ACF OPRE: Early Head Start Research and Evaluation Project (EHSRE)

III. PARTICIPATION IN EARLY HEAD START SERVICES

Early Head Start is a complex intervention program that is challenging to implement. As a first step toward understanding the intervention’s impacts on children and families, we document program accomplishments and the services families received. Did the 17 research programs provide a fair test of the Early Head Start concept? Evidence from the implementation study shows that, overall, the research programs succeeded in implementing Early Head Start services and delivering core services to most families while they were enrolled in the program (Administration on Children, Youth and Families 2002).

To set the context for examining program impacts, this chapter describes in detail families’ participation in program services and levels and intensity of service use during 28 months, on average, after families’ enrollment in Early Head Start. The chapter also describes variations in program participation and the intensity of services received by families across program types and patterns of program implementation. The first section describes the data sources we used and the terms we use to discuss levels of service use and intensity during various time periods. The sections that follow describe families’ levels of overall program participation and participation in specific child development and family services. The final section summarizes our conclusions about the levels and intensity of program participation. The next chapter contrasts the services that program families received with those received by control group families.

A. DATA SOURCES

We drew on the following data sources to analyze families’ participation in program services:

Head Start Family Information System application and enrollment forms completed at the time of enrollment.
Parent services follow-up interviews targeted for 6, 15, and 26 months after program enrollment (and completed an average of 7, 16, and 28 months after enrollment). We included in our analyses families for whom data were available for all three of these follow-up periods (71 percent of program group members).
Exit interviews conducted when children were approximately 3 years old and families became ineligible for Early Head Start.1
Ratings of each family’s engagement with the program provided by program staff in summer 2000, after most families had left the program.
Data from the implementation study on Early Head Start programs’ three main approaches to providing child development services—home-based, center-based, and mixed-approach (combination of home- and center-based).2
Ratings of program implementation developed as part of the implementation study, in which programs were classified as early implementers (fully implemented in 1997 and 1999), later implementers (fully implemented in 1999 but not in 1997), or incomplete implementers (not fully implemented in 1997 or 1999).3

The length of the follow-up period and children’s ages at the time of the interviews varied over a wide range for each wave of parent services interviews. The length of followup ranged from 4 to 15 months after enrollment for the first follow-up interview, 9 to 27 months for the second interview, and 24 to 59 months for the third interview.4 Because the interviews were conducted according to the length of time since families enrolled, the ages of the children in the research sample at the time of the interview also varied. On average, focus children were 10 months old when the first follow-up interview was completed, 20 months old at the second interview, and 32 months old at the third interview.

In this and the next chapter, we report primarily on cumulative levels of service use across all three follow-up periods covered by the parent services follow-up interviews. We use the term “combined follow-up period” to refer to the entire period covered by these cumulative measures. We also report some measures of service receipt and intensity of services received 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. Unless otherwise noted, the measures are based on parent reports.

B. LEVELS OF OVERALL PARTICIPATION IN PROGRAM SERVICES

Almost all program group families participated in Early Head Start at least minimally during the combined follow-up period. Overall, 91 percent of program families received at least one Early Head Start home visit, participated in Early Head Start center-based child care, met with an Early Head Start case manager at least once, and/or participated in Early Head Start group activities (group parenting education, group parent-child activities, or parent support group). Moreover, nearly all these families (90 percent of program group members) participated beyond this minimum level, receiving more than one home visit or case management meeting, center-based child care, and/or group parenting activities.5

Although participation levels exceeded 90 percent in 15 of the 17 research programs, two center-based programs had lower participation rates (64 and 75 percent). Several factors contributed to these low rates. In one program, some families needed full-time child care before the program expanded to offer it. In the other, a very rapid initial recruiting process and a delay in opening one center may have led some program families to find child care elsewhere.

