS k I Agitation HOSTHJTY ► I SELF-ACCEP TANCE 4 Agency NTER SAT HAP PY WITH R E L PATINO CO MPE TEN C E DYAWCADJ ► [ R EL PARENTS ► I R EL FAMLY RE L K R O S IT Y LAWABOANCE / Intimate \ I Relations I Social \ Support J. / Social X (conformity) /L icit Drug\ I Use j ^ 0 Of Abortion* [ 0 Of CMdren - > [ CIG AR ET TE U S E ► [ ALCOHOL U S T M Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4 Summary of Model Fit Indexes and Nested Tests Model X * df x 2 / df P NNFI* CFIb CFAC 1238.07 700 1.76 <.001 .90 .92 Initial STRd 1238.07 700 1.76 <.001 .90 .92 Final STR 1219.1 713 1.71 <.001 .90 .92 Multiple-group analyses All Free 501.17 252 1.99 b o O 00 00 FL=ef 530.79 269 1.97 - .78 .88 Difference Test 29.63 17 - <.05 - - FL and covariances 566.03 297 1.91 - .77 .87 Difference test 35.24 28 - <.01 - - Note. Dashes indicate that data were not applicable. aNNFI=non-normed fit index (Bentler, 1989; this is identical to the Tucker-Lewis fit index). b CFI= comparative fit index (Bentler, 1990). c CFA= confirmatory factor analysis. d STR= structural or path model. eFL= factor loading. fAn equal sign indicates that these parameters were constructed to equality between social support groups. 0 0 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 5. Intercorrelations Among the Latent Factors From the Initial Confirmatory Factor Analysis Model Latent Factors Latent Factors 1 Dysphoria9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 2 Agitation9 .86c -.75c -.91c -.51c -.15* -.03 -.01 .12* .54c .52c -,43c -.49c -.47c -.16 -.01 .00 -.08 3 Agency9 - -.65c -.76c -.41c -.34c .11 .04 .00 .49c .61c -.40c -.36c -.34c -,34c .12 -.09 .01 4 Intimate Rel9 - ,74c .45c .24c -.04 .04 -.15b -.46c -.44c .63c .41c .43c .19b -.05 .02 .01 5 Social Supprt9 - .64c .09 .05 .04 -.14 -.66c -.52c .54c .82c .62c .10 -.05 .02 .10 6 Social Confor9 - .22b -.03 -.01 .00 -.35c -.33c .28c .31c .69c .22b -.01 -.01 .08 7 Licit Drug Use9 - -.75c -.33c .06 -,20b -,25b .19b .18* .18b .89c -,60c -.21b .08 8 # of Abort9 - .23c -.20b -.04 .04 -.03 .02 .05 -6 2 c .79c .05 .-10 9 # of Child9 - .11* -.02 -.05 .01 -.01 .01 -.27c .14* .05 .13* 10 Dysphoria 13 .02 -.04 -.11* -.02 -.09 .11 -.07 -.03 .19c 11 Agitation 13 - .73c -.85c -.68c -.58° -,38c .12* .02 -.02 12 Agency13 - -.58c -.50c -.42c -.37c ,18b .02 -.06 13 Intimate Rel13 - .58c .53c .33c -.14" -.02 -.09 14 Social Supprtl 3 - .43c .33c -.13* -.05 .08 15 Social Conforl 3 - .29c -.02 -.10 .05 16 Licit Drug Usel 3 - -.57c -.22b .12 17 # of Abortl 3 - .08 -.16b 18 # of Childl 3 - -.09 Two-tailed level of significance: a=p< .05. b = P < .01. c= p < .001. - 45 Path Models Following the confirmation of the factor structures in the CFA model, the across-time path models were constructed. The same factor structures confirmed in the CFA model were used as the basis for the path models. All within-time constructs or their related disturbance terms were allowed to freely correlate. All Year 9 constructs were initially allowed to predict all Year 13 outcomes. A saturated model was established. All paths were tested, including those which were not hypothesized, in order to avoid model misspecification. Nonsignificant paths were deleted. The results of the path model indicate that Social Conformity was the only significant predictor of abortion. Higher Year 9 Social Conformity was related to lower Number of Abortions in Year 13 (-. 10, p < .05). The results also indicated only one significant consequence of abortion. Number of Abortions was related to Number of Children (.09, p <.01), indicating that a higher Number of Abortions in Year 9 was related to a higher Number of Children in Year 13. No other predictors or consequences of abortion were found. Year 9 Intimate Relations negatively predicted Year 13 Dysphoria (-.61, p <001). Those with stronger Intimate Relations in Year 9 were lower on Dysphoria in Year 13. This relationship was greater than that of Dysphoria in Year 9 with Dysphoria in Year 13. Intimate Relations also predicted Social Support (.21, p <001), indicating better Intimate Relations in Year 9 was related to more Social Support in Year 13. Also suggested was that Year 9 Social Support might lead to less Year 13 Agitation (-. 10, p < 10). However, due to the significance level of this finding, it must be considered exploratory. The final path model is presented in Figure 2. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 2 YEARS Dysphoria Agitation FINAL PATH MODEL YEAR 13 ^ Agency^ / Intimate \ x I R elations/^ [ ~ T o i ^Dysphoria XI Agitation Agency / Intimate Relations Social I Support / Social 87*** X Conformity i o f Abortions i o f Qwdren / Social \ I Support ) / Social (Conformity / Licit Drug\ I Use i 0 o f A bortion *-p< 05. **-p<.01. ***-p<.001. +«p<. 10. All tests are two-talled except the path between Social Conformity and # of Abortions, which is one-tailed. -f* cn 47 Specific or Nonstandard Effects Common or shared aspects of two or more variables are captured by the latent constructs (Newcomb, 1994). An analysis of the unique or specific aspects of the individual measured variables was performed in order to examine any specific predictors or consequences independent from the common latent construct. As no psychological or relational effects were found for Number of Abortions, additional analyses were performed to search for any effects. This further analysis was important in order to rule out the possibility that effects for abortion were not captured when analyzing the constructs instead of individual variables. The unique predictors and consequences of Number of Abortions were targeted and Number of Children was included to provide comparisons. Specific effects were added to the model using the Langrangian Multiplier test. The possible paths included three types of across-time paths that were not restricted to those between latent variables. These possible paths were (a) from earlier measured variables or their residuals to later latent constructs; (b) from earlier latent constructs to later measured variables; or (c) from earlier measured variables or their residuals to later measured variables (Newcomb, 1994). Utilizing these possible paths enabled an examination of the antecedents or consequences of the individual measured variables that made up each of the constructs. The models were overfit to capture all possible associations, as recommended by MacCallum (1986) and Bentler ( 1990b). Nonsignificant paths were then deleted using the Wald test (Chou & Bentler, 1990). The independent constructs and residuals of the outcome constructs were allowed to freely correlate. Two nonstandard or specific effects were suggested for the Number of Abortions variable. One of these nonstandard effects was significant and one was not. Religiosity, over and above the fact that one is socially conforming, significantly predicted Number of Abortions (-. 