Voice and Support: Programs for Children Experiencing Parental Separation and Divorce


There are a variety of existing programs for children experiencing parental separation and divorce, most of them intended to meet the needs identified by the research discussed in section 1. But while there are programs for children in most Canadian provinces (see Appendix A for a sample) they do not appear to be widely available and, according to providers, do not currently meet the demand.

This section describes the kinds of supports provided by children's programs, the kinds of programs available, and the research on their effectiveness. Most of the North American literature describes programs in the U.S. Most evaluations are also of American programs, so these figure prominently in the discussion. Appendix C provides a sample of prominent American programs, most of which have been evaluated. Appendix A provides a sample of Canadian court and community programs, with evaluations if available.

3.1 Program Goals and Content

Programs for children experiencing family breakdown generally offer one or more of the following supports.

3.2 Program Structure and Delivery

Children's programs may be for groups or individuals. Programs for individuals are typically family and individual counselling programs providing intensive therapeutic support to children in crisis or with severe problems. Canadian family service agencies providing group programs for children experiencing parental separation and divorce usually also offer individual programs to these children and their parents (e.g. the Families in Transition programs provided by the Family Services Association of Metropolitan Toronto). In the court system, an integrated court services pilot program currently operating in Corner Brook, Newfoundland and Labrador, combines individual short-term counselling for children, or children and parents, with mediation and other services for families involved in custody and access disputes, including access denial (Reynolds, pers. comm., see Appendix D).

Most group programs appear to offer education, information, emotional support, and coping skills in some combination, although some group programs also provide intensive therapy. Group programs typically serve four to ten children, usually in groups based on age, through a course of weekly meetings. They cost less than individual supports, but researchers and providers also endorse their therapeutic value. Research shows that discussing the divorce with other children of divorce helps normalize the experience for children and gives them a potentially supportive network (Kalter et al. 1988, Pedro-Carroll and Cowen 1987, cited in Grych and Fincham 1992; Vallant 1999). Also, many children are more comfortable discussing difficult, sensitive issues in groups rather than with unfamiliar adults alone (Pedro-Carroll and Cowen 1987, cited in Grych and Fincham 1992). Even for children experiencing more severe difficulties, group sessions with families and children may enhance individual counselling by helping therapists to see the whole picture of a family and children that an individual therapy setting prevents (Gertner, pers. comm., see Appendix D).

Several Internet sites also offer facts and questions and answers for children. For example, iConnect is an interactive Web site for youth 12 to 15 years old, run by University of Illinois academics (http://www.aces.uiuc.edu/). Banana Splits (New York, N.Y.) provides emotional support for 5 to 11 year olds experiencing parental separation and divorce, through group art therapy (http://www.divorcesource.com/NY/DS/rosenberg.html).

3.2.1 Community and Court Group Programs

Most of the Canadian children's programs identified in this research are community programs. Many are offered by social service agencies with counselling capacity, or by mental health institutions. Some are provided by lay volunteers: for example, the Rainbows program in Canada and the United States which operates in schools, churches, social agencies and other community venues.

Most programs in the United States are also community-based, and most are provided in schools (Grych and Fincham 1992). School-based programs in the U.S. may serve children participating in court-mandated programs for parents and children in separation and divorce proceedings, as well as other families. Most American court-connected programs are also delivered by community agencies (Geasler and Blaisure 1999). American school-based programs are often run by school counsellors: for example, the Rollercoasters program (see Appendix C). Basing programs for children in schools is thought to have expanded the number of children who can participate, since school programs are usually free. A school is also familiar ground, and can provide a child with a natural support network (Cowen et al. 1989, cited in Grych and Fincham 1992). One of the most rigorously evaluated American programs for children, the Children of Divorce Intervention Program, operates in primary and middle schools, including 50 in the Rochester, N.Y., area (see Appendix C).

A few Canadian provincial courts provide programs for children. In Manitoba, the court-based Caught in the Middle helps children, whose parents are before the courts over separation and divorce issues, to work through their feelings and concerns about the separation, and helps them avoid being caught in their parents' conflict (see Appendix A for details). St. John's Family Court in Newfoundland and Labrador provides a counsellor-led group program aimed at helping children normalize their feelings and develop coping strategies. The similar Confidences program offered by the Centres jeunesse de Montréal is intended for children with parents participating in court-ordered mediation. A provincially funded Vancouver program, linked to the Burnaby-New Westminster Family Justice Centre, provides a counsellor-led group aimed at helping children express and understand their feelings about parental divorce or separation.