C. DURATION OF PROGRAM PARTICIPATION

According to staff, program group families participated in Early Head Start for an average of 21 months, with nearly half of the families participating for at least two years (Table III.1). Families in the research sample could have enrolled in Early Head Start at any time after the mother became pregnant with the focus child until the child’s first birthday. Thus, families’ length of eligibility for program services varied, ranging from more than three years (if the family enrolled before the focus child’s birth) to about two years (if the family enrolled when the focus child was nearly a year old). Therefore, families who participated in Early Head Start for less than 24 months (49 percent of program families) left the program before their eligibility ended.

Research families left the programs for a variety of reasons. When staff rated the families’ engagement in summer 2000 (see section III.H. below, on program engagement ratings), they indicated the reasons families left the programs. Of the three quarters of families who had left the program by summer 2000, approximately one-third had graduated or transitioned out of the program when their eligibility ended. One-fourth had moved out of the service area before completing the program. Nearly one-third were terminated by staff because of poor attendance or lack of cooperation, or they asked to be removed from the program rolls. Home-based programs were much more likely to report that they terminated families’ enrollment for poor attendance or lack of cooperation, while center-based and mixed-approach programs were more likely to report that families had asked to be removed from program rolls.

TABLE III.1

DURATION OF EARLY HEAD START PARTICIPATION BY PROGRAM FAMILIES, FOR THE FULL SAMPLE AND KEY PROGRAM SUBGROUPS
	Full Sample	Center-Based	Home-Based	Mixed- Approach	Early Implementers	Later Implementers	Incomplete Implementers
	Program Approach				Pattern of Implementation
Percentage of Families Who Were Enrolled in Early Head Starta:
Less than 1 month	2	1	2	2	2	2	2
1 to less than 6 months	11	18	10	7	9	13	10
6 to less than 12 months	13	13	13	12	13	13	12
12 to less than 18 months	11	11	11	13	10	10	13
18 to less than 24 months	12	9	13	13	12	11	12
24 to less than 30months	20	22	16	19	22	18	19
30 to less than 36 months	17	12	20	18	18	17	16
36 months or longer	16	14	15	16	13	17	17
Average Number of Months Families Were Enrolled in Early Head Start	21	20	22	23	22	19	22
Sample Size	1,214	252	554	408	455	410	283
SOURCE: Exit interviews conducted near the time of the focus child’s third birthday and information on dates of last contact with the family obtained from program staff in Summer 2000. Excludes one site that did not provide a date of last contact for most families. NOTE: The percentages are average percentages across programs in any given group. aBased on the date of the program’s last contact with the family as reported by program staff in summer 2000 if the family had left the program by then or if no exit interview was completed, and on the number of months of participation reported by the family in the exit interview if the family left the program after the summer of 2000 and completed an exit interview.(back)

D. LEVELS OF PARTICIPATION IN CORE CHILD DEVELOPMENT SERVICES

The Early Head Start programs took three main approaches to providing core child development services. Home-based programs provided these services primarily through home visits. Center-based programs provided child development services primarily through child care in Early Head Start centers. Mixed-approach programs provided home-based services to some families, center-based services to some families, and a mix of home- and center-based services to some families. Thus, home visits and child care in Early Head Start centers were the programs’ primary vehicles for delivering child development services.6

Nearly 9 in 10 program group families received core child development services—either home visits, Early Head Start center care, or both (Table III.2). This percentage may underestimate the proportion of families who received core child development services, because some families received child development services in other child care settings under contract with an Early Head Start program during the combined follow-up period and our measure of core child development services captures only the services provided by Early Head Start directly.

The Head Start Program Performance Standards require programs to provide child development services through weekly home visits, at least 20 hours per week of center-based child care, or a combination of the two. Nearly two-thirds of families received core child development services at the required intensity during at least one of the three follow-up periods, and one-quarter received these services throughout the combined follow-up period. Families in home-based and mixed-approach programs were the most likely to receive core child development services at the required intensity level for at least one follow-up period (70 percent), compared with families in center-based programs (53 percent).