11, p <05). Specifically, the higher in Religiosity in Year 9, the lower in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48 Number of Abortions in Year 13. The other specific effect relating to abortion was not significant at the p <.05 level and is presented only as an exploratory finding. Number of Abortions in Year 9 predicted lower Agitation (-.08, p<. 10) in Year 13. All other specific or nonstandard effects not related to abortion were related to Number of Children. Intimate Relations and Dependency both predicted Number of Children (. 1 3, p <.05 and . 17, p <.001). Stronger Intimate Relations and higher Dependency predicted a higher Number of Children. Number of Children predicted Religiosity (.09, p <.001), Happy with Relationship (. 10, p <.05), (less) Suicidality (-. 11, p <.01), and (less) Social Support (-.06, p <.001). Those who were higher on Number of Children were also likely to be more religious, happier in their relationships, lower in suicidality, and lower in social support. The specific or nonstandard effects are presented in Table 6. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 6 Specific or N onstandard Effects Year 9 predictors Year 13 outcomes Measured Latent Measured Latent Standardized Variable Variable Variable Variable Parameter Estimate Religiosity # of Abortions -.1 1 * Dependence # of Children .1 7 *** Intimate Relations # of Children .13 * # of Children Happy with Relationship .10* # of Children Religiosity .0 9 *** # of Children Social Support -.0 6 *** # of Children Suicidality -.1 1 ** # of Abortions Agitation -.08+ *= p < .0 5 . ** = p<.01. ** * = p<.001. + = p<.10; two-tailed level of significance - t* L D 50 Multiple-Group Analyses Multiple group analyses were performed in order to evaluate social support as a moderator in the abortion response. According to the buffering model of social support, those high on social support would have less negative psychological and relational functioning than those low on social support following abortion. Multiplc-group analyses also served to further search for any effects of abortion. While the previous analyses indicated that Number of Abortions had no effect on psychological or relational variables, it appeared possible that effects might exist but be cancelled out in the analyses because of the existence of separate groups high or low in social support. To rule out this possibility, multiple-group analyses were performed (Bentler, 1989; Newcomb, 1990b; Byrne, 1994) using the Year 9 data. Prior to testing the hypothesis related to invariance, a baseline model was first established (see Byrne, 1994). The confirmatory factor analysis of the Year 9 data was used for these analyses. Baseline models were created by allowing all parameters in each group to be free. For the second model, all of the correlations among factors and the variables of interest were constrained to equality between the two support groups. The two groups examined were High Social Support and Low Social Support. Identical sets of variables were examined for each group. The multiple-group analyses examined how both Number of Abortions and Number of Children related to all the other factors, depending on whether or not there was High Social Support or Low Social Support. The fit indices and nested group difference tests are presented in Table 4. An omnibus test subtracting the baseline model from the constrained to equality model was executed. This test indicated a significant difference existed. The factor loadings were examined. There were no significant differences on any of the factors between the high and low social support groups for Number of Abortions or Number of Children. However, a difference was suggested between the variables of Number of Abortions and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51 Number of Children. This difference indicated that Number of Abortions and Number of Children correlated differently for those low in social support. For those high in social support, there was no association between Number of Children and Number of Abortions. This finding may be merely a matter of chance. However, as it was the only significant piece of information found related to the research hypotheses, possible interpretations for this finding will be provided in the Discussion section. There were no other differences found between the high and low support groups for Number of Children or Number of Abortions. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 CHAPTER IV DISCUSSION Speckhard and Rue (1992) argue for the existence of a “Post Abortion Syndrome” similar to Post Traumatic Stress Disorder. The results of the current study do not support this notion. Indeed, much less severe effects on a diagnostic continuum were not found. The results of this study indicate that abortion does not have psychological or relational effects on the variables examined. This finding is consistent with that of the APA panel which reviewed the abortion research in 1990 (Adler et al., 1992). Speckhard and Rue (1992) further state that definitive conclusions about the psychological effects of abortion cannot be made due to the methodological flaws in the research. As many (e.g., Adler ct al., 1992; David, 1994) have pointed out, the extant abortion research does have many significant Haws, including sampling bias, high attrition rates in abortion samples, lack of baseline data on subjects, and very limited follow-up. The current study addressed some of the major criticisms of previous abortion research and still did not find evidence of psychological or relational disturbance following abortion. One major criticism of previous research is that those who volunteer for abortion research might have more positive attitudes toward abortion. A community sample not recruited for abortion research was utilized for this study. However, those who volunteer for research might tend to have less conservative attitudes in general (Joe, Jones, & Ryder, 1977), which might, in turn, predict more favorable attitudes toward