Table of Contents | Previous | Next |
3. Characteristics, Living Situations, and Maltreatment of Children Involved with the Child Welfare System
This chapter begins the presentation of the findings from our analyses of the characteristics of children who are in families investigated by CWS following reports of child abuse and neglect.4 These children may continue to reside at home with their families or may reside in out-of-home care. If they are residing at home, their families may be receiving no formal child welfare services; that is, their case was closed at intake or their family may be receiving in-home services. If they are in out-of-home care, the children may be living with relatives in “kinship care,” with nonrelatives in nonkinship care, or in group or residential care. If children were placed with relatives following the investigation, but the placement was not identified as a foster care placement in the caseworker or caregiver interview, these children fall into the group designated the in-home group.
The demographic distribution of cases that comes to the attention of CWS has received substantial analysis, based on administrative data records reported as part of the National Child Abuse and Neglect Data System (NCANDS) and the Adoption and Foster Care Analysis and Reporting System (AFCARS). While not all state data are included in these national-level statistics, evidence about the race, age, type of maltreatment, and gender of children who are reported for abuse and neglect, and whose cases are subsequently substantiated, is now available at a level not previously attained (DHHS, 2003). The emphasis of this report is on describing the NSCAW sample to provide a basis for understanding subsequent, and more unique, analyses related to children’s development and service use and the relationship between these two. We will make some comparisons to the NCANDS and AFCARS data, although these comparisons are limited by the differing methods used by each data system.
Child welfare personnel have a fundamental responsibility to make fair decisions that respect the rights of children for protection from harm and the rights of families to experience minimum levels of intrusion. Much discussion has followed the findings that a modest proportion of all cases reported for child maltreatment will go on to be substantiated and receive services (see, e.g., Besharov, 1985) and a very small proportion of those cases will be considered serious enough to require out-of-home care (Berrick et al., 1998). This pattern of service provision is one of the reasons that some child welfare agencies are endeavoring to find other ways to address the needs of the many children and families who are investigated by CPS but do not proceed on to CWS (Schene, 1998; Waldfogel, 2000). Yet little is known about the families and children who have their cases closed following an investigation (Wolock et al., 2001).
The decision about whether to serve children at home or to place them into out-of-home care is a complex one. The analyses in this chapter bring to bear a few basic child characteristics to describe which services were provided and where. First, basic demographics of the children involved with CWS (age, gender, race, and ethnicity) are discussed, as well as the settings in which these children are currently living (i.e., in-home vs. out-of-home). For children remaining at home, the report also discusses whether they have received services from child welfare services. (It is assumed that all children in out-of-home placements have received services from the agency.) The discussion includes a definition of receipt of child welfare services.
Next, this chapter introduces the types of maltreatment that brought the children to the attention of the child welfare agencies for the “current” investigation (i.e., the investigation that led them to be included in this study). Distributions of the most serious abuse type for each child, as identified by the child welfare worker, are presented overall and by age, gender, race/ethnicity, and setting. Data on the subtypes of abuse, multiple types of abuse, and severity and time since onset of abuse are also presented. The maltreatment discussion is followed by a section on substantiation of CWS reports.
In addition, children’s exposure to violence in the home is discussed, as measured by the VEX-R for children aged 5 and older. Results of the child report version of the Parent Child Conflict Tactics Scale (CTS-PC) are then presented for children aged 11 and older. The CTS-PC elicits the children’s report of exposure to parental discipline and maltreatment, including nonviolent discipline, psychological aggression, and physical assault. The chapter concludes with the in-home caregiver’s self-report of discipline and child maltreatment, as measured by the corresponding Parent-to-Child version of the CTS-PC, which includes the aforementioned dimensions, in addition to neglect and sexual maltreatment.
3.1 Characteristics of Children Involved with the Child Welfare System
Knowing the demographic characteristics of children involved with CWS is an important foundation for interpreting subsequent analyses of these children. Whether they tend to be younger or older or of a particular gender or racial/ethnic group provides a basis for understanding the more complex details of the lives of these children and their families. Eventually, this knowledge could help focus policies and programs to address more precisely the needs of subgroups of children and families.