TABLE III.2

LEVELS OF PARTICIPATION IN CORE EARLY HEAD START CHILD DEVELOPMENT SERVICES, FOR THE FULL SAMPLE AND KEY PROGRAM SUBGROUPS
	Full Sample	Center- Based	Home- Based	Mixed- Approach	Early Implementers	Later Implementers	Incomplete Implementers
	Program Approach				Pattern of Implementation
Percentage of Families Who Received:
At least minimal core servicesa	89	80	91	92	94	88	83
More than minimal core servicesb	87	77	90	90	93	86	80
Percentage of Families Who Received Core Services at the Required Intensity Level forc:
At least 1 follow-up period	66	53	70	70	79	62	56
At least 2 follow-up periods	43	40	47	40	57	37	33
Combined follow-up period	25	26	26	23	37	19	18
Sample Size	940-1,020	166-205	476-478	298-338	310-351	330-359	301-311
SOURCE: Parent Services Follow-Up Interviews completed an average of 7, 16, and 28 months after enrollment. NOTE: The percentages are average percentages across programs in any given group and are weighted for nonresponse. aAt least one Early Head Start home visit and/or at least two weeks of Early Head Start center-based child care.(back) bMore than one Early Head Start home visit and/or at least two weeks of Early Head Start center-based child care.(back) cWeekly Early Head Start home visits for home-based sites, at least 20 hours per week of Early Head Start center-based child care for center-based sites, and weekly Early Head Start home visits or at least 20 hours per week of Early Head Start center-based child care for mixed-approach sites.(back)

The 75 percent of families who did not receive core child development services at the required intensity throughout the combined follow-up period does not necessarily indicate program failure to comply with the performance standards. The combined follow-up period covers the 28 months, on average, after families enrolled in Early Head Start. Families’ length of participation in the program, however, averaged 21 months, with nearly half of the families participating for less than 24 months (Table III.1). Thus, the majority of families who did not receive core child development services at the required intensity during all three follow-up periods (the combined period) were not actually enrolled in the program throughout this entire period.

Early, full implementation appears to be associated with receipt of core child development services at the required intensity level. Early implementers provided these services to 79 percent of families for at least one follow-up period, compared to 62 percent of families in later implementers and 56 percent in incomplete implementers. Likewise, early implementers provided these services to nearly 40 percent of families throughout the combined follow-up period, compared to less than 20 percent of families served by later and incomplete implementers.

1. Early Head Start Home Visits

All Early Head Start programs are required to complete home visits, whether they are home-based, center-based, or provide a mix of services. In center-based programs, services are delivered primarily in Early Head Start child care centers, but staff are required to complete home visits with children and their families at least twice a year. They may meet with families in other places if staff safety would be endangered by home visits or families prefer not to meet at home.7 Home visitors are required to visit families receiving home-based services at home weekly, or at least 48 times per year. In mixed-approach programs, some families receive home-based services, some receive center-based services, and some receive a combination of the two.

Across all three program types, 84 percent of families received at least one Early Head Start home visit, and almost all of these families received more than one visit (Table III.3). As expected, families in home-based and mixed-approach programs were most likely to receive at least one home visit (90 and 89 percent, respectively, compared with 65 percent of center-based programs). Across programs with different patterns of implementation, early implementers were most likely to provide at least one home visit (90 percent), followed by later implementers (84 percent), and incomplete implementers (76 percent).

Most families received home visits at least monthly. More than two-thirds received home visits monthly or more often during at least one of the three follow-up periods, and one-third received home visits at least monthly throughout the combined follow-up period. In home-based programs, 86 percent received monthly visits during at least one follow-up period, and nearly half received visits at least monthly during the combined follow-up period. Almost all of these visits lasted an hour or longer.

According to the revised Head Start Program Performance Standards, programs serving families through home-based services must provide weekly home visits to families. As noted in the implementation study, however, programs found it very challenging to complete visits with some families weekly.8 Nevertheless, home-based programs were able to deliver weekly home visits to many families. Seventy percent of families in home-based programs reported receiving weekly visits during at least one follow-up period, nearly half received weekly visits during at least two periods, and one-quarter received weekly visits throughout the combined follow-up period.9

TABLE III.3

LEVELS OF PARTICIPATION IN EARLY HEAD START HOME VISITS, FOR THE FULL SAMPLE AND KEY PROGRAM SUBGROUPS
	Full Sample	Center- Based	Home- Based	Mixed- Approach	Early Implementers	Later Implementers	Incomplete Implementers
		Program Approach			Pattern of Implementation
Percentage of Families Who Received: Any Early Head Start home visits	84	65	90	89	90	84	76
More than 1 Early Head Start home visit	78	46	89	86	82	81	70
Percentage of Families Who Received Early Head Start Home Visits at Least Monthly for:
At least 1 follow-up period	69	23	86	82	72	71	65
At least 2 follow-up periods	53	12	72	58	56	51	52
Combined follow-up period	32	4	47	34	38	27	32
Percentage of Families Who Received Early Head Start Home Visits at Least Weekly for:
At least 1 follow-up period	52	5	70	62	56	51	48
At least 2 follow-up periods	31	0	47	34	38	28	27
Combined follow-up period	16	0	26	16	23	12	14
Percentage of Families Who Received Monthly Early Head Start Home Visits Lasting At Least One Hour for:
At least 1 follow-up period	64	17	80	78	66	64	62
At least 2 follow-up periods	45	6	61	53	52	40	43
Combined follow-up period	25	2	36	29	33	18	24
Percentage of Families Who Received Weekly Early Head Start Home Visits Lasting At Least One Hour for:
At least 1 follow-up period	48	3	64	60	54	45	45
At least 2 follow-up periods	29	0	43	31	37	24	25
Combined follow-up period	15	0	24	15	21	11	12
Sample Size	1,007-1,029	206-210	470-480	331-339	346-351	353-364	308-314
SOURCE: Parent Services Follow-Up Interviews completed an average of 7, 16, and 28 months after enrollment. NOTE: The percentages are average percentages across programs in the given group and are weighted for survey nonresponse.

Based on the frequency of home visits families reported receiving during each of the three waves of follow-up interviews, we estimate that families received roughly 52 Early Head Start home visits, on average, during the 26 months after program enrollment (not shown).10 As expected, families in home-based programs received the most home visits, on average (71 visits), followed by families in mixed-approach and center-based programs (65 and 11 visits). While these estimates are useful in providing a rough sense of the number of home visits families typically received, caution should be used in interpreting their precision. The estimates are based on families’ reports of the typical home visit frequency during the relevant follow-up period, not on respondent reports or program records on the completion date of each home visit.

To better understand the reasons for variation in home visit frequency across families, local research partners at the University of Washington and the University of Missouri-Columbia examined associations between home visit frequency and various family characteristics. Boxes III.1 and III.2 describe their findings.

2. Early Head Start Center-Based Child Care

The revised Head Start Program Performance Standards require programs serving families through the center-based option to provide center-based child development services to children for at least 20 hours a week. This section describes families’ participation in this core child development service during 26 months after they enrolled in the program. The next section describes program families’ use of all types of child care, including care provided by Early Head Start and other providers in the community. Because 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. 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.

During their first 26 months in the program, 28 percent of all program group children received care in an Early Head Start center, including 71 percent of children in center-based programs and 30 percent of children in mixed-approach programs (Table III.4).11 For 21 percent of all families in the sample, an Early Head Start center was their child’s primary child care arrangement (the arrangement the child was in for the greatest number of hours during the 26 months after program enrollment). In center-based programs, 57 percent of families used an Early Head Start center as their child’s primary arrangement. Early Head Start centers served as the primary arrangement for 20 percent of families in mixed-approach programs.

On average, program group children received 450 hours (about 4 hours a week) of care in an Early Head Start center.12 As expected, children in center-based programs received more than three times as many hours of Early Head Start center care—1,391 hours (about 12 hours per week), on average. In mixed-approach programs, children received 336 hours (about 3 hours a week, on average) of Early Head Start center care. In addition to receiving more hours of Early Head Start center care, on average, children enrolled in center-based programs were more likely to receive this care continuously. Nearly a third of families in center-based programs used care in an Early Head Start center continuously during the 26 months after enrollment, and more than half used it for at least half of this period.

Children served by early implementers were most likely to receive care in an Early Head Start center (38 percent), compared to later implementers (27 percent) and incomplete implementers (17 percent). Children served by early implementers also received more than twice as many hours of care in an Early Head Start center, on average, than children served by later and incomplete implementers.13 In the two center-based programs that were early implementers, children received an average of 2,028 hours of Early Head Start center care (about 18 hours per week, on average).

BOX III.1

PATTERNS OF PARTICIPATION IN HOME-BASED SERVICES

Fredi Rector and Susan Spieker
University of Washington

We examined home visitor records to determine whether this suburban, Pacific Northwest Early Head Start program showed particular patterns of program participation. Of the 90 families recruited for the research program, 76 (84 percent) participated in three or more home visits (more than simply enrollment). All participants were expected to take part in weekly home visits. However, when participation results were analyzed, we identified two groups. The low-participation group (n = 46) had at least one visit per month for an average of 10.33 (SD = 5.41) months, while the high-participation group (n = 30) had at least one home visit per month for an average of 25.43 (SD = 6.76) months. Only 17 of these families, however, remained active until the focus child was 36 months old.

Content analysis of home visit records revealed 14 target content topics for home visits.1 The percent of home visits that focused on target content topics varied between the high- and low-participation groups. For example, 58 percent of the home visits to the high-participation group included specific content on the growth and development progress of the focus child, while only 33 percent of the low-participation group visits focused on this topic (p<.01). Similarly, 47 percent of the home visits to the high-participation group included child play activities, compared with 21 percent of the home visits to the low-participation group (p<.01). The topic of housing was also associated with longevity in the program. In the high-participation group, 9 percent of home visits included discussions of housing issues, compared with 18 percent in the low-participation group (p<.05).

A primary goal of this program was to facilitate a secure parent-child attachment relationship. To that end, the research team and the home visitors developed 10 parent-child communication intervention (PCCI) protocols, which the home visitors delivered. They delivered these protocols in home visits to 44 percent of the low-participation group and 32 percent of the high-participation group (p<.05). However, the high-participation group completed more PCCI protocols than did the low-participation group (3.3 versus 1.5, p<.01). In addition, caregivers whose adult attachment representations, as measured by the Adult Attachment Interview2were classified as insecure and unresolved due to trauma/loss (28 participants) completed fewer PCCI protocols than did caregivers who were not unresolved, regardless of their security classification (1.8 versus 2.7, p<.05).

Caregiver adult attachment classification and housing needs at the time of enrollment were both related to patterns of program participation. Caregivers who had insecure attachment (see Hesse 1999) were more likely to be in the low-participation group (p<.05), as were participants who initially identified housing as a need (p<.1). However, housing needs identified at enrollment were not significantly related to the discussion of housing issues during home visits. These findings suggest that assessment at enrollment should include a measure of post-traumatic stress, since 37 percent of the sample was coping with unresolved trauma and loss, and this factor was related to their level of participation in the program. Early Head Start programs also need to address effectively the issue of safe, adequate housing. Further research is needed to understand the relationships between unresolved trauma and loss, housing problems, and program participation.

1The target content topics were observations of child growth and development, child play/recreation, child health, child assessment, child care, parent development, PCCI protocols, employment, caregiver health, caregiver assessment, education, family recreation, housing, and information and referral.(back)

2Hesse, E. “The Adult Attachment Interview: Historical and Current Perspectives.” In Handbook of Attachment: Theory, Research, and Clinical Applications, edited by J. Cassidy and P.R. Shaver. New York: The Guilford Press, 1999, pp. 395-433.(back)

BOX III.2

RELATIONS AMONG MOTHER AND HOME VISITOR PERSONALITY TRAITS, RELATIONSHIP QUALITY, AND AMOUNT OF TIME SPENT IN HOME VISITS

Elizabeth A. Sharp, Jean M. Ispa, Kathy R. Thornburg, and Valerie Lane
University of Missouri-Columbia

In response to the low frequency of home visits in programs across the country (Gomby et al. 1999), the current study examined associations between mother and home visitor personality, the quality of mother-home visitor relationships, and the amount of time spent in home visits. We hypothesized that the quality of the mother-home visitor bond mediates links between their personality characteristics and time in home visits.

The participants were 41 African American, low-income, first-time mothers enrolled in an Early Head Start program in a large, Midwestern city, and five home visitors. Most of the mothers were in their late teens or early 20s and had limited education.

The mothers and home visitors completed the Multidimensional Personality Questionnaire, Form NZ (Tellegen 1982). Home visitors also completed the Bond Subscale of the Working Alliance Inventory (Short Form) (Horvath and Greenberg 1989). The dependent variable was participation, defined as the mean number of minutes per month spent with each mother in home visits, based on three months to two years of visits.

The results of hierarchical linear modeling did not support our mediational hypotheses. However, significant associations emerged among the personality, relationship, and participation variables. Maternal personality traits that showed orientation toward control and achievement were negatively related to home visit participation. On the other hand, maternal tendencies to feel vulnerable or taken advantage of (for example, high stress reaction and alienation) were positively related to participation. Maternal stress reaction and alienation were also positively linked to home visitor ratings of bond quality. Home visitor stress reaction was negatively related to participation. Finally, the quality of mother-home visitor bond was positively related to participation.

One explanation for these findings may be that home visitors thought home visits were especially important for highly stressed mothers who are low in control and in striving for achievement, because the services come to the mothers; the mothers do not have to take the initiative to go to the services. Moreover, highly stress-prone mothers may have been more likely to draw the home visitors into personal relationships because they had more issues to address. If home visitors perceived achievement-oriented mothers as more capable of meeting their own needs, they may have made fewer attempts to schedule visits to them.

The stress-proneness of home visitors may be related to their skills in establishing relationships. It may be especially important for social service providers whose work takes them into high-stress situations to have a positive, less stress-prone personality. Individuals with a negative, more stress-prone personality may find the difficult circumstances of parents like those in our sample overwhelming.

References

Gomby, D.S., P.L. Culross, and R.E. Behrman. “Home Visiting: Recent Program Evaluations—Analysis and Recommendations.” Future of Children, vol.9, 1999, pp. 4-26.

Horvath, A.O., and L.S. Greenberg. “Development and Validation of the Working Alliance Inventory.” Psychological Assessment, vol. 1, 1989, pp. 207-210.

Tellegen, A. Brief Manual for the Multidimensional Personality Questionnaire. Unpublished manuscript. University of Minnesota, Minneapolis, 1982.

TABLE III.4

LEVELS OF PARTICIPATION IN EARLY HEAD START CENTER-BASED CHILD CARE, FOR THE FULL SAMPLE AND KEY PROGRAM SUBGROUPS
	Full Sample	Center- Based	Home- Based	Mixed- Approach	Early Implementers	Later Implementers	Incomplete Implementers
		Program Approach			Pattern of Implementation
Percentage of Children:
Who ever received Early Head Start care	28	71	1b	30	38	27	17
For whom Early Head Start care was the primary arrangementa	21	57	1	20	30	18	12
Percentage of Children Who Were In Early Head Start Care For:
0 hours per week	75	31	99	76	66	76	85
1-9 hours per week	10	25	0	12	12	11	7
10-19 hours per week	7	16	0	8	9	6	4
20-29 hours per week	4	14	0	2	7	4	1
30-39 hours per week	2	6	0	0	3	1	1
40 or more hours per week	2	6	0	1	3	0	1
Average hours per week	4	12	0	3	6	3	2
Average Total Hours That Children Were In Early Head Start Care	450	1,391	9	336	701	347	273
Percentage of Children Who Were in Early Head Start Care For:
0% of the follow-up period	72	29	99	70	62	74	83
1-19% of the follow-up period	3	4	0	5	4	2	3
20-39% of the follow-up period	3	8	0	6	3	5	4
40-59% of the follow-up period	3	5	0	4	4	3	1
60-79% of the follow-up period	3	7	0	4	5	2	2
80-99% of the follow-up period	7	15	0	8	6	8	4
100% of the follow-up period	9	31	0	4	16	6	3
Sample Size	918-1,010	175-211	444-458	297-341	308-346	316-375	284-302
SOURCE: Parent Services Follow-Up Interviews completed an average of 7, 16, and 28 months after enrollment. NOTE: The percentages are average percentages across programs in the given group and are weighted for survey nonresponse. aArrangements that children were in for the largest number of hours during the follow-up period.(back) bThese child care arrangements were identified by respondents as Early Head Start center-based child care. The arrangements may be slots that were provided by an Early Head Start program through contracts with community providers or slots in another Early Head Start program in the community, or respondents may have reported them as Early Head Start arrangements in error.(back)

E. USE OF CHILD CARE SERVICES

Rates of child care use were high across all three program types and patterns of implementation. Almost all program group families used child care (86 percent) for the focus child at some point during the 26 months after enrollment in Early Head Start (Table III.5) In this section we describe program families’ use of center-based care; use of multiple care arrangements; types of primary care providers; care during nonstandard work hours; total hours children were in child care; duration of child care use over the 26-month follow-up period; and out-of-pocket costs of child care to families.

More than half of program group children received care in a child care center for at least two weeks during the 26 months after enrollment. As expected, families in center-based programs were most likely to receive center-based care (79 percent), followed by those in mixed-approach programs (52 percent) and home-based programs (33 percent). Families served by early implementers were also more likely to use center-based care (56 percent), compared with later and incomplete implementers (50 and 45 percent).

During the 26 months after enrollment, 64 percent of children received care in more than one child care arrangement, and over half received care in more than one arrangement concurrently. On average, program group children received care in two child care arrangements during their first 26 months in Early Head Start. Fifty-two percent received care in more than one arrangement concurrently at some point during this period. Nearly three-quarters of the children in center-based programs were cared for in concurrent arrangements, suggesting that Early Head Start centers did not provide care during all the hours that families needed it.

TABLE III.5

CHILD CARE ARRANGEMENTS USED BY EARLY HEAD START FAMILIES, FOR THE FULL SAMPLE AND KEY PROGRAM SUBGROUPS
Percentage of Children:
		Program Approach			Pattern of Implementation
	Full Sample	Center- Based	Home- Based	Mixed- Approach	Early Implementers	Later Implementers	Incomplete Implementers
Who received any child care	86	93	80	89	87	82	90
Who received any center-based child care	51	79	33	52	56	50	45
Percentage of Children Who Received Care in the Following Number of Arrangements:
0	14	7	20	11	13	18	10
1	22	15	24	23	21	20	25
2	25	25	24	26	24	23	28
3	20	26	17	20	21	18	21
4 or more	20	27	16	20	21	22	16
Average number of arrangements used	2	3	2	2	2	2	2
Percentage of Children Who Received Care in More than One Arrangement Concurrently	52	73	39	53	56	48	50
Percentage of Children Whose Primary Child Care Arrangement Wasa:
Not in child care	14	7	20	11	13	19	10
Early Head Start/Head Start	21	57	1	20	30	18	12
Other child care center	17	10	20	17	16	17	19
Nonrelative	14	5	19	14	12	10	22
Parent or stepparent	8	5	9	7	8	5	10
Grandparent or great-grandparent	18	12	19	21	16	20	19
Other relative	8	3	10	9	3	11	8
Parent at school or work	0	0	1	0	1	0	0
Percentage of Children Whose Primary Arrangement Included Care During:
Evenings	30	25	31	32	31	29	31
Early mornings	49	42	53	50	47	49	52
Weekends	17	11	18	21	15	17	21
Overnight	22	12	33	16	13	37	15
Sample Size	920-988	189-206	410-459	306-323	311-330	330-388	273-293
SOURCE: Parent Services Follow-Up Interviews completed an average of 7, 16, and 28 months after enrollment. NOTE: The percentages are average percentages across programs in the given group and are weighted for survey nonresponse. aArrangements that children were in for the largest number of hours during the follow up period.(back)