abortion. Nevertheless, utilizing a community sample is likely to have eliminated the issue of abortion research volunteer bias. It does not seem likely that the reason so few effects were found for abortion was due to any sample bias in this study. Another criticism of previous research addressed in this study is that those who are most affected by abortion may drop out of abortion research. The lack of results relating to abortion found in this study is not likely to be due to this factor. This possibility was Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53 examined with a series of attrition analyses. Previous examinations of the attrition effects for the sample suggest that loss of subjects from Year 1 to Year 13 was minimally affected by any systematic selection process over the twelve years of the study (Newcomb, 1992; Stein, Newcomb, & Bentler, 1993). Furthermore, most of the loss of subjects was due to the inability to obtain a current address for them (Newcomb. 1997). For this study, additional attrition analyses were performed to determine whether or not those who participated in the study in both Year 9 and Year 13 were different from those who dropped out of the study al ter Year 9. Correlations between each of the 21 variables and attrition were examined and none, including abortion, were found to be significant. A stepwise multiple regression analysis indicated that abortion was not a predictor of leaving the study. Relationship with Family was the best predictor of dropping out and even it was not significant. Next, a simultaneous multiple regression was performed. Abortion was not related to leaving the study. Based on these attrition analyses, it appears highly unlikely that the lack of findings relevant to abortion was due to having subjects who were affected by abortion dropping out of the study. A final criticism of previous research at least partially corrected in this study is the limited follow up of subjects. It could be argued that abortion has no immediate psychological or relational effects but that significant effects might arise years al ter the event. The time frame in this study covered four years, a significant improvement over much of the previous research with time frames of about one hour to about two years. Therefore, the lack of negative psychological and relational results related to abortion in this study is not likely to be due to an insufficient follow-up. However, the possibility does exist that some women might experience negative psychological effects 10 or 20 years after an abortion. For example, perhaps a 40 year old woman having difficulty carrying a wanted pregnancy to term might have feelings of regret about an abortion 20 years earlier. In such a case, a previous abortion might interact with current situational and psychological Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54 variables, resulting in psychological distress. Nevertheless, it seems unlikely that such distress would be directly related to the abortion alone or that it would warrant a psychiatric diagnosis. Several issues were addressed in ANOVA analyses. For example, it appeared possible that nonlinear effects might exist for this data. Women who had one abortion might be lower on a variable such as Anxiety than either women who had no abortions or more than one abortion. Therefore, ANOVAs were run to look for main effects, one for each of the 20 variables being studied. The means for each of the variables in relation to Number of Abortions were examined. The results indicated no significant differences on any of the variables based on Number of Abortions. No differences were found between the group having no abortions and the groups having one or more abortions. Further, there were no within group differences among women having one, two, or three or more abortions. The possible combinations were considered and no consistent pattern was found. It also appeared possible that few effects were found because women who had never been pregnant were included in the analyses. All women in the sample were included in this study. To consider this possibility, additional ANOVAs were run for the Year 9 and the Year 13 data. The means for three groups were examined. The three groups were a) never pregnant, b) pregnant, no abortions, and c) at least one abortion. All of the variables were included in the analyses. No significant differences were found for the different groups on any of the psychological and relational outcome variables, suggesting the inclusion of all women in the sample did not affect the lack of findings pertaining to psychological and relational distress. Differences were found for both Year9 and Year 13 on the variables Religiosity, Law Abidance, Cigarette Use, and Alcohol Use, suggesting small but significant differences may exist between the three aforementioned groups on these variables. Further research might expand on these findings. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55 According to Lazarus and Foikman’s (1984) model of stress and coping, social support is one of the stable resources one brings into a stressful situation. In the current study, level of social support did not have an impact following abortion. Those high in social support did not have better psychological and relational functioning following abortion than those low in social support. Number of Abortions was not significantly related to any of the other factors whether or not there was high or low social support. These results do not suggest a problem with Lazarus and Foikman’s model (1984) or with the view of social support as a stress buffer. Facing a stressful situation does not necessarily lead one to psychological or relational distress. In fact, managing a stressful situation might lead to positive growth. This may be the case for the majority of women who obtain abortions. Further research would be needed to determine positive psychological and relational effects of abortion. Another possible explanation for the intriguing findings in this study may have been due to considering abortion as a stressor. Perhaps it would be more accurate to consider unwanted pregnancy as the stressor. Social support might then serve as a buffer against the stress of unwanted pregnancy. Abortion, then, would present one possible resolution of the stress of unwanted pregnancy. In some cases, abortion might be a positive, stress- reducing action. This possibility