freydís jóna freysteinsdóttir 2 Characteristics of families who are repeatedly reported to child protection services in Iceland

Transkript

1 LEDER Detta nummer av Nordisk Socialt Arbeid inleds med två artiklar som båda fokuserar på barn och deras villkor. Freydís Jóna Freysteinsdóttir diskuterar i sitt bidrag vikten av att tidigt kunna identifiera familjer i vilka barnmisshandel förekommer och sådana riskfaktorer som om de elimineras kan få stopp på dessa missförhållanden. Mia Hakovirta och Mari Broberg uppmärksammar därefter de barn vars föräldrar inte finns närvarande i deras vardag och sådana förhållanden som ökar möjligheten till en god kontakt i lägen där barn och föräldrar inte bor tillsammans. Huruvida kombinationen könsmakt och profession inom psykiatrin får som konsekvens att manliga arbetssökande tenderar utestängas problematiseras av Rita Sommerseth medan Karin Haarberg Aas diskuterar möjligheter att i utbildningssystemet stärka utvecklingen av vad hon benämner professionell relationskompetens. Hon konstaterar att utbildningssystemen i Norden lägger stor vikt på undervisning och träning i kompetenshöjande syfte men att såväl studerande som nyligen utexaminerade uppger sig ha svårigheter att knyta samman praktik och teori inom detta område. Heidi Gautuns artikel slutligen handlar om samspelet mellan välfärdsstat och individ. Hon driver tesen att en ökad förståelse av detta samband kräver hänsyn till sådana kulturella sammanhang som ofta förbises i europeisk idétradition. Numret avslutas som vanligt med anmeldelser och bokomtaler. INNHOLD freydís jóna freysteinsdóttir 2 Characteristics of families who are repeatedly reported to child protection services in Iceland mia hakovirta & mari broberg 19 Parenting from a Distance: Factors Connected to the Contact between Children and a Non-resident Parent rita sommerseth 34 Menn i psykisk helsearbeid utstøtes de som har sosialfaglig bakgrunn? karin haarberg aas 45 Utvikling av relasjonskompetanse. Det er vanskelig å sette ord på det, men det ligger liksom i bakhodet ett eller annet sted heidi gautun Velferdsstatens forming av individers livsløp 56 anmeldelser 64 bokomtaler 80 kalender 87 Lena Dahlgren Huvudredaktör för Nordisk Socialt Arbeid

2 REFEREEBEDØMT Antaget til publicering januar 2007 Freydís Jóna Freysteinsdóttir Characteristics of families who are repeatedly reported to child protection services in Iceland FREYDÍS JÓNA FREYSTEINSDÓTTIR professor in Social Work at the University of Iceland It is important to identify families who are at risk of repeatedly maltreating their children early in the process in order to develop specific ways to help these families stop maltreating their children or to work on termination of parental rights. Thus, identifying factors that contribute to repeated child maltreatment is essential. The focus of this study was to investigate risk factors for repeated maltreatment, specifically neglect, of children in Iceland. Keywords: repeated child maltreatment child neglect risk factors ecological model child protection services NORDISK SOSIALT ARBEID NR VOL 7 SIDE Universitetsforlaget 2-18.is

3 characteristics of families who are repeatedly reported to child protection services in iceland Incidence of child maltreatment During 1999, at the time the study was conducted, a total of 1,067 cases of child maltreatment (children who had been abused, neglected or engaged in risk behavior) were reported and required some kind of action by the child protection authorities in Iceland (Government Agency for Child Protection, 2000). The total number of children under 18 years of age in Iceland in 1999 was (Statistics Iceland, 2004). Thus the incidence rate was roughly 1.4%. Most of the reports (738) included children with risk behavior (not victims) or children s risk behavior, such as difficulties in school (such as school attendance) or child puts own health or maturity in danger (Government Agency for Child Protection, 2000). Even though these cases include children (mostly adolescents) with risk behavior, rather then as victims, their severe symptoms may suggest that many of them have had a history of difficulties in childhood, including maltreatment (Freydís Jóna Freysteinsdóttir, 2005), since some children develop risk behavior following maltreatment (Miller-Perrin & Perrin, 1999). Fewer reports included children as victims of maltreatment (329). Neglect was the most common form of maltreatment in Iceland in Child neglect can be defined as a failure to fulfill a child s basic needs (Miller-Perrin & Perrin, 1999) and has been divided into four main categories in Iceland; physical neglect, inadequate supervision, educational neglect and emotional neglect (Freydís Jóna Freysteinsdóttir, e.d.). Ofcourse, neglect has different forms depending on the age of children. For example, a three year old child left home alone is neglected (inadequate supervision). However, a 13 year old child left alone would not be considered neglected, unless it was home alone during a long period of time. Sixty-eight percent of the victims suffered from neglect in Iceland (Government Agency for Child Protection, 2000). In addition, roughly 15% of victims had parents with a recorded alcohol or drug abuse problem, which was a specific category of child maltreatment in Iceland (Government Agency for Child Protection, 2000). However, parental alcohol or drug abuse problem should be viewed as a risk factor for child maltreament. Ten percent of the victims suffered from sexual abuse, 4% from psychological/emotional abuse and 3% from physical abuse (Government Agency for Child Protection, 2000). The effects of child maltreatment Maltreated children are at risk of experiencing physical, cognitive, social and emotional problems (Miller-Perrin & Perrin, 1999) and are more likely to have mental health problems than other children (Turner, Finkelhor & Omrod, 2006). Results from a recent Icelandic study seems to support these findings (Halldór Guðmundsson, 2005). The potential negative effects of maltreatment on children can be divided into two broad categories, internalized and externalized behavior problems. Examples of internalized behavior problems are anxious behavior and withdrawal. Examples of.is

4 freydís jóna freysteinsdóttir externalized behavior problems are aggression and delinquent behavior (Markward, 1997; Saunders, 1994; Zielinski & Bradshaw, 2006). Externalized behavior problems are more common among boys, whereas internalized problems are more common in girls (Jouriles & Norwood, 1995). Symptoms such as anxiety and behavior problems that are frequent or prolonged often lead to the diagnosis of a psychiatric disorder (Finkelhor & Berliner, 1995). For example, excessive withdrawal may be a sign of depression and excessive delinquent behavior may lead to conduct disorder (American Psychiatric Association, 1994). Since the effects of maltreatment have been documented, it has become clear that repeated child maltreatment is more likely to lead to severe symptoms (Markward, 1997; Saunders, 1994). Severe behavior problems and excessive withdrawal may reflect a disorder (American Psychiatric Association, 1994). Thus, when the maltreatment is repeated, even after it has been reported and intervention has occurred, children may be more likely to develop disorders (Cicchetti & Lynch, 1995; Knutson, 1995). Because child maltreatment can have serious consequences, it is important to study repeated child maltreatment. The ecological model as a conceptual basis This study was based on the ecological framework of Belsky (1980), which was grounded on Bronfenbrenner s (1977) ecological model. Belsky s (1980) ecological model organized different risk factors by levels. The first level consists of individual factors (ontogenetic factors). The second level consists of the micro system, which is what takes place in the household. The third level includes a larger social system in which the family is embedded (exo system) and the fourth level includes cultural beliefs and values (macro system) that influence the micro system as well as the exo system. Child maltreatment is a phenomenon, which is likely to occur as the result of a complex interplay of different stress-producing factors (Belsky, 1980). For example, an eighteen year old single mother who takes drugs, rejects her infant who has been born prematurely, is disconnected from her family, is unemployed and lives in a society where it is socially acceptable to use physical punishment is at greater risk of maltreating her child than a twenty five year old married and employed middle-income woman who has a supportive husband as well as supportive relatives that live close by in a culture where it is not acceptable to use physical punishment. Several risk factors have been found at various levels of Belsky s (1980) ecological model related to parents (Berger, 2005; Cicchetti & Lynch, 1995; Sidebotham & Heron, 2006), and children (DePanfilis & Zuravin, 1999; Wu et al., 2004), as well as family factors (Belsky & Vondra, 1989; Berger, 2004; Coohey, 2006; Sidebotham & Heron, 2006), social factors (Connell-Carrick & Scannapieco, 2006, Freydís Jóna Freysteinsdóttir, 2005) and cultural factors (Belsky, 1980; Belsky & Vondra, 1989; Shor, 1998). The general research question in this study was: What risk factors at different eco-.is

5 characteristics of families who are repeatedly reported to child protection services in iceland logical levels are associated with repeated child maltreatment, specifically neglect, in Iceland. Method A case comparison design was used to compare two groups on a variety of factors at different ecological levels, which have been related to child maltreatment. Families where one incident of confirmed child maltreatment had taken place were compared with families where two or more confirmed child maltreatment incidents had taken place. All cases had at least one confirmed report of neglect as the first incident, so the main variable studied was neglect. Neglect was divided into the four main categories discussed above. Reported incidents of neglect varied ofcourse with different age of children. For example, a very young child left alone might be an example of inadequate supervision. However, a teenager wandering outside late at night because the parents have not set rules about when to come home, is inadequately supervised as well. Thus, all the cases fulfilled criteria for maltreatment and neglect in first incident using the four categories of neglect discussed above. The groups were compared on a variety of factors at different ecological levels to assess differences between them. In both groups, only cases with no prior registered incidents were included, which is a strength of this design. By studying clean cases, that is, cases with no prior incidents, it is possible to detect risk factors that increase the likelihood of further child maltreatment. The sampling frame included all case records of child maltreatment incidents reported to two social service agencies in Reykjavík (capital city of Iceland), from 1. January 1998 through 31. August 2000 and one social service agency in Hafnarfjörður, from 1. January 1998 through 31. December Hafnarfjörður is located within 10 miles of Reykjavík and can be considered as a part of the Reykjavík metropolitan area geographically. Families constituted the units of analysis in this study, rather than children. Cases were divided into two groups: 1) Single incident group (cases with one reported incident of neglect), 2) Repeated incidents group (cases with two or more reported maltreatment incidents, and neglect at the first incident). The following exclusion criteria were established, in order to increase the internal validity of the study. First, only cases with children as victims of neglect in the first incident were included in this study, although other types of maltreatment (physical, emotional abuse or sexual abuse) could have involved in repeated incidents. In all cases, neglect was the first incident and cases with other types of first incidents were excluded from this study because each type of maltreatment has specific characteristics and thus it can be problematic to combine all types in a single category when trying to understand repetition among neglecting families (Knutson, 1995). Second, as already mentioned, only cases with no known prior incidents (before 1. January 1998) were included in this study. Third, the study only included cases with families 5.is

6 freydís jóna freysteinsdóttir that lived within the city limits of Reykjavík on the one hand and Hafnarfjörður on the other hand when the first incident was reported and when the data were gathered. This factor is important, since the cases were followed for a specific time period and the families could otherwise have been reported for maltreatment in other jurisdictions. If it appeared that families had been living in different areas for many years they were excluded from this study, except for families where there was a strong reason to believe from case notes that the family had not been involved with child protection services (CPS) before. This exclusion, however, does limit the study to families that live a rather stable live in terms of residence. Fourth, foster families and adoptive families were excluded, to ensure that other factors, such as an earlier removal from a family, were not interfering with the factors studied. Thus, only families with the biological children of one or both parents were included in this study. Fifth, the target child was 16 years old or younger at the time of first incident. This requirement was made in order to allow time to follow cases for 18 months, with children still under 18 years of age at the end of the time period. Sixth, only cases that consisted of a parent as responsible for the maltreatment in one or more incidents were included. Content analysis was used as a method for translating information from the child protection worker s reports into quantitative data. Content analysis is an unobtrusive method (Rubin & Babbie, 1997) and can be defined as a systematic, replicable technique for compressing many words of text into fewer content categories based on explicit rules of coding (Stemler, 2001, p.1). Coding sheets were created that included various questions about the incident and the family. Coding guidelines were also used, including a coding system in a coding manual, to increase the reliability of the coding. A coding manual can be defined as a set of instructions or rules on how to systematically observe and record content from text (Neuman, 2000, p.294). The data were analyzed using SPSS statistical software. Results Descriptive statistics: A total of 517 cases that fulfilled certain criteria were reviewed (in the child protection agencies in Reykjavík, a case is recorded for each child, however in the child protection agencies in Hafnarfjörður, a case is recorded for each family). Because of extensive exclusion criteria only a total of 77 cases (one case for each family) were coded in this study, 38 single incident cases and 39 repeated incidents cases. The police 1 was the most frequent reporter in single incident cases, followed by school/playschool/child care provider, neighbors and parents (usually without custody). However in most cases the reporter was other in repeated incident cases (unknown how the reporter is connected to the family, followed by the police, parents (usually without custody) and neighbors. 1 The police is required to report domestic violence to child protection services if a child is located at home when it occurs..is 6

7 characteristics of families who are repeatedly reported to child protection services in iceland Age of parents at first child maltreatment incident was similar in the repeated incidents group and in the single incident group. The mean age of mothers in the repeated incidents group was 33 and the mean age of mothers in the single incident group was 32. The mean age of fathers was 36 in both groups. The parents in this study had a total of 152 children. All of the children were living in the home at the time of the first incident except one, who lived with relatives. The range of number of children was from 1 to 5 total, but 1 to 4 living at home. Interestingly, parents in the repeated incident group had fewer children than the parents in the single incident group. More parents in the repeated incidents group had two children than in the single incident group, but fewer parents had three children in the repeated group than in the single incident group. Age of the children was similar in the groups and there was a similar number of boys and girls in the two groups. In more cases information was missing in the single incident group on where the child stays during the day. When missing information was excluded, a similar ratio of children in both groups were in playschool or day care and a similar ratio of children were staying at home. In 18 cases the child/ren had a different family type 18 months after the first incident (11 repeated incidents cases and 7 single incident cases). Fewer children in the 7.is

8 freydís jóna freysteinsdóttir repeated incidents group were living with both biological parents than in the single incident group. More children in the repeated incidents group were living in a single parent family, than in the single incident group. Finally, fewer children in the repeated incidents group were living in a step family compared to the single incident group. Thus, in fewer cases children were living with two parents in the repeated incidents group compared to the single incident group Type of housing did not differ for the two groups. However, more families had moved during the twelve months preceding the first incident in the repeated incidents group compared to the single incident group. For a higher proportion of single incident cases it was unknown if the family had moved or not. In most cases parents had only completed elementary school. However, education was unknown in a number of cases, especially for mother s former partners. More mothers had only completed elementary schooling in the repeated incidents group than in the single incident group. A similar number of mothers had completed upper secondary school and a similar number had completed a university degree in both groups. Education of fathers was similar in both groups (Table 1). Table 1. Education of parental figures. Single incident group Education Mother Father in home Father out of home Mother Repeated incident group Father in home Father out of home Elementary Trade school Upper secondary school University Unknown Total In the repeated incidents group, mothers were employed in over half of the cases, and were unemployed/disabled/homemaker in slightly fewer cases. In the single incident group, fewer than half of the mothers were employed, and a third of the mothers were unemployed/disabled/ homemaker. The fathers in the repeated incidents group were employed/students in over half of the cases and unemployed/disabled in 2 cases. In the single incident group, over half of fathers were employed and 5 were unemployed/disabled. Information about this important factor was missing in many cases (Table 2)..is 8

9 characteristics of families who are repeatedly reported to child protection services in iceland Table 2. Employment status of parents. Employment Single incident group Repeated incident group Father in Father out Father in Father out Mother home of home Mother home of home Self employed In paid job Parental leave Homemaker Student Disabled/Sick leave Unemployment Prison Unknown Total More than two thirds of the families were experiencing financial problems at the time of the first incident and nearly half had been on welfare at some time prior to the first incident. In the repeated incidents group, 27 (of 39 total) were experiencing financial problems and 21 had been on welfare. In the single incident group, 23 (of 38 total) were experiencing financial problems and 16 had been on welfare. A total of 153 child maltreatment incidents were reported for the groups. In 98 of the incidents, the mother was responsible for the maltreatment incident. In 46 of the cases the father was responsible for the incident, in 22 cases the mother s former partner was responsible, and finally in 4 cases someone else was responsible for the maltreatment incident. In thirteen cases children left the home or were removed by someone else, such as an extended family member (7) or by Child protection services (6). Lack of supervision/inadequate supervision was the leading type of child maltreatment. In 65 incidents the type was supervisory neglect. In 20 incidents, the children had been exposed to domestic violence. In a total of 35 incidents their basic needs had not been met, the type was unspecified, or the child had suffered from two or more types of maltreatment. Interestingly, general supervisory neglect (a failure to monitor a child) was the leading type of maltreatment in all incidents and was the only type involved in fifth and sixth incidents (Table 3). In all cases except three it was unknown when the reported maltreatment had started or it was believed it had started when reported or a few days earlier..is

10 freydís jóna freysteinsdóttir Table 3. Type of maltreatment. Type Denial of critical care - unspecified Inc. 1- single Inc. 1-rep Inc. 2 Inc. 3 Inc. 4 Inc. 5 Inc. 6 Total Physical neglect School neglect 1 1 Inadequate supervision Emotional maltreatment including domestic violence Physical abuse Sexual abuse More than one type Total The children in the groups were experiencing various problems that may be a result from the maltreatment they were receiving, since child maltreatment leads to various problems (Miller-Perrin & Perrin, 1999). School problems (for example, having low grades and/or frequently absent from school) were the problems most rated in both groups, followed by internal problems (such as depression or anxiety). However, since inter-rater reliability was poor for internal problems, that variable was not included in the analysis. A few children in the repeated incidents group had harmed other children sexually, compared to none in the single incident group. However, other problems were similar between the groups (Figure 2)..is 0

11 characteristics of families who are repeatedly reported to child protection services in iceland Bivariate statistics The demographic variables analyzed in this study, were family type, family size, age of parents at first reported child maltreatment incident, the education level and employment status of parents, type of housing and financial status. Ecological variables were also analyzed, which were parental problems, child vulnerabilities, family dynamics and social support deficits. It should be noted that some variables can fit into different categories, for example, financial status can be viewed as a demographic variable and a social support deficit (lack of adequate income). However, in this study, only informal social support was included in the social support variables. Finally, parental cooperation and types of services the families were provided were analyzed. A significance level of.10 was chosen as criteria, because a traditional significance level of.05 often fails to identify important variables when the sample is small, which is the case in this study (Hosmer & Lemeshow, 1989). Furthermore, when a sample size is small, the power to detect relation is week (Hays, 1994). Three demographic variables were significantly related to the dependent variable. The families in the repeated incidents group were more likely to include a single parent and to have moved within the last 12 months. In addition, the mother figures in the repeated incidents group had a lower education level than mothers in the single incident group. Individual factors: Parental problems and child vulnerabilities The mothers in the repeated incidents group had higher frequency of alcohol problems, depression and other mental illness. Furthermore, higher number of mothers in the repeated incidents group had mental deficiency than mothers in the single incident group. Finally, more mothers in the repeated incidents group had been maltreated in their own childhood than in the single incident group (Table 4). An index was created for problems related to the mother (one variable that includes the sum of all mother figure problems) and for all problems related to the father (one variable that includes the sum of all father figure problems). The mothers in the repeated incidents group were experiencing a significantly higher number of problems than the mothers in the single incident group. The fathers in the repeated incidents group had been maltreated in their own childhood and moved in or out of the household more often than fathers in the single incident group. Other problems related to fathers were not significant. There was a significant difference between the groups on two child vulnerability variables out of nine. More families in the repeated incidents group had a child with a mental deficiency and/or a child who was noncompliant than in the single incident group. However, there was no significant difference between the groups on number of child vulnerabilities..is

12 freydís jóna freysteinsdóttir Table 4. The Bivariate relation between the parental risk factors and repeated child maltreatment Risk factors: Parental risk factors: Mother figure problems: Number Single incident Repeated incidents p value Alcohol abuse*** Other drug problems Depression*** Other mental health problems**** Mental deficiency** Chronic physical illness Maltreated in own childhood** Has committed crimes Has moved in or out of the household the last 12 months *p<.10. ** p<.05. *** p<.01.**** p<.001. Family and social factors: Family dynamics and social support Coercive parenting or discipline problems were more common in the repeated incidents group than in the single incident group, as well as conflict between parental figures (Table 5). Thus, the families in the repeated incidents group were experiencing significantly more problems in family dynamics than the single incident group. A similar number of families received social support from extended families, friends and groups in the repeated incidents group and in the single incident group. The repeated incidents group received social support from community/neighbors in more cases than the single incident group. The social support deficit index (a variable that includes the sum of all social support deficits variables) was not significant in the bivariate statistics. Table 5. The Bivariate relation between family dynamics and repeated child maltreatment Family dynamics: Coercion pattern/discipline problem between a parent and a child*** Conflict in the parental relationship** *p<.10. ** p<.05. *** p<.01.**** p<.001..is

13 characteristics of families who are repeatedly reported to child protection services in iceland Parental cooperation and services Majority of mothers were available for all interviews with the child protection worker at the child protection services at the agreed upon times. However, less than half of fathers were available for all interviews at the agreed upon times. In only two cases fathers not living at home were available for any interview. Very few parents minimized other s abusive/neglecting behavior. Rather few parents admitted that s/he was responsible for the maltreatment. It seemed that they had difficulties facing that they had done something wrong. However, majority of parents in both groups were willing to do something so that the maltreatment would not recur. No difference was found between the groups on any of the parental non-cooperation variables or the parental non-cooperation indices. More parents were offered/referred to services in the repeated incidences group than the single incident group. When offered services, more parents in the repeated incidents group than in the single incident group accepted the services. Of those who did not accept services, majority of mothers and half of fathers gave a valid explanation according to the child protection worker for not accepting services, such as that they had already seeked a support of some kind. Interestingly, only in two cases (single) were victims referred to a service because of domestic violence. In addition, in 8 cases (2 single, 6 repeated) it was recorded that the mothers had ever used a domestic violence service. All families were seen for at least one interview with a child protection worker and received counseling on child upbringing and child care. More families in the repeated incidents group than the single incident group received some other types of services: family adviser, supportive family, child/day care, child placed at a farm during summer and therapy/treatment for parents. The groups received other types of services in similar numbers, such as youth adviser and counseling or therapy for the child (Table 6). Fewer cases in the single incident group were submitted to the child protection committee than in the repeated incident group. Cases are introduced to or submitted to the child protection committee, when they are perceived as serious and/or the parents and/or a child that is 15 to 17 years old do not cooperate with the child protection services. Three other types of mandated services were given to families in the repeated incidents group but not the single incident group. However, only two mandated services (services not approved by parents) were significant: case submitted to child protection committee and child removed from home. The social service index (a variable that includes the sum of all social service variables) was significantly related to repeated child maltreatment in the bivariate statistics..is

14 freydís jóna freysteinsdóttir Table 6. The Bivariate relation between repeated child maltreatment and social services Social services Supportive services Family adviser*** Youth adviser Supportive family* Child receives counseling or therapy Childcare (day care) provided*** Housing provided (or waiting list) Child placed on a farm during summer* Parent(s) receive therapy or treatment*** Child receives short term treatment Child receives long term treatment Child temporal foster care Child permanent foster care *p<.10. ** p<.05. *** p<.01.**** p<.001. Multivariate statistics Three demographic variables were selected to be entered into a logistic regression model, since they were significantly related to the dependent variable at a.10 level: family type (single parent vs. two parent families), education level of mother figure and number of times the family has moved the last 12 months. When the three variables were entered simultaneously into the equation at step one, family type and number of times the family has moved the last 12 months were not significant and were not in the final model. Indices were created for mother figure problems, father figure problems, child vulnerabilities, family dynamics, social support deficits, mother figure non-cooperation, father figure non-cooperation and social services. At step two, significant risk indices were entered into the equation, first mother figure problems and then family dynamics. Child vulnerabilities, father figure problems and social deficits were not significant and therefore not in the final model. At step three, services were entered into the model, since that index was significant. However, when that index was entered into the model, it was not significant and therefore not in the final model. Mother non-cooperation and father non-cooperation were not entered into the model since these indices were not significant. As can be seen in Table 7, the final model explains a total of.443 of the variance in the dependent variable: repeated child maltreatment. In addition, the final model correctly classifies 71.8% of repeated incidents cases and 76.3% of single incident cases or a total of 74% (Table 8). The log likelihood was and the model s chi-square was is

15 characteristics of families who are repeatedly reported to child protection services in iceland Table 7. Hierarchical stepwise logistic regression (N=77) β Chi Square p Value Odds Ratio R 2 Demographics:.064 Education of mother Risk factors:.443 Mother figure s problems Family dynamics Table 8. Comparison of actual maltreatment incidents with prediction of maltreatment Predicted Single incident Predicted Repeated incidents Total Single incident Repeated incidents Sensitivity 71.8 % Specificity 76.3 % False positive rate 23.8 % False negative rate 28.2 % Positive predictive value 74.0 % Discussion In this study, a strong relation was found between maternal problems at the individual level and conflict in the parental relationship at the family level on one hand and repeated child maltreatment on the other. Thus, evaluation of problems related to the mother and implementation of services that help mothers deal with their personal problems, such as alcoholism and depression, seem to be very important. In addition, an evaluation of the mother s relationship with her husband or partner is important as well, in cases where the mother has one. These findings also show that it is important to support young people in receiving education, especially mothers, since mothers who had low education level were significantly more likely to repeatedly maltreat their children than mothers with higher education level. It is also important to support single parents, since they seem to be more likely to repeat maltreatment than families that include two parents, especially when they suffer from stress, at different levels such as unemployment or depression. Furthermore, stable housing seems to be an important factor, since children in families who had moved within the last 12 months before first incident were more likely to be repeatedly maltreated. Finally, at the social level, poverty seems to be a major risk factor for child maltreatment since two thirds of both groups had financial problems and half of both groups was or had been on welfare. A recent study showed that the.is

16 freydís jóna freysteinsdóttir poverty rate in Iceland is higher than in the other Nordic countries and that many have to deny themselves of things that most people think of as usual, such as buying birthday presents and going to the cinema (Sigríður Jónsdóttir, 2002). Ofcourse it is important to keep in mind that risk factors increase the likelihood of child maltreatment. Not all parents who experience risk factors maltreat their children, but they are more likely to do so. The families in the repeated incidents group were receiving more support from neighbors and more services than families in the single incident group. They were ofcourse facing more risk factors as well and therefore in more need of support and services. Thus, it seems that child protection workers are directing services to the families that are in more need of services. These results are consistent with the results of two Norwegian studies, showing relation between risk factors and lack of support on one hand and need for services on the other (Andenæs, 2004; Aamodt, Jörgensen & Aamodt, 2002). However, this finding is not consistent with other studies (DePanfilis & Zuravin, 1999). It might be speculated that informal social support is not as important in Iceland as in the U.S., because citizens in Iceland might receive more informal social support than U.S. citizens (Stefán Ólafsson, 1999). It is important to note that since this study was based on child protection records, it is limited by what was written in the records. The detail and quality of the record keeping seemed to vary with the child protection workers involved. Even basic information, such as education level and employment status was missing or difficult to find in some cases, since these variables were not recorded consistently. This is consistent with findings of other studies conducted in Iceland using the method of content analysis of child protection records (Guðrún Kristinsdóttir, 1991; Sigrún Júlíusdóttir, 2002). Furthermore, a lack of father involvement can be seen in this study, since less than half of them were available for interviews compared to majority of mothers. This is consistent with a finding from an earlier Icelandic study, showing a lack of father s involvement in the child protection services (Guðrún Kristinsdóttir, 1991). However, in this study, it is not clear if lack of involvement is due to a lack of interest of fathers themselves or a failure of child protection services to include them or both. By assessing risk factors, it is possible to predict whether a child is likely to be maltreated in the future (Pecora, 1991). This study is important as a first study on repeated child maltreatment in Iceland. Some clear predictors were isolated, especially problems related to the mother and problems in family dynamics. The importance of further study to assess cross-cultural differences and to isolate predictive factors for repeated maltreatment in Iceland and other Nordic countries more definitely is important..is

17 characteristics of families who are repeatedly reported to child protection services in iceland References Aamodt, Jörgensen B. & Aamodt A. (2002). Enslige mødre familier med store psykososiale belastninger. Nordisk Socialt Arbeid, 22 (4), Andenæs, A. (2004). Hvorfor ser vi ikke fattigdommen? Fra en undersøkelse om barn som blir plassert utenfor hjemmet. Nordisk Socialt Arbeid, 22 (1), American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders; DSM IV (4th ed). Washington DC: American Psychiatric Association. Belsky, J. (1980). Child maltreatment: An ecological integration. American Psychologist, 35 (4), Belsky, J. & Vondra, J. (1989). Lessons from child abuse: The determinants of parenting. In D. Cicchetti & V. Carlson (Eds.), Child Maltreatment: Theory and Research on the Causes and Consequences of Child Abuse and Neglect (pp ). New York: Cambridge University Press. Berger, L.M: (2004). Income, family structure, and child maltreatment risk. Children and Youth Services Review, 26 (8), Berger, L.M. (2005). Income, family strucutre, and child maltreatment risk. Child Abuse & Neglect, 29 (2), Bronfenbrenner, U. (1977). Toward and experimental ecology of human development. Child Abuse & Neglect, 29, Connell-Carrick, K. & Scannapieco, M. (2006). Ecological correlates of neglect in infants and toddlers. Journal of Interpersonal Violence, 21 (3), Cicchetti, D. & Lynch, M. (1995). Failures in the expectable environment and their impact on individual development: The case of child maltreatment. In D. Cicchetti & D. J. Cohen (Eds.), Developmental Psychopathology (Vol. 2, pp ). New York: John Wiley & Sons, Inc. Coohey, C. (2006). Physically abusive fathers and risk assessment. Child Abuse & Neglect, 30 (5), DePanfilis, D. & Zuravin, S.J. (1999). Predicting child maltreatment recurrences during treatment. Child Abuse and Neglect, 23 (8), Finkelhor, D. & Berliner, L. (1995). Research on the treatment of sexually abused children: A review and recommendations. Journal of the American Academy of Child and Adolescent Psychiatry, 34 (11), Freydís Jóna Freysteinsdóttir (2005). Risk Factors for Repeated Child Maltreatment: An Ecological Approach. Reykjavík: University of Iceland Press. Freydís Jóna Freysteindóttir. (e.d.). Definitions and category system in Child Protection [Skilgreiningar-og flokkunarkerfi í barnaverndarmálum]. Reykjavík: Government Agency for Child Protection [Barnaverndarstofa]. Government Agency for Child Protection [Barnaverndarstofa]. (2000). Skýrsla um starfssemi Barnaverndarstofu árin 1995 till 1999 og barnaverndarnefnda á Íslandi árin [Icelandic Welfare Council: A report on the Council s Operation from 1995 to 1999 and child protection committees in Iceland from 1996 to 1999]. Reykjavík: Government Agency for Child Protection [Barnaverndarstofa]. Guðrún Kristinsdóttir. (1991). Child Welfare and Professionalization. Umeå: Umeå Universitets Tryckeri. Hays, W.L. (1994). Statistics (5 th ed.). Fort Worth: Harcourt Brace College Publishers. Hosmer, D.W. & Lemeshow, S. (1989). Applied Logistic Regression. New York: A Wiley-Interscience Publication. Jouriles, E.N. & Norwood, W.D. (1995). Physical aggression toward boys and girls in families characterized by the battering of women. Journal of Family Psychology, 9 (1), Knutson, J. F. (1995). Psychological characteristics of maltreated children: Putative risk factors and consequences. Annual Review of Psychology, 46, Markward, M. J. (1997). The impact of domestic violence on children. Families in Society: The Journal of Contemporary Human Services, 1, Miller-Perrin, C.L. and Perrin, R.D. (1999). Child Maltreatment: An Introduction. London: SAGE Publications. Neuman, W. L. (2000). Social Research Methods: Qualitative and Quantitative Approaches, (4th ed.). Boston: Allyn and Bacon. Pecora, P.J. (1991). Investigating allegations of child maltreatment: The strength and limitations of current risk assessment systems. Child and Youth Services, 15 (2), Rubin, A. and Babbie, E. (1997). Research Methods for Social Work (3rd ed.). Pacific Grove: Brooks/Cole Publishing Company. Saunders, D. G. (1994). Child custody decisions in families experiencing woman abuse. Social Work, 39 (17), is

18 freydís jóna freysteinsdóttir.is Shor, R. (1998). The significance of religion in advancing a culturally sensitive approach towards child maltreatment. Families in society: The Journal of Contemporary Human Services, 79 (4), Sidebotham, P. & Heron, J. (2006). Child maltreament in the children of the nineties : A cohort study of risk factors. Child Abuse & Neglect, 30 (5), Sigriður Jónsdóttir. (2002). The life situation of long-term recipients of social assistance. Nordisk Socialt Arbejde, 22 (3), Sigrún Júlíusdóttir. (2002). Ævintýri á fjöllum: Rannsókn á reynslu unglinga af starfi með Hálendishópnum á tímabilinu Reykjavík: Háskólaútgáfan. Statistics Iceland [Hagstofa Íslands]. (2000). Landshagir: Statistical Yearbook of Iceland Reykjavík: Statistics Iceland [Hagstofa Íslands]. Stefán Ólafsson. (1999). [Íslenska leiðin Almannatryggingar og velferð í fjölþjóðlegum samanburði. Reykjavík: Tryggingastofnun ríkisins. Stemler, S. (2001). An overview of content analysis. Practical Assessment, Research & Evaluation, [on line], 7(17). Available: getvn.asp?v=7&n=17. Turner, H.A., Finkelhor, D. & Omrod, R. (2006). The effects of lifetime victimization on the mental health of children and adolescents. Social Science & Medicine, 62 (1), Zielinski, D.S. & Bradshaw, C.P. (2006). Ecological influences on the sequelae of child maltreatment: A review of the literature. Child Maltreatment, 11 (1), Wu, S.S., Ma, C., Carter, R.L., Ariet, M., Feaver, E. A., Resnick, M. & Roth, J. (2004). Risk factors for infant maltreatment: A population-based study. Child Abuse & Neglect, 28 (4), Summaries Characteristics of families who are repeatedly reported to the child protection services in Iceland The purpose of this study was to identify risk factors for repeated child maltreatment. The method used in this study was content analysis. Seventy seven cases that had never been reported to the child protection services before were included in this study. Each case was followed for 18 months from the first reported incident. In all cases the first reported incident was neglect. A group of cases that was only reported once was compared with another group of cases that was reported repeatedly. Families in the repeated incidents group had more risk factors than families in the single incident group. Problems related to the mother and family dynamics were found to be more important in predicting repeated child maltreatment than problems related to fathers and child vulnerabilities. Sérkenni fjölskyldna sem endurtekið eru tilkynntar til barnaverndaryfirvalda á Íslandi Markmiðið með rannsókn þessari var að bera kennsl á áhættuþætti varðandi endutekna misbresti í uppeldi barna á Íslandi. Í rannsókninni voru skoðuð og innihaldsgreind 77 mál, sem ekki höfðu komið við sögu áður hjá barnaverndaryfirvöldum. Hverju máli var fylgt eftir í 18 mánuði frá fyrstu barnaverndartilkynningu. Í öllum tilfellum var um vanrækslu að ræða í fyrstu tilkynningu. Þau mál sem tilkynnt var um einu sinni voru borin saman við mál sem voru tilkynnt tvisvar eða oftar til barnaverndaryfirvalda. Fjölskyldur sem höfðu verið tilkynntar endurtekið upplifðu meiri áhættuþætti heldur en fjölskyldur sem höfðu verið tilkynntar einu sinni. Vandkvæði tengd mæðrum og fjölskylduvandkvæði spáðu fremur fyrir um endurtekinn misbrest í uppeldi barna heldur en vandamál tengd feðrum og veikleika hjá börnum. Toistuvasti lastensuojelupalveluille raportoitujen perheiden ominaispiirteitä Islannissa Tämän tutkimuksen tavoitteena oli identifioida lasten kaltoinkohtelun uusiutumiseen vaikuttavia riskitekijöitä. Tutkimuksen metodina oli sisällönanalyysi. Tutkimukseen valittiin 77 lastensuojelupalveluille ensimmäistä kertaa raportoitua tapausta, ja jokaista tapausta seurattiin 18 kuukauden ajan ensimmäisestä raportoidusta tapahtumasta lähtien. Jokaisessa tapauksessa ensimmäinen raportoitu tapahtuma oli laiminlyönti. Tutkimuksessa verrattiin vain kerran raportoituja tapauksia useasti raportoituihin tapauksiin. Tutkimuksessa havaittiin, että usein raportoiduilla perheillä oli enemmän riskitekijöitä kuin vain kerran raportoiduilla perheillä. Äitiin ja perhedynamiikkaan liittyvien ongelmien havaittiin olevan merkittävämpiä lapsiin kohdistuvan kaltoinkohtelun uusiutumisen ennustamisessa kuin isään ja lasten haavoittuvuuteen liittyvien ongelmien.

19 REFEREEBEDØMT Mia Hakovirta and Mari Broberg Antaget til publicering januar 2007 Parenting from a Distance: Factors Connected to the Contact between Children and a Non-resident Parent MIA HAKOVIRTA Ph.D. Research Fellow, Department of Social Policy, University of Turku, Finland. miahak@utu.fi MARI BROBERG M.Ed. Researcher, Department of Education, University of Turku, Finland. marbro@utu.fi MIA MARI A growing number of children are living in households with only one of their parents, usually the mother. Moreover, the adults living with the children are not always the children s biological parents. The changing nature of family life has meant that parenting roles, expectations and responsibilities are in transition. The responsibility for children is not based on the marriage and partnership but rather on the parental obligation towards the children (Maclean & Eekelaar 1997). Keywords: non-resident parent children lone parent families stepfamilies NORDISK SOSIALT ARBEID NR VOL 7 SIDE Universitetsforlaget.fi

20 mia hakovirta and mari broberg Interest in family structure and cross-household ties between parents and children are motivated by the observed effects of contact arrangements on children s economic and psychological well-being. Many studies indicate that the interest of children in the post-separation situation is generally best served when children can maintain continuing and frequent contact with the non-resident parent (Amato & Gilbreth 1999; Pryor & Rodgers 2001; Bauserman 2002; Dunn et al. 2004). The child s visiting right is therefore an important tool for improving the well-being of those children whose parents are not living together. In this article we analyse the children s contact with the non-resident parent. We focus on the families in which the child has undergone the parental separation, either divorce or breakdown of consensual union, or the biological parents have never lived together. The article studies how the background of the family, the current family situation, the relationship with biological parents and the child s relationship with the non-resident parent are connected to the contacts between the child and the nonresident parent. The information on the functioning child-parent contact after parental separation can help to improve family policy and social work practices which are aimed at supporting parents and children after separation. Sharing the responsibility of children: the Finnish context In Finland, as in all Nordic countries, the number of lone parent families and stepfamilies has increased. At the beginning of this millennium, in Finland, of all families with children 19 % were lone parent families and 8 % stepfamilies (Statistics Finland 2002). This means that a quarter of Finnish children live apart from one of their biological parents (Kartovaara & Sauli 2001). In Sweden and Norway, 21 % of all working-age parents with dependent children were lone parents in 2001 (Bradshaw & Finch 2002). In Sweden, four out of ten children live at least part of their childhood in a lone parent family (Gähler 2001), and nearly one out of ten children live in a stepfamily (Statistics Sweden 2006). In the Nordic countries, policy making has emphasised biological parenthood and the right of the child to parental support. This idea has shaped the laws on child custody after parental separation in all Nordic countries. The aim is to give high priority to supporting the contact between the child and the non-resident parent after parental separation, and the legal provisions that regulate the visitation rights are congruent with each other in all Nordic countries. (Corden 1999; Bergman & Hobson 2002.) In Finland, the parental responsibility is enacted in two laws. The Child s Custody and Right to Access Act reflects the views that the sharing of parental responsibility should be encouraged and that the child has a right to meet both parents after separation. According to the Act, it is in the best interest of children to have contact with their biological parents, and joint legal custody is also seen as an opportunity to encourage.fi 0