Program families used a wide range of providers for their primary child care arrangement (the arrangement focus children were in for the greatest number of hours) during the 26 months after program enrollment. Thirty-eight percent of families used a child care center as their primary child care arrangement, including 21 percent who used an Early Head Start center and 17 percent who used other child care centers in the community. One-third of families used child care provided by a relative—most often, a grandparent or great-grandparent—as their primary child care arrangement. Fourteen percent of families used care provided by an unrelated family child care provider. Finally, 14 percent of families did not use any child care for the focus child during the 26 months after program enrollment.

Families reported that a substantial proportion of the primary child care arrangements they used offered care during nonstandard work hours. Nearly half of the primary child care arrangements used by program families offered care during early morning hours. Nearly a third offered care during evenings hours. Smaller proportions offered care during weekends (17 percent) and overnight (22 percent).

Most program children received child care for substantial amounts of time during the 26 months after program enrollment. On average, program group families used 1,483 hours (about 14 hours per week) of child care during the 26 months following enrollment; one-quarter used child care for at least 20 hours a week (a total of 2,253 hours) during this period (Table III.6).14 Program group children received 688 hours of center care, or about six hours per week, on average. As expected, families in center-based programs used the most child care (2,354 hours, or 21 hours per week), on average, followed by mixed-approach programs (1,458 hours or 14 hours per week), and home-based programs (1,007 hours or 9 hours per week). Families in center-based programs also used the most center care, on average (1,580 hours).15

TABLE III.6

INTENSITY OF CHILD CARE USE BY EARLY HEAD START FAMILIES, FOR THE FULL SAMPLE AND KEY PROGRAM SUBGROUPS
	Full Sample	Center- Based	Home- Based	Mixed- Approach	Early Implementers	Later Implementers	Incomplete Implementers
	Program Approach				Pattern of Implementation
Average Number of Hours That ChildrenWere In:
Any child care	1,483	2,354	1,007	1,458	1,466	1,251	1,782
Any center-based child care	688	1,580	266	586	894	549	609
Percentage of Children Who Were in Any Child Care For:
0 hours per week	14	7	21	12	13	19	10
1-9 hours per week	39	28	45	41	42	39	37
10-19 hours per week	22	24	19	23	18	22	26
20-29 hours per week	11	15	8	12	11	12	9
30-39 hours per week	6	11	4	6	8	4	8
40 or more hours per week	7	15	4	7	8	5	10
Average hours per week	14	21	9	14	14	12	16
Percentage of Children Who Were in Any Center-Based Child Care For:
0 hours per week	58	23	79	58	50	61	65
1-9 hours per week	18	27	12	21	21	17	16
10-19 hours per week	12	22	5	12	13	13	9
20-29 hours per week	6	13	2	6	8	7	3
30-39 hours per week	2	8	1	1	4	1	3
40 or more hours per week	3	8	1	2	4	1	4
Average hours per week	6	14	3	5	8	5	5
Percentage of Children Who Were in Any Child Care For:
0% of the follow-up period	15	7	22	12	13	20	10
1-19% of the follow-up period	7	2	11	6	7	9	5
20-39% of the follow-up period	10	4	11	14	10	11	10
40-59% of the follow-up period	10	7	11	12	12	9	10
60-79% of the follow-up period	11	9	15	14	12	13	14
80-99% of the follow-up period	18	16	15	21	18	14	23
100% of the follow-up period	26	53	16	20	28	22	29
Percentage of Children Who Were in Any Center-Based Child Care For:
0% of the follow-up period	50	21	68	48	44	52	55
1-19% of the follow-up period	8	4	10	8	7	7	9
20-39% of the follow-up period	8	8	7	10	9	8	9
40-59% of the follow-up period	6	5	5	10	7	8	5
60-79% of the follow-up period	8	8	6