was suggested, though not confirmed, in the statistically nonsignificant exploratory finding that those higher in Number of Abortions in Year 9 were lower in Agitation in Year 13. Correlations were run between Number of Abortions and the other 20 variables in the study. These correlations indicated small but significant associations between Year 9 Number of Abortions and the Year 9 variables of Dating Competence, (less) Religiosity, (less) Law Abidance, Cigarette Use, and Alcohol Use. These results suggest women who have abortions may be more competent at dating, less religious and law abiding, and may have more licit drug use than women who do not have abortions. The combination of these Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56 variables suggest some relationship between abortion and women who are perhaps more deviant in their attitudes and licit drug use. Five significant associations were found between abortion and the other variables. However, no direction of effects or causality can be inferred from the correlational analysis. Therefore, structural modeling was next used to determine if there is some effect over time. In the path model, only one construct predicted Number of Abortions. As expected. Licit Drug Use did not predict abortion occurrence. Social Conformity was the only significant predictor of abortion, with those higher on conformity having fewer abortions. This result supports the hypothesis regarding Social Conformity and Number of Abortions. This finding also makes intuitive sense and is consistent with the findings of Costa, Jessor, & Donovan (1987), who found greater psychosocial unconventionality, greater political liberalism, and less religiosity to be correlated with women who have had an abortion. The reason for this relationship may be twofold. Those who are less conforming may be more likely to engage in increased sexual activity outside of wedlock, as well as more likely to utilize abortion to resolve an unplanned pregnancy. No other predictors of abortion were found. The path model indicated one consequence of Number of Abortions. Number of Abortions was found to be significantly related to Number of Children. Those with a higher Number of Abortions in Year 9, were more likely to have a higher Number of Children in Year 13. Potential explanations for this finding might include a “guilt factor,” whereby someone who has had an abortion(s) decides to have the baby when next pregnant due to feelings of guilt. Perhaps a more plausible explanation would simply be that those who are more likely to have an abortion(s) in Year 9 are more sexually active or more fertile than their counterparts. These women might be likely to both get pregnant and to choose to bear children four years later when more stable financially or relationally. No other consequences of abortion were suggested. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57 An interesting finding, unrelated to the abortion research questions, was that Year 9 Intimate Relations was negatively related to Year 13 Dysphoria. This relationship was stronger than that of Year 9 Dysphoria to Year 13 Dysphoria. This implies that the quality of Intimate Relations may be a more important predictor of Dysphoria than previous Dysphoria Intimate Relations also predicted Social Support, indicating stronger Intimate Relations in Year 9 was related to more Social Support in Year 13. Another finding, which was statistically nonsignificant and therefore must be considered as exploratory , was that Year 9 Social Support might be related to less Agitation in Year 13. This finding would support both the beneficial and the buffering hypotheses regarding stress and social support. As so few results were found relating to abortion, further analyses were performed to rule out the possibility that some effects could be found for individual variables that were not statistically relevant for the constructs as a whole. However, even with further sifting of the data, few results were discovered. An analysis of the unique aspects of the individual measured variables suggested two nonstandard or specific effects for the abortion variable. One of these effects was significant and one was not. Religiosity, over and above the fact that one is socially conforming, significantly predicted less abortion. This finding appears logical for a variety of reasons. For example, higher religiosity might lead a woman with an unintended pregnancy to consider fewer options for pregnancy resolution. Abortion might be against the dictates of her religion. Interestingly, abortion rates indicate similar percentages of Catholic women and Protestant women have had an abortion (Costa et al., 1987). The reasons for this might be twofold. Firstly, type of religion as a category does not indicate the strength of one’s religiosity or the degree to which one adheres to the dictates of a particular religion. Secondly, Catholic women may be equally likely to abort an unintended pregnancy because the ramifications for bearing a child out of wedlock might be more severe than for a Protestant woman. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58 Though not significant at the p <05 level, Number of Abortions predicted lower Agitation. This is somewhat similar to the finding by Zabin et al. (1989) that teenagers who abort an unwanted pregnancy are less anxious than those who do not. When viewing unwanted pregnancy from a stress and coping framework, it is possible that abortion could act as a stress reducing method of coping, thereby reducing one’s level of agitation. Further research is necessary to determine whether or not abortion can have a positive impact on one’s level of agitation. No other specific effects relating to abortion were found. Other specific effects were found which related to Number of Children. Stronger Intimate Relations predicted a higher Number of Children. This finding is not surprising, as many couples would be more likely to choose to reproduce within a satisfying dyad than within a poorer relationship. Dependency also predicted Number of Children. This may simply be due to independent women being more significantly involved in the workforce than dependent women. Both personal interest and practical necessity might limit the number of children an independent person would want. On the other hand, dependent women may achieve more of their self-worth through homemaking and raising children than through any