DOCUMENT RESUME ED 397 365 AUTHOR TITLE
PUB DATE NOTE PUB TYPE
EDRS PRICE DESCRIPTORS
CG 027 221 Gastil, Richard W. The Process of Divorce Recovery: A Review of the Research. May 96 69p.; Doctoral Research Paper, Biola University. Doctoral Dissertations (041) Dissertations/Theses Information Analyses (070)
MF01/PC03 Plus Postage. *Adjustment (to Environment); Affective Behavior; Developmental Psychology; *Divorce; *Emotional Development; Emotional Response; Family Problems; Literature Reviews; Psychologicai Patterns Kubler Ross (Elisabeth)
Many researchers have spec61ated over the nature of the divorce recovery process. Is the process similar to Kubler-Ross's stages of grief or does divorce recovery follow a unique process? This paper examines the current body of empirical research in an attempt to answer these questions. From the 91 sources analyzed, it was discovered that most of the literature is either theoretical or clinical. The theoretical literature suggested that recovery from divorce involves a progression through certain time stages: predecision, legal divorce, and postdivorce. While these chronological stages divide the process into specific transition periods, they do not help in identifying the processes of effective adjustment either across or within the time stages. Furthermore, a lack of resources and methodological problems have hampered empirical research, although several process dynamics have been identified, such as the observation that divorcees usually improve over time. However, unlike Kubler-Ross' stages of gr:ef adjustment, no processes or psycho-social stages of divorce recove:y have been confirmed by current research. Even so, researchers have learned more about the role that adjustment factors, such as gender, locus of control, and social involvement, play in a divorce. Future research should try to clarify definitions of adjustment, instrumentation, procedures, and sampling. Contains 42 references. (RJM)
*********************************************************************** Reproductions supplied by EDRS are the best that can be made from the original document. ***********************************************************************
THE PROCESS OF DIVORCE RECOVERY: A REVIEW OF THE RESEARCH
Richard W. Gastil
(°je 10't_Z Cher
Patr cia L. ike,
Mg COPY AVA1LABU 2
THE PROCESS OF DIVORCE RECOVERY: A REVIEW OF THE RESEARCH
A Doctoral Research Paper
Presented to the Faculty of the Rosemead School of Psychology Biola University
In Partial Fulfillment of the Requirements for the Degree Doctor of Psychology
Richard W. Gastil May. 1996
Copyright © 1996 by Richard W. Gastil
THE PROCESS OF DIVORCE RECOVERY: A REVIEW OF THE RESEARCH by
Richard W. Gastil
The general body of research was examined to ascertain whether a process of divorce recovery was identifiable. The research question was whether or not divorce recovery could be seen to follow some process of
phases or stages, such as Kubler-Ross' (1969) stages of grief. From the 91 reference sources collected, it was discovered that most of the literature
in the field is theoretical or clinical. The theoretical views were summarized and 18 empirical articles related to the process of divorce recovery over time were identified and individually critiqued. Numerous
methodological limitations were discovered and explained. Chronological stages of divorce were seen to be most commonly those of predivorce. legal
divorce, and postdivorce. Some psycho-social stages were proposed, but
empirical verification was limited. However, in the research various demographic and psycho-social variables were identified as correlated
with a more effective recovery. Finally, some suggestions for future research are offered.
TABLE OF CONTENTS PAGE
DOCTORAL RESEARCH PAPER
Theoretical Contributions to the Topic
Definitions of Terms
Adjustment or recovery
Locus of control
Psycho-Social Stages of Recovery
Chronological Stages of Recovery
Limitations of the Instrumentation
Transitional Models or Process Research
Empirical Research on the Process
Crosby, Gage, and Raymond, (1983)
Crosby, Lybarger, and Mason, (1986)
Kressel and Deutsch, (1977)
Spanier and Casto, (1979)
Spanier and Thompson, (1984)
Summary of the Process Research
Longitudinal, Stratified, or Retrospective Research iv
Remarriage and Adjustment
Spanier and Furstenberg, (1982)
Saul and Scherman, (1984)
Summary of remarriage and adjustment
Zeiss, Zeiss, and Johnson, (1980)
Bloom and Caldwell, (1981)
Colburn, Lin, and Moore, (1992)
Summary of gender differences
Psychological and Interpersonal Adjustment
Kolevson and Gottlieb, (1982)
Pettit and Bloom, (1984)
Summary of the psychological and interpersonal research
Current State of the Literature
Directions for Future Research
THE PROCESS OF DIVORCE RECOVERY: A REVIEW OF THE RESEARCH
Introduction Divorce is an unfortunate reality which directly or indirectly
affects a large segment of the population of this county. Both divorcees and therapists attempt to understand and facilitate effective adjustment following divorce. Much theoretical speculation has been
made as to the specific nature of this recovery process. For instance, is it similar to Kubler-Ross' (1969) stages of grief? If not, what are the
unique stages of this critical life transition? This paper looks to the current body of empirical research in an attempt to answer these questions. It's purpose is to increase the general body of knowledge regarding the process of divorce recovery.
Theoretical Contributions to the Topic
At best, divorce is difficult. Most often, it becomes a painful and
traumatic life crisis. However, it has been noted that all crises pose opportunities for both danger and opportunity (Kanal. 1975). Either a higher or lower level of functioning will eventually be seen in the
reorganization of the person's life. Adjustment to the disruption of one's marriage offers the divorcing individual new opportunities for personal growth and for the development of an even more satisfying lifestyle (Wiseman, 1975). For instance, Wiseman viewed divorce as an
opportunity to rework unresolved adolescent conflicts, thereby arriving
at new resolutions involving identity and intimacy. He further explained this developmental aspect of divorce as a stage which involves
reexamination of sexual identity. relationships, career choices, role models, and moral and ethical values. Krauss (1979) noted that the adjustment to the crisis of divorce was affected by many variables. Yet
he theorized this transition to have three potential outcomes: new levels of reorganization, strengthened functioning, or psychopathology.
The purpose of this paper is to examine the empirical research in
an attempt to better understand this critical life transition. However, because most of the work in this field has been theoretical, some
background in this regard is important. Self-help books, clinical specialists, and divorce recovery groups are plentiful. Yet even in the
theoretical area, there is much variation. The definition of key terms differs among theorists. Psycho-social models vary in their attempts to describe the process. Some chronological stages lend clarification, but
are not entirely consistent with one another. Therefore, the first section will provide a theoretical overview of the relevant terms, the psychosocial models of recovery, and the chronological stages of recovery. Definitions of Terms
Definitions of terms were scarce and inconsistent in the literature. Therefore, for the purposes of this paper, the following definitions are discussed and clarified. To begin with, what is meant by the process of divorce recovery must be explained. This paper is defining
process as a potential model, normative phases. or stages of transition. This process is theorized to include many variables: these must also be
differentiated. These variables may be demographic (e.g., gender, age,
number of children, length of marriage, initiator status), resources, (e.g., finances), perceptions of the situation, divorce history, stressors, personality structure (e.g., locus of control), and sense of well-being
(Buehler, et al., 1985) Divorce refers to the psycho-social and legal dissolution of the marital relationship. Recovery is generally defined as the adjustment or transition process from divorce to a new life identity.
Terms which require further examination are adjusiment, initiator .status, well-being, and locus otcontrol. Adjustment or Recovery. Adjustment and recovery are terms
which are used synonomously. Seminal research in the field proposed
the theoretical assumption that the index of positive divorce adjustment was remarriage (Goode, 1956). This assumption continued for decades,
without empirical validation using personality measures of adjustment.
It was assumed that the formerly married would need to transcend their divorce grief before courting. This belief implied that divorced
individuals who remained single were still grieving or maladjusted
(Bohannon, 1970; Goode, 1956; Wiseman, 1975). These assumptions persisted without any empirical investigation of the prevalence of remarriage while still in a grieving, or not yet adjusted, stage. In later years, adjustment or recovery began to be seen in more
psychological terms. Kressel and Deutsch (1977) attempted to theoretically identify the criteria for and obstacles to a "constructive divorce" (p. 413). A constructive divorce was described as finishing the
process of psychic divorce. This psychic adjustment included "attitudes and behaviors of the former spouses toward one another, the welfare of
the children, and the level of functioning of each of the e;:-mates as a newly single person" (p. 419). Later. Zeiss, Zeiss and Johnson (1980)
hypothesized that positive adjustment involved a decrease in such
factors as suicidal ideation, tension, depression, anger, fatigue, and general emotional distress: along with an increase in independence,
stability, and generally positive functioning and expectations. To facilitate the assessment of adjustment, Bloom and Caldwell (1981) developed a measure entitled Composite Symptom Checklist. Recovery was measured related to level of fatigue, the feeling of impending
breakdown, inability to accomplish life's demands, feeling tense or
nervous, anxiety, and general ill health and weakness. Given these developments, adjustment or recovery is more
appropriately considered to involve psychological dynamics, rather than being evidenced by remarriage. Kitson and Raschke (1981) offered a
definition which, in general, reflects the current view in the field. They
stated that, while good adjustment is sometimes indicated by remarriage, it seems more appropriate to identify it as "an ability to
develop an identity for oneself that is not tied to the status of being married or the ex-spouse and ability to function adequately in role responsibilities of daily life - home, family, work and leisure time" (p.16).
Initiator_Status_. Initiator status raises another question in regard to definition. Terms of categorization such as initiatornoninitiator, active-passive, divorcer-divorcee, plaintiff-defendant or
presser-pressee have been used. Unfortunately, these terms are not used consistently. Technically, the person who actually filed or
petitioned the court for the divorce would be the plaintiff or divorcer.
However, this is an unclear distinction inasmuch as the actual divorcer may or may not have been the more active one in pressing for, initiating the dissolution, or destroying or leaving the relationship (Crosby, Gage. & Raymond, 1986; Federico, 1979). For the sake of clarity, this paper will attempt to follow what seems the most commonly accepted
definition of initiator status. It will consider the initiator, or presser to be the more active party who wanted the dissolution of the relationship. The individual who actually filed or moved out first will be viewed as
irrelevant. Although more descriptive, this distinction of wanting the divorce also necessitates the adding of a third category of mutuals, which would describe those parties who equally desired the separation.
WelLbeing Another general area dealt with in this review is well-
being. Different instruments, usually self-report, are used to measure well-being. Some variables included in such measures are life
satisfaction, satisfaction with health, self-esteem, affective balance, psychosomatic symptoms, and consistency of habits. Well-being has been a common variable in the general field of psycho-social research,
and it has recently been utilized in the context of divorce (Spanier & Furstenberg, 1982).