3.2.2 Linked and Stand-Alone Group Children's Programs

Children's programs may be stand-alone, but are often linked with parallel programs for parents. Peer support groups providing emotional support are usually stand-alone (for example, Rainbows in North America, or Relateen in the United Kingdom). However, many court-connected American programs for children are tied to court-mandated parent education programs for separating and divorcing parents. None of the Canadian provincial jurisdictions that currently offer parent education programs provides a children's component, though British Columbia and Alberta are considering it. Community agencies in the U.S. and Canada also offer children's programs linked with parent education and support programs.

Parents' and children's programs can be linked in various ways. One common approach is for children and parents to attend parallel sessions with complementary curricula (for example, Focus on Children in Separation or FOCIS in Jackson Country, Missouri). Another is for children and parents to attend parallel sessions most of the time, with parents' and children's groups meeting periodically, or at the end (for example, the court-mandated, community-based U.S. program Rollercoasters).

Individual children and their parents may also meet periodically, or at the end, with each other and the program counsellor. The aim of the meetings is usually to consolidate and reinforce what they have learned, and perhaps set follow-up goals for parents, based on what the child has expressed (for example, Toronto's Jewish Child and Family Services' in Picking Up the Pieces program). Even when no parallel parent program is offered, parents can be linked to the children's program through family sessions at various points during the program course (e.g. Healthcare Corporation's It's Still O.K. in St. John's, Newfoundland and Labrador).

Therapeutically intensive programs for children are also often linked to intensive therapy programs for parents. In the group mediation program in Alameda County, California, for example, a children's program providing therapeutic emotional support is linked with intensive impasse mediation for high-conflict parents in chronic litigation. These children's programs are thus also linked directly to court proceedings for parents in custody and access disputes.

Some stand-alone Canadian programs for children also link directly to custody and access court proceedings. Mediators participate whenever possible in the final family session of the Centres jeunesse de Montréal's Confidences program. Quebec is also considering a children's component as part of a revamped parent information session currently under discussion (Tanguay, pers. comm., see Appendix D). Agency staff in the Vancouver peer support group program liaise regularly with the Burnaby-New Westminster Family Justice Centre, which refers most of the families using the program. Parents of children participating in Toronto's Families in Transition’s core program, who meet with the children and their case manager at the end of the program to set goals and identify new needs, often proceed to mediation to develop or revise parenting plans, incorporating the results of the children’s program (Freeman, pers. comm., see Appendix D).

3.2.3 Restricted and Open-Entry Programs

Community programs providing education, information and support, like Rainbows or British Columbia's Circle of Friends (see Appendix A), are typically open to all children who are experiencing, or have experienced, family breakdown (see Appendices A and C for details). Therapy-based community programs may either be open or restricted to high-conflict or litigating families. The Family Services Association of Metropolitan Toronto's Families in Transition program for children, for example, accepts all children who the counsellors feel will benefit from a group setting, including high-conflict families that are still open to insight and change (Freeman, pers. comm., see Appendix D).

Some programs exclude children in violent families, referring them to individual therapy, because these children will not be able to communicate with their parents about their feelings. The Family Centre of Winnipeg's Giving Children Hope is restricted to high-conflict parents who are also entrenched in court disputes.

Most provincial court-provided programs are targeted, or explicitly restricted, to children with parents in custody and access proceedings (for example, Manitoba's Caught in the Middle program, Vancouver's ARK Child Services Society's peer support group, and the Centres jeunesse de Montr’al's Confidences program).

Since children's programs in the U.S. are often linked to mandatory parent education programs, many American programs are restricted to children whose parents are currently before the courts with separation or divorce issues. In California, children's programs are often attached to impasse mediation or other intensive mediation programs for litigating parents.

3.2.4 Program Duration

Most group programs are short-term, with education and information programs sometimes spanning only one or two sessions, and emotional support programs sometimes lasting only three or four sessions. Many providers feel that short-term emotional support interventions around the time of separation or divorce are enough to help most children understand what is happening, accept reality, and acknowledge and manage their own feelings about it (Nichols, pers. comm., see Appendix D). Most parents whose children complete Toronto's Jewish Child and Family Service's One Family, Two Homes, for example, do not proceed to the more intensive Picking Up the Pieces because they feel they and their children have got what they needed (Gertner, pers. comm., see Appendix D). However, the Marriage Council of Philadelphia provides up to four month-long sets of sessions for children, and, by allowing them to attend more than one group series, can follow them for up to a year (see Appendix C for details). Maryland's Children of Separation and Divorce Center involves children and parents long-term by training them to participate in parenting seminars and to serve as peer counsellors (Davis et al. 1997).

3.3 Program Availability

American court-connected programs for children experiencing parental separation and divorce increased rapidly during the 1990s. A national survey (Geasler and Blaisure 1999) found the number of U.S. counties and cities offering such programs jumped from 10 percent in 1994 to 21 percent in 1998. The courts responding to the survey identified programs for children as the single most important innovation they would like to make to their existing divorce education program. The growth of children's programs corresponds to a tripling of court-connected parenting education programs in the United States between 1994 and 1998 (Geasler and Blaisure 1999).

Nearly all Canadian provinces now also offer parent education programs through the courts, or purchase of this service from community providers, although many of these programs are very recent (Bacon and McKenzie 2001). Broad-reaching, publicly-sponsored parent education programs have grown the most since 1997, when a national survey located about 140 such programs in the country (Bacon and McKenzie 2001). As already indicated though, no court-sponsored program offers a complementary children's program.

Given their links to parent education programs, many of the new American programs for children provide education and information, although many also offer emotional support (for example, Jackson County, Missouri's Focus on Children in Separation). In 1998, most of the American court-connected children's programs were offered to elementary school children (99), followed by children in middle school (85), high school (62) and pre-school (21). Community providers delivered most of the programs (up from 42 percent in 1994), and about two thirds were mandatory for children of separating or divorcing parents (Geasler and Blaisure 1999).

Stand-alone community programs providing mainly education, information or peer support seem to be expanding in Canada and the United States. The volunteer-run, peer-support program Rainbows, for example, has licences to operate in all provinces except Saskatchewan (see Appendices A and C). Still, availability seems spotty; Rainbows is currently available in most schools in Durham, Ontario, just east of Toronto, for example, but has only two sites in Prince Edward Island. Yukon is also considering the school system as a medium for providing education programs for children (McLeod, pers. comm., see Appendix D).

Counsellor-run emotional support and more intensive therapeutic programs are also available in major cities in many Canadian provinces, usually linked to other family services and provided by community agencies, hospitals or mental health institutions.Footnote 1 It is not known whether they are expanding. Providers report services are limited and under-resourced (pers. comm., community service providers across Canada). Similarly, individual-based therapy and counselling is available in all jurisdictions, but often only in major centres, and to a limited degree (see Appendix A).

Court-connected therapeutic programs appear to have expanded in British Columbia recently, but an intensive Ontario program for children with high-conflict parents in litigation—For Kids' Sake—recently closed. The United Kingdom and Australia also have some court-connected and community-based children's programs, but their extent is unknown. Children's programs are often offered as part of a package of mediation and conciliation services in England, Wales and Scotland. Programs linked with mediation and conciliation services are more likely to include some therapeutic emotional support to draw out children's feelings, improve communication between children and parents, and help children begin to deal with their feelings and experiences. Australia's Family Court also offers children's programs in some locations as part of its widely available mediation services (Strategic Partners 1999).

Overall, most court officials and community service providers indicate a need for more children's programs of all kinds.

3.4 Types of Programs and Evaluations of Effectiveness

It is difficult to categorize existing programs, given their enormous diversity. They are discussed here in five broad groupings, based on their goals, policy rationale, content, delivery and degree of therapeutic or clinical services. Few rigorous evaluations exist even for the major programs, although exit surveys of education and emotional support programs tend to show high client satisfaction.

3.4.1 Education and Information Programs

Among the fastest growing kinds of children's programs are those linked with, or corollary to, parent education programs. These programs vary widely, typically including information and education as well as an emotional support component. These two are difficult to distinguish in practice for very young children, because very young children respond primarily through their feelings. However, many of the children's programs linked with parent education, especially the court-mandated programs, emphasize education and information about the legal and emotional implications of divorce for children (for example, the Kids First programs in Hawaii, described in Appendix C). Most also seem to include some training in practical techniques for dealing with parents' inappropriate behaviour or their own acting out). However, courses may be short, perhaps one or two sessions, and may include as many as 35 to 40 children at a time, as in the Kids First programs (Anaya, pers. comm., see Appendix D). Education about the legal dimensions of divorce may be quite concrete. The first activity in a Kids First program is a tour of a courtroom, where children are encouraged to sit in the judge's seat and pound the gavel (Anaya, pers. comm., see Appendix D) (Di Bias 1996).

Where children's programs linked with parent education programs do include emotional support, their content will likely complement the parent program's goals of making parents' more aware of their children's needs and feelings, and focussing parents' actions on the child's best interests, rather than their own. Part of the purpose in helping children to feel and express their feelings about the separation is to increase parent-child communication, especially about the divorce, and so increase parents' awareness and attentiveness. The practical coping strategies offered in these courses are intended to help protect children when the parents fail to learn from their own program.

For example, the Toronto Jewish Child and Family Service's One Family, Two Homes is linked with concurrent parent-education workshops. A major goal of this program is to improve parent-child communication by helping the children express themselves, and helping the parents listen and attend to what their children express. The program was started two years ago because of long waiting lists for the more therapeutic Picking Up the Pieces program. Both programs aim to provide safe places for children to express feelings and explore coping strategies; the key distinction between them is that in One Family, Two Homes the issues are discussed generally, whereas in Picking Up the Pieces counsellors work with each child individually, focussing on what they are feeling and doing (Gertner, pers. comm., see Appendix D). The more intensive program also excludes pre-schoolers.)

As indicated above, education-oriented programs can be stand-alone (for example, the St. John's, Newfoundland and Labrador, court-based program is stand-alone and mixes education and emotional support). Some families who graduate from the short One Family, Two Homes program proceed to the more intensive program, but others say the first program gave them what they needed (Gertner, pers. comm., see Appendix D). Evaluations

Evaluations of education-oriented children's programs are sketchy and few. No evaluations were found that measured knowledge, attitudinal or concrete behavioural outcomes for children participating in education-oriented programs. In their absence, evaluations of parent education programs may offer some insight. Studies show non-didactic parent education programs—didactic programs have no impact (Arbuthnot et al. 1997)—may be effective in giving parents new knowledge that they retain, and in changing their attitudes (Arbuthnot and Gordon 1996; McKenzie and Guberman 1997, cited in Kirby 1998). Parents reported changes in their behaviours in several studies (e.g. Gray et al. 1997). However, one study also measured their actual behaviours and found most of the problem behaviours had not changed (Arbuthnot and Gordon 1996).

An evaluation of the mandatory Families in Transition program in Louisville, Kentucky (see Appendix C) measured short-term outcomes for graduates of their children's and parents' programs, using the Divorce Adjustment Inventory, and found that most graduates were "adjusting satisfactorily" to the divorce (Brown et al. 1994). The program combines education and emotional support. The evaluators also reported that fewer than 10 percent of families completing the mandatory program re-litigated child-related issues afterwards. However, without a control group for comparison, the program's effect remains unclear, since most families adjust satisfactorily to divorce over time anyway. Relatively few divorcing and separating parents re-litigate in any case, for example, and the families attending the courses are likely to be the most cooperative and concerned about their children's well-being.

One problem for American children's programs that are tied to mandatory parenting education programs is low attendance (see, for example, Jackson County, Missouri's Focus on Children in Separation, in Appendix C).

3.4.2 Programs Providing Therapeutic Emotional Support

Another common type of program uses therapeutic techniques to help children learn the coping skills required to work through emotional responses such as anxiety, blaming, anger, acting out and depression—responses to divorce that not only cause most pain, but could arrest a child's development. Besides providing information and education, these programs give children a safe place to express and share their feelings with others and with supportive adults. They are best seen as part of a continuum with the education-based programs described above; indeed, individual programs in the two categories may be indistinguishable on paper.

These programs may be offered by family service agencies, mental health institutions, counselling centres or school counsellors, and may be located in schools or at these agencies. Some may be linked with parent education programs, or parent programs that also use therapeutic techniques. Many are stand-alone. Programs in the United States

In the U.S., the Rollercoasters program is run largely by school counsellors, using a curriculum loosely based on Wallerstein's six "tasks" for children following separation and divorce discussed in section 1 (Fischer 1997).

The Children of Divorce Intervention Program (CODIP)is also a school stand-alone program that combines counselling techniques with emotional support, information and education, and teaches practical coping skills (Pedro-Carroll and Cowen 1985; Pedro-Carroll et al. 1986). This program, conceived as a preventive mental health measure, is well known for its extensive evaluations and positive reported outcomes (see below). Its curricula are widely used in such school programs across the country. The program was initially for pre-schoolers, but different versions have now been adapted for older children up to grade eight.

CODIP was built on the earlier, promising Divorce Adjustment Program (DAP), designed for psychologically healthy 7 to 13 year-old children experiencing parental separation or divorce and for their parents (Stolberg and Garrison 1985, cited in Shaw and Ingoldsby 1999). DAP's 12 Children's Support Group sessions were divided evenly between the discussing of divorce-related topics, and teaching coping skills to resolve problems, control anger, promote communication and put children at ease. CODIP reduced the anger control component and added a more general emotional support component (Pedro-Carroll and Cowen 1985). DAP also included parents in community-based single parent support groups and groups involving children and parents.

Another U.S. program, Kid's Turn, covers the same ground as CODIP and Rollercoasters, but links its children's program with a parallel parents' program (both parents attending when possible, but in separate sessions). Parents and children join in a potluck supper at the end of the course. Programs in Canada

Programs providing therapeutic emotional help and support to help children develop coping skills were encountered often in Canada during research for this report. Examples include the Family Services Association of Metropolitan Toronto's Families in Transition core program, It's Still O.K. in St. John's Newfoundland and Labrador, and the Manitoba government's Caught in the Middle. The first two programs are open to all children, but directors say that children entering these programs tend to be more distressed than the average (Sinclair, Freeman, pers. comm., see Appendix D), especially in Toronto.

Most of the programs encountered engaged parents individually in the program at some stage, reflecting providers' widespread opinion that parents need to be involved for the programs to have therapeutic effect, that is, to bring about the children's well-being and emotional and behavioural adjustment. Several programs had extensive intake processes of several hours (for example, the Family Services Association of Metropolitan Toronto's Families in Transition core program). When parents' and children's programs are linked, families may have case managers who follow up with parents, and leaders of both programs meet regularly to ensure that the right issues are raised in the right way in the groups.

Providers agree their programs are more effective when both parents are involved. However, where both parents are involved in a parental program of any kind, it is usual for them to attend separately. Providers also emphasize the need to mix genders in the parent groups. Evaluations

The evaluations of these programs are among the best available for children's programs.

Children of Divorce Intervention Program

The Children of Divorce Intervention Program's (CODIP) two pilot program evaluations are among the few to show positive results (Pedro-Carroll and Cowen 1985; Pedro-Carroll et al. 1986). Pre- and post-test studies found that children reported lower anxiety (and less anxiety than children in a control group) and fewer negative self-attitudes and attitudes about the divorce. There was no change in the children's perception of their competence and self-esteem. However, parents, teachers and program leaders reported that the children were less shy or anxious at school, had fewer school problems, were more competent (that is, showed less frustration, were more sociable, more compliant with rules and more appropriately assertive) and were less self-blaming, as well as less anxious overall (Pedro-Carroll and Cowen 1985).

A second pilot comparing children of separated or divorced families with children in intact families found that the children in the program again overwhelmingly showed less anxiety. Parents, teachers and facilitators again reported improvements on most fronts. With respect to many school and other external measures, the children in the program had caught up with those in the control group, whom they had been behind at the pre-test (Pedro-Carroll et al. 1986).

A recent follow-up of program participants two years later found that CODIP children were still less anxious than other children of divorce in a control group (Pedro-Carroll and Sutton 1999). Parents also reported increases in their children's coping skills and abilities to effectively handle divorce-related concerns. Children in the divorce control group had more behavioural problems and visited the school nurse more often.


Evaluation of the Rollercoasters program found that 85 percent of the parents reported improvements on at least one of five measures: children's general level of communication, their level of communication about the divorce, their willingness to express feelings, their acting-out behaviour, and their self-esteem (Fischer 1997). Children who were reported to be more communicative before the program were more likely to be reported as having become even more communicative afterwards. Parents were more likely to report improvements with regard to acting out, expression of feelings and self-esteem for children who were below average on these measures before the program. However, teachers reported no changes in children's negative behaviours (using the Behaviour Problems Index). The evaluation did not include children's self-reports.

Divorce Adjustment Program

The Divorce Adjustment Program (DAP) evaluation also found significant improvements in self-concept and adaptive social skills among the children in its program, during post-treatment and five-month follow-ups (Stolberg and Garrison 1985, cited in Shaw and Ingoldsby 1999). However, DAP found that results differed for its stand-alone children's program, the combined parent-child program, and its parents-alone program. Only children in the stand-alone program improved, and only parents who participated in the parents-only program improved. Neither parents nor children improved in the combined program (Stolberg and Garrison 1985, cited in Pedro-Carroll et al. 1986). A study conducted in the 1980s by the founder of Toronto's Families in Transition program also tested various combinations of programs and found the most effect for programs involving parent and child, with programs for parents alone more effective than programs for children alone (Freeman, pers. comm., Appendix D).

A later DAP study explored effects of its program on more troubled children (one half had clinically significant problems) and added workbook assignments to help the children transfer the skills they were learning in the program to their actual lives (Stolberg and Mahler 1994, cited in Shaw and Ingoldsby 1999). This study showed similar results for the children. But the workbook component made no difference to the gains. Moreover, the only positive outcome still in effect a year later was the improvement in children's behaviour in the home (Stolberg and Mahler 1994, cited in Shaw and Ingoldsby 1999).

Despite mildly promising results of some of the control group studies, other control group studies of similar types of programs have found no positive impacts (Lee et al. 1994; see also citations in Pedro-Carroll and Cowen 1985). Similarly disappointing effects were found for seven studies of parent interventions that used control groups (Lee et al. 1994). Studies using no control groups have shown similar mixed results (e.g. see citations in Crosbie-Burnett and Newcomer 1990).


Although it did not measure child outcomes, a recent evaluation of the Centres jeunesse de Montréal's Confidences program showed that 80 percent of the 112 child participants felt the program had helped them. Thirty-six percent said talking about their parents' separation had helped them, and 15 percent reported that talking had helped them understand the separation (Vallant 1999). Parents said that their children's sense of security had been increased by seeing other children in the same circumstances, and nearly as many thought the program helped their children express themselves more easily about the separation. A quarter of the parents reported that their child was calmer, and one fifth felt their children expressed themselves more readily about the separation (see Appendix A).

3.4.3 Lay Programs Providing Peer Emotional Support

Another group of programs also provides children with a safe space in which to explore their feelings at the time of parental divorce—especially their feelings of loss, fear and even hopelessness—with other children under the guidance of trained volunteers. The programs range from simple "rap sessions" to more structured activities. They may include teaching coping skills to help children stay out of the middle, and education about the legal terms and processes concerning separation and divorce, and their effects on children. In Brighton, England, for example, Relateen program counsellors provide loosely structured discussion sessions for local 11 to 18 year olds whose parents have separated and divorced (http://www.brightonrelate.org.uk).

In Canada and the United States, the Rainbows programoffers 12 weekly sessions for groups of similarly aged children plus two multi-group days. It is open to all children, including those whose parents have been separated or divorced for many years, and serves children suffering loss by death in the family as well. It is run by teachers, sympathetic adults and sometimes guidance counsellors who undergo brief training. In Canada, Rainbows is licensed to operate at 1070 sites, that is, that number of organizations have completed the training and been approved to provide the program. About 9,000 children's workbook/journals were sent to various program localities in 2000.

In British Columbia, Vancouver's ARK Child Services Society has also developed a curriculum for peer support groups, Connections, to be run by professionals trained in listening skills. The Boys and Girls Clubs in the province also run support groups for school-age children and teenagers suffering loss, including family break-up.

An evaluation of 97 fourth, fifth and sixth graders attending Rainbows programs at 28 schools near Chicago found positive responses but few gains for these children, compared to a control group (Kramer and Laumann 2000). In high-conflict families, pre- and post-tests found some improvement in the perceptions of children of one of the coping skills: positive reappraisal, i.e. the ability to look on the bright side. There was no change in children's perceptions of their adjustment, the quality of their relationship with the parent they spend most time with, and other coping skills such as getting support from peers, avoiding strategies that reflect hopeless or blaming attitudes, and seeking support from other adults. Children's perceptions of their well-being did not change, although it declined for children of high-conflict families in the control group. This suggests that the program might provide some stabilizing support to its children.

Counsellors express some concern that lay facilitators in the lay programs may not be able to read the signals indicating that children in high-conflict families, or otherwise in distress, are reacting adversely to what is happening in the group. One strength of these programs seems to be their ease in normalizing the divorce and separation process.

3.4.4 Focussed Skills Teaching Programs: Children in the Middle

Separating and divorcing parents frequently put their children in the middle of their conflicts by demanding that they choose sides and work with one parent against the other. Teaching children strategies to deal with being caught in the middle is a popular component of children's programs. Many community agencies make the video-based Children in the Middle program materials the primary focus of their program. Since the program has both parents' and children's components, it is also often used as the basis for parenting education programs that include a children's program component. The children's program has been distributed to more than 500 service providers around North America.

Therapy and mediation programs targeted to high-conflict litigious parents are also likely to deal with children being caught in the middle, since this is more typical in situations involving high-conflict parents. But the problem in high-conflict families usually reaches a different order of magnitude than in other families. One common feature of high-conflict parents is that they have extreme difficulty hearing and listening to their children, or to each other. The education-oriented Children in the Middle is not designed for these families.

Manitoba's court-based Caught in the Middle is targeted to children whose parents are in custody and access legal disputes (Bewski, pers. comm., Appendix D). Although it deals with being caught in the middle, it is designed to provide more general therapeutic emotional support and to teach coping skills for children akin to those provided in the programs discussed above.

There has been considerable evaluation of specific Children in the Middle programs, although mostly of the parent-education version of the program (see Arbuthnot et al. 1997; and Arbuthnot and Gordon 1996, for some of the most rigorous ones). However, one small study of 33 fourth, fifth and sixth graders found significant improvements in children's stress levels (Kearnes et al. 1991). In a four-week follow-up, children reported the frequency and stress of situations in which they felt caught in the middle. The children reported experiencing significantly less stress than those in a control group who watched a non-skills-oriented divorce video When Mom and Dad Break Up. Improvements were clinically significant for 50 percent of the children. The program appeared to have no impact on frequency.

3.4.5 Therapy-Based or Clinical Interventions

These kinds of programs are distinguished by their therapeutic intensity, whether in the children's program itself or in a parents' program. Children's groups will often include two therapists (or one with an assistant), and focus on therapeutic emotional support and therapeutic training in coping skills to manage feelings, such as despair and depression, and reduce anger and acting out. When the children's program is linked with a parental program, the intensity would be equally located in the parents' program, where small groups of couples are attended by two therapists working to identify and resolve entrenched destructive behaviours among the parents.

These intensive programs are usually directed at children in obvious distress, or children in high-conflict divorcing and separating families. The programs are often linked with individual and family therapy services (parents or children may be receiving both concurrently, or may move between the program and individual therapy) and may be provided by family service agencies, counselling and therapy centres, and mental health institutions. Clinicians who work with these families emphasize how typically intransigent and hostile the parents are, and how difficult it is to get them to attend to or recognize their children's suffering (Hood, pers. comm., Appendix D).

Programs that link children's and parents' programs may also be court-connected. Families are often referred to these programs because they are locked in court battles, frequently involving violence and abuse. The ultimate goal of the parents' program may be to resolve those disputes using a combination of intensive group therapy and mediation. In these cases, the children's program provides a venue for children to receive therapeutic emotional help and to learn coping skills. One primary function, though, is to give the parents feedback about their children's needs and concerns, and to reinforce and intensify the therapeutic push to change the parents' understanding and behaviour (Johnston and Campbell 1988, cited in Brown 1995). These programs provide what many experts consider to be the least threatening way for children to voice their concerns about their parents' conflict and separation to their parents (Brown 1995).

California's center for Families in Transition has been a leader in developing court-connected versions of these therapy-intensive interventions in the state (Johnston and Campbell 1988, cited in Brown 1995). Family court services in Alameda County, for example, runs linked parent and child programs for families before its court. To enter the program, children must be clearly suffering and the parents must have failed mediation at least twice. The children's programs provide peer support to the children. After carefully preparing the children and parents, the two groups are eventually brought together for a session about halfway through the eight-session program (Schepard 1998). In these kinds of programs, children convey their messages to their parents as a group through such devices as stories, pre-recorded videos, role-playing or videos (Brown 1995).

These safe techniques for providing feedback to parents are also used in less clinically intensive programs that include concurrent parent and children's programs, and in which parents and children join in a group meeting at some point. In many other programs, children prepare these messages together as part of their activities (a group letter is popular) and deliver them individually to their parents.

The Family Centre of Winnipeg's Giving Children Hope is a targeted program for parents and children in high-conflict families (usually also in litigation) that combines therapy and mediation for the adults. Its children's sessions, which start a few weeks in advance of the parent sessions, involve several therapists. Parents in litigation must suspend it during the course. An evaluation of the program is in progress.

No evaluations of other intensive therapy programs for children were found. Evaluations of the hybrid mediation-therapy programs measured parental outcomes only. A two to three year follow-up of two studies of high-conflict families found that two thirds were able to keep or renegotiate their own agreements regarding custody and access, and hence stay out of court (see citations in Johnston 1994).

3.4.6 Evaluation Overview

Overall, there has been little evaluation of children's programs, and existing evaluations tend to be compromised by small sample sizes and by the use of limited research designs and methods, often due to the external parameters of the evaluations (Fischer 1997; Grych and Fincham 1992). There seem to be no long-term follow-ups.

Many of the program evaluations encountered in the literature stop at recording participants' satisfaction with courses and their willingness to recommend the program to others. On this measure, children's programs tend to score highly. So do parent education programs. However, there is only some evidence to show that parental education programs can actually change parents' attitudes, and little or no evidence that they change parental behaviours (see above). It is even more difficult to assess the impact of children's programs on their emotional states, attitudes and behaviours when, in many studies, assessments are based on parents', teachers' or group facilitators' reports of how the children are feeling or adjusting. Children's self-reports, on the other hand, may be no more accurate than parents' self-reports about the extent to which they are aware of their children's needs and feelings, are communicating well with their children, or are putting their children in the middle of their parental conflicts.

In an overview of existing studies in the early 1990s, researchers raised several other concerns (Grych and Fincham 1992). First, they pointed out that most evaluations do not assess whether the goals of the group have been achieved (e.g. whether children have a better understanding of divorce) before assessing whether the program improved children's functioning. Only one study referred to above (Roseby and Deutsch 1985, cited in Grych and Fincham 1992) specifically tested whether children had understood the significance of divorce, for example, before testing for adjustment outcomes.

Secondly, the teachers and parents who rated the children's adjustment were usually fully aware that the children had been in the program, and so may have been biased (the halo effect). In the Children of Divorce Intervention Program evaluations, this possible effect for the adults' assessments was offset by the children's own mixed assessment of changes in their adjustment.

Thirdly and finally, researchers argue that evaluation studies should distinguish between children living in single-parent homes and those in stepfamilies, since experiences and problems of these two groups of children may be very different (Grych and Fincham 1992). Similarly, some of the studies do not distinguish children whose parents have recently separated from those who may have been divorced for several years.

Another problem raised by researchers is the lack of good instruments for measuring children's adjustment outcomes (e.g. Freeman 1995; Pedro-Carroll and Cowen 1985). The psychometric properties of several of the measures commonly used have yet to be determined. Moreover, some researchers have expressed concern that behavioural measures, in particular, tend to focus on negative behavioural outcomes. There are few measures to assess gradations of positive behavioural outcomes (e.g. Amato 1994).

The dearth of evaluations on children's programs limits assessment of their value. It also means there is no research to guide policy makers in deciding, for example, whether it is more effective to resource programs for parents (perhaps parent education, possibly support groups) or programs for children, or only programs involving both, in order to alleviate children's distress and enhance their well-being.

Similarly, although courts in the U.S. clearly tend to tie children's programs to parent education programs, the impression is that Australian and British courts are expanding children's programs primarily as adjuncts to mediation or conciliation. There is no research yet to say which works best, for children directly, for their parents, or for custody and access decision making committed to securing children's best interests.

Another important issue is whether targeted, more intensive interventions—possibly to children exposed to many post-separation stressors and lacking supportive parents or social supports (Wolchik et al. 2000)—are more needed than broad-based interventions for children. After all, the research predicts children will rebound fairly quickly from the acute distress of separation and divorce, if their parents also rebound.

In spite of the lack of evaluation of programs for children experiencing parental divorce and separation (cited as the main reason for so few Canadian program evaluations), there appears to be strong support for these programs among community social service providers, other community sectors and parents. Providers believe strongly these programs work for children. Most appear to believe that parents should also be involved, but not necessarily through a linked or parallel program. In many programs, parents are very keen for their children to participate, and are eager to know counsellors' assessments at the end of the sessions (Filion, Gertner, Freeman, pers. comm., Appendix D).

Providers of lay-led emotional-support programs have expressed some concerns about counsellor-led programs, and counsellors leading programs expressed concerns about lay-led programs. These two types of programs share the goal of helping children understand the significance of divorce, that is, coming to terms with the fact that their parents will probably not reconcile, but that they will not be abandoned by both parents. However, a support group provides less structured support. The relative effectiveness of the different program approaches, and the groups of children for which the different approaches might be effective, is another gap in the research. Moreover, as indicated earlier, the one study that explored the impact of this understanding (Roseby and Deutsch 1987, cited in Grych and Fincham, 1992) found that understanding did not improve children's adjustment.