jobs outside of the home. A more dependent woman might also be more passive in her approach to birth control, as well as more passive in her approach to an unwanted pregnancy than an independent woman. A more independent women might take an active approach to birth control as well as an active approach to an unwanted pregnancy by choosing an abortion. There were no other specific predictors of Number of Children. Number of Children predicted Religiosity, Happy with Relationship, (less) Suicide Ideation, and (less) Social Support. Number of Children predicting Religiosity is an interesting finding. Perhaps having children leads one to develop or renew one’s faith in order to provide one’s children with a religion or a religious community. Children predicting happier relationships is another unusual finding, which contradicts Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59 commonsense. It would be expected that more children would lead to more strain in a relationship due to fewer available resources (e.g., time, money). On the other hand, child rearing might encourage a couple to become a more united, stronger dyad. The remaining specific effects relating to Number of Children were less surprising. Being responsible for another person, especially a child, is likely to reduce any suicide ideation. This may be due to a shifting of focus from one’s ow n problems or emotions to taking care of a dependent child. Less social support would most likely be related to less time to socialize or to reciprocate support due to hav ing children. Number of Children was not associated with any other variables. More analyses of the data were done to search for any effects of abortion. Multiple- group analyses indicated that neither Number of Abortions nor Number of Children were significantly related to any of the factors, whether or not there was high or low social support. However, a significant difference was found in one area. For those low in social support. Number of Abortions and Number of Children were correlated differently. For those high on social support, there was no association between Number of Children and Number of Abortions. This finding is probably due to chance. Other possible explanations for this result might be that women who are lower in social support may be more likely to conceive, to have abortions, and to have children. In the path model, Number of Abortions was significantly related to Number of Children. One of the potential explanations given for this result was a “guilt factor,” whereby women who had previous abortions chose to have the child when finding themselves pregnant again. The finding from the multiple-group analysis could be similarly related to a “guilt factor” but only relevant for those low in social support. Another possible explanation might be that women who are low in social support may attempt to fulfill unmet needs for affiliation and affection through sexual activity as well as through child-rearing. These women may be less likely to use birth control and less likely to be involved in stable relationships. In their Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 attempts to fulfill unmet needs, these women may be more likely to get pregnant, to have abortions, and to have children. Women high on social support would have more direct means to meet their needs. The overall finding from this study is that abortion does not have any severe psychological or relational effects for the majority of women on the variables examined. This finding is consistent with that of the APA panel which reviewed the abortion research in 1990 (Adler et al., 1992). The major conclusion from the present study is that abortion has no relevance as a public mental health concern on the variables studied. However, it is still important to consider the impact abortion might have for the statistically insignificant few. Perhaps there were a few women in this study who did suffer psychological or relational consequences to abortion. It could be damaging, or at least ineffective, for mental health workers to expect a woman to have no serious psychological response following abortion. Equally damaging could be expecting that a woman w ho has had an abortion will experience severe psychological effects. Individual differences in abortion experience need to be taken into account. Otherwise, either discounting or pathologizing a woman’s experience of abortion might occur. The results of this study indicate that abortion does not have any severe psychological or relational sequelae on the variables in this study. However, subclinical responses also need to be considered. While psychiatric symptoms may not be a concern for most women, perhaps milder forms of distress may occur. For example, feelings of regret, loss, or sadness may occur. In such cases, these feelings might naturally resolve with time. Nevertheless, sensitive pre- and post abortion counseling might be useful in such situations. Future research might increase its focus on subclinical responses to abortion. It would also be useful to determine how abortion might interact with pre-existing psychiatric problems. This study indicates that the vast majority of women do not experience significant difficulty adjusting to abortion. However, case studies (e.g., Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 61 Cavenar et al., 1978; McCraw, 1989) often report dramatic psychiatric symptoms following abortion. It is likely that pre-existing psychiatric symptoms are involved in each of these case studies. However, no information is available regarding how women with a particular disorder (e.g., schizophrenia) might respond following abortion. This issue might have particular relevance when considering that some women take psychiatric medications which can cause birth defects. If such a woman discovered she was pregnant, she might abort a wanted child and experience a relapse or increase in symptoms. A less dramatic example might be a woman with panic disorder, who is afraid to have a child out of fear of not being able to care for it if in a state of panic. She might choose to have an abortion. While she might feel relief following the abortion, she might also experience an increase in her level of anxiety. Further research would be needed to identify individuals who might have difficulty adjusting to an abortion due to pre-existing mental health problems. The results of this study suggest that abortion is not a public mental health concern. No negative psychological or relational effects for abortion were found on the variables studied. Social Conformity, with the variable of Religiosity in particular, predicted abortion occurrence. Number of Abortions predicted Number of Children. There was some suggestion, though not statistically significant, that abortion might be related to lower agitation. Further research would need to examine this possibility. Future research might also examine how abortion interacts with pre-existing psychiatric conditions. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES Adler, N.E. (1975). Emotional response of women following therapeutic abortion. American Journal of Orthopsychiatry, 45, 446-454. Adler, N.E. (1976). Sample attrition in studies of psychological sequelae of abortion: How great a problem? Journal of Applied Social Psychology. 6(3), 240-259. Adler, N.E.. David. H.P.. Major. B.N., Roth, S.H., Russo, N.F., & Wyatt, G.E. (1990). Psychological responses after abortion. Science. 248. 41-44. Adler, N.E., David, H.P., Major, B.N., Roth, S.H., Russo, N.F., & Wyatt, G.E. (1992). Psychological factors in abortion: A review. American Psychologist. 47( 10), 1194-1204. Aneshensel, C.S., Clark, V.A., & Frerichs, R.R. (1983). Race, ethnicity, and depression: A confirmatory analysis. Journal of Personality and Social Psychology. 44, 385-398. Armsworth, M.W. (1991). Psychological response to abortion. Journal of Counseling & Development. 69. 377-379. Baluk, U. & O’ Neill, P. (1980). Health professionals perceptions of the psychological consequences of abortion. American Journal of Community Psychology. 8(1), 67-75. Barnett, W., Freudenberg, N., & Reinhard, W. (1992). Partnership after induced abortion: A prospective controlled study. Archives of Sexual Behavior. 21.(5), 443-455. Bentler, P.M. (1995). EOS structural equations program manual. Encino, CA: Multivariate Software. Bentler, P.M. ( 1990a). Comparative fit indexes in structural models. Psychological Bulletin. 107. 238-246. Bentler, P.M. ( 1990b). Latent variable structural models for separating specific from general effects. In L. Sechrest, E. Perrin, & J. Bunker (Eds.), Research methodology: Strengthening causal interpretations of nonexpert mental data (pp. 61-83). Rockville, MD: Department of Health and Human Services. Blatt, S.J., Quinlan, D.M., Chevron, E.S., McDonald, C., & Zuroff, D. (1982). Dependency and self-criticism: Psychological dimensions of depression. Journal of Consulting and Clinical Psychology. 50. 113-124. Bracken, M.B. (1978). A causal model of psychosomatic reactions to vacuum aspiration abortion. Social Psychiatry. 13. 135-145. Bracken, M.B., Phil, M., Hachamovitch, M., & Grossman, G. (1974). The decision to abort and psychological sequelae. The Journal of Nervous and Mental Disease. 158(2). 154-162. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 Brewer, C. (1977). Incidence of post-abortion psychosis: A prospective study. British Medical Journal. !, 476-477. Byrne, B.M. (1994). Multiple-group analyses. In Foster. D. S. (Ed.). Structural Equation Modeling with EOS and EOS/Windows, (pp. 159-191). Thousand Oaks, CA: Sage. Butler, C. (1996). Late psychological sequelae of abortion: Questions from a primary care perspective. The Journal of Family Practice. 43(4), 396-401. Castro, F.G., Maddahian, E., Newcomb, M.D., & Bentler, P.M. (1987). A multivariate model of the determinants of cigarette smoking among adolescents. Journal of Health and Social Behavior. 28, 273-289. Cavenar, J., Maltbie, A., & Sullivan, J. (1978). Aftermath of abortion: anniversary reaction and abdominal pam. Bulletin of the Mennineer Clinic. 42, 433-438. Christensen, C.P. (1988). Issues in sex therapy with ethnic and racial minority women. The Haworth Press. Chou, C. P., & Bentler, P.M. (1990). Model modification in covariance structure modeling: A comparison among likelihood ratio, LaGrange Multiplier, and Wald tests. Multivariate Behavioral Research. 25. 115-136. Clare, A.W. & Tyrell, J. (1994). Psychiatric aspects of abortion. Irish Journal of Psychological Medicine. j_[(2). 92-98. Cohen, L. & Roth, S. (1984). Coping with abortion. Journal of Human Stress. Fall. 140-145. Cohen, S. & Wills, T.A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin. 98, 310-357. Congleton, G.K. & Calhoun, L.G. (1993). Post-abortion perceptions: A comparison of self-identified distressed and nondistressed populations. The International Journal of Social Psychology. 39(4), 255-265. Conklin, M.P. & O’ Connor, B.P. (1995). Beliefs about the fetus as a moderator of post-abortion psychological well-being. Journal of Social and Clinical Psychology. 14(1). 76-95. Costa, F., Jessor, R., Donovan, J.E. (1987). Psychosocial correlates and antecedents of abortion: An exploratory study. Population and Environment. 9, 3-22. Cozzarelli, C. (1993). Personality and self-efficacy as predictors of coping with abortion. Journal of Personality and Social Psychology. 65. 1224-1236. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 64 Cozzarelli, C., Karrasch, A., Sumer, N., & Major, B. (1994). The meaning and impact of partner* s accompaniment on women’s adjustment to abortion. Journal of Applied Social Psychology. 24(22), 2028-2056. Cozzarelli, C. & Major, B. (1994). The effects of anti-abortion demonstrators and pro-choice escorts on women’s psychological responses to abortion. Journal of Social and Clinical Psychology. 13(4), 404-427. Dagg, P.K.B. (1991). The psychological sequelae of therapeutic abortion - denied and completed. American Journal of Psychiatry. 148:5. 578-585. David, H.P. (1985). Postabortion and postpartum psychiatric hospitalization. In R. Porter & M. O’Connor (Eds.), Abortion: Medical progress and social implications (Ciba Symposium No. 45; pp. 150-161). London: Pitman. David, H.P. (1994). Reproductive rights and reproductive behavior. American Psychologist. 49(4), 343-349. Dcrogatis, L.R. (1983). SCL-( 10)-R Manual-I. John Hopkins University School of Medicine. Towson, MD: Leonard R. Derogatis. Dugger, K. (1998). Black women and the question of abortion. In Beckman, L.J. & Harvey, S.M. (Eds.), The new civil war: The psychology, culture, and politics of abortion. ( pp. 107-131). Washington DC: American Psychological Association. Dunn, O.J. (1961). Multiple comparisons among means. Journal of the American Statistical Association. 56, 52-64. Erikson, R.C. (1993). Abortion trauma: Application of a conflict model. Pre- & Peri-Natal Psychology Journal. 