Locus_of_Controt Locus of control is generally seen as an
internal personality factor relating to internal versus external control of one's life. Internals have a greater tendency to contemplate and struggle internally with the consequences of their decisions before acting. Externals are generally seen as more likely to respond to external
stimuli, and to adjust to the internal consequences later. Rotter's
Locus of Control provides one common measure of this internal versus external categorization (Mirels, 1970).
Another locus of control variable is actually more closely
correlated with what has been termed initiator status. One study (Barnet. 1990) has termed this variable the MaritaLLocuaDICantrol.
Marital locus of control is evident throughout the marriage, and is not solely related to whom was most active in the divorce. It includes identifying who initiated or took certain actions entering, during and exiting the marriage. Psycho-Social Stages of Recovery Models of divorce recovery have theorized several different
processes. They are based on grief, sociological, or psychological
models. A brief overview will provide a theoretical introduction to the process of divorce recovery. Waller (1938) proposed one of the original
psychological transition models. He suggested four stages that included: (a) breaking old habits, (b) beginning of reconstruction of life,
(c) seeking new love objects, and (d) readjustment completed. These
stages included an inherent assumption, typical of most early theorists,
that adjustment to divorce involved remarriage. One of the simplest has been Weiss' (1976) model: Phase I: Transition and Phase 2: Recovery.
Another early theorist was Paul Bohannon (1970), who saw the
divorce transition process as including six stations: (a) the emotional divorce transition involves the disengagement from the marital connection: (b) the legal divorce or dissolution of the marriage includes
the legal agreements and arrangements necessary to the process; (c) the
economic divorce transition includes potential changes in financial
responsibilities, employment or housing; (d) the co-parental divorce
transition is additionally present when raising children necessitates renegotiation of responsibilities and values in this area; (e) the community divorce transition involves changes in friends, in-laws, and
social relationships; and (f) the psychic divorce transition relates to regaining individual autonomy and establishing new attitudes and expectations. Kaslow (1984) saw merit in Bohannon's stages from a clinical viewpoint and expanded upon them by assigning to each stage a
series of personal actions, adjustment tasks, and potential therapeutic interventions. The most commonly theorized position has proposed the psychological components of the divorce recovery process to be the same or similar to Kubler-Ross' five stages of grief (Crosby, Gage, & Raymond,
1983; Herman, 1974). This model includes the transition phases of denial, anger. bargaining, depression and acceptance. It may be argued
that these transitions may be more cyclical than linear, and that certain phases may be specific to given aspects of the divorce process.
In other words, an individual may be in the anger phase in regards to
the financial negotiations, while in the acceptance stage regarding the dissolution of the marriage. Kubler-Ross' (1969) model has provided a viable beginning to understanding the process of divorce, yet her stages are not universally accepted. Wiseman (1975) identified many similarities between Kubler-
Ross's stages and divorce recovery, but introduced some variation with
regard to the rejection of the lost object and the rebuilding of a new life. These modifications resulted in Wiseman's proposed stages: (a) denial,
(b) loss and depression, (c) anger and ambivalence. (d) reorientation of life-style and identity, and (e) acceptance and new level of functioning. Krantz ler (1975) and Weiss (1975) reduced Kubler-Ross' model to three
"classical stages of the divorce process" (Kalb, 1987, P. 54). These stages
are: Stage 1feelings of shock and denial; Stage 2feelings of anger, resentment, bitterness, lowered feelings of self-esteem, withdrawal, and
distancing; and Stage 3feelings of acceptance, adjustment, and comfort in being single. Steefel (1992) conducted theoretical research
and concluded that although the number of stages may vary, the process follows a similar pattern to Kubler-Ross'. She proposed a modified four stage process of (a) shock and denial, (b) anger and depression from the acceptance of reality, (c) testing of new behaviors
and interpretations, and (d) establishment of new actions and attitudes. Steefel noted that moving through these stages involves emotional and behavioral change and often chaos. Utilizing a series of unstructured interviews, Vaughan (1979)
observed the most significant phenomenon to be not chaos, but an underlying order. The observable and subjective experience seemed
disorderly and chaotic, yet from her interviews, she hypothesized a
consistent process which she called uncoupling. Hers was a more sociological, or psycho-social model. She attempted to summarize her involved view of this process as follows:
Beginning within the intimacy of the dyad, the initial objectivation occurs as the secret of the troubled marriage that the initiator has held is shared with the significant other. With this, the meaning has begun to move from the subjective to the objective. Definition negotiation begins. While attempting to negotiate a common definition, the initiator acts to increase the validation of his identity and place in the world by the use of
accompanying reconstructions of reality. The autonomy of the initiator increases as he finds self-validation outside the marriage, and an ideology that supports the uncoupling. The increased autonomy of the initiator brings the significant other to accept a definition of the marriage as troubled, and they enter into the stage of trying. The process continues as counseling and separation further move the new definition into the public sphere. The telling of, the symbolic physical signs of the uncoupling, and the initiation of formal legal proceedings validate the increasing separation of the partners as they negotiate a new reality which is different from that constructed private sphere which validated their identity as a couple. Eventually, a redefinition of the mutual identity occurs in such a way that the joint biography is separated into two separate autonomous identities. (pp. 447-448) Vaughan further emphasized in this process the necessity of language or
conversation to social reconstruction and validation. A final observation she made was that, in even the most conflict-ridden of divorces, there was some ongoing level of caring.
Another psycho-social model has been presented by Ahrons (1980, 1983), who adopted a family systems approach that viewed the process
of divorce recovery as five transitions: Transition 1individual cognition (where external roles of the family remain intact, but an emotional
divorce is in process); Transition 2family metacognition (where the family as a whole begins to share the realization that the marriage is
disintegrating); Transition 3separation (the transition of one partner physically moving out of the relationship); Transition 4family reorganization (boundaries and child rearing responsibilities are
renegotiated); and Transition 5family redefinition (possibilities of single parenting or blended families are introduced).
The above models suggest that much speculation has occurred regarding the process of divorce recovery. Because these proposed
models have been theoretical or clinical in nature, further empirical
validation is necessary and appropriate. However, these theoretical views provide important background for a review of the empirical
Chronological Stagesofileravery Stages or phases in divorce are terms which refer to very different
phenomenon. The previous section addressed the psycho-social stages
that have been introduced in the literature. Chronological stages have also been proposed, which suggest that divorce is a process which,
regardless of the emotional or social adjustments, progresses through
some phases that can be identified in the context of time. Spanier and Casto (1979) defined two separate, but overlapping,
adjustments: (a) the dissolution of the marriage, and (b) setting up a new life-style. Bloom and Caldwell (1981) also divided the process into
the two stages of: (a) preseparation and (b) postseparation. Others have expanded this time transition to include as many as four phases. Kressel and Deutsch (1977) proposed these to be: (a) the pre-divorce decision period. (b) the decision period, (c) the period of mourning, and (d) the period of reequilibrium. However, the most common chronological framework divides the
divorce process into three stages. Crosby. Gage, and Raymond (1983)
describe these as (a) first serious thought to separation, (b) separation to final decree, and (c) final decree to penultimate closure. Kaslow (1984). in a similar formulation, defined them as (a) predivorce (a time of disequilibrium and despair); (b) during divorce (a time of legal
involvement); and (c) postdivorce (a time of exploration and
reequilibrium). Barnet (1990) reduced them to (a) predecision. (b)
divorce proper, and (c) postdivorce. These three time periods seem to have face validity regarding the chronological process of divorce recovery.
Yet, these time periods raise additional questions with regard to understanding divorce as process. Are the psycho-social adjustments different for each time period? Are the transition phases for each time
period similar, yet with different adjustment tasks? To date, there has been relatively little empirical research to challenge or validate the
theoretical or clinical constructs proposed in the literature. Therefore, it is important to examine what has been done empirically to validate the proposed models and stages of divorce recovery. Prior ,o a review of
the empirical literature, methodological issues will be addressed. Methodological Considerations In the relatively underexplored area of divorce recovery research,
several methodological concerns were consistent. To establish a proper
context from which the reader may understand this research, it becomes
necessary to also address its limitations. Rather than repeat general concerns inherent in each study, some consistently problematic issues will be summarized in this section. It should be remembered, however,
that in fields of research such as divorce, strict methodological standards are virtually impossible. Even if unlimited resources were available, to attempt to control for all confounding variables would
destroy the realism and ultimately unpredictable nature of the human experience. Therefore, each new study is received with gratitude for the
increased understanding it attempts to offer. Nevertheless, the reader
needs to consider the general limitations of the sampling,
instrumentation, procedures, and statistics. Sampling_Limitations
In the study of divorce recovery, one methodological limitation
involves the restrictions and availability of the subjects studied. By definition, the lives of divorcing individuals are in crisis and transition. Willingness to take on any new projects is low. Personal survival is often viewed as overwhelming. Children, financial concerns, daily living
and personal direction may already lack the attention they need, therefore, devoting attention to a research project may not be a priority. And even for those willing to participate in a study, divorce often means relocation, and tracking of subjects becomes difficult.
As a result, convenience samples were often chosen. Subjects were found based on their availability, and often were contacted through
singles or parent groups, or through clinical settings. Yet, involvement in such groups demands specific social skills, functioning ability,
personal need, and time. As a result, such samples describe only this type of divorcee, and cannot automatically be assumed to reflect the general population or overall process of divorce. However, better
sampling sources are readily available. Because divorce documents are
a matter of public record, the overall population is identifiable in any given county.
Yet, the drawing of representative samples from county records
still has difficulties. There is a cyclical variation in the number of divorces filed each month (Kitson & Raschke, 1981). With this
accounted for, another bias is introduced based on those willing to
participate. In fact, divorce studies rarely surpass 50% participation from subjects originally. identified. One study (Goode, 1956), attempted
to counteract this difficulty by having an interviewer show up
unannounced at the subject's door. Although coming closer to a random sample, the sometimes annoyed refusals raised the same
question. It should further be remembered that each city or county in which the study was done may have unique characteristics that must be considered. How the characteristics of any given study's sample may have compared to national means was only examined twice in the research reviewed (Goode, 1956; Kolevson & Gottlieb, 1982). Therefore,
results should be accepted with the recognition that their ability to generalize beyond the sampled population is often unknown.
Also, most of the studies tended not to evaluate the effects of the
heterogeneity of their samples. In other words, although the demographics of a given sample (e.g., age, gender, time since divorce,
time married, number of children) may be descriptively reported, their
effect on the results is often not analyzed or controlled for. Exceptions to this are either obvious, or noted in the current review. They will be obvious by the inherent design of some studies (e.g., studies involving
women under 38 years old with children). Such studies have more descriptive power, but less ability to generalize to divorce as a whole.
Other exceptions will be noted in the few studies that performed factor analyses of the effects of the different demographic variables. In summary, in all studies reviewed, the above sampling limitations should be kept in mind.
It is difficult to objectively and accurately identify the dependent
variables of interest in this field (e.g., adjustment). In all cases, some form of self-report is used. Either the subject described his or her experience, or, in one instance, therapists offered their opinions (Kressel
& Deutsch. 1977). Such self report has some inherent limitations. Assuming subjects were telling the truth. it was still the individual's
subjective perspective, as he or she understood it at that point, and as much as they were willing to disclose it. In the many studies which utilized retrospective self report. these issues are compounded by the question. "Do they remember?" How someone was feeling a year ago can
easily be either forgotten or reframed in order to cope. Also, to correlate or differentiate based on subjective data may create inequitable
comparisons. Gender or cultural biases are good examples. In other words, did the men's scaling of distress or anxiety create the same numerical variable as the women's? Nevertheless. although jectivity
would be ideal, it is seldom possible. and self report is sta dard in this type of research
The above limitations are compounded by the general lack of
standardization of the measures used. Without standardization, the reliability and validity of the results are questionable. All that is demonstrated is face validity. In most cases, measures which suited the purposes were not available. Most researchers developed their own untested questionnaires. When previously tested instruments were employed, they were usually shortened to allow a better response rate
from mailings. And when mailings were not used, new potential biases
may have been introduced by the style or expertise of the interviewer.
Additionally, the potential demand characteristics of being studied
could seldom be eliminated. Unfortunately, in published studies, few of the researchers discussed limitations arising from either sampling or
instrumentation. Zeiss, Zeiss, and Johnson (1980) summarized the
situation well by stating that "in the strictest sense, research in this field should probably be viewed as presenting information about the covariation of self-reporting variables in the divorce adjustment of those who are most geographically stable and most willing to disclose information" (p. 32).
Some general procedural limitations also need to be noted. Most
studies reported data collected at a single point. In the area of adjustment, longitudinal data would have been preferable and was acquired in some studies. Yet, tracking over time is difficult, especially
given this transient population. By and large, the resources for this more preferable design simply weren't available. Yet this raises another procedural difficulty: the lack of any base information, or pre-tests. In other words, subjects' relative level of adjustment prior to the study is
unknown. In addition, subsamples. unless noted otherwise, were neither matched to one another, nor necessarily representative r.f the general population.
A final consideration is the lack of control groups. This component was often not practical. and identifying a conceptually suitable control becomes difficult. Kitson and Raschke (1981) suggested drawing a matched sample of married persons in order to compare and
contrast the characteristics of the two groups. However, it becomes
questionable in the area of adjustment whether it is appropriate to match divorced persons with marrieds. Might divorced persons be better
compared with singles? If so, where shall a control match for those with children be found? Although numerous procedural limitations are probably unavoidable, they should nevertheless be considered in
assessing the research.
Statistical Limitations In studying divorce recovery, usually with nonstandardized
measures, the data itself raises statistical difficulties. Even when numerical values can be assigned, non-standardized subjective reporting raises questions as to the equivalence of the numbers between subjects.
Thus the data itself may confound the statistical manipulation. Given this limitation, the power of the statistics in this field of literature is
also limited. And more often than not, the statistical methods employed are reported only minimally in the published articles. This
leaves the reader with the two options of either assuming that the study utilized appropriate statistical methods, or rejecting the conclusions of
he study based on a lack of statistical documentation. Because most of the source publications in this field seem to be more clinically than scientifically inclined, concerns are increased regarding the soundness
of the scientific methodology. Finally, it should be remembered that, at
best, these studies can infer correlated relationships, not causal relationships. In this paper, specific statistical values are reported when available, but this information was not always the case.
Transitional Models or Process Research
Initially, literature searches for this paper centered on transitional models of, or the processes of, divorce recovery.
Unfortunately, as stated earlier, this uncovered mostly theoretical papers. Therefore, extensions of the search were conducted in an effort to identify empirical studies that might have application to the process of transition after divorce. A total of 18 studies were deemed to provide
relevant to the current research question. The first ones to be examined will be those directly focused on a process or model of divorce recovery.
Empirical Research on the_Process Out of 91 references eventually collected and reviewed, there were
only five empirical studies that directly attempted to answer the
question asked by this paper. The first two studies, chaired by John Crosby (1983, 1986) at the University of Kentucky, attempted to identify
the stages, or the process, of divorce recovery. In the third, Kressel and Deutsch (1977) interviewed 21 therapists in New York and Boston in an
attempt to identify a divorce recovery process. Finally, Spanier and
Casto (1979) studied 50 cases and results were reported through an initial journal article and a later book (Spanier & Thompson, 1984). These studies all had significant methodological limitations, yet their findings did contribute some interesting observations to an understanding of the process of divorce. Their shortcomings provided
insight into possible design changes for future research. Crosby._Gage. andRaymonc1J198.3). This pilot study, conducted
from 1977 to 1980, was one of the few that sought directly to find a process, or identify psychological stages, in divorce. It's research
question asked was whether there was a sequence of phases associated with the grief work of divorce that would resemble the stages of grief
postulated by Kubler-Ross (1969). The authors hypothesized that it would. They expected to find "a sequence of feelings and behaviors
which progress in general sequence from denial to anger to bargaining to depression to acceptance" (p. 8). Although precisely the kind of
question most relevant to this paper, the methodology of this pilot project had several limitations. The sample consisted of 17 divorced
persons with no restrictions placed on demographic characteristics such as age, race, socioeconomic status, educational level, or religious preference. And variables regarding sex of respondent, presence of children, or length of marriage were simply combined. These potentially
confounding variables and the relatively small size of the sample call for
initial caution in interpretation of the results. The instrumentation of this study added additional concerns. Although lack of standardization seems relatively common in this area
of research, the instrument developed in this case was not able to
isolate the variables proposed in the hypothesis. Each subject was asked to write an essay, following twelve guideline questions, describing
the feelings and experiences of their divorce. Then two of the researchers independently grouped feelings, cognitions, and behaviors by
a frequency count. No actual statistics were performed or significance verified, but rather trends of the findings were interpreted. The authors proposed many interesting theories about the process of divorce and guardedly referred to potential similarities with the model of Ku bier-
Ross. However, the numerous design limitations of this study seemed
to limit its ability to offer much assistance with the original question. Nevertheless, as pilot research, it did provide useful information for the design and instrumentation of the follow-up study. S. e"
: 6). In this follow-up study,
Crosby was joined by Lybarger and Mason and they conducted their research from 1982 to 1985. Unfortunately, all reference to Kubler-
Ross' stag -s were dropped. In fact, no specific research hypothesis was reported. Their sample was greatly expanded (N = 141), and new
instrumentation was developed. The increase in sample size would be assumed to improve its power, yet some limitations might still be noted. The sample's continued demographic variability was not controlled for cm examined, making its effect unclear. Also, the sample was all
Caucasian, heavily Protestant (N = 98). and largely female (N = 108) conditions which may impose limitations on the generalizability of the
findings. The pilot's instrumentation was redesigned by the creation of a self-report questionnaire which could better capture the information desired. Typical concerns about the accuracy of self-report were compounded by applicants attempting to recall cognitions, affects and
behaviors for three distinct time periods. However, the authors' did recognize that reliability and validity questions regarding their results remained unanswered.
The methodology and interpretation of this study was designed around a new 3 x 3 x 3 taxonomy, which had been developed from the
pilot study. It presented a matrix of three chronological stages (predivorce, during divorce. postdivorce); three filing categories (actives,
passives, mutuals); and three variables (affect, cognition, behavior).
One group of responses was used to classify subjects as actives (N = 56),
passives (N = 42) or mutuals (N = 43). Another section of questions involved picking from a predetermined listing of potential affects,
cognitions. and feelings experienced during each of the time periods.
The statistical method that was added since the last study was a chi-square. with variables reported which reached either a .05 or .001 significance level. The chi-square was most likely chosen due to the
nominal nature of the data, yet the statistical design also raises some additional questions which were not addressed. The 3 x 3 x 3 taxonomy divided the sample into 36 subcategories, thus reducing the power of the
sample size. Using the chi-square, some variables showed statistical
significance when as little as four to seven respondents had chosen that characteristic. The 36 subcategories, each contained three or four significant characteristics, also created difficult summarization problems.
Given these limitations, the results of this research can be summarized as follows. For active agents. Stage 1 (from the first awareness of problems to a decision for divorce) showed significant
characteristics of fear, hurt, anger, approach. avoidance, and reconstruction. Actives in Stage 2 (from separation or filing to final decree) identified hope, fear, acceptance, approach. and reconstruction. And actives in Stage 3 (from final decree to new life) saw hope, apathy,
acceptance, analysis, reconstruction, or acceptance. Passive agents in
Stage I reported hurt, apathy, avoidance, approach, and negotiation. In Stage 2, passives experienced fear, hurt, acceptance, ambivalence,
negotiation, and acceptance. In Stage 3. they reported hope, hurt, 0 h"
acceptance, analysis, and reconstruction. Mutual agents in Stage 1 reported fear, hurt, approach, acceptance, and negotiation. Mutuals in Stage 2 showed hope, fear, acceptance, analysis, and reconstruction.
Stage 3 mutuals reported hope, acceptance, analysis and reconstruction.
From these results, it seems useful to note which significant feelings were identified by respondents in each of the chronological
stages. It could be summarized that active and mutual agents arrived at acceptance and hope sooner than passives. Passives felt more hurt throughout, whereas actives and mutuals felt more fear. Yet by Stage 3; hope, acceptance, and reconstruction were the highest ranking affective, cognitive, and behavioral categories for all three initiator groups. All subcategories tended to move on to establish a new lifestyle over time.
liresseland_Dentachalaza This study involved semistructured, in-depth interviews with 21 therapists. all of whom had a specific specialty and at least five years experience in working with divorcing
patients. It attempted to focus "on the process set in motion when a marriage is being terminated, rather than on the causes of the divorce" (p. 414). However, the methodology of the study did present some
limitations. The design of the instrumentation did not include the categorization of specific variables. By nature of the interviews, concepts were rarely articulated in consistent ways, leaving it to the researcher to identify and classify relevant themes. Although the reporting was done by professionals in this field, perspectives varied
greatly and all reports were third party observations. No statistical analysis was performed to determine the significance of the findings.