At the time of the baseline interviews, the children in this sample range in age from 1 month to 15 years.5 The average age of the children is 7, as is the median age (mode ≤ 1 year) (first column of Table 3-1). Children age 6 to 10 years old make up the largest portion of children involved with CWS (36%). Another 25% are 11 years of age or older. Twenty-one percent of the children are age 3 to 5, and 19% are age 2 or younger. The children involved with CWS are evenly divided between males and females (first column of Table 3-2a).
| Age | TOTAL | Setting | ||||||
|---|---|---|---|---|---|---|---|---|
| In-Home | Out-of-Home | |||||||
| No Services | Services | TOTAL In-Home | Foster Care | Out-of-Home Kinship Foster Care | GroupCare | TOTALOut-of-Home ^ | ||
| Mean | 7 | 7 | 7 | 7 | 6 | 6 | 12 | 7 |
| Median | 7 | 7 | 7 | 7 | 6 | 7 | 13 | 7 |
| Mode | <1 | <1 | <1 | <1 | <1 | <1 | 14 | <1 |
| Range | <1-15 | <1-15 | <1-15 | <1-15 | <1-15 | <1-15 | <1-15 | <1-15 |
| ^ Includes children in other out-of-home placement settings. (back) |
To classify children for this analysis, we used a blend of race and ethnicity, considering ethnicity before race, such that those children identified as Hispanic/Latino were classified as such regardless of their race6. Those who were identified as non-Hispanic/Latino were classified by their race (i.e., African American, White, or other). Using these classifications, we find that 47% of children are White/non-Hispanic, 28% are African American/non-Hispanic, 18% are Hispanic/Latino,7 and 7% are classified as other races (first column of Table 3-2a). Differences in child characteristics across service settings are discussed in Section 3.4.
For the reader’s reference in this and subsequent chapters, Table 3-2b presents the same data as in Table 3-2a, but unweighted. These unweighted data are presented to give the reader perspective on the sample sizes involved and their power to answer questions of concern; all other tables in the report present weighted data.
Also for the reader’s reference, Table 3-2c presents the weighted Ns overall and broken down by the various child characteristics. These Ns are national estimates of children involved with the child welfare system produced by applying the weights (as described in Chapter 2) to the sample. They are presented to give the reader perspective on the population sizes involved in each of the categories.
| Characteristic | TOTAL | Setting | |||||||
|---|---|---|---|---|---|---|---|---|---|
| In-Home | Out-of-Home | ||||||||
| No Services | Services | TOTAL In-Home | Foster Care | Kinship Foster Care | Group Care | TOTAL Out-of-Home ^^ | |||
| Percent ^/ (SE) | |||||||||
| Age | 0-2 | 18.8 (1.0) |
17.3 (1.3) |
18.5 (1.5) |
17.6 (1.2) |
36.5(a) (3.4) |
28.8 (4.7) |
2.3 (1.4) |
27.7(b) (2.8) |
| 3-5 | 20.3 (1.1) |
21.0 (1.5) |
22.8 (1.9) |
21.5 (1.2) |
8.3 (1.9) |
15.1 (3.1) |
2.3 (1.6) |
11.0 (1.5) |
|
| 6-10 | 36.3 (1.4) |
38.7 (2.2) |
31.7 (1.9) |
36.8 (1.7) |
34.2 (3.7) |
34.3 (4.4) |
22.8 (11.2) |
32.3(c) (2.7) |
|
| 11+ | 24.6 (1.1) |
22.9 (1.9) |
27.0 (2.5) |
24.0 (1.3) |
21.0 (3.1) |
21.8 (4.5) |
72.6(d, e) (11.0) |
29.1(f) (2.9) |
|
| Gender | Male | 49.8 (1.8) |
49.5 (2.4) |
52.2 (2.2) |
50.2 (1.9) |
52.0 (3.9) |
39.4 (5.1) |
57.8 (10.8) |
46.9 (2.8) |
| Female | 50.2 (1.8) |
50.5 (2.4) |
47.8 (2.2) |
49.8 (1.9) |
48.0 (3.9) |
60.6 (5.1) |
42.2 (10.8) |
53.2 (2.8) |
|
| Race/Ethnicity | African American | 28.1 (2.5) |
26.0 (2.6) |
30.9 (3.1) |
27.3 (2.6) |
38.4 (5.6) |
33.7 (4.3) |
18.0 (5.9) |
34.6 (3.8) |
| White | 46.9 (3.7) |
47.9 (4.1) |
45.4 (3.8) |
47.2 (3.7) |
38.9 (6.9) |
47.7 (5.1) |
61.9 (9.5) |
44.8 (4.1) |
|
| Hispanic | 18.0 (2.9) |
19.3 (3.4) |
16.6 (3.1) |
18.6 (3.1) |
14.9 (4.5) |
13.1 (3.2) |
12.0 (4.5) |
14.0 (2.8) |
|
| Other | 6.9 (0.8) |
6.8 (1.0) |
7.2 (1.3) |
6.9 (0.8) |
7.8 (2.2) |
5.6 (1.8) |
8.1 (3.9) |
6.7 (1.4) |
|
| TOTAL | 100 | 64.7 (1.6) |
24.0 (1.5) |
88.6 (1.2) |
4.4 (0.6) |
5.1 (0.6) |
1.0 (0.2) |
11.4 (1.2) |
|
|
a Children 0-2, out-of-home are more likely than children 11+, out-of-home, to be in nonkinship foster care (X2 = 7.3, p < .01). (back) b Children 0-2 are more likely than children 3-5 to be in an out-of-home placement (X2 = 14.7, p < .001). (back) c Children 6-10 are more likely than children 3-5 to be in an out-of-home placement (X2 = 7.6, p < .01). (back) d Children 11+, out-of-home are more likely than children 0-2, out-of-home, to be in a group home (X2 = 12.6, p < .001). (back) e Children 11+, out-of-home are more likely than children 3-5, out-of-home, to be in a group home (X2 = 9.3, p < .01). (back) |
| Characteristic | TOTAL | Setting | |||||||
|---|---|---|---|---|---|---|---|---|---|
| In-Home | Out-of-Home | ||||||||
| No Services | Services | TOTAL In-Home | Foster Care | Kinship Foster Care | Group Care | TOTAL Out-of-Home ^^ | |||
| Percent^/ n | |||||||||
| Age | 0-2 | 36.3 (n=1998) |
32.9 (n=569) |
34.4 (n=796) |
33.8 (n=1365) |
49.5 (n=364) |
43.2 (n=247) |
8.7 (n=9) |
43.2 (n=633) |
| 3-5 | 15.2 (n=834) |
17.6 (n=304) |
16.2 (n=375) |
16.8 (n=679) |
9.7 (n=71) |
13.5 (n=77) |
2.9 (n=3) |
10.6 (n=155) |
|
| 6-10 | 27.1 (n=1492) |
29.8 (n=514) |
27.3 (n=630) |
28.3 (n=1144) |
24.1 (n=177) |
24.8 (n=142) |
15.3 (n=16) |
23.7 (n=348) |
|
| 11+ | 21.4 (n=1180) |
19.6 (n=338) |
22.1 (n=511) |
21.0 (n=849) |
16.9 (n=124) |
18.5 (n=106) |
73.1 (n=76) |
22.6 (n=331) |
|
| Gender | Male | 49.6 (n=2729) |
50.1 (n=865) |
50.4 (n=1164) |
50.3 (n=2029) |
49.3 (n=363) |
44.2 (n=253) |
51.0 (n=53) |
47.7 (n=700) |
| Female | 50.4 (n=2775) |
49.9 (n=860) |
49.7 (n=1148) |
49.7 (n=2008) |
50.7 (n=373) |
55.8 (n=319) |
49.0 (n=51) |
52.3 (n=767) |
|
| Race/Ethnicity^^^ | African American | 32.2 (n=1767) |
26.9 (n=463) |
31.1 (n=718) |
29.3 (n=1181) |
40.4 (n=295) |
39.8 (n=227) |
36.9 (n=38) |
40.2 (n=586) |
| White | 43.1 (n=2364) |
48.5 (n=834) |
42.9 (n=991) |
45.3 (n=1825) |
37.8 (n=276) |
36.3 (n=207) |
39.8 (n=41) |
37.0 (n=539) |
|
| Hispanic | 17.4 (n=956) |
17.2 (n=296) |
18.5 (n=427) |
17.9 (n=723) |
14.5 (n=106) |
18.4 (n=105) |
12.6 (n=13) |
16.0 (n=233) |
|
| Other | 7.3 (n=400) |
7.4 (n=127) |
7.5 (n=173) |
7.5 (n=300) |
7.4 (n=54) |
5.4 (n=31) |
10.7 (n=11) |
6.9 (n=100) |
|
| TOTAL | 100.0 (n=5504) |
31.3 (n=1725) |
42.0 (n=2312) |
73.4 (n=4037) |
13.4 (n=736) |
10.4 (n=572) |
1.9 (n=104) |
26.7 (n=1467) |
|
|
|
| Characteristic | TOTAL | Setting | |||||||
|---|---|---|---|---|---|---|---|---|---|
| In-Home | Out-of-Home | ||||||||
| No Services | Services | TOTAL In-Home | Foster Care | Kinship Foster Care | Group Care | TOTAL Out-of-Home ^ | |||
| n /(SE) | |||||||||
| Age | 0-2 | 450,297 (37,231) |
268,802 (27,124) |
106,249 (10,482) |
375,051 (32,578) |
38,864 (7,428) |
34,918 (8,448) |
546 |
75,246 (14,178) |
| 3-5 | 487,410 (48,669) |
326,141 (37,296) |
131,403 (15,474) |
457,545 (46,619) |
8,802 (1,775) |
18,234 (5,179) |
527 (357) |
29,865 (5,885) |
|
| 6-10 | 869,833 (85,468) |
599,500 (68,955) |
182,627 (19,338) |
782,127 (81,330) |
36,455 (7,938) |
41,519 (7,981) |
5,339 (3,261) |
87,706 (14,519) |
|
| 11+ | 589,964 (62,546) |
355,505 (49,071) |
155,341 (23,203) |
510,846 (57,128) |
22,377 (4,878) |
26,421 (5,512) |
17,009 (4,246) |
79,118 (10,811) |
|
| Gender | Male | 1,194,912 (108,527) |
767,194 (83,132) |
300,311 (28,897) |
1,067,506 (98,751) |
55,367 (10,990) |
47,753 (7,346) |
13,545 (4,449) |
127,407 (17,983) |
| Female | 1,202,592 (113,853) |
782,755 (82,334) |
275,309 (29,452) |
1,058,064 (102,677) |
51,131 (8,229) |
73,339 (14,567) |
9,877 (3,096) |
144,529 (21,909) |
|
| Race/Ethnicity | African American | 672,248 (84,530) |
401,021 (57,487) |
177,511 (26,453) |
578,532 (75,070) |
40,790 (10,097) |
40,792 (6,541) |
4,194 (1,166) |
93,715 (16,073) |
| White | 1,121,468 (96,026) |
738,718 (74,064) |
261,157 (26,400) |
999,875 (90,321) |
41,347 (7,013) |
57,654 (10,816) |
14,461 (5,022) |
121,594 (17,002) |
|
| Hispanic | 431,074 (95,500) |
297,830 (70,665) |
95,317 (21,791) |
393,147 (87,922) |
15,828 (6,406) |
15,808 (5,602) |
2,802 (1,037) |
37,927 (11,434) |
|
| Other | 164,381 (23,226) |
104,831 (17,235) |
41,502 (8,162) |
146,333 (21,378) |
8,310 (2,542) |
6,710 (2,052) |
1,890 (979) |
18,048 (3,997) |
|
| TOTAL | 2,397,504 (205,682) |
1,549,949 (148,106) |
575,620 (52,925) |
2,125,569 (185,224) |
106,498 (17,402) |
121,092 (18,120) |
23,421 (5,562) |
271,935 (37,000) |
|
3.2 Discussion of Child Characteristics
In general, children who have become involved with CWS span all age groups, with the greatest proportion (see Table 3-2a) of children younger than 3 years old—more than 18% of all children are 1 or 2 years old at the time of investigation.8 (Almost half of all children in this population are White/non-Hispanic, with African American/non-Hispanic children making up over one-quarter (28%), and Hispanic children being less than one-fifth of this population. Although this means that African American/Non-Hispanic children are overrepresented among children who are investigated (as compared with children in the general American population), this underrepresentation is far lower than is seen in national statistics for foster care. The proportion of African American children in foster care at the end of 2000 was 43%; Whites were only 36% of the children in foster care, and Hispanics were 15% (AFCARS, 2003). Although the reasons for the difference in proportion of children taken into foster care and remaining there are complex, the significant finding here is that the numerical disproportionality is not as great at the entrance into child welfare services as it is among children in out-of-home care. We can observe these discrepancies in our data, as the proportion of children in out-of-home care who are African American is just under 35%. Our NSCAW-based findings are comparable to those in the 2000 NCANDS report, which indicate that about 25% of the victims of child maltreatment were African American/non-Hispanic, about 51% of victims were White/non-Hispanic, and about 14% were Hispanic. There is no predominance of males or females in the population of children involved with CWS in NSCAW, nor was there such a predominance in the NCANDS data although females are slightly more numerous. Among children in out-of-home care in NSCAW, 53% are female and 47% are male; this is the reverse of what we observe in AFCARS, which is explainable if females remain in foster care longer.
3.3 Living Situations
As with demographic characteristics, knowing the living situations of the children involved with CWS is critical to understanding subsequent analyses. Whether the child has ongoing contact with the child welfare agency and whether he or she is living apart from or together with a permanent caregiver has a multidimensional influence on the child. In this context, “living situation” means whether the child remained at home following the CWS investigation or was placed in out-of-home care, and for those remaining at home, whether the case was closed after investigation or there is ongoing receipt of services. The data presented reflect the living situation of the children at the time of the initial NSCAW interview.
Overall, 89% of the children are identified as living at home with their permanent primary caregiver, while 11% have been removed from the home and are living in an out-of-home placement at the time of the current caregiver interview (Table 3-2a). Of all children receiving services (35% of the total), 32% are in out-of-home care. Of the in-home group, 96% are living with at least one of their parents9 and 4% are living with relatives (i.e., the child’s parent is not their primary caregiver). Of the children in out-of-home care, the largest group (45%) is in kinship foster care, while another 39% of children are in a nonkin foster home. Nine percent of the children in out-of-home care are in group care and 8% are in other out-of-home placements.10 From the perspective of the total population of children involved with CWS, the proportion of children in each out-of-home placement type is as follows: kinship foster care (5%), nonkin foster home (4%), group care (1%), other (1%) (Table 3-2a). The vast majority of children whose families are investigated for child abuse and neglect will remain at home.
To further classify the “in-home” children, we looked at whether the child or his or her family had received services from the child welfare agency. Such services may include, but are not limited to, counseling (for caregiver and/or child); assistance in obtaining food, clothing, or other necessities; income support; substance abuse treatment (for caregiver and/or child); mental health treatment (for caregiver and/or child); parenting classes; family support services; domestic violence services; and legal services. There were three sources for determining whether or not the child or his or her family had received child welfare services: the NSCAW sampling frame, the child welfare worker, and the caregiver. These data were collected at different points in time.
The sampling frame from which this sample of children was selected included a variable indicating whether or not the child or his or her family had received services from the child welfare agency. In other words, subsequent to the investigation in question, was the child’s case opened for some period (as opposed to no action being taken on the case following the investigation)? If the case had been opened for any period (even if it was closed shortly after being opened), the sampling frame data reflect that the child or his or her family had received services. These data were typically provided to the sampling team via electronic data transfer, although in about 15% of cases this determination was made following manual entry of the data into a spreadsheet.
Child welfare workers were asked if “any services have been provided to or arranged for the family.” This item referred to services provided or arranged prior to the baseline interview, regardless of the outcome of the investigation. For each service indicated to have been provided to or arranged for the family, the child welfare worker was then asked if the services had been “provided by the agency, arranged, or referred.” We considered the child welfare worker’s response affirmative if they indicated that a service had been provided or arranged and paid for by the agency (i.e., a service that was referred out to another provider but not paid for by CWS did not qualify for a child welfare service received by the child or his or her family).
Caregivers were asked if “your caseworker or someone else from the child welfare agency met with you and your family to talk about how best to deal with your family’s needs, concerns, and/or problems.” We interpreted an affirmative answer to this question to mean that the family had received services from the child welfare agency. Although this is a fairly liberal interpretation of whether or not services were provided, we felt inclusion of data from the caregiver was valuable. Had this variable been excluded, the proportion of children classified as remaining at home and receiving services would have increased by 4%, indicating that inclusion of the caregiver variable made our estimate of service receipt more conservative.
While using the sampling frame, child welfare worker, and caregiver as information sources provided more data as well as the reassurance of having data confirmed by more than one respondent, there were also cases in which data from these various sources were in conflict. In fact, in 51% of the unweighted cases there was a discrepancy between two of the respondents. In order to resolve these discrepancies, we devised and applied the following set of rules to most of the discrepant cases:
-
If there were responses from all three sources, a similar response given by two of the three sources was used (n = 1586, 77% of discrepant cases).
-
If there were responses from only two sources:
– Data from the sampling frame were given precedence over data from the caregiver (n = 95, 5% of discrepant cases).
– Data from the child welfare worker were given precedence over data from the sampling frame (n = 240, 12% of discrepant cases).
An exception to the final bullet above was instituted for 7% (n = 146) of the discrepant cases, for which the sampling frame indicated that services were received and the child welfare worker indicated that services were provided or arranged for the family but did not identify specific services that were provided or arranged (e.g., they may have been referred out). In these cases, because there was one definite yes answer (sampling frame) and one partial yes answer (child welfare worker), we erred on the side of identifying the case as having received services.
After applying the above rules to the cases with conflicting services data (and what was used in our analyses), the resulting proportions indicate that 73% of in-home children have not received services, while 27% have received services. There are no significant differences in the proportion of children living with parents who received services (27%) versus the proportion of children living with relatives who received services (34%). From the perspective of the total population of children involved with CWS (including the children in out-of-home care), 65% of the children remain at home with no services, and 24% remain at home and have received services (Table 3-2a).
For the remainder of this report, the two in-home subgroups will be identified as “in-home, no services” and “in-home, services” to differentiate between those who did not receive child welfare services and those who did receive such services. Note that while the “in-home, no services” group was determined to have not received services from the child welfare agency, it is possible that they received and/or are receiving services from one or more outside agencies.
3.3.1 Discussion of Living Situations
The vast majority of children who become involved with CWS remain at home following an investigation by child welfare services, and almost three-quarters of these children are not receiving services from the child welfare agency while at home. The proportion of children placed in out-of-home care following an investigation is relatively small, a fact that may not be understood by the general public, that may identify child welfare services closely with the placement of children into foster care. Indeed, CWS is sometimes referred to as the “foster care system.” Still, these removals are not inconsequential for children and families, as almost all of these removals are court-ordered and begin a formal involvement with child welfare services that has the potential to markedly change the lives of all involved. At the same time, nearly half of all children in out-of-home care are in kinship foster homes, so they experience some significant level of continuity. About two-fifths of the children in out-of-home care are in nonkin foster homes and a small percentage, mostly older children, is in group care.
3.4 Setting and Services by Child Characteristics
The next step in interpreting the children’s living situations is assessing the impact of various demographic characteristics. Determining how the proportions of children in the various age, race/ethnicity, and gender groups vary across settings, and whether any differences are significant, can help to create a better profile of how cases are funneled through the system and inform programs and policies to enhance services.
In general, average age remains consistent across the setting categories. The mean, median, and modal age for the subgroups of all children remaining at home and all children in out-of-home care are the same as for the total population of children involved with CWS, as described earlier (mean age = 7, median age = 7, modal age ≤ 1 year) (Table 3-1). The largest deviation from these numbers is for the group-home subpopulation, which has a mean age of 12, a median age of 13, and a modal age of 14. Children in each of the setting subgroups span the entire age range (< 1 to 14).
Although bivariate associations were run between each of the child demographic variables (age, gender, and race/ethnicity) and the setting variable, age is the only characteristic that appears to be associated with the setting in which the child is living. With regard to whether a child is living at home or in an out-of-home placement, those who are aged 3 to 5 years are the least likely to be in an out-of-home placement—children in all other age groups are more likely than those aged 3 to 5 years to be in an out-of-home placement (Table 3-2a).
Among children living in out-of-home placements, the oldest children are more likely than children in each of the two youngest age groups to be in a group home. In fact, almost three-quarters of the children in group care are aged 11 or older, while the 0- to 2-year-olds and 3- to 5-year-olds together make up only 4.6% of the group-care population. In addition, children aged 11 or older are less likely than children aged 0 to 2 to be in nonkinship foster care. As with gender and race/ethnicity, age does not appear to be associated with whether or not a child living at home has received services (Table 3-2a).
Logistic regression was used to determine the likelihood that a child was placed in out-of-home care, controlling for age, gender, race/ethnicity, and most serious abuse type. The results supported the bivariate analyses indicating that of these four variables only age is a predictor, with 3- to 5-year-olds being significantly less likely than 0- to 2-year-olds (OR = 2.90, p < .001), 6- to 10-year-olds (OR = 1.73, p = .01), and children 11 and older (OR = 2.30, p < .01) to be in out-of-home care.
Logistic regression was also used to establish whether or not a child still living at home was receiving services, controlling for age, gender, race/ethnicity, and most serious abuse type; the multivariate analyses again supported our previous analysis that there was a lack of association among these variables. That is, the finding of no bivariate differences between in-home services recipients and in-home no services recipients was maintained even when relationships between variables were mathematically controlled. In later chapters, we examine the case characteristics that are associated with the level of out-of-home care into which children are placed.
3.4.1 Discussion of Setting and Services by Child Characteristics
In summary, of the three demographic variables examined (age, gender, race/ethnicity), only age is a predictor of placement into out-of-home care (3- to 5-year-olds are the least likely to be in out-of-home care), while none of them predicts receipt of services among children remaining at home. Analyses presented later in this report examine what other factors may predict placement and/or service receipt. Regarding out-of-home placement type, the children in group care are most likely to be the oldest children, which is not surprising given the difficultbehavioral and emotional issues that older children are more likely to bring with them into care, while the children in nonkinship foster care are most likely to be the youngest children.
3.5 Types of Maltreatment
This section describes the types of abuse or neglect that brought the children to the attention of the child welfare agency for the current investigation. Although the children may have experienced other types of abuse prior to or since the current investigation, these data provide a benchmark to examine the most prevalent types of abuse and neglect, as well as child and case characteristics associated with various types of abuse and neglect. This section also discusses the frequency of children experiencing multiple types of abuse or neglect concurrently so as not to presume that the presence of these abuse and neglect types is mutually exclusive.
Types of maltreatment involved in the current investigation were classified using the modified coding scheme described by Manly, Cicchetti, and Barnett (1994) and used extensively by the LONGSCAN group (Runyan et al., 1988). The child welfare worker who was interviewed used information from the case record to report type, severity, and time since the onset of maltreatment. This approach has three major advantages over conventional means of gathering data on types of maltreatment from administrative records. First, it allows for more than one type of maltreatment to be indicated. Second, inclusion of the severity and duration of the maltreatment allows us to distinguish important differences among experiences of children with the same type of maltreatment. Third, the child welfare worker can describe the actual case characteristics rather than force the case to fit a category for court or administrative purposes. (For this reason, some cases were classified as “other.” When asked to choose one of the nine types of maltreatment in the instrument, some workers declined, indicating that they wanted to use this opportunity to describe the case as it occurred, not as paperwork dictated.) As a result, these maltreatment findings have more uniformity and specificity than would be found in administrative records.
Child welfare workers were asked to identify the most serious type of abuse and any other types that were reported present. The most serious type of abuse for almost half of the children involved with CWS is neglect, which encompasses both failure to supervise and failure to provide. Physical maltreatment is the most serious type of abuse for more than another quarter of the children (Table 3-3). The 4.5% of children in the other category were referred to CWS for reasons other than abuse or neglect (e.g., for mental health services or domestic violence). Even though these children had been classified as abused or neglected in the official CWS records—and, therefore, eligible for inclusion in the study—interviews with the child welfare workers indicated that other reasons were responsible for their involvement with the child welfare agency.
A direct comparison between the NSCAW maltreatment type proportions presented here and the proportions presented in NCANDS (2000) is imprecise, as the NCANDS proportions allow for multiple maltreatment types for each victim and are based on state-reported data. Whereas NSCAW also collected data on all reported maltreatment types (see discussions on number of main abuse types and most serious and additional types of abuse later in this section), we have generally limited our analyses to the most serious abuse type identified for each child.
|
||||||||||||||||||||||
|
Sexual, emotional, and particularly physical abuse appear to be more common in NSCAW, with proportions of 11%, 7%, and 27%, respectively, than in NCANDS, with proportions of 10%, 8%, and 19%, respectively. The NSCAW proportions for each type would differ, however, if allowing for multiple maltreatment types per victim.
For the remainder of this report, we refer to five major categories of abuse and neglect when looking at abuse type and its relationship to other characteristics. Physical abuse, sexual abuse, failure to provide, and failure to supervise were retained, with the abandonment cases subsumed in the latter category. The less common types of abuse (i.e., emotional maltreatment, educational maltreatment, moral/legal maltreatment, and exploitation) were combined into a maltreated-other (types of abuse) category. However, the maltreated-other category was not included in most analyses that employ an abuse type because the interpretation for this group is too complex. These analyses also excluded cases with abuse types described as nonmaltreated-other (signifying other reasons for placement, as discussed above), as well as cases with don’t know, refused, or missing responses (don’t know, refused, and missing responses account for a weighted 7% of the total population). The distribution of the most serious abuse types following this recoding of the data is presented in the last row of Table 3-4, which also presents most serious type of abuse by age, gender, and race/ethnicity.
Bivariate analyses indicate that the most serious type of abuse suffered by children involved with CWS differs significantly depending on the child’s age. In general, the youngest children are the most likely to have a most serious abuse type of neglect (i.e., failure to provide or failure to supervise) and the least likely to have a most serious abuse type of physical or sexual maltreatment, or one of the other abuse types. The oldest children are the most likely to have a most serious abuse type of sexual maltreatment (Table 3-4).
Gender also plays a significant role in the most serious abuse type, with males being significantly more likely to be victims of physical maltreatment, and females being significantly more likely to be victims of sexual maltreatment (p < .001) (Table 3-4). Our analyses did not reveal a bivariate association between race/ethnicity and most serious abuse type.
| Characteristic | Physical Maltreatment | Sexual Maltreatment | Failure to Provide | Failure to Supervise | Other | TOTAL | |
|---|---|---|---|---|---|---|---|
| Percent^/ (SE) | |||||||
| Age | 0-2 | 22.6 (2.2) |
6.1 (1.7) |
29.9 a b (2.5) |
36.6 c (3.0) |
4.8 (1.7) |
100 |
| 3-5 | 23.6 (2.9) |
12.8 (2.8) |
23.8 (3.8) |
30.3 (2.6) |
9.5 (2.3) |
100 | |
| 6-10 | 31.2 d (2.6) |
11.1 (2.4) |
18.9 (2.4) |
26.1 (2.4) |
12.7 e (2.1) |
100 | |
| 11+ | 32.7 (3.1) |
14.9 f (2.1) |
12.7 (2.3) |
29.7 (2.5) |
10.0 (1.8) |
100 | |
| Gender | Male | 32.3 g (2.0) |
5.9 (1.7) |
21.0 (2.0) |
32.1 (2.1) |
8.6 (1.3) |
100 |
| Female | 24.5 (1.8) |
17.0 h (1.7) |
19.8 (1.8) |
27.5 (2.4) |
11.2 (1.7) |
100 | |
| Race/Ethnicity | African American | 24.1 (3.0) |
9.2 (2.2) |
22.0 (2.4) |
36.4 (2.4) |
8.2 (1.9) |
100 |
| White | 29.2 (2.0) |
12.5 (1.7) |
21.2 (2.0) |
27.1 (2.2) |
10.1 (1.9) |
100 | |
| Hispanic | 33.3 (3.4) |
11.6 (2.7) |
15.2 (3.1) |
27.2 (4.9) |
12.7 (2.1) |
100 | |
| Other | 25.8 (3.9) |
13.5 (5.0) |
23.3 (5.6) |
29.9 (3.7) |
7.5 (2.5) |
100 | |
| TOTAL | 28.4 (1.5) |
11.5 (1.2) |
20.4 (1.5) |
29.8 (1.5) |
9.9 (1.2) |
100 | |
|
a Children 0-2 are more likely than children 6-10 to have a most serious abuse type of failure to provide (X2 = 7.4, p < .01). (back) b Children 0-2 are more likely than children 11+ to have a most serious abuse type of failure to provide (X2 = 25.9, p < .001). (back) c Children 0-2 are more likely than children 6-10 to have a most serious abuse type of failure to supervise (X2 = 8.7, p < .01). (back) d Children 6-10 are more likely than children 0-2 to have a most serious abuse type of physical maltreatment (X2 = 6.9, p ≤ .01). (back) e Children 6-10 are more likely than children 0-2 to have a most serious abuse type of other (X2 = 7.4, p < .01). (back) f Children 11+ are more likely than children 0-2 to have a most serious abuse type of sexual maltreatment (X2 = 11.9, p < .001). (back) g Males are more likely than females to have a most serious abuse type of physical maltreatment (X2 = 9.5, p < .01). (back) |
Table 3-5 presents data on the most serious abuse type by child’s living situation. Bivariate tests of association indicate that the child’s setting and receipt of services do not differ significantly based on his or her most serious abuse type.
When the most serious type of abuse was reported to be physical maltreatment or neglect, the child welfare worker was asked about the subtypes of alleged abuse and the most serious of the subtypes that occurred. When the most serious type of abuse was reported to be sexual maltreatment, the child welfare worker was asked to indicate all types of sexual maltreatment that were reported. Although the child welfare worker was not asked to identify the most serious type of sexual maltreatment, we created a hierarchy of the types from least to most serious in order to select the most serious type for the purpose of these analyses. Table 3-6 presents the most serious subtypes of abuse overall and by child setting.
| Type of Abuse | TOTAL | Setting | ||||||
|---|---|---|---|---|---|---|---|---|
| In-Home | Out-of-Home | |||||||
| No Services | Services | TOTAL In-Home | Foster Care | Kinship Foster Care | Group Care | TOTAL Out-of-Home ^^ | ||
| Percent^ / (SE) | ||||||||
| Physical Maltreatment | 28.4 (1.5) |
30.0 (2.1) |
26.5 (1.8) |
29.1 (1.6) |
22.8 (4.2) |
16.3 (3.7) |
24.1 (11.1) |
22.4 (3.8) |
| Sexual Maltreatment | 11.5 (1.2) |
11.4 (1.7) |
12.4 (1.8) |
11.7 (1.4) |
7.0 (1.8) |
10.2 (2.5) |
23.9 (10.3) |
10.0 (2.2) |
| Failure to Provide | 20.4 (1.5) |
19.7 (2.0) |
21.4 (2.8) |
20.1 (1.7) |
24.9 (2.9) |
26.0 (5.7) |
7.7 (3.7) |
22.9 (2.7) |
| Failure to Supervise | 29.8 (1.5) |
28.4 (2.0) |
30.5 (2.1) |
29.0 (1.6) |
36.6 (4.7) |
40.5 (4.4) |
31.0 (11.9) |
36.6 (3.1) |
| Other | 9.9 (1.2) |
10.5 (1.5) |
9.2 (1.5) |
10.1 (1.3) |
8.7 (2.6) |
7.1 (2.1) |
13.3 (5.8) |
8.1 (1.7) |
|
|
For children with a most serious abuse type of sexual maltreatment or failure to supervise, the most serious subtype of abuse differs significantly between children remaining at home and those in out-of-home care. With regard to sexual maltreatment, children remaining at home are over twice as likely to be in the least severe of the subtype categories (fondling or molestation without genital contact, or other less severe subtype such as exposure to sex or pornography) than children in out-of-home care (59% vs. 23%). With regard to failure to supervise, children remaining at home are twice as likely than children in out-of-home care to be in the “environment” (failure to ensure child is playing in safe area) category (31% vs. 15%), while children in out-of-home care are five times as likely to have been abandoned (20% vs. 4%; p < .001).
As noted above, child welfare workers were asked to identify all of the types of maltreatment inflicted on the child as alleged in the current report. For the following analysis, we looked specifically at the presence of more than one of the four main abuse types. As shown in Table 3-7, almost three-quarters of the children are victims of just one of these four types (most often physical maltreatment or failure to supervise), while one-fifth of the children experience more than one of these four types. There are differences in this variable based on child setting, with children remaining at home significantly more likely than those in out-of-home care to have experienced none of the main abuse types (7% vs. 3%) or just the main abuse type of physical maltreatment (26% vs. 17%) in the current report. Children in out-of-home care are significantly more likely than those remaining at home to have experienced two of the main abuse types in the current report (29% vs. 17%; p < .001).
| Type of Abuse | TOTAL | Setting | |||||||
|---|---|---|---|---|---|---|---|---|---|
| In-Home | Out-of-Home | ||||||||
| No Services | Services | TOTAL In-Home | Foster Care | Kinship Foster Care | Group Care | TOTAL Out-of-Home ^^ | |||
| Percent | |||||||||