8( 1), 33-42. Forrest, J., & Henshaw, S. (1993). Providing controversial health care: Abortion services since 1973. Women’s Health Issues. 3(3), 152-157. Freeman, E.W. (1977). Influence of personality attributes on abortion experience. American Journal of Orthopsychiatry. 47(3), 503-513. Greer, H.S., Lai, S., Lewis, S.C., Belsey, E.M., & Beard, R.W. (1976). Psychosocial consequences of therapeutic abortion: King’s termination study III. British Journal of Psychiatry. 128. 74-79. Heise, D.R. (1969). Separating reliability and stability in test-retest correlation. American Sociological Review. 34. 93-101. Henshaw, S.K. (1998). Barriers to access to abortion services. In Beckman, L.J. & Harvey, S.M. (Eds.), The new civil war. The psychology, culture, and politics of abortion, (pp. 61-80). Washington, DC: American Psychological Association. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65 Henshaw, S., & Kost, K. (1996). Abortion patients in 1994-95: Characteristics and contraceptive use. Family Planning Perspectives. 28. 140-147, 158. Honigman, B., Davila, G., & Petersen, J. (1993). Reemergence of self-induced abortions. The Journal of Emergency Medicine. 11. 105-112. Joe, V.C., Jones, R.N., & Ryder, S. (1977). Conservatism, openness to experience and sample bias. Journal of Personality Assessment. 41.(5), 527-531. Lazarus, R.S. & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer. Levenson, R.W. & Gottman, J.M. (1978). Toward the assessment of social competence. Journal of Consulting and Clinical Psychology. 46. 453-462. Lynxwiler, J. & Gay, D. (1994). Reconsidering race differences in abortion attitudes. Social Science Quarterly. 75(1), 67-84. MacCallum, R.C. (1986). Specification searches in covariance structure modeling. Psychological Bulletin. 100. 107-120. Major, B., Cozzarelli, C., Sciacchitano, A.M., Cooper, M.L., Tesla, M., & Mueller, P.M. (1990). Perceived social support, self-efficacy, and adjustment to abortion. Journal of Personality and Social Psychology, 59(3). 452-463. Major, B. & Cozzarelli, C. (1992). Psychosocial predictors of adjustment to abortion. Journal of Social Issues. 48(3). 121-142. Major, B., Mueller, P., & Hildebrandt, K. (1985). Attributions, expectations, and coping with abortion. Journal of Personality and Social Psychology. 48(3), 585-599. Masters, W.H., Johnson, V.E., & Kolodny, R.C. (Eds.). (1992). Birth control and abortion. In Human Sexuality (4th ed.). (dp. 167-177). New York: Harper Collins. McCarthy, W.J. & Newcomb, M.D. Two dimensions of perceived self-efficacy: Cognitive control and behavioral coping ability. In R. Schwarzer (Ed.), Self-efficacv: Thought control of action, pp. 39-63. Washington, DC: Hemisphere Publishing. McCraw, R.K. (1989). Obsessive-compulsive disorder apparently related to abortion. American Journal of Psychotherapy. 43(2), 269-276. McGee, L. & Newcomb, M.D. (1992). General Deviance Syndrome: Expanded hierarchical evaluation at four ages from early adolescence to adulthood. Journal of Consulting and Clinical Psychology. 60, 766-776. Mellinger, G.D., Balter, M.B., Uhlenhuth, E.H., Cisin, I.H., Manheimer, D.I., & Rickets, K. (1983). Evaluating a household survey measure of psychic distress. Psychological Medicine, 13, $37-621 . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66 Melton, G.B. (1987). Legal regulation of adolescent abortion: Unintended effects. American Psychologist. 42(1). 79-83. Mertus, J.A. (1990). Fake abortion clinics: The threat to reproductive self-determination. Women & Health. J 1 6 , 95-113. Miller, W.B. (1992). An empirical study of the psychosocial antecedents and consequences of induced abortion. Journal of Social Issues. 48, 67-93. Miller, W.B., Pasta, D.J., & Dean, C.L. (1998). Testing a model of the consequences of abortion. In The new civil war The psychology, culture, and politics of abortion, (pp.235-267). In Beckman, L.J. & Harvey, S.M. (Eds.), Washington DC: American Psychological Association. Moseley, D.T., Follingstad, D.R., Harley, H., & Heckel, R.V. (1981). Psychological factors that predict reaction to abortion. Journal of Clinical Psychology. 37(2), 276-279. Mueller, P. & Major, B. (1989). Self-blame, self-efficacy, and adjustment to abortion. Journal of Personality and Social Psychology. 57(6). 1059-1068. Newcomb, M.D. (1986). Nuclear attitudes and reactions: Associations with depression, drug use, and quality of life. Journal of Personality and Social Psychology. 50, 906-920. Newcomb, M.D. ( 1990b). What structural modeling techniques can tell us about social support. In I.G. Sarason, B.R. Sarason, & G.R. Pierce (Eds.), Social support: An interactional view (pp. 26-63). New York: Wiley. Newcomb, M.D. (1992). Understanding the multidimensional nature of drug use and abuse: The role of consumption, risk factors, and protective factors. In M.D. Glantz & R. Pickens (Eds.), Vulnerability to drug abuse (pp. 255-297). Washington DC: American Psychological Association. Newcomb, M.D. (1994). Drug Use and Intimate Relationships Among Women and Men: Separating Specific from General Effects in Prospective Data Using Structural Equation Models. Journal of Consulting and Clinical Psychology. 62(3), 463-476. Newcomb, M.D. (1997). Psychosocial predictors and consequences of drug u se: A developmental perspective within a prospective study. Journal of Addictive Diseases. 16(1), 51-86. Newcomb, M.D. & Bentler, P.M. (1986b). Loneliness and social support: a confirmatory hierarchical analysis. Personality and Social Psychology Bulletin. 12. 520-535. Newcomb, M.D. & Bentler, P.M. ( 1988a). Consequences of adolescent drug use: Impact on the lives of voung adults. Newbury Park, CA: Sage. Newcomb, M.D. & Bentler, P.M. ( 1988b). Impact of adolescent drug use and social support on the problems of young adults: A longitudinal study. Journal of Abnormal Psychology. 97. 64-75. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 67 Newcomb, M.D. & Harlow, L.L. (1986). Life events and substance use among adolescents: Mediating effects of perceived loss of control and meaninglessness in life. Journal of Personality and Social Psychology. 51(3). 564-577. Newcomb, M.D., Maddahian, E., Skager, R., & Bentler, P.M. (1987). Substance abuse and psychosocial risk factors among teenagers: Associations with sex, age, ethnicity, and type of school. American Journal of Drug and Alcohol Abuse. 