Nevertheless, given these cautions, some of their observations were
relevant to the transitional theme of this paper. The researchers posited a "process of psychic divorce" (p. 416)
consisting of four stages. each with specific characteristics. The first was the pre-divorce decision period, including marital dissatisfaction,
failed attempts at reconciliation, decline in intimacy, and a breakdown in the facade of solidarity. The second was the period after the decision
to divorce had been firmly made. It included anxiety, panic, attempts to renew marital intimacy, and renewed fighting that reinforced the acceptance of the inevitability of divorce. The third period was the period of mourning, seen as the most complex and critical of the stages. It involves feelings of guilt, self-reproach. anger, negotiation, and
eventually acceptance. The final stage was called the period of reequilibrium. It was reported to involve a period of heightened self owth and diminished dwelling on the marriage: Transitional factors seen to improve this recovery process included
a willingness to negotiate realistic new attitudes and behaviors toward the ex-spouse; and avoidance of strong, unrelenting feelings of self-
disparagement or failure. Obstacles to recovery were cited as an unequal acceptance or motivation for the divorce, and the involvement
of third parties. These third party involvements included a new romantic involvement by either of the parties, or an overly adversarial
conditions escalated by attorneys, or both. Spanies_and_Casto. (19791 This study, conducted in 1976,
utilized a sample drawn from the Pennsylvania court records of 50 people who had filed for divorce within the previous two years. 0 II
Unstructured interviews were conducted with a focus on the process and problems of the post-separation period. Demographic differences of the respondents were reported, and some analysis was done of the
statistically significant differences. The four hypotheses of the study examined (a) the effects of lingering attachment to the former spouse. (b) the degree of social interaction outside the home, (c) the role of
dating relationships, and (d) the relative effects of sudden and
unexpected separations. By virtue of the reporting format of the results,
it can be assumed that a chi square was again utilized to identify reported factors as statistically significant. Methodological concerns arise in this study relating to the validity and reliability of a
retrospective, self-report instrument. Another instrumentation issue was the accuracy of the authors' interpretation of the unstructured interviews. Nevertheless, the results do shed some light on the hypothesized questions.
The results related to the four hypotheses as follows. Attachment to the ex-spouse. as assessed from the interviews, was not found to provide statistically significant differences in distress levels. The more
social interaction the separated individual had (with relatives, friends, and the community), the fewer were the reported adjustment difficulties, X2(1, N = 50) = 6.93, p < .01. Separated individuals who participated in
heterosexual dating or cohabitation relationships had fewer adjustment problems than those who did not, X2(1, N = 50) = 10.28, p < .01. The
degree to which the separation was sudden and unexpected was positively related to the d( gree of initial emotional problems, X2(2, N
50) = 17.5, p < .001. The study also claimed to have found, without
reporting statistical significance, that creating a new life-style was more
important, and more difficult, than was dealing with the dissolution of the marriage. Spanier_anthThompaon,119841. This book, published in 1984,
examined additional variables from the 50 unstructured interviews done
in 1976. Significance was established at p < .05, yet statistical results were not reported in the same detail as the above article. However, the book reported additional conclusions of relevance to this review.
The results indicated that the women were more likely than the men to express either distress or relief. Women also showed significantly poorer adjustment related to blaming of the spouse or an outsider. Men or women who were distressed were the least accepting of the breakup. while those who were relieved were the most accepting.
Loneliness was also positively correlated with distress and negatively
correlated with relief. Dating was reported as significantly helpful in
social adjustment. Yet, dating was not reported as being connected with well-being. Nevertheless, remarriage or having definite plans to remarry were reported to be connected with greater levels of well-being. However, these reports on remarriage were later contradicted by
Spanier's more extensive study (Spanier & Furstenberg. 1982). which is reviewed in the next section.
aummary_ofiheTrocess Research In his second project, Crosby (1986) identified transitional
characteristics for each of the three stages of divorce he specified. He also discovered slight differences in the negotiation of these stages
among actives, passives, or mutuals. This type of process, or stage.
research is the focus of this review. Unfortunately. his taxonomy has not been replicated in other studies. Kressel and Deutsch (1977) added to the understanding of the process by proposing four stages of recovery.
Spanier, Casto. and Thompson (1979, 1984) showed that certain social
and interpersonal factors effected the adjustment of their subjects. However, additional research was sought in an effort to confirm, differentiate and generalize the nature of the process of divorce
transition. In order to find further empirical support for such processes, it was necessary to look anew into the current body of research. Unfortunately, most of the research involved heterogeneous samples
surveyed a particular point. In other words, characteristics of particular samples were reported without differentiation as to where subjects fell chronologically following their divorce. As this did not allow
examination of the process, these studies were not helpful to the
current research question. Furthermore, although the divorce process conceptually includes the preseparation or decision to divorce period;
studies from this time period were considered also to be outside the scope of this review because their results identify the reasons why marriages fail and not the process of divorce recovery. However, further
review of the literature eventually differentiated those empirical studies with relevance to divorce as a process.
Longitudinal, Stratified, or Retrospective Research
Studies examining the characteristics of divorcing persons over time provided additional empirical evidence as to the divorce process.
Although the main purpose of each study may not have been examining
stages of transition, the consideration of effects at different time periods provided additional insight into the divorce process. This included three types of methodological designs. They were (a) retrospective reports on
different time periods. (b) stratification of subjects by time since divorce,
or (c) actual longitudinal analysis. This criteria resulted in 13 additional articles for review. These have been organized in three categories based on either the major variables studied, or relevance. The categories are (a) remarriage and adjustment, (b) gender differences, and (c) psychological and
interpersonal adjustment. Under each of these headings, research is presented chronologically. However, several studies yielded results
covering more than one of the above categories Within the above organization, studies will be presented and critiqued on an individual
basis. In this way. results can be considered in light of the
methodological limitations of each. Yet this necessitates that, at times, certain results will be included under each study that may not fit exactly under these four organizational headings.
RemarriageandAdjustment A question debated for several decades is what effects remarriage
has on the process of adjustment. Some have theoretically argued that psychological adjustment is uncorrelated to remarriage. Others have
assumed that to be married again was the goal of the adjustment process. Three studies provided direct relevance to this question and are
presented below. In addition, their results provided information on the effects of cohabitation, dating and general well-being. The first of these
studies also provided important baseline considerations for the general development of research in this field.
Goode, (1956) Beginning in 1948, William Goode was the first to
collect in-depth empirical data on the process of divorce adjustment. His seminal findings were published in book form in 1956 and remained
the most extensive available for several decades. Regrettably, his work
was limited to divorced mothers, age 38 or less. He did not examine the divorce process for older women, women without children, or males. In order to study the process of divorce, his subjects (N = 425) were
stratified into four time groups ranging from 2 months to 26 months. He took great care to eliminate sampling bias, although the population was limited to Wayne County, Michigan. His researchers collected unstructured responses in lengthy interviews, guided by a 26 page
schedule of questions. Although his measure was not standardized, it was deemed the best available for the research purposes.
His lengthy section on the necessity and limitations of the methodology offered guidance to much of the subsequent work in the
field. He concluded that the only complete list of the divorced
population exists in county courthouses, and that random sampling from this population could be considered a defensible plan. He recognized a substantial form of bias to be self-selection, which is
generally unavoidable given the factor of choice to participate. In an
attempt to avoid this, his interviewers approached each subject without prior permission or advanced knowledge of the study. However, his refusals raised fewer, but some of the same, difficulties in obtaining a truly random sample. 3 zi
He also examined the issue of whether accurate data could be
collected based on the necessity of respondents reporting the truth. He concluded from previous social research that, in an extended interview with empathic professionals, very few people deliberately and
consistently lied about themselves. Nevertheless, the question of whether they would actually know the truth remained an issue. Adjustment and personality factors demand some repression or
subjective reframing of personal accounts. In this regard, it was recognized that gaps would exist without direct observation or
interviewing of the ex-partner. Yet, the position was taken that "we are
trying to investigate the adjustment which the spouse made to the reality as she saw it and sees it" (Goode, 1956, p. 25). This adjustment of subjects to their own reality is another research position which has
remained true in all subsequent studies. Several of his results have baseline relevance to the continued
study of the process of divorce adjustment. Prior to this study, it had commonly been thought that higher status occupational groups had higher divorce adjustment rates. Goode (1956) found that occupational
status and level of education had little effect on trauma, distress, or adjustment. However, income was related to adjustment. The higher the level of income, the better the adjustment and the lower the levels of distress. Higher social participation was also significantly correlated
with adjustment. He further discovered that the least trauma occurred when the decision to divorce was a mutual one and that noninitiators were likely to be the most highly traumatized. Other significant results showed that the divorce experience can lead to problematic behaviors
such as sleeping difficulties, physical and health related concerns, low work efficiency, and memory difficulties. His subjects confirmed the
generally accepted hypothesis that divorce poses a significant threat to an individuals' well-being.
He further discovered that the option of divorce was almost
universally recognized as a regrettable necessity. The actual decision to divorce was generally a slow, reluctant one, with the greatest trauma
occurring at the time of the separation. It was suggested that the intensity of emotional stress is directly affected by time in general. and by the specific phase of divorce in which an individual finds him or
herself. An unfortunate lack in his reporting was that, although levels of statistical significance were established, only descriptive percentages
were reported in most cases. Finally, it is important to note that Goode's main index of positive adjustment was more of a societal or
theoretical assumption. He assumed adjustment to be remarriage, and organized his research and reporting accordingly. This position would be challenged in later years.
Spanier_ansLEursienberg._(1982) Data from a larger longitudinal
study of divorce and remarriage was examined to ascertain whether remarriage appeared to be helpful in enhancing one's self esteem following divorce. Subjects (N = 181) were initially identified through
nonprobability. purposive sampling of court records in Pennsylvania. Of the persons initially identified, an impressive 61% were included in the
final sample. Structured interviews were conducted lasting an average
of 2 hours and 15 minutes. Included in the interview structure were three widely accepted measures of well-being, which yielded several
adjustment variables. These included life satisfaction, satisfaction with health, self esteem, affective balance, psychosomatic symptoms, and
changes in habits. Initial measures (Time 1) were repeated 2.5 years later (Time 2). Typical methodological concerns apply. Nevertheless, the
study's relatively large sample size and its longitudinal data collection
allowed these results to shed valuable light on the process of
adjustment. The extent of remarriage, time until remarriage, and percentage
cohabitating was remarkably similar to general population statistics. In transitioning from divorce to remarriage, no significant differences were
found at Time 2 in regards to age, income, occupational status, education, religion, religiosity, gender, or children. One factor was
significantthose who initiated the divorce were more likely to be remarried at Time 2. (p < .05). It was further discovered that subjects in
general showed an enhanced sense of well being at Time 2, with
demographics variables having no significance. This reflected the commonly held notion that, relatively speaking. divorcees get better over
time. In regards to remarriage and well-being, again no significant
correlation was found. In other words, adjustment was not related to remarriage. Neither was it related to cohabitation. Understandably, the only exception to this was the variable change_oLhabits (p < .05).
Although not hypothesized, other significant points of interest were discovered. Those who reported a higher life satisfaction at Time 1 were more likely to be remarried at Time 2, X2(2, N = 181) = 11.4,
p < .01. Therefore, although well-being did not appear to change with remarriage, the likelihood that one would remarry was related to a
positive sense of well-being during the postdivorce period. Furthermore,
among those who remarried, the quality of the remarriage was also positively related to well-being. However, given the self reporting nature
of the data, the results may not be entirely comparable. It might be woadered whether divorced and remarried persons used differing
reference points to judge their current well-being. Regardless, the
results offer evidence to dispel the old notion that adjustment is related to remarriage (Goode, 1956).
Saul and Scherman (1984). This study also hypothesized that cultural factors had shifted and, contrary to results from thirty years ago, predicted that there would be no significant differences on measures of grief or adjustment between those who remarried and those who remained single. The subjects (N = 144) were volunteers from
various social organizations. To eliminate possible variations from heterogeneity, all were between the ages of 25 to 35 and none had
undergone any intensive therapy. Yet, the power of the sample size was diluted significantly by the creation of twelve subsamples of between 8
to 10 subjects each. Subsample assignment was based on: 6-18 months since divorce, 18-36 months since divorce, single, remarried, males with
no children, females with children, and females with no children. Means and standard deviations were examined for each of the twelve groups related to several dependent variables. MANOVAs were performed
(p = .05) to analyze main and interactive effects. The instrumentation included previously established measures, deemed to possess adequate validity. The dependent variables gathered from the measures included
demographics, self worth, emotional disengagement from the ex-spouse,
anger toward the ex, grief work, social trust, and behavioral adjustment. Some cautions were necessary when interpreting the results of
this study. Most significant among them was the limitation of the sample population to subjects who were sufficiently socialized as to be
involved in singles' organizations and be willing to volunteer. In light of these, the results did offer some interesting conclusions for
consideration. There was no statistical significance on measures of divorce grief, personal adjustment, time since divorce, or demographics.
Furthermore, no differences on main or interactive effects surfaced when
subjected to separate analysis of variance. In other words, contrary to earlier assumptions, remarriage did not effect adjustment. However, neither did the presence of children, time, or gender. These results confirmed the hypotheses of the researchers, and provide additional
insight into the divorce adjustment process. Summary_DLRemartiage_ancLAdjustment The above studies
tended to confirm the currently held societal assumption that adjustment to divorce does not seem to be related to remarriage. It also introduced the question of whether gender, time, or other variables were
major factors. These variables will be examined further through the other studies reviewed. Gender Differences
For both clinicians and divorcees, questions regarding gender differences seem to arise. Does one gender tend to initiate divorce more
than the other? Do women or men end up better off financially? Is it the man or the woman who has the greatest lifestyle adjustment? And
how do the perceptions of each gender match the realities? In answering such questions, subjectivity is rampant. One's own gender or the experiences of significant others often cloud the general realities.
Yet empirical research has brought some clarification to these
questions. Zeiss4 Zeiss,_ and Johnson. (1980). On the basis of prior research,
this study hypothesized that both men and women would report that women were more likely to initiate separation leading to divorce and
that women would show better adjustment than men following such
separation. Beginning in January 1974, subjects were recruited from those filing for divorce in Eugene. Oregon. A total of 133 subjects were
selected at four different sampling times over an eight month period. Each sampling yielded equal numbers of between 12 to 18 men and women in each of five categories. These were time since filing categories
of 6 months, 1 year, 1.5 years, 2 years, and 2.5 years. Although the relative homogeneity of the subsamples may be an unexamined issue, the time since filing differences allowed for adjustment over time to be explored.
Along with basic demographic information, the study employed
four different instruments, each of which had previously been cliAermined to possess sufficient validity and reliability.
Intercorrelations among the measures were performed to assess the
significant predictive value on the definition of adjustment. Positive adjustment was hypothesized to involve a decrease in such factors as decrease in suicidal ideation, tension, depression, anger. fatigue, or general emotional distress; along with an increase in independence,
stability, and generally positive functioning and expectations. Individual differences between gender as well as general changes in
adjustment were reported.
Respondents did confirm the hypothesis that both men and women would report that women more frequently were the initiators of separation leading to divorce. (X2 = 11.48, p = < .001). In regard to
adjustment to divorce, men reported a lower level of overall adjustment when compared with women (t = 2.43, p = .02). Although both genders
were less bothered by suicidal thoughts than they were prior to the separation, women were significantly better off than men (t = 2.23, p = .05). Overall, men in this study reported having more difficulty moving away from the past marriage, and they experienced less improvement in
their general adjustment than did the women. However, in 1974, these women did report significantly less income (t = 3.93, p < .001), and more
tension (t = 1.98, p = .05). It should be remembered, though, that specific geographic characteristics or men's relative repression of tension were not accounted for. Bloom_and_Caldwell, (1981). This study researched the
adjustment differences between men and women during the
preseparation and postseparation periods. Their hypothesis was that women would report greater personal disruption during the pre-
separation phase, while men would report greater personal disruption
during the early post-separation period. They utilized data from four independent samples of newly separated or divorced persons, all of
which had completed the same self report instrument. These samples included a total of 179 women and 104 men. Some validity questions
are inherent in the use of self-report instruments, especially in regards to the degree to which gender may influence disclosure. Nevertheless,
the methodology of this study was better operationalized than many in
this field and its results provide useful information. The researchers gave thoughtful consideration to the development
and testing of their instrument. It was a brief 22-item self-report measure which was a compilation of four previously validated scales. Data collected from 153 newly separated persons were factor analyzed
and three clusters of variables were identified. Cluster I was neurasthenia, denoting chronic fatigue, the feeling of impending breakdown, inability to accomplish life's demands, feeling tense or
nervous and general ill health and weakness. With a Cronbach's alpha of .81 and a degree homogeneity of .39, it was considered adequately
reliable. The second cluster was anxiety, and with an internal consistency of .72 and a homogeneity ratio of .35. it was also considered
reliable. The third cluster, vascular. was assessed to be too neurologically oriented and its results were not considered reliable.
Although validity and standardization concerns still exist for this instrument, its development and testing was commendable in this specialized field of research.
Results of the study confirmed the hypothesis. Females did have, in the pre-separation period, significantly higher scores on the
neurasthenia cluster, t (151) = 4.14, p < .01. During the postseparation period, males scored significantly higher on the neurasthenia cluster, t (133) = 2.78, p < .01, as well as the anxiety cluster, t (133) =
3.78, p < .01. While assertions about adjustment differences over time
could have been made more confidently from longitudinal as opposed to
cross-sectional data, the consistency of these findings is noteworthy. During the preseparation period, women reported substantially poorer
psychological adjustment than men. In contrast, during the early postseparation period, men reported substantially poorer adjustment than women. The authors offered two additional hypotheses concerning
these results. The first was that the findings could be considered consistent with the common observation that women appear to have a generally higher readiness than men to admit personal symptoms and
difficulties. The second was that, shortly after separation, women may be feeling the positive effects of their separations, while men may be just
beginning to come to grips with the loss.
Ke1t:4119851 In this study, longitudinal data was used to compare the objective and subjective financial well-being of older, divorced or separated men (n = 114) and women (n = 251) at the
beginning and end of a ten year period. Data were collected in both 1969 and 1979 from interviews of subjects who continued to be divorced
or separated in both time periods. Demographically, the women had somewhat higher levels of education than the men (t = 2.22. p < .05). Men and women were equally involved in the labor force, and there were
no gender differences in employment status at either time period. A nonstandardized self-report instrument was used at both time periods.
Although the instrumentation raises inherent cautions regarding such measures, the nature of the variables seemed to possess face validity. It was assumed that actual income would be objectively reported and the perception of its adequacy would be a subjective response which could
be well captured through self-report. Multiple regression analyses were
conducted, thus imposing controls over time and allowing a higher
reliability than simple differences. Given these thoughtful designs, this study provided some helpful information to the field. Consistent with the convergence of gender differences emerging in
other studies, men and women did not differ at either interview on satisfaction with level of living (t = 0.59; t = 0.28. time 2). At Time 1,
the men's median income was significantly higher than the women's (t = 2.82, p = <.01). Yet at Time 2, median incomes did not differ significantly (t = 0.778). However, men am i women did not differ significantly on perceived adequacy of income at either interview (t =
0.32, t = 1.41). Nevertheless, cross tabulations of the direction of change (increase, decrease) in both satisfaction and financial adequacy revealed that, when men's responses did change, they were more likely
than women to change to negative assessments. Women who were employed felt their finances were more adequate
(r = .20) which, in turn, influenced satisfaction with level of living. (r =
.36). As previously mentioned, the median of women's actual income
increased over the decade to be the same as the men's. Yet men's perceptions of income adequacy was more strongly correlated with
actual income (r = .41) than those of the women (r = .22). In summary, women's actual level of income increased following divorce, as did their perceived levels of income adequacy and satisfaction with level of living. Although the gender differences became more convergent following
divorce, the men's perceived adequacy and satisfaction decreased. However, whether these correlated changes were due to the divorce, or to
societal changes over the decade studied was a possible confound for
which there was no control. An alternative explanation could be that, regardless of marital status, gender differences, in general, are converging in our society, with women perceiving more satisfaction with
this change than men. Colburn. Lin, and Moore,119a21. The object of this study was
also to examine the extent to which males and females are converging in their divorce experience. It utilized a sample of 268 persons who had
divorce decrees filed in the Indianapolis area. The 29% return rate included 41% male (n = 111) and 59% female (n = 157). It utilized a
non-standardized measure consisting of 189 closed and open-ended
questions. Retrospective self-reports were gathered related to married life, life after divorce and adjustment to divorce. Methodological concerns common to this type of design would apply. For instance, the
generalizability of the results would be limited to those with similar background and demographics as respondents from Indianapolis. The limited number of persons willing to respond to the lengthy survey may
have further defined the sample and limited its generalizability. There are validity and reliability concerns about the instrument and the
limitations of retrospective self report. Nevertheless, this study's findings were helpful.
Most relevant to the topic was this study's section on adjustment to divorce. It was also in this area that this research showed the most
significant convergence in gender differences. As background, the study
noted that more of the females (73.9%) than the males (41.8%) had
petitioned for divorce. Results showed that the females tended to agree
more than the males that a greater exercise of independence served as a means of coping after divorce (t = -2.24, p < .05). Involvement in a support group, joining a singles' club, or devoting time to work were not
found to be significantly different in males' and females' perception of
coping (t = -1.80, p = .073). Furthermore, there was no indication that females experienced downward economic mobility as a result of the
divorce more than did males (t = 0.86, p= .388). Also, contact with one's ex-spouse did not differ significantly in males' and females' reports (t =
-0.88, p = .379). Finally, neither males' nor females' differed
significantly in the perception of the overall sexual adjustment to divorce (t = -0.52, p = .604).
Additionally, the study found that, the younger the age of the
respondents. the greater the tendency to cite the helpfulness of activities in the creation of a new identity (r = -0.3593, p = .000). Also, the younger the respondents, the greater the importance of involvement with children as a mechanism for coping (r = -0.2431, p = .001). Other correlations suggested that the younger the respondent, the greater the likeliness for there to be agreement about issues related to gender roles. Ultimately, the statistical and methodological concerns should be remembered in considering the above significances. Nevertheless, the subjects studied did seem to report dramatic convergence in gender
differences as compared to the traditional attitudes of earlier studies. (Goode. 1956).
Summary of Gender Differences. In the above studies, results indicated that women tended to initiate divorce more often than men (Colburn, IAn, & Moore, 1992: Zeiss, Zeiss, & Johnson, 1980: also see
Jordan, 1988) In the early 1980's, Bloom and Caldwell (1981) found
that, during the pre-separation period, women tended to experience more emotional distress. Yet following the divorce, men were found to have more difficulty. However, Zeiss, Zeiss, and Johnson (1980) found that the Women's general adjustment was more difficult following
divorce. In the area of finances, Keith reported women's actual level of income increased following divorce, along with their sense of life
satisfaction. Nevertheless, there was no significance between genders on their perceptions of income adequacy (Keith, 1985). A later study found gender differences were not significant on downward mobility
(Colburn, Lin, & Moore, 1992). Another study done by Barnet (1990,
reviewed later in this paper), revealed that it was the area of social
adjustment that was significantly more difficult for the men. Despite some discrepancies in the findings, there was increasing consensus that adjustment issues based on gender tend to be converging. RsychologicaLandInterperaonalAdjuatment
In this section, empirical research that focused on the psychological, sociological or interpersonal factors of divorce will be
reviewed. In this area, several additional questions become relevant. Is there a period of greatest distress? What variables are shown to be most significant in facilitating the divorce recovery process? Does this research validate any psycho-social or chronological stages of divorce recovery? Each of six studies will be described, followed by a summary
addressing the above questions. Kolevson_ and _Gottlieb. (19.82). Utilizing time_since.__divorce as the
major independent variable, this study examined correlations with the
dependent variables of depression, hostility, and abilitimate contac s at different time periods. The sample (N = 157) was drawn from
the mailing list of Parents Without Partners, with the obvious demographics of this organization imposing some limitations in generalization of the results. Along with a demographic questionnaire, two standardized measures were selected for their reliability and
validity; one measuring depression and hostility, and the other the
ability to form intimate contacts. Statistical analysis utilized Pearson Product Moment Correlation Coefficients to look at the relationships
between the variables, and a multivariate analysis to obtain additional specification between the original bivariate relationships and a series of control variables. While remembering that results were indicative of divorcees with children willing to be involved in a support organization,
this was a well designed and executed study that provided additional empirical evidence on the process of divorce adjustment. Results showed that, os.q.-:r time, there were significant decreases in
levels of depression (p = .01) and hostility (p = .04). Ability to form
intimate relationships, however, showed no relationship to the variable of tim,!. Normative data for the measures also indicated that there were higher levels of depression, hostility, and an ability to form intimate
relationships in these divorcing individuals than in the general
population. Consistent with other research, the women were statistically more likely to initiate the divorce (p < .005). Multivariate
analysis of control variables showed that female respondents manifested a significant decrease in levels of depression and hostility, with no significant change in the men over time. Depression and hostility also 4
showed significant decreases for those involved in counseling, those who
did not initiate the divorce, those who resided in the same area as the ex-spouse, and those who retained custody of the children. In each of these areas, no significant change occurred for the opposing
characteristic (e.g.. not involved in counseling). Lastly. the ability to form intimate relationships significantly decreased for those who retained custody of the children.
Pettit and Bloom. (1984). Demonstrating the advantages of a
longitudinal study, this research utilized initiator status as the independent variable correlated with 15 dependent variables. While
stating that the one who actually files for divorce is not always the
active party in forcing the action, they found little alternative in the
available data but to operationally define initiator status as the ones who reported that it was their decision to divorce. This included mutuals, identified through multivariate analysis as showing no significant difference from initiators. Subjects (N . 144) randomly
obtained from archival data were tested initially, and at 6 and 18 months. Several instruments, deemed from previous studies to have sufficient validity, were employed, along with a demographic
questionnaire. Although two-thirds of the participants were randomly assigned to a treatment program, comparison with the one-third control group showed no significant difference. Thus the entire sample was
included in reporting of results. The overall design of the study,
especially its longitudinal nature, provided results that are far above average in their relevance to the process of recovery.
Not surprisingly, a higher number of initiators were female, Initiators younger, and had been married for a shorter period of time. higher in were also less family oriented, t (144) = 2.20, p < .03, scored orientation to selfself-realization (t = 27.3, p < .007), and had a greater within realization (t = 3.04, p < .003). As predicted, in terms of attitude remained more in favor both genders and at all time periods, initiators However, over time, of the decision to separate than did noninitiators. At the 6 month noninitiators did endorse more favor with the decision. reconciliation, yet by 18 months, point, initiators were less in favor of this difference no longer had significance. By 18 months, previous In differences in quality of life issues were also no longer significant. divorce where fact, in terms of all problems typically encountered after shortinitiator status does have effect, the duration of the effect was by 18 months. lived, with no significant differences being manifested postseparation And contrary to expectations, there were no significant guilt with respect to differences regarding stress related symptoms or initiator status. In summary, it was concluded that only a small for proportion of the variance in adjustment to divorce was accounted
by initiator status. investigate whether certain Kalb. (19_87). This study attempted to three stages of the key biographical factors differentially affected the examine the within divorce adjustment process. The author hoped to research, which has tended to stage differences, as opposed to previous (N = 88) consisted of focus on the between stage differences. His sample attending an outpatient recently divorced men and women who were Separation Index recovery group. They were administered the Behavioral
(BSI) on four separate occasions: (a) upon entering the group. (b) at the
end of group. (c) at a 3 month reunion, and (d) at the 6 month reunion. They wished to examine the participants emotional states at the four time periods relative to the three stages of divorce proposed by Weiss
(1975). These stages, mentioned earlier in this paper, are: Stage 1 feelings of shock and denial; Stage 2feelings of anger, resentment, bitterness, lowered feelings of self-esteem, withdrawal and distancing;
and Stage 3feelings of acceptance, adjustment and comfort in being single. Because the measure chosen did not specifically yield results for these stages, two psychologists categorized each of the measures' 51
response items into one of the three stages (inter-judge agreement = .93, .91, .95 for each stage). The methodological concerns inherent in the design of this study will be noted following a summary of the results.
In regard to the Stage 1 dynamics of shock and denial, factors
significantly correlated at the time of initial testing were maintaining a feeling of love for the spouse (p < .05) and wanting to reunite (p < .01).
These remained consistent factors for significantly greater levels of
shock and denial through the fourth testing session over 7 months later (p < .10). Demographically, less educated subjects reported more shock and denial initially (p < .10), but those with higher levels of education
reversed the significance at the three month reunion (p < 10). Initiator status regarding the divorce was found not to be significant. The
authors noted that the biggest changes in shock and denial happened during the time of the group, after which they seem to plateau. They do not consider however, whether this effect may be more specific to having
been in the group than to the process of adjustment.
Stage 2 examined responses of anger, resentment, and withdrawal. Although women reported a significantly greater amount of anger at the beginning of the group (p < .05), there were no significant
gender differences at later points. Less educated clients were also significantly higher at the beginning of the group (p < .05). There were no other significantly correlated differences initially and no differences
at all by the end of the group or at the follow-up periods. This lack of statistical differences included age, length of marriage, number of
children, desire to reconcile, falling out of love, and initiator status. In regard to Stage 3. feelings of acceptance, adjustment, and comfort in being single, there were only three significant differences. By
the 3 month reunion, less educated clients showed significantly more efforts at rebuilding a new single life (p < .10). Younger clients also
made significantly more efforts at starting over by the end of the group
(p < .10). Subjects without children made significantly more efforts at
rebuilding at the sixth month follow-up poirt. Although the above biographical factors shed light on several possible factors affecting the process of recovery, some methodological concerns should be noted.
It should be remembered that this was a convenience sampling, thus limiting generalization beyond those motivated to seek group
therapy. In regard to the measure used, arriving at the three stages from this measure raised validity cautions. The four time periods were relatively limited in length. given that most research over time tends to extend from one to two years. Also, the time since divorce was not
analyzed as a potential variable. Although none of these limitations were addressed by the author, the greatest potential confound was the
group itself. How many of the changes in denial, anger, or acceptance might or might not be group specific results should be considered.
Jordan. (1988). This study utilized retrospective measures to examine the reactions and experiences of separated men in their personal, psycho-social, and practical areas of life. Consistent for inclusion in this paper, each of the measures was caiswered retrospectively.for (a) the time just preceding marital separation, (b)
during or immediately after the separation, and (c) at the time of the survey (one to two years following the separation). Instrumentation was above average for this field. The measures included a demographic
questionnaire, two standardized measures assessing psychological wellbeing and psycho-physical well-being, and two further measures to
assess attachment to wife and children, general coping, and living
conditions. Statistical significance was reported to be p < .05 and a MANOVA was utilized to conduct a discriminant analysis of the
contrasting variables. However, several of the results were reported only by descriptive percentages. Although random sampling from court files provided a good degree of generalization to the general population of
divorced men, the study was done in Australia, thus raising cultural questions regarding generalization.
In the areas of well-being, health and practical living problems.
these men reported the period of greatest distress around the time of the separation with significant improvement one to two years post
separation. The effect of the separation was pervasive, and could be
measured on all of the scales. Most dramatic difficulties at separation were seen for the variable of social initiative, including doing household
chores, taking social initiative, work, and finances. Although the impact of separation was quite marked, the men not only showed a trend toward recovery after one to two years, but also reported gains in three areas. These were (a) an increase in positive feeling or experiences, (b) reduced difficulty with practical daily living chores, and (c) fewer financial difficulties.
An interesting result was found by correlating bad feelings before
separation with good feelings at the survey time 1 to 2 years later. Those men who reported more good feelings before separation had higher
bad feelings and health complaints 1 to 2 years later. This may suggest that men who were least aware or affected by problems before the
separation were least likely to recover. In analyzing the covariates, the
most significant factor of this separation distress was whether or not the separation was wanted by the men. A related demographic variable was that 65% of the men said it was the wife's decision to divorce.
Other significant factors included the facts that (a) no pre-separation conflict was recalled. (b) reconciliation was attempted. and (c)
occupational status was low. A final trend noted was that respondents who were not living alone one to two years later were significanily more
likely to report more positive signs of adjustment. B_arnet(199Ial. This article provides relevant process results-due
to its measures' retrospective accounts of the three major time phases (decision phase, divorce proper, and postdivorce). Its sample (N = 107)
was stratified among those who had been divorced 6, 12, or 18 months.
Statistically, it attempted to improve on previous research through the use of path analysis over the different time periods. Due to the applied
nature of the study. paths with a standardized beta weight of more than .10 (absolute) were defined as practically significant. The study examined two locus of control variables (Rotter Locus of Control, Marriage Locus of Control), three demographic variables
(gender, number of children, and duration of marriage), time since divorce, decision time, number of divorce difficulties, and six measures
of stress and adjustment. Marriage Locus of Control was operationalized as who initiated or took certain actions entering, during, or exiting the marriage. Rotter's Locus of Control measured the more general personality style. From these, respondents were classified
as internals or externals. In an attempt to maximize questionnaire response rate, all measures were modified to include a minimum number of questions, an
instrumentation decision which introduced validity concerns. It was postulated that scales would be validated automatically in the data analysis, because the path analysis reported expected relationships (convergent validity) and expected nonrelationships (divergent validity).
Due to the number of factors analyzed, reporting of results became somewhat cumbersome. The results most relevant to this review
related to post divorce stress and adjustment. Postdivorce stress increased with externality, Rotter Locus of Control .18, E (1, 94) = 4.13, p = <.05; Marriage Locus of Control (.11 total); children (.13 total);
number of divorce difficulties (.19 total); time to peak stress, .20, F (1,94) = 5.18, p < .05; increases in divorce stress intensity, .28, E (1,94) = 5.18, p = < .05; increases in divorce stress duration (.18); and divorce recency (-.13). Unexpectedly. postdivorce stress increased with longer
decision times (.15. direct) and predecision stress (.12 direct, .09 total), but decreased with marriage length, -.20, F(1,94) = 4.69, p < .05.
Postdivorce social maladjustment also increased with externality (Rotter Locus of Control .15 total; Marriage Locus of Control .21 total), number of divorce difficulties. .23, F(1.94) = 5.56, p < .05: .31 total, stress
intensity (.13), stress duration (.16). and time from the inception of divorce problems up to the point of maximum stress, .28, E(1,94) = 9.99,
p < .01. Men reported more postdivorce social maladjustment than
women (-.11 total). Interestingly, stress, but not social maladjustment, decreased from 6 to 18 months postdivorce. Those with internal locus of control, when compared with externals, reported that stress peaked earlier and was higher during the predecision period, yet less during the divorce proper and postdivorce.
aursikala91). This longitudinal study sought to clarify those factors related to women's adjustment at different times during the divorce process. The sample was obtained from public divo'rce records in
the Boston area and consisted of 104 women who had been physically
separated from their husbands for at least eight months. The sample was divided into three subsamples: women without children (N = 36), women with young children (N = 35), and women with adult children (N = 33). These women were interviewed and tested initially (Time 1), and
18 months later (Time 2). The instruments used were carefully chosen
as the best available for the task and included one to two hour interviews, numerous self-report questionnaires, and projective
measures. The attrition of participants between time periods was insignificant. Pearson correlation coefficients were used to measure
significance and ANOVAs were implemented to measure covariation. In
summary, the common methodological problems related to sampling,
instrumentation and statistics were minimized greatly in this study. The study proposed seven specific hypotheses, and the results add
substantial new understanding to the divorce recovery process, at least as it relates to women. Organized in context of the researcher's original hypotheses, the results were as follows:
1. Women who were younger would show higher levels of
postseparation adjustment. No significance was found for this hypothesis on any of the Time 1 or Time 2 variables.
2. Women who were married shorter periods of time would show
higher levels of postseparation adjustment. No significant results were found.
3. Women with fewer children would show higher levels of
postseparation adjustment. This hypothesis was not supported. 4. Women with older children would show higher levels of
postseparation adjustment. This also showed no significance. In fact, contrary to previous assumptions, measures of covariance showed little significant difference on adjustment related to demographic factors such as age, length of marriage, or children.
5. Women with nontraditional sex roles attitudes would show higher levels of postseparation adjustment than those with traditional sex role attitudes. This was confirmed in all the subsamples at both Time 1 and Time 2. Higher scores on a masculinity scale were
significantly related to well being (r = .53, p <.01) and emotional health (r =
p < .05). Scores on a femininity scale showed no significance.
6. Women who had and utilized a social support network would show higher levels of postseparation adjustment than those women who were socially isolated. At Time 1, neither social participation nor social
isolation was significantly associated with adjustment in any subsample. At Time 2, there was no significance on measures of social
support or social participation. However, social isolation was negatively correlated with emotional health (r = -.42, p < .05) and well-being (r = -.44, p < .01). It could, therefore, be concluded that although
interactions with family and friends did not seem to lessen distress,
social isolation did decrease postseparation adjustment. 7. Women who maintain a civil or amicable relationship with the ex-spouse would show higher levels of adjustment than those whose
relationships with the ex-spouse are hostile and strained. The trends were strong in this direction across all subcategories. For example, for women without children, a civil relationship with the ex-spouse was significantly associated with physical health (r = .47, p < .05) and also significantly related to emotional health for the sample of women with young children (r = .50, p = .01).
Summary_of the Psychological andinterpersonal Research. From the studies.above, it can be concluded that divorcing individuals experience higher levels of depression, hostility, and inability to form
intimate relationships than that seen in the general population (Kolevson & Gottlieb, 1982). Distress was greatest immediately
following the separation or divorce: yet 1 to 2 years later, adjustment
variables showed significant improvement (Jordan, 1988; Kalb, 1987:
Pettit & Bloom, 1984). The potential effects of certain factors on the divorce process were clarified by these studies.
The results supported earlier findings that initiators are more likely to be female (Jordan, 1988; Kolevson & Gottlieb, 1982). Yet
initiator status was determined to account for little of the significant differences in adjustment to divorce (Kalb. 1987; Pettit & Bloom, 1984).
And where depression or hostility was experienced by non-initiators, it decreased significantly over time (Kolevson & Gottlieb, 1982). Two
studies reported that women initially endorsed greater levels of
emotional distress than men, yet, after time had passed, anger or depression varied little across genders (Kalb. 1987; Kolevson & Gottlieb,
1982). Other factors found not to be significant included age, length of marriage, number of children, and age of children (Bursik, 1991; Kalb,
1987). However, Barnet (1990) reported that children did increase stress and social maladjustment for divorcees.
Barnet (1990) also found increases in stress and social maladjustment related to number of divorce difficulties, higher divorce stress intensities, higher divorce stress duration, and external locus of
control. Actually, those with an internal locus of control experienced more difficulties initially, while those with an external locus of control did worse later in the process. Additionally, Kolveson and Gottlieb (1982) found significant decreases in levels of depression and hostility for those involved in counseling, those who resided in the same area as the ex-spouse, and those who retained custody of the children. They
also noted that divorcees' ability to form intimate relationships did not change significantly over time.
Two gender specific studies added insight to possible male or
female variables. In studying men in Australia, Jordan (1988) found
that the most dramatic difficulties were encountered at the time of separation. Of greatest significance was the factor of not wanting the separation. Other variables tha,b increased adjustment difficulties for .men included (a) no recollection of preseparation conflict, (b) failed attempts at reconciliation, and (c) low occupational status. However, after two years, the only significant detriment was living alone. Among women, Bursik (1991) found that adjustment benefits were related to
non-traditional gender attitudes (i.e., more masculine), and civil or amicable relationships with the ex-husband. And although social support or participation had no significance for these women, social
isolation was detrimental to postdivorce adjustment.
Unfortunately, the only research structured around potential psychological stages was that conducted by Kalb (1987). In what was termed Stage 1 (shock and denial), recovery was significantly decreased
by feelings of love for the spouse and wanting to reunite, a significance which remained through the other two stages. In Stage 2 (anger,
resentment, and withdrawal) and Stage 3 (acceptance, adjustment to singleness), no additional significances were noted. However, less educated subjects made less effort at rebuilding a new life and women with children made more efforts at reestablishing a new life. So, while
the longitudinal research shed new light into many factors effecting the
54 divorce recovery process, only Kalb's work provided additional insight
into any psycho-social processes.
The purpose of this paper was to ascertain whether a process of divorce recovery could be etnpirically validated from the research. From
this review, it has been discovered that limitations in the current research have prevented an adequate empirical verification of a process
of divorce recovery. Much theoretical and clinical organization has been
proposed around the question. However, lack of resources and methodological problems have hampered the empirical efforts.
Nevertheless, progress is consistently being made in this field. The general body of knowledge has dramatically advanced since the first
studies. Although a general process is yet to be verified, several process dynamics have been identified in specific populations. And, probably most helpful, various factors have been repeatedly shown to have
significant relevance to the successful negotiation of the adjustment process.
Current State ()film Literature
In the theoretical literature, there is general agreement that recovery from divorce involves progression through certain time stages.
The chronological stages of the process have most often been divided into the three periods of (a) predecision, (b) decision for divorce until
legal termination, and (c) postdivorce. Although no direct empirical validation has been done of these stages. their face validity seems appropriately delineated by the date of decision to divorce and by the
date of legal dissolution of the marriage. These chronological stages divide the divorce adjustment process into specific transition periods
involving the dissolution of the old relationship and the adjustment to a new life. However, they do not address the research question of identifying the processes of effective adjustment either across or within
the time stages. Empirical research of the postdivorce adjustment period of recovery has shown that, in general, subjects improve over time. In defining adjustment, studies reviewed in this paper strongly suggest that effective adjustment is not tied to remarriage, but involves various psychological and/or social processes. Yet, unlike Kubler-Ross' (1969) observations of grief adjustment, no processes or psycho-social stages have been confirmed by the current body of divorce research. Only three
of the studies reviewed attempted to examine any actual processes of divorce (Crosby, Gage & Raymond, 1983; Kalb. 1987; Kressel & Deutsch, 1977). Rather than focusing on a general process of divorce recovery,
the research has tended to focus on the identification of certain adjustment factors which may correlate with more effective recovery,
Current research has offered a better understanding of several adjustment factors such as gender. initiator status, locus of control, connection to the ex-spouse. social involvements, or expression of affect. Nevertheless, many of the results have been found in only one or two studies. Given the identified methodological concerns. thes: results
need cross-validation through additional research. Potentially significant factors need to be replicated and verified in future studies. Whether certain aspects of adjustment correlate within and/or across
stages of the divorce recovery process needs to be determined. Additionally, possible trends in recovery need to be verified across
different demographic groups. In summary, the process of divorce recovery remains a subject in need of increased empirical study. Directions for F_uture Research
Methodological issues to be addressed in the future include
definitions of adjustment, instrumentation, procedures, and sampling. Additionally, the process of divorce is confounded by many personal and
sociological variables, which also need to be clarified Future research might consider the following design issues.
Whether remarriage occurs or not, adjustment seems to involve a psychological and social transition to a new life-style. This involves the moving away from the old relationship and establishing an
autonomously functioning and socially related sense of self. Within this framework, dependent variables of adjustment need to be further
defined. Some variables that past studies have suggested to have significance have been: (a) disconnection from the old relationship, (b)
functional autonomy in the new role, (c) avoiding isolation, and (d)
working through the emotional components of the transition. These and other identifying adjustment variables will need to be defined and verified in greater detail in future research. The greater difficulty becomes the ability to measure the relevant
divorce adjustment variables. As mentioned earlier, few scales in this field have been standardized. Of particular importance would be the development of a psychometrically sound measure related to one of the
psycho-social models of transition. For example, a measure identifying
the relative areas of denial, anger, questioning, or acceptance would be critical to the question of the relevance of the Kubier-Ross (1969) model
to divorce. Additionally, measures could be structured to assess the adjustment variables determined to be most relevant during different phases of adjustment and recovery. Procedurally, some form of design is necessary to examine
subjects at different time phases in the divorce process. Research might be limited to the predecision, divorce proper, or postdivorce periods. But
even within these periods, the research design needs to examine time variables to determine the existence of a process over time. Longitudinal research would accomplish this by examining homogenous
samples at differing time periods. Stratified sampling could divide homogenous subsamples in demographically similar groups at differing
periods in the process. Although less desirable, self-report measures could continue to compare retrospective accounts relative to the different time periods of the process.
The use of control groups in this field is a luxury yet to be
employed. In fact, finding a suitable control group seems an impracticality. However, comparison to norms would be a feasible
addition. With the development of a standardized instrument, comparisons could be made to the normative data. In this way. subjects' responses could be compared to the responses for similar
individuals in the population at large. The best source of sampling to be identified from the literature is
county court records. Using defensible sampling strategies. this would provide the most comprehensive sample of the divorced population for a
given area. Using a longitudinal design with this population would provide results regarding the process of divorce recovery, yet tracking
subjects over time becomes difficult. An alternative would be a
stratified sampling strategy that matched statistically comparable
subsamples from the divorced population. In this design, adjustment variables could be correlated to matched groups that had been divorced for varying lengths of time.
This paper has sought to provide a summary of the current body of literature in the area of divorce recovery, and to suggest directions for
future research. Many adjustment variables have been identified empirically, but the verification of a process of divorce recovery is less
clear. Future efforts focused on improving the body of knowledge in this
area of human experience will only benat the life and health of future generations.
Ahrons, C. (1980). Divorce: A crisis of family transition and change. Fami1y_Relations-29, 533-540.
Ahrons, C. R. (1983). Divorce: Before, during and after. In H. McCubbin & C. R. Fig ley (Eds.), StresaancLilie_family (pp. 102-115). New York: Brunner/Mazel.
Barnet, H. S. (1990). Divorce stress and adjustment model: Locus of control and demographic predictors. Journal of Divorce. 13, 93109.
Barnet. H. S. (1990). Divorce stress and adjustment model: Locus of control and demographic predictors: Errata. Journal of Divorce and_Remardage. 14. 143-145.
Berman, W. H. (1988). The relationship of ex-spouse attachment to adjustment following divorce. LaurnaLof Family_Psyehology._ L. 312-328.
Berman, W. H. (1988). The role of attachment in the post-divorce experience. Journal.ofTemonality.and SociaLEsychology, 54. 496503.
Bloom. B. L.. & Caldwell, R. A. (1981). Sex differences in adjustment during the process of marital separation. tiournal_ofMarriage_& Family. 43, 693-701.
Bohannon. P. (1970). The six stations of divorce. In P. Bohannon (Ed.), Divor_ce_and. after (pp. 29-55). Garden City, N.Y.: Doubleday.
Buehler, C. A., Hogan, M. J., Robinson. B. E.. & Levy. R. J. (1985). The parental divorce transition: Divorce-related stressors and wellbeing. Journal of Divorce. 9. 61-81.
Bursik, K. (1991). Correlates of women's adjustment during the separation and divorce process. JournaLof Divorce and Remarriage._14. 137-162.
Colburn, K., Lin, P. L., & Moore, M. C. (1992) Gender and divorce experience. JournaLofDivorce and_Remarriage. 17, 87-108.
Crosby. J. F., Gage. B.. & Raymond. M. (1983). The grief resolution process in divorce. Journal of Divorce, 7. 3-18. Crosby, J. F., Lybarger. S. K., & Mason. R. L. (1986). The grief resolution process in divorce: Phase II. Journal of Divorce, 10, 17-40.
Federico, J. (1979). The marital termination period of the divorce adjustment process. Journal of Divorce. 3. 93-106. Goode, W. J. (1956). After__divorce Glencoe, IL. : Free Press. (Re-
released in 1978 as Women in divorce. Westport, CT: Greenwood Press.)
Herman, S. J. (1974). Divorce: A grief process. Perspectiveln Psychiatric Care.___12.. 108-112.
Jordan. P. (1988). The effects of marital separation on men JaurnaLof Divorce,__12. 57-82.
Kalb, M. (1987). The effects of biography on the divorce adjustment 53-64. J.., process.
Kanal. K. (1996). Aguide_to crisis intervention. Belmont, CA: Wadsworth. Kaslow, F. W. (1984). Divorce: An evolutionary process of change in the family system. Journal of Divorce. 7. 21-39.
Keith, P. M. (1985). Financial well-being of older divorced/separated men and women: Findings from a panel study. J_ournaLof Divorce,,_9.. 61-72.
Kitson, G., & Raschke, H. (1981). Divorce research: What we know, what we need to know. Journal of Divorce, 4. 1-37. Kolevson, M. S.. & Gottlieb, S. J. (1982). The impact of divorce: A multivariate study. JournaLof Divorce. 7. 89-98.
Krauss, S. (1979). The crisis of divorce: Growth promoting or pathogenic? Journal of Divorce-1. 107-119. Kressel, K.. & Deutsch, M. (1977). Divorce therapy: An in-depth survey of therapists' views. Eamily_Process1_16 413-443. Kubler-Ross. E. (1969). On_death_and_dying. New York: McMillan. Kessler, S. (1975). The_American_way of_divorce: Prescriptions for
change. Chicago: Nelson-Hall.
Mirels, H. (1970). Dimensions of internal versus external control. Journal of Consulting _and_ Clinical_Psychology, _34. 226-228.
Pettit, E. J., & Bloom, B. L. (1984). Whose decision was it? The effects
of initiator status on adjustment to marital disruption Journal of_MarriagelLthe Family 46. 587-595.
Salts, C. J. (1979). Divorce process: Integration of theory. Journal of Divorce. 2, 233-240.
Salts, C. J. (1985). Divorce stage theory and therapy: Therapetitic implications throughout the divorce process. JournaLof Psychotherapy and the Family._1(3), 13-23. Saul, D. C., & Scherman, A. (1984). Divorce grief and personal adjustment in divorced persons who remarry or remain single. Journal of Divorce. 7, 75-85. Spanier, G., & Furstenberg, F. (1982). Remarriage after divorce: A longitudinal analysis of well being JournaLof_Marriage_ancl_the Family, 44. 709-720.
Spanier, G., & Thompson, L. (1974). Parting: The aftermath of separation and divorce. Beverly Hills, CA: Sage Publications. Spanier, G., & Thompson, L. (1983). Relief and distress after marital separation. Journal of Divorce. 7. 31-49.
Spanier, G.B., & Casto, R.F. (1979). Adjustment to separation & divorce: An analysis of 50 case studies. Journal of Divorce. 2. 241-253.
Steefel. N. M. (1992). A divorce transition model. Paythological Regatta...1% 155-160. Vaughan, D. (1979). Uncoupling: The process of moving from one lifestyle to another. Alternative Lifestyles. 2. 415-442. Waller, W. (1938). The family. A_dynamic_interpretation, New York: The Cordon Company Inc. (Republished 1951. New York: Holt, Rinehart & Winston).
Weiss, R. (1976). Marital Separation New York: Basic Books. Wiseman, R. S. (1975). Crisis theory and the .process of divorce. Social
Caseworka1 205-212. Zeiss, A. M., Zeiss, P. A., & Johnson, S. M. (1980). Sex differences in initiation of and adjustment to divorce. Journal of_Divorce.,4. 21-33.
RICHARD W. GASTIL EDUCATION:
Rosemead School of Psychology Clinical Psychology
Rosemead School of Psychology
California Christian Institute
California State University, Fullerton Business Administration, Management
California Christian Institute Social Work
Marriage, Family and Child Therapy
1995 Southern California Edison, Health Department San Onofre Nuclear Generating Station. California EMPLOYMENT OVERVIEW:
Richard W. Gastil. MFCC Private Practice, Orange, California
Health and Human Resources Center Employee Assistance Professional
California State University, Fullerton Part-time Faculty Instructor
Trinity School of Graduate Studies, Orange Part-time Faculty Instructor
Western Medical Center Hospital, Anaheim Inpatient Therapist Priveleges
Orange Psychological Associates Group Practice, Psychological Assistant (1990)
Hotline Help Centers, Inc., Anaheim Executive Director