Table of Contents | Previous | Next |
IV. EARLY HEAD START IMPACTS ON SERVICE RECEIPT
In Chapter III, we described services families received; here we compare services received by program and control group families. Although control group families could not receive Early Head Start services, they were free to seek other similar services in their communities. If most control group families received similar services, and if these services were as intensive as the services received by Early Head Start families, we might find few significant impacts on child and family outcomes, even if the Early Head Start research programs were highly successful in achieving their desired outcomes. Thus, for understanding program impacts on child and family outcomes, it is important to examine the differences in service receipt between program and control group families.
Our analysis of Early Head Start programs’ impacts on service receipt shows that, even though many control group families received some similar services from other community service providers, program families were much more likely to receive key child development and case management services during the combined follow-up period (28 months after program enrollment, on average). Early Head Start programs’ impacts on service receipt were large and statistically significant in most of the service areas we examined. The pattern of impacts on service receipt was generally similar to the pattern reported when families had been in the program for 16 months, on average.1
This chapter presents our analyses of program impacts on families’ service receipt. The first section describes global impacts of the Early Head Start programs on service receipt and service intensity during 28 months, on average, after random assignment.2 The second section summarizes the variations in these impacts among key subgroups of programs. The final section discusses the implications of these analyses for the analyses of impacts on children and families.
A. GLOBAL IMPACTS ON SERVICE RECEIPT AND SERVICE INTENSITY
Early Head Start program families were significantly more likely than control families to receive any key services (home visits, case management, center-based child care, and group parenting activities) during the combined follow-up period. The Early Head Start programs increased receipt of any key services by 14 percentage points (from 82 to 96 percent). While Early Head Start significantly increased services to program families, most control families received some services from other providers in the community.
The following subsections describe the global impacts of Early Head Start programs on families’ receipt of specific services, including any core child development services (home visits or center-based care), home visits, child care, parenting education and parent-child group socialization activities, child health services, services for children with disabilities, case management, family health services, and family development services.
1. Impacts on the Receipt of Core Child Development Services
As described in Chapter III, Early Head Start programs provided child development services primarily through home visits and child care in Early Head Start centers. The Early Head Start programs’ impact on receipt of these core child development services was large and statistically significant. Nearly all program families received at least minimal core services (93 percent), compared with 58 percent of control families (Figure IV.1).3 While almost all program families received more than minimal core services (more than one home visit or at least two weeks of center-based child care), only half of control families received more than minimal core services.
The programs’ impact on receipt of core child development services was larger when service intensity is taken into account. Program families were substantially more likely than control families to have received core child development services at the intensity required by the revised Head Start Program Performance Standards (weekly home visits, at least 20 hours a week of center-based child care, or a combination of the two). Nearly three-quarters of program families received the required intensity of services during at least one of the three follow-up periods, and half received them during at least two follow-up periods. Among control families, however, only 14 percent received core services at the required intensity during at least one follow-up period, and only 7 percent received them during at least two follow-up periods.
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. The percentage of program families who received core child development services is slightly larger than in Table III.3, because that table includes only home visits and center-based child care provided directly by the Early Head Start programs. Because some control families received these services from other community providers, the percentages here include home visits and center-based child care received from any source. A small percentage of program families also received these services from other community providers. aAt least one home visit and/or center-based child care. bMore than one home visit and/or at least two weeks of center-based child care. cWeekly home visits for home-based sites, at least 20 hours per week of center-based child care for center-based sites, and weekly home visits or at least 20 hours per week of center-based child care for mixed-approach sites. *Significantly different from zero at the .10 level,
two-tailed test. |
a. Impacts on Receipt of Home Visits
As described in Chapter III, all Early Head Start programs are expected to visit families at home on a regular basis. Home-based programs are expected to visit families weekly, and center-based programs must visit families at least twice a year. Mixed-approach programs are expected to provide families with weekly home visits, center-based child care, or a combination of the two.
The Early Head Start programs had large impacts on families’ receipt of home visits. During the combined follow-up period, 87 percent of program families received at least one home visit, compared with 34 percent of control families (Figure IV.2).4 Not only were program families much more likely to have received any home visits, they were also much more likely to have received home visits at least monthly. Nearly three-quarters of program families received home visits at least monthly during at least one follow-up period, compared with 15 percent of control families. Likewise, very few control families received home visits at least weekly during at least one follow-up period, while more than half of program families received home visits at least weekly during at least one follow-up period. Nearly all families in both groups who received home visits reported that they received child development services during the visits. Thus, the Early Head Start programs’ impacts on receipt of home visits are similar to impacts on receipt of child development services during home visits.
Based on the frequency of home visits families reported receiving during each of the three waves of follow-up interviews, we estimated that program families received roughly 56 home visits, on average, during the 26 months after program enrollment, while control families received an average of six visits (not shown).5 Thus, while a third of control families received some home visits, program families received many more visits, on average. Although these estimates are useful for providing a rough sense of the number of home visits families typically received, caution should be used interpreting their precision. As described in Chapter III, these estimates are based on families’ reports of the typical home visit frequency during the relevant follow-up period, not on their reports of numbers of home visits or program records on the date of each home visit.
|
[D] |
| Source: Parent Services Follow-Up Interviews
conducted approximately 7, 16, and 28 months after random assignment.
Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. The percentage of program families who received home visits is slightly larger than in Table III.4, because that table includes only home visited provided directly by the Early Head Start programs. Because some control families received home visits from other community providers, the percentages reported here include home visits received from any source. A small percentage of program families also received home visits from other community providers. *Significantly different from zero at the .10 level,
two-tailed test. |
b. Impacts on Receipt of Child Care Services
The Early Head Start programs significantly increased families’ use of child care. Most families in both groups used some child care during their first 26 months after random assignment, but program children were significantly more likely than control children to have received some child care—86 compared to 80 percent (Figure IV.3).6 The programs increased families’ use of center-based child care more substantially. Half of program families used center-based child care during their first 26 months after random assignment, compared with 36 percent of control families.
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. *Significantly different from zero at the .10 level,
two-tailed test. |
Program families were significantly more likely than control families to use concurrent child care arrangements (more than one child care arrangement at a time). Program families may have had a greater need for multiple arrangements to cover all the hours during which they needed child care because they used significantly more center-based care than control families. Centers may have been less likely than some other providers, such as relatives or family child care providers, to offer care during nonstandard work hours such as evenings and weekends.
Not only did the Early Head Start programs increase the percentage of families using any child care, they also increased the amount of child care that children received (Figure IV.4). Program children received significantly more hours of child care than control children during the 26 months after enrollment (1,544 compared to 1,224 hours, on average) and significantly more hours of center-based child care (687 compared to 357 hours, on average) during the 26 months after random assignment.7
Program families paid significantly less money out of pocket for child care, on average, than control families. Program families paid $326 less for child care, on average, during the 26 months following enrollment—nearly a 40 percent reduction in out-of-pocket child care costs (Table IV.1). Some of the Early Head Start programs provided child care to some or all families free of charge. Others helped families make child care arrangements with other community providers and paid some or all of the cost of care. Early Head Start programs, however, did not significantly affect the percentage of families who reported obtaining individual subsidies or vouchers to pay for child care during the 26 months after random assignment.8
|
[D] |
| Source: Parent Services Follow-Up Interviews
conducted approximately 7, 16, and 28 months after random assignment.
Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. *Significantly different from zero at the .10 level,
two-tailed test. |
| Program Group | Control Group | Estimated Impact per Eligible Applicant |
|
|---|---|---|---|
| Average Total Out-Of-Pocket Child Care Costs | $490 | $816 | -$326*** |
| Percentage
of Families Who Ever Received a Child Care Subsidy for: |
|||
| Any child care arrangement | 29.6 | 32.1 | -2.5 |
| A center-based child care arrangement | 16.7 | 16.6 | 0.1 |
SOURCE: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. NOTE: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. *Significantly different from zero at the .10 level,
two-tailed test. |
|||
2. Impacts on Receipt of Other Child Development Services
In addition to home visits and center-based child care, Early Head Start programs provided a range of other child development services. In this section, we report impacts on receipt of parenting education and parent-child group socialization services, child health services, and services for children with disabilities.
a. Impacts on Receipt of Parenting Education and Parent-Child Group Socializations
The Early Head Start programs significantly increased the likelihood that families received parenting education services, including discussions with case managers about parenting and group parenting activities. Nearly all program families (94 percent) received some parenting education, compared with 64 percent of control families (Figure IV.5).
Although the Early Head Start programs found it very challenging to achieve high participation rates in group parenting activities (parenting classes, parent-child group socialization activities, and parent support groups), they significantly increased program families’ participation in these services relative to control families’ participation in similar activities in the community. Seventy-one percent of program families participated in a group parenting activity during the combined follow-up period, compared to 37 percent of control families. The impact of the program on participation in parent-child group socialization activities was also substantial. Forty-two percent of program families participated in these activities during the combined follow-up period, compared with only 14 percent of control families.
In 12 of the Early Head Start research sites, when children were approximately 3 years old, interviews were conducted with fathers about their receipt of child development services. Box IV.1 summarizes the impacts the program had on fathers’ receipt of child development services.
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. *Significantly different from zero at the .10 level,
two-tailed test. |
|
FATHER PARTICIPATION IN PROGRAM-RELATED ACTIVITIES Early Head Start programs have increasingly devoted energies to involving men in program activities, and also to encouraging biological fathers and father figures to be more active participants with their children and families. The Early Head Start father studies began at a time when the majority of the research programs had not implemented specific father involvement components and did not target father outcomes as areas of expected change. Direct assessment of fathers and father outcomes were not included in the original evaluation design, but Father Studies were added to the research to provide descriptive information about the role of fathers or father figures (social fathers) in the lives of their children and to explore how father involvement in children’s lives is related to child outcomes. Here we describe father participation in program-related activities. Data about fathers’ participation in program-related activities were collected from fathers in the 12 father study sites. As described in Chapter 2, our findings about fathers are drawn from father interviews conducted when the children were approximately 36 months old. The father study samples, measures, and constructed variables are described in Appendix C. Early Head Start programs affected fathers’ program participation in important ways. In interviews with fathers, we asked about their participation in five types of activities: home visits, dropping off/picking up child at a child development or child care center, attending parenting classes or events, attending parent-child activities, and attending meetings or events just for fathers. 1
Patterns of father participation varied only slightly by program approach. Regardless of program approach, more fathers and father figures of Early Head Start children reported participating in home visits than control-group fathers/father figures did. There were no differences by program approach for dropping off and picking up the child from a child development program or center (see Box IV.1, Figure 1). Center-based and home-based programs affected father attendance at parenting classes or activities and participation in parent-child activities, but mixed-approach programs did not. Overall program implementation was related to father participation in program-related activities. Overall program implementation (especially among sites reaching full implementation later) increased father and father figure participation in most (but not all) program-related activities (see Box IV.1, Figure 2). Programs may be able to engage more fathers and engage them more frequently if they implement the performance standards and consider the unique needs of their fathers and father figures, along with existing barriers to their involvement in the context of overall family partnerships. 1The 12 father interview study sites included all 4 center-based programs, 5 of the home-based programs, and 3 mixed-approach programs. The pattern of implementation across the 12 sites included 5 sites in the early group, 4 sites in the later group, and 3 sites in the incomplete group.(back) |
| Program Group | Control Group | Estimated Impact per Eligible Applicanta | |
|---|---|---|---|
| Ever Engaged in Activity | |||
| Home Visit | 33.7 | 4.5 | 29.1*** |
| Dropped
Off/Picked up Child from a Child Development/Child Care Center |
45.4 | 40.7 | 4.7 |
| Parenting Classes or Events | 25 | 11.4 | 13.6*** |
| Parent-Child Activities | 20.1 | 8.4 | 11.7*** |
| Meetings or Events Just for Fathers | 9.6 | 5.9 | 3.7* |
| Engaged above Threshold in Activity | |||
| Home Visit Once per Month or More | 22.6 | 1.3 | 21.3*** |
| Dropped Off/Picked up Child from a Child Development/Child Care Center 10 or More Times | 11 | 1.7 | 9.3*** |
| Parenting Classes or Events Three or More Times | 16.2 | 8.3 | 7.9*** |
| Parent-Child Activities Three or More Times | 9.9 | 4 | 5.9*** |
| Meetings or Events Just for Fathers Three or More Times | 4.4 | 2.6 | 1.8 |
| Sample Size | 326 | 311 | 637 |
|
SOURCE: Father interviews conducted in the father study sites when children were approximately 36 months old. NOTE: All impact estimates were calculated using regression models, where data were pooled across sites. aThe estimated impact per eligible applicant is measured as the difference between the regression-adjusted means for all program and control group members.(back) *Significantly different from zero at the .10 level,
two-tailed test. |
|||
|
[D] |
Source: Father interviews conducted in the 12 father study sites when the children were approximately 36 months old. Notes: All percentages are regression-adjusted means estimated using models that pool across site. The differences between program and control families are estimated impacts per eligible applicant. * Program impact is significantly different from zero
at the .10 level, two-tailed test. |
|
[D] |
Source: Father interviews conducted in the 12 father study sites when the children were approximately 36 months old. Notes: All percentages are regression-adjusted means estimated using models that pool across site. The differences between program and control families are estimated impacts per eligible applicant. * Program impact is significantly different from zero
at the .10 level, two-tailed test. |
b. Impacts on Receipt of Child Health Services and Child Health Status
All children in both groups received some health services during the combined follow-up period, which reflects the accessibility of health services afforded by Medicaid and State Children’s Health Insurance Programs (Table IV.2). It also reflects the fact that many of the Early Head Start research programs recruited families at health clinics or WIC offices, where families were linked to health services before applying to Early Head Start. Few impacts on receipt of specific child health services were statistically significant during the combined follow-up period, because most families in both groups received services. Likewise, parents’ reports of the health status of their children when they were 3 suggest no statistically significant differences in the health status of program and control children.
Nevertheless, the Early Head Start programs increased children’s receipt of a few health services. Early Head Start programs had small but statistically significant impacts on the percentage of children who visited a doctor for treatment of illness (83 compared to 80 percent) and on the percentage of children who received immunizations (99 compared to 98 percent) during the combined follow-up period. The programs had a larger, negative impact on the likelihood of hospitalization for an accident or injury in the child’s third year (0.4 compared to 1.6 percent).
c. Impacts on Receipt of Services for Children with Disabilities
The Early Head Start programs had a pattern of small, significant impacts on eligibility for and receipt of early intervention services, as well as on the incidence of disability indicators. The programs increased the percentage of children who were ever identified by their parents as eligible for early intervention services (7 compared with 6 percent) during the combined follow-up period (Figure IV.6). The percentage of children who, according to parents, ever received early intervention services was also slightly higher among program families (5 compared with 4 percent).
| Outcome | Program Group |
Control Group |
Estimated Impact per Eligible Applicant |
|---|---|---|---|
| Average Percentage of Children Who Received Any Health Services | 100 | 99.8 | 0.2 |
| Percentage of Children Who Visited a Doctor: | |||
| For any reason | 98.9 | 98.4 | 0.5 |
| For a check-up | 95 | 95.1 | -0.1 |
| For treatment of an acute or chronic illness | 82.9 | 80.2 | 2.8* |
| Average Number of Doctor Visits: | |||
| For check-ups | 6.6 | 6.3 | 0.3 |
| For treatment of an acute or chronic illness | 6.2 | 5.8 | 0.4 |
| Percentage of Children Who Visited An Emergency Room | 54 | 53.5 | 0.5 |
| Average Number of Emergency Room Visits: | |||
| For any reason | 1.6 | 1.8 | -0.2 |
| For treatment due to accident or injury | 0.1 | 0.1 | 0 |
| Average Number of Hospitalizations During Child's Third Year | 0.1 | 0.1 | 0 |
| Average Number of Nights Hospitalized During Child's Third Year | 0.3 | 0.5 | -0.3 |
| Child Ever Hospitalized in Third Year for Accident or Injury | 0.4 | 1.6 | -1.3*** |
| Average Percentage of Children Who: | |||
| Visited a dentist | 28.3 | 26.2 | 2.1 |
| Received immunizations | 98.8 | 97.8 | 1.1* |
| Average Percentage of Children Who Received: | |||
| Any screening test | 66.8 | 66.5 | 0.2 |
| A hearing test | 40.2 | 40.1 | 0.1 |
| A lead test | 28.4 | 30.5 | -2.2 |
| Average Parent-Reported Health Status of Child at 36 Monthsa | 4 | 4 | 0 |
| Percentage
of Children Who Were Reported by Parents To Be in Fair or Poor Health at 36 Months |
8.2 | 8.7 | -0.5 |
| Sample Size | 966-1,104 | 915-1,010 | 1,966-2,106 |
SOURCE: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. NOTE: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. aPrimary caregivers rated their children’s health status on a scale of 1 (poor) to 5 (excellent).(back) *Significantly different from zero at the .10 level,
two-tailed test |
|||
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. Level 1 diagnosed conditions indicate eligibility for early intervention services and include a diagnosed hearing problem, severe or profound hearing loss, difficulty hearing or deafness, vision problem, difficulty seeing or blindness, speech problem, mobility problem, mental retardation, emotional disturbance, cleft palate, or a serious condition that showed up at birth or soon after, such as Down Syndrome, Turner’s Syndrome, or spina bifida. Level 2 diagnosed conditions, which may indicate eligibility for early intervention services, include crossed eyes or nearsightedness, epilepsy or seizures, hyperactivity, or a developmental delay. Functional limitations include possible hearing and vision problems, communication problems, trouble with arm/hand or leg/foot, and use of special equipment to get around. *Significantly different from zero at the .10 level,
two-tailed test. |
Based on parents’ reports, the extent of eligibility for early intervention services (reported eligibility or incidence of first-level diagnosed conditions) was also greater among program families (16 compared with 13 percent by the third followup). This increase probably reflects greater awareness or willingness among program families to report eligibility for early intervention services or diagnosed conditions or a higher likelihood among program children that conditions were diagnosed, but it could also reflect a higher incidence of the conditions among program children.
In contrast, the incidence of functional limitations or second-level diagnosed conditions reported by parents was smaller among program families (20 compared with 23 percent). This may reflect differences in program parents’ perceptions of functional limitations, differences in actual functional limitations due to help the program provided to families in obtaining health care to address the limitations, or differences in children’s development brought about by the Early Head Start programs.
Through a series of case studies, the local research team at Catholic University examined Early Head Start’s role in supporting families in obtaining services for young children with disabilities. These case studies are summarized in Box IV.2.
3. Impacts on Receipt of Family Development Services
Early Head Start programs helped families access a range of family development services, either by providing them directly or through referral to other community service providers, and significantly increased families’ receipt of many services. The following subsections describe the programs’ impacts on receipt of case management, health care, education-related services, employment-related services, transportation, and housing services.
|
BOX IV.2 EARLY HEAD START SUPPORTS FAMILIES
IN OBTAINING SERVICES Shavaun M. Wall, Nancy E. Taylor, Harriet
Liebow, Christine A. Sabatino, Although young children in low-income families face a higher risk of delays and disabilities, these families are less likely to obtain early intervention services than are more affluent ones. We conducted two studies to (1) determine whether Early Head Start enhances the likelihood that low-income families will obtain early intervention services when needed, and (2) identify how Early Head Start collaborates with families toward that goal. The first study used case studies of 32 research families with children suspected of needing early intervention to investigate whether Early Head Start facilitates referral, identification, and early intervention service provision. The families lived in a poor section of a generally affluent, densely and diversely populated, suburban area. Suspected need was defined as a recommendation by medical or community providers, Early Head Start staff, or researchers (as part of notification of low Bayley scores) that parents contact early intervention services. The case studies used in-depth interviews of mothers and staff members and a review of program and research records. A larger number of Early Head Start families were notified of a suspected need to refer (19, versus 13 in the control group), probably because Early Head Start staff members working with their children thought it necessary (see Table 1 in Volume III). With the active encouragement of Early Head Start staff, 18 of 19 (94 percent) Early Head Start families followed through to make the referral to the Part C or Part B office, compared with only 7 of 13 (54 percent) control families. A greater proportion of Early Head Start children were evaluated (89 versus 46 percent) and found eligible for services (79 versus 31 percent). The Early Head Start children represented a wider range of types of disabilities and severity levels, which suggests that Early Head Start programs may empower families to notice their children’s developmental challenges and obtain services, not only for medically related disabilities, but also for developmental delays. In the second study, researchers analyzed four case studies to determine how Early Head Start service providers supported families in obtaining early intervention services. As Early Head Start staff members began to work with the focus child, they earned trust and established relationships with the parents by helping with problem solving and resource identification to address basic family needs. Early Head Start workers were then able to help parents focus on the less familiar challenges central to their children’s development. In very different ways, according to parents’ abilities and emotions, Early Head Start staff helped parents understand child development, recognize and accept their children’s unique challenges, comprehend that early intervention services might have something to offer, and learn how to navigate the complex early intervention system. |
a. Impacts on Receipt of Case Management
Program families were significantly more likely than control families to receive case management services during the combined follow-up period—87 percent compared with 55 percent (Figure IV.7). Program impacts on the receipt of case management services at least monthly were large and similar to the impacts on receipt of home visits at least monthly. As was the case for home visits, approximately one-fourth of control families met with a case manager at least monthly during at least one follow-up period, compared with more than three-quarters of program families.
b. Impacts on Receipt of Family Health Care Services and Primary Caregiver’s Health Status
Nearly all program and control families reported that at least one family member (excluding the focus child) received some health services during the combined follow-up period (97 and 98 percent, respectively, received health services), and the program impact was not significant (Table IV.3). Likewise, we found no statistically significant impact on primary caregivers’ self-reported health status when their children were 3 years old.
c. Impacts on Receipt of Family Mental Health Services
The Early Head Start programs also did not have a significant impact on families’ receipt of mental health services. Twenty-three percent of program families reported receiving mental health services during the follow-up period, compared to 22 percent of control families.
d. Impacts on Receipt of Other Family Development Services
An important focus of Early Head Start services was supporting families’ progress toward self-sufficiency goals. The programs significantly increased families’ receipt of services designed to promote self-sufficiency, including education-related services, employment-related services, and transportation services. The programs increased primary caregivers’ receipt of education-related services (participation in school or job training or discussion about education services with a case manager). Eighty-seven percent of program families received these services, compared with 59 percent of control families (Figure IV.8). Likewise, programs increased families’ receipt of employment-related services (job search assistance or discussion about finding a job with a case manager). Seventy-seven percent of program families received these services compared with 46 percent of control families. Programs also increased families’ receipt of transportation services. One-third of program families received these services compared to 23 percent of control families. Early Head Start programs had no statistically significant impact on families’ receipt of housing services, including subsidized housing, rental assistance, help finding housing, energy assistance, and emergency housing.
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. *Significantly different from zero at the .10 level,
two-tailed test. |
| Program Group |
Control Group |
Estimated Impact per Eligible Applicant |
|
|---|---|---|---|
| Percentage of Families Who Received Any: | |||
| Family health services | 97.3 | 97.9 | -0.6 |
| Mental health services | 22.5 | 21.5 | 1 |
| Average Self-Reported Health Status of Parent or Guardian When Child Was 36 Months Oldb | 3.4 | 3.5 | 0 |
| Sample Size | 1,061-1,093 | 1,000-1,009 | 2,062-2,093 |
SOURCE: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment and Parent Interviews conducted when children were approximately 14, 24, and 36 months old. NOTE: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. aFamily health care services include services received by all family members except the focus child.(back) bPrimary caregivers rated their own health status on a scale of 1 (poor) to 5 (excellent).(back) *Significantly different from zero at the .10 level,
two-tailed test |
|||
B. DIFFERENCES IN PROGRAM IMPACTS ON RECEIPT OF SERVICES ACROSS SUBGROUPS OF PROGRAMS
It is important to go beyond overall impacts on service receipt described in the previous sections and explore variations in impacts on service receipt among targeted subgroups of programs. Variations in program impacts on service receipt may help explain differences in program impacts on child and family outcomes for subgroups of programs, and may highlight successes and challenges that particular groups of programs experienced in providing services to families. This section describes key differences in impacts on service receipt across subgroups of programs.
Caution must be used in interpreting the variations in impacts on service receipt among subgroups of programs. Most subgroups are defined on the basis of a single program characteristic, but the groups may differ in other characteristics. These other unaccounted-for variations in program characteristics may also influence the variation in impacts on service receipt. Thus, in our analyses, we focus on patterns of impacts across outcomes and consider the potential role of other differences in characteristics that may have influenced the outcomes being examined (Appendix Tables E.IV.1 and E.IV.12 show the configuration of family characteristics across the research sites and for select subgroups).
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The differences between program and control families are estimated impacts per eligible applicant. *Significantly different from zero at the .10 level,
two-tailed test. |
The program subgroup analyses show that the impacts of the Early Head Start programs on service receipt were broad-based and not limited to a particular subset of programs. The estimated impacts on families’ receipt of key services were large and statistically significant in nearly all the program subgroups we examined. Although the impacts on service receipt were large for all groups of programs, the magnitude of the impacts varied among subgroups, usually in expected directions. The variations in impacts on service use among subgroups of Early Head Start programs discussed in the sections that follow can inform our understanding of which program features may promote higher levels of participation and service receipt. The following subsections describe the differences in program impacts by program approach and pattern of implementation. We also examined some other site-level subgroups to explore whether Early Head Start impacts on service use varied among urban and rural locations or among programs located in states with and without welfare regulations requiring parents to engage in work activities while their youngest child was under 1 year old. Since the latter analyses did not suggest that these were important ways of classifying programs to understand impacts on services or on children and families, we do not discuss these subgroups here, but tables presenting the impacts for these subgroups are included in Appendix E.IV.
1. Difference in Impacts on Service Receipt by Program Approach
As described in Chapter I, the Early Head Start programs adopted different approaches to providing child development services, based on the unique needs of the children and families in their communities. In 1997, four programs offered center-based services only, seven offered home-based services only, and six took a mixed approach by offering both home- and center-based services.9
We expected to find differences in program impacts on service receipt that reflected the different approaches these programs took to serving children and families. In general, the differences in impacts are consistent with our expectations. Home-based programs had the largest impacts on receipt of any home visits, weekly home visits during at least one follow-up period and during all three follow-up periods, and parent-child group socialization activities (Figure IV.9 and Appendix Table E.IV.3). Center-based programs had the largest impacts on use of center-based child care and on the weekly out-of-pocket cost of care. Center-based programs also had a large, negative impact on the use of individual child care subsidies or vouchers, probably because they provided center-based child care for free and did not require most families to obtain individual child care subsidies or vouchers to pay for the care. Mixed-approach programs tended to produce impacts that were between those of home- and center-based programs but were often closest in magnitude to the impacts of home-based programs.
Overall, home-based and mixed-approach programs had the largest impacts on the receipt of any key services, and home-based programs had the largest impacts on receipt of core child development services. These differences reflect both lower levels of service receipt by program families in center-based sites and greater receipt of services by control families in those sites. Home-based and mixed-approach programs had the largest impacts on receipt of a range of family development services, including case management, education-related services, employment-related services, and transportation. Only center-based programs, which were located in areas where control families were much less likely to report receiving housing assistance, significantly increased receipt of housing services (Appendix Table E.IV.4).
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The difference between program and control families are estimated impacts per eligible applicant. aAny home visits or center-based child care. bWeekly home visits during the combined follow-up period. *Significantly different from zero at the .10 level,
two-tailed test. |
2. Differences in Program Impacts on Service Receipt by Implementation Status
Based on the ratings developed for the implementation study, the research programs differed in their patterns of overall program implementation. As summarized in Chapter I and reported more fully in Pathways to Quality (Administration on Children, Youth and Families 2002), six programs were rated as fully implemented in fall 1997 (early implementers), six were not rated as fully implemented in fall 1997 but were rated as fully implemented in fall 1999 (later implementers), and five were not rated as fully implemented at either time (incomplete implementers). The incomplete implementers either emphasized family support (with less emphasis on child development) or faced difficult implementation challenges (such as early staff turnover in leadership positions or partnerships that did not work out well).
The implementation ratings were based in part on staff reports of the frequency of services delivered, so we expected that the level and intensity of service receipt reported by program families would be highest among the early implementers. Thus, if levels of service receipt among control families in the early, later, and incompletely implemented program sites were similar, we would also expect the impacts on service use to be largest among the early implementers. The findings generally conform to this expected pattern.
Early implementation was associated with larger impacts on receipt of core child development services—home visits and center-based child care. Although programs in all three groups significantly increased receipt of these services, the impacts were consistently largest among programs that became fully implemented early (Figure IV.10 and Appendix Table E.IV.4). The large impacts of early implementers on receipt of core child development services were generally due to higher levels of service receipt in the program group, not lower levels in the control group.
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The difference between program and control families are estimated impacts per eligible applicant. aAny home visits or center-based child care. bWeekly home visits or at least 20 hours a week of center-based child care during the combined follow-up period. cWeekly home visits during the combined follow-up period. *Significantly different from zero at the .10 level,
two-tailed test. |
Impacts on receipt of core child development services at the intensity required by the revised Head Start Program Performance Standards were also largest among programs that became fully implemented early. For example, families served by early implementers were much more likely than program families in the other programs to receive core child development services at the required intensity in at least one follow-up period and throughout the combined follow-up period.
The overall implementation ratings used to form subgroups of early, later, and incomplete implementers take into account program implementation in all areas—child development, family partnerships, staff development, community partnerships, and program management. Because implementation of child and family development services may have the strongest linkages to child and family outcomes, we also examined subgroups based on the implementation ratings in these key areas. We formed two groups—those that reached full implementation in both child and family development in both periods (fall 1997 and fall 1999), and those that did not. The group that reached full implementation in child and family development in both periods consists of four of the six early implementers described at the beginning of this section.
The programs that reached full implementation in child and family development in both time periods had larger impacts on receipt of a range of services. For example, they had larger impacts on receipt of any key services, core child development services, home visits, center-based child care, and several family development services (see Appendix Table E.IV.5). In addition, these fully implemented programs had larger impacts on most measures of service intensity, such as receipt of core child development services at the required intensity, weekly home visits, and weekly case management. The programs that were not fully implemented in child and family development in both time periods had slightly larger impacts on group parenting activities.
It is possible that other factors might explain differences in impacts by implementation pattern. For example, differences in program approaches or family characteristics could be confounded with implementation pattern. Within the mixed-approach and home-based programs, however, it is possible to examine differences in impacts by implementation pattern while holding program approach constant.10 The results provide evidence confirming that fully implementing the performance standards makes a difference in the magnitude of impacts on service use. In the following subsections, we describe differences in impacts by implementation pattern for mixed-approach and home-based programs.
a. Differences in Impacts on Service Receipt for Mixed-Approach Programs by Implementation Status
Among the six programs that took a mixed approach to service delivery, three were early implementers (rated as fully implemented in fall 1997 and 1999), two were later implementers (rated as fully implemented in fall 1999 but not in 1997), and one was an incomplete implementer (not rated as fully implemented in either time period). One of the mixed-approach early implementers provided center-based services through contracts with community child care centers; the other two provided care to small numbers of program children in Early Head Start centers. The incomplete implementer and one of the later implementers provided Early Head Start center care to a large proportion of program families, and the other later implementer provided Early Head Start center care to a smaller number of families. Thus, program families in the mixed-approach programs that were later or incomplete implementers were more likely to receive Early Head Start center care, compared with families served by mixed-approach early implementers.
Program impacts on service use and intensity, by implementation pattern within mixed-approach programs, suggest that early, full implementation of key elements of the performance standards resulted in larger impacts on service receipt among families in mixed-approach programs. The mixed-approach early implementers had larger impacts on receipt of any key services, any core child development services (home visits or center-based child care), and core child development services provided at the intensity required by the Head Start Program Performance Standards (weekly home visits or 20 hours per week of center-based child care) (Figure IV.11 and Appendix Table E.IV.6). The group of mixed-approach later and incomplete implementers had a larger impact on receipt of any home visits, because control families in those sites were much less likely than control families in the early-implemented, mixed-approach sites to receive home visits. However, the mixed-approach early implementers had much larger impacts on receipt of home visits at least weekly.
The mixed-approach later and incomplete implementers also had a much larger impact on receipt of case management services. Their larger impact reflects the fact that control families in sites where mixed-approach early implementers were located were much more likely than those in the other sites to receive case management services.
b. Differences in Impacts on Service Receipt for Home-Based Programs by Implementation Status
Among the seven programs that took a home-based approach to service delivery, one was an early implementer, three were later implementers, and three were incomplete implementers. To have sufficient programs in each subgroup to conduct the analysis, we combined early and later implemented home-based programs into one subgroup and compared them to the home-based incomplete implementers.
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The difference between program and control families are estimated impacts per eligible applicant. aAny home visits or center-based
child care. *Significantly different from zero at the .10 level,
two-tailed test. |
Programs in both subgroups had large impacts on receipt of services in most of the service areas we examined, and differences in the size of impacts across the two subgroups were, in most cases, small (Figure IV.12 and Appendix Table E.IV.6). An exception to this pattern was the difference in the programs’ impact on participation in parent-child group socialization activities. The early- and later- implemented home-based programs had a substantially larger impact on participation in parent-child group socialization activities (49 percentage points in the early/late group compared to 16 in the incomplete group). This difference was due to differences in the proportion of program families who received these services, rather than to differences in service receipt among control families.
Several factors may account for the similarities in patterns of service use impacts in early and later compared to incompletely implemented home-based programs. First, only one of eight home-based programs achieved early implementation—the group in which we would expect to see the largest rates of participation in program services. Second, home-based programs that were not fully implemented often had strong family support components and provided frequent home visits and case management services. Other factors, such as the content of home visits and an insufficient emphasis on child development relative to other issues during the visits, prevented these programs from being rated as fully implemented. These other factors (such as topics covered during home visits), however, were not captured in our measures of service use and intensity. Thus, our measures may not incorporate some features of fully implemented programs that could account for differences in impacts on child and family outcomes across home-based programs with different patterns of implementation.
|
[D] |
Source: Parent Services Follow-Up Interviews conducted approximately 7, 16, and 28 months after random assignment. Note: All percentages are regression-adjusted means estimated using models that weight each site equally. The difference between program and control families are estimated impacts per eligible applicant. aAny home visits or center-based child care. bAny parent education classes, parent support groups, or parent-child group socialization activities. *Significantly different from zero at the .10 level,
two-tailed test. |
C. IMPLICATIONS FOR CHILD AND FAMILY OUTCOMES
The Early Head Start programs succeeded in greatly increasing the extent to which families received key program services, especially core child development services. Moreover, they provided much more intensive services than control families received from other sources in their communities. The estimated program impacts on the receipt of key program services and core child development services (home visits and center-based child care) were significant, large, and broad-based during the combined follow-up period.
Programs that fully implemented key Head Start Program Performance Standards early achieved the largest impacts on receipt of core child development services and on the receipt of intensive services. Because early, full implementation of the performance standards was associated with the delivery of intensive services to more families, the differences in impacts on child and family outcomes by implementation pattern can provide insights into the effects of “higher dosages” of Early Head Start services. In other words, if the early implemented programs have the largest impacts on child and family outcomes, then it is likely that part of the difference in the impacts associated with early implementation can be attributed to the more intensive services that families in those programs received, and the magnitude of the difference in impacts is in part an indicator of the importance of service intensity in producing the larger impacts.
In several service areas, the estimated impacts on service receipt were small, and most were not statistically significant. In particular, because nearly all children and families received some health services, the Early Head Start programs generally did not have a significant impact on health care receipt; even when impacts on health care receipt were significant, they were very small. Consistent with the lack of large differences in health care receipt, the estimated impacts on broad measures of the overall health status of children and primary caregivers were not significant. Finally, estimated impacts on identification of children eligible for early intervention services and receipt of early intervention services were statistically significant but small, and the proportion of identified children was fairly low in both the program and control groups.
The Early Head Start programs also did not have a significant impact on families’ receipt of mental health services. All of the programs made referrals to mental health services when they identified needs, and some provided some mental health services directly. While nearly a quarter of program families reported receiving mental health services, a similar proportion of control families also reported receiving mental health services. Thus, it appears that outreach to families with mental health service needs by other service providers was effectively reaching control families, or the programs were not able to enhance families’ access to mental health care.
The following chapters explore whether these impacts on service receipt led to impacts on child and family outcomes.
1See Building Their Futures: How Early Head Start Programs Are Enhancing the Lives of Infants and Toddlers in Low-Income Families (Administration on Children, Youth and Families 2001) for more details about these interim impacts on service receipt.(back)
2To analyze the Early Head Start programs’ impacts on service receipt and service intensity, we drew primarily on data from the Parent Services Follow-Up Interviews. These interviews were targeted for 6, 15, and 26 months after program enrollment and completed an average of 7, 16, and 28 months after enrollment. As described in Chapter III, we report primarily on cumulative levels of service use across all three follow-up periods. We use the term “combined follow-up period” to refer to cumulative levels of service receipt derived from the three waves of the parent services follow-up interviews. We also report some cumulative levels of service receipt and intensity that occurred in at least one or two of the three follow-up periods. Occasional deviations from the use of these terms are explained in the text.(back)
3The percentage of program families who received core child development services is slightly larger in Figure IV.1 than in Table III.2, because Table III.2 includes only home visits and center-based child care provided directly by the Early Head Start programs. Figure IV.1 includes home visits and center-based child care received from any source for both the program and control groups. A small percentage of Early Head Start families also received core child development services from community service providers.(back)
4The percentage of program families who received home visits is slightly larger in Figure IV.2 than in Table III.3, because Table III.3 includes only home visits provided directly by the Early Head Start program. Figure IV.2 includes home visits received from any source for both the program and control groups. A small percentage of program families also received home visits from other community service providers.(back)
5We calculated this estimate by adding together the estimated number of home visits received during each of the three follow-up periods and then prorating the estimate to 26 months after random assignment (by multiplying the estimated number of home visits by 26 divided by the actual length of the combined follow-up period). Estimates for each follow-up period were derived by multiplying the estimated number of home visits per unit of time based on the reported frequency of home visits by the length of the follow-up period in the same units of time.(back)
6Because the parent services follow-up interviews collected detailed information on families’ use of child care services, including dates of arrangements, we constructed a 26-month timeline that contains information on all the child care arrangements reported during the three waves of parent services follow-up interviews. Summary measures of child care use were developed using the timeline. Thus, the follow-up period for child care services is 26 months (the period covered for nearly all families who completed the interviews) for all families unless otherwise noted.(back)
7These averages include families who did not use any child care.(back)
8On follow-up surveys, parents were asked if they received a special check or voucher to pay for each child care arrangement. Thus, the percentages reported here include child care subsidies that parents received in the form of vouchers, but do not include subsidized child care provided through slots contracted directly by the state or free care provided by Early Head Start or other sources.(back)
9Over time, many of the home-based programs increased their efforts to ensure that the child care used by program families was of good quality, and some began offering a small number of child care center slots. However, few research sample members used these slots.(back)
10We were unable to examine differences in implementation within the center-based programs, because only 4 of the 17 research programs were center-based. In addition, the analysis of implementation within the mixed-approach and home-based programs required dividing the implementation patterns differently in order to have enough programs in each subgroup for the analysis. Thus, within mixed-approach programs, we compared early implementers with later and incomplete implementers. Within home-based programs, we compared early and later implementers with incomplete implementers.(back)
| Table of Contents | Previous | Next |