13, 413-433. Newcomb, M.D. & Rickards, S. (1995). Parent drug use problems and adult intimate relations: Associations among community samples of young adult women and men. Journal of Counseling Psychology, 42(2), 141-154. Osofsky, J.D. & Osofsky, H.J. (1972). The psychological reaction of patients to legalized abortion. American Journal of Orthopsychiatry. 42( 1). 48-60. Peppers, L.G. (1987). Grief and elective abortion: Breaking the emotional bond? Omega. 18(1), 1-12. Petrie, K & Chamberlain, K. (1983). Hopelessness and social desirability as moderating variables in predicting suicide behavior. Journal of Consulting and Clinical Psychology. 51, 485-487. Pliner, A.J. & Yates, S. (1992). Psychological and legal issues in minors’ rights to abortion. Journal of Social Issues. 48(3), 203-216. Radloff, L.S. (1977). The CES-D scales: A self-report depression scale for research in the general population. Applied Psychological Measurement. !, 385-401. Robbins, J.M. (1979). Objective versus subjective responses to abortion. Journal of Consulting and Clinical Psychology. 47(5), 994-995. Robbins, J.M. & DeLamater, J.D. (1985). Support from significant others and loneliness following induced abortion. Social Psychiatry. 20. 92-99. Romans-Clarkson, S.E. (1989). Psychological sequelae of induced abortion. Australian and New Zealand Journal of Psychiatry. 23, 555-565. Russo, N.F. & Dabul, A.J. (1997). The relationship of abortion to well-being: Do race and religion make a difference? Professional Psychology: Research and Practice. 28(1), 23-31. Russo, N.F. & Zierk, K.L. (1992). Abortion, childbearing, and women’s well being. Professional Psychology: Research and Practice. 23(4), 269-280. Shainess, N. (1968). Abortion: Social, psychiatric, and psychoanalytic perspectives. New York State Journal of Medicine. December. 3070-3073. Simon, N.M. & Senturia, A.G. (1966). Psychiatric sequelae of Abortion: Review of the Literature, 1935-1964. Archives of General Psychiatry. 15. 378-389. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68 Smith, E.M. (1973). A follow-up study of women who request abortion. American Journal of Orthopsychiatry. 43(4), 574-585. Sorrel, W.E. (1967). Abortion: Its psychodynamic effects. Psvchosomatics. 8. 146-149. Spanier, G.B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. Journal of Marriage and the Family. 38, 15-28. Speckhard, A.C. & Rue, V.M. (1992). Postabortion syndrome: An emerging public health concern. Journal of Social Issues. 48(3). 95-119. Stein, J.A., Newcomb, M.D., & Bentler, P.M. (1993). Differential effects of parent and grandparent drug use on problem behaviors of male and female children. Developmental Psychology. 29, 31-43. Stein, J.A., Newcomb, M.D., & Bentler, P.M. (1994). Psychosocial correlates and predictors of AIDS risk behaviors, abortion, and drug use among a community sample of young adult women. Health Psychology. 1 _ 3 , 308-314. Turell, S.C., Armsworth, M.W., & Gaa, J.P. (1990). Emotional response to abortion: A critical review of the literature. Women & Therapy, 9(4), 49-68. Zabin, L.S., Hirsch, M.B., Emerson, M.R., & Raymond, E. (1989). When urban adolescents choose abortion: Effects on education, psychological status, and subsequent pregnancy. Family Planning Perspectives. 21. 248-255. Zakus, G. & Wilday, S. (1987). Adolescent abortion option. Social Work in Healthcare. 12(4), 77-91. Zolese, G. & Blacker, C.V.R. (1992). The psychological complications of therapeutic abortion. British Journal of Psychiatry. 160. 742-749. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Supervisor self -disclosure as a predictor of supervisee self-disclosure, satisfaction with supervision, and perceptions of supervisor credibility
PDF
Psychological impact of acculturation on Armenians living in the United States
PDF
Science learning: A self-efficacy study in higher education
PDF
The relationship among attachment style, self-esteem, and violence among incarcerated male inmates
PDF
Female parenting styles: associations with psychological distress, interpersonal relations, conscientious-assertiveness, and general deviance
PDF
Concordant and discordant drug use in intimate relationships: A longitudinal study
PDF
The relationship between disordered eating behaviors and coping
PDF
The relationship between reading achievement and Piaget's phase of formal operations for ninth grade students
PDF
Predictors and consequences of drug abuse in adulthood: Associations with psychological distress, intimacy, work adjustment and criminal behavior
PDF
The relationship between psychological expertise and clinical assessment: A comparison of novice and expert performance
PDF
The ecology of the change process: Art, science, and relationship in psychotherapy
PDF
A comparison of values among counseling psychologists characterized as scientist-practitioners, practitioners, or scientists
PDF
Frequency of trainees' nondisclosures to their supervisors as affected by relationship quality and trainee reactance
PDF
Gender role conflict, cultural identity, and self-esteem among African-American men
PDF
Relationality and socio-historical situatedness: A hermeneutic perspective on multicultural counseling
PDF
Sexual harassment on campus: Refinement of a measure of students' perceptions
PDF
Gender and expectancies of negative mood regulation as predictors of coping: A study of the 1993 Malibu firestorm survivors
PDF
Therapists' perceptions of marital counseling for older couples and the relation of perceptions to selected therapist demographics
PDF
A structural model of problem-solving ability, self-efficacy, effort, worry, and achievement in calculus
PDF
The psychosocial determinants of voluntary HIV testing: A utilization of the Health Belief Model for predicting intentions to test and HIV testing behavior among university students
Asset Metadata
Creator
O'Neil, Sharon Helene
(author)
Core Title
The effects of abortion on psychological distress and intimate relationships in a community sample
Degree
Doctor of Philosophy
Degree Program
Education (Counseling Psychology)
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,Psychology, clinical
Language
English
Contributor
Digitized by ProQuest
(provenance)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c17-571112
Unique identifier
UC11351238
Identifier
9987588.pdf (filename),usctheses-c17-571112 (legacy record id)
Legacy Identifier
9987588.pdf
Dmrecord
571112
Document Type
Dissertation
Rights
O'Neil, Sharon Helene
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA