Table of Contents | Previous | Next |
VIII. THE QUALITY OF SELECTED CHILD DEVELOPMENT SERVICES
An important dimension of program implementation is the degree to which programs offer high-quality services. The Early Head Start program guidelines specifically require programs to provide high-quality early education services, home visits, and parent education, and to ensure that infants and toddlers who need child care receive high-quality care. The guidelines also require programs to ensure that the full range of family-oriented services is of high quality.
Our examination of quality focuses on two important child development services—child care and child development home visits—because these are core Early Head Start services, and measurement tools existed or could be developed for assessing their quality. We begin this chapter by describing our methods for assessing the quality of core child development services, and then report on the progress programs made in improving the quality of these services between fall 1997 and fall 1999.
A. METHODS FOR ASSESSING QUALITY
We used two main methods for assessing the quality of core child development services. First, we assessed the quality of child care used by Early Head Start families using data from observations of the child care settings used by Early Head Start children. Second, we developed rating scales and a rating process similar to those used for assessing implementation (see Chapter IV) to rate inputs to the quality of child care in Early Head Start centers, programs’ efforts to assess and monitor quality in community child care settings and to support child care providers, and inputs to the quality child development home visits. In this section we describe the data sources and analytic methods we used to rate inputs to the quality of child care and home visits and to assess child care quality.
1. Rating Inputs to Quality
We developed scales for rating the “inputs to quality” of child care and child development home visits. The literature on child care research indicates that researchers take a variety of approaches to defining quality in child care (Love, Schochet, and Meckstroth 1996). Some define quality as including such factors as staff qualifications and retention or stability (Ferrar 1996; Ferrar et al. 1996; and Phillips and Howes 1987); others consider these as contributors to program quality (for example, Layzer et al. 1993). We adopted the latter approach in the Early Head Start evaluation, considering elements that support what happens in classrooms or in home visits to be “inputs to quality.” For child care, the inputs we rated were curriculum, assignment of primary caregivers, educational attainment of teachers, and teacher turnover. For home visits, the inputs we rated were supervision, home visitor training, home visitor hiring, planning home visits, frequency of home visits, emphasis on child development, and integrating home visits with other services. We also developed a scale for rating all programs on the extent to which they monitored the quality of child care arrangements and provided training and support for child care teachers.
We used data from site visits conducted in fall 1997 and fall 1999 to assign ratings to programs. To facilitate the assignment of ratings, we assembled site visit data into checklists organized according to the inputs to quality we rated (Appendix A).
2. Observations of Child Care Quality
We used data from observations of Early Head Start children’s child care settings (including Early Head Start centers, community child care centers, and family child care homes) conducted when they were 14 and 24 months old to assess the quality of child care that children received.1 These observations include data collected using a slightly shortened version of the Infant-Toddler Environment Rating Scale (ITERS; Harms et al. 1990) and the Family Day Care Rating Scale (FDCRS; Harms and Clifford 1989), as well as observed child-teacher ratios and group sizes. These scales are widely used and consist of 35 items to assess the quality of care.2 These scales produce scores on each item ranging from 1 to 7, in which 3 is described as minimal care, 5 as good care, and 7 as excellent care.
To compute average ITERS scores for Early Head Start centers, we began by averaging the observations for each classroom.3 Classrooms were observed as often as once per quarter (or more often if staff or children had changed since the last observation), depending on when Early Head Start children were in care. We then averaged the classroom scores for each center. If a program operated multiple centers, we averaged the center scores to generate an average program score. Thus, the average ITERS scores reported here do not reflect the average quality of care received by individual children. Rather, they represent the average quality of Early Head Start centers, determined at the classroom level.
To compute average ITERS scores for community child care centers, we computed an average score for each center, and then averaged the center scores to compute an average site score. Likewise, to compute average FDCRS scores, we computed an average score for each family child care home, and then averaged these home scores to compute an average site score.
Observed child-teacher ratios and group sizes were calculated based on child and adult counts taken during structured observations of child care settings.
B. INPUTS TO CHILD CARE QUALITY
In fall 1999, more than half of the 12 Early Head Start research programs with child care centers received good or high ratings on several inputs to child care quality, including curriculum, assignment of primary caregivers, and educational attainment of Early Head Start teachers (Figure VIII.1). However, only two programs received a good or high rating on staff turnover.
To receive a good rating for curriculum as an input to child care quality, Early Head Start centers had to use a curriculum strongly integrated into the center’s daily activities and appropriate for the population served. Centers that individualized their curriculum for each child received a high quality rating. Eight out of the 12 research programs with centers received a good or high quality rating on this dimension in fall 1999 (Figure VIII.1).
To receive a good rating for assignment of primary caregivers, Early Head Start centers had to assign primary caregivers to children and adhere to these assignments throughout the day. In addition, primary caregivers had to conduct almost all routine care activities for the children in their group. To receive a high rating, primary caregivers had to communicate regularly with parents and plan the activities for children in their group. Eight out of the 12 research programs with centers met the criteria for a good or high quality rating on this dimension in fall 1999.
To receive a good rating for educational attainment of teachers, most teaching staff in Early Head Start centers had to have a CDA, an associate’s degree, or a bachelor’s degree, or be in CDA training. To receive a high rating, all teaching staff had to have a CDA, an associate’s degree, or a bachelor’s degree, or be in CDA training. Seven out of the 12 research programs with centers met the criteria for a good or high quality rating on this dimension in fall 1999. On average, 58 percent of center staff had their CDA or higher degree in fall 1999, and an additional 19 percent were working on obtaining a CDA.
![[FIGURE VIII.1 NUMBER OF PROGRAMS WITH CENTERS IN WHICH INPUTS TO QUALITY WERE RATED AS GOOD OR HIGH]](images/pways_fig8_1.gif)
[D] |
Only two programs with centers received a good or high rating in the area of teacher retention, which required the centers’ teacher turnover rate to be below 20 percent for the previous year. On average, about 39 percent of full-time and part-time staff working directly with children in Early Head Start centers left and were replaced during the year prior to the fall 1999 site visits.
We rated all 17 research programs on two types of inputs to child care quality—quality monitoring and training and support for providers—and in these areas, between one-fourth and one-half of the programs received a good or high rating (Figure VIII.2). These ratings encompass monitoring and teacher training and support in both Early Head Start centers and other community child care settings. To receive a good rating for quality monitoring, Early Head Start centers had to carry out ongoing quality assessments and give feedback to staff about the care they were providing.
To receive a high rating, the approach to quality improvement had to be systematic. To receive a good rating, programs in which some or all children received child care in community centers or family child care homes had to (1) assess the quality of child care settings before referring children, and (2) monitor child care quality regularly for most children in care, whether or not the program placed the children in their child care settings. To receive a high rating, these programs had to take a comprehensive approach to assessing quality and had to monitor quality regularly for all children in child care. Seven out of the 17 research programs, including the 4 center-based programs, met the criteria for a good or high quality rating on this dimension in fall 1999.
To receive a good rating for training and support of child care providers, programs had to provide regular training to nearly all child care teachers and family child care providers caring for Early Head Start children, and if some children were in relative care, the program had to provide support and training to some of them as well. To receive a high rating, the program had to provide the training according to individual needs and to base training activities on an individualized training plan. Four out of the 17 research programs (the four center-based programs) met the criteria for a good or high quality rating on this dimension in fall 1999.
![[FIGURE VIII.2 EARLY HEAD START INPUTS TO CHILD CARE QUALITY THAT WERE RATED GOOD OR HIGH FALL 1999]](images/pways_fig8_2.gif)
[D] |
C. OBSERVED CHILD CARE QUALITY
Child care observations were conducted in three settings that represent the range of arrangements that Early Head Start children were in: (1) Early Head Start centers, (2) community child care centers that Early Head Start children attended, and (3) family child care homes (both regulated and unregulated) that Early Head Start children attended. In this section, we report on child care observations conducted between October 1997 and September 1999 in Early Head Start centers in 9 research sites, community child care centers in 16 research sites, and family child care homes in 14 sites.
1. Quality in Early Head Start Centers
Our analysis indicates that the quality of care provided by Early Head Start centers during their first three years of serving families was good (Table VIII.1). All programs scored above 4 on average, on the ITERS, with the average being 5.3 (in the good range) in both the first and second years after the fall 1997 site visits.4
In most programs, the average ITERS score changed by only a few points, but in one program it fell substantially (from 6.3 in the first year to 5.6 in the second year, still well within the good range), and in one program it rose substantially (from 5.5 to 6.3). Early Head Start centers in several programs received average ITERS scores of 6 or above, which indicates good to excellent care. Comparisons with other child care quality studies show that Early Head Start centers were doing very well. For example, the Cost, Quality, and Child Outcomes Study Team (1995) found that the average ITERS score across infant-toddler classrooms in the four states studied was only 3.4, and 40 percent of the classrooms in that study received ratings below 3.0; no Early Head Start center had an average score below 4.2 in 1997-1998 or 4.4 in 1998-1999 (Table VIII.1).
| Program | Early Head Start Centers (ITERS) | Community Child Care Centers (ITERS) | Family Child Care Homes (FDCRS) | |||
|---|---|---|---|---|---|---|
| 10/97-9/98 | 10/98-9/99 | 10/97-9/98 | 10/98-9/99 | 10/97-9/98 | 10/98-9/99 | |
| A | -- | -- | -- | 4.4 (1) | 4.0 (7) | 4.0 (5) |
| B | -- | -- | 3.7 (6) | 4.7 (7) | -- | 2.6 (2) |
| C | 4.6 (14) | 4.5 (11) | 2.7 (1) | 4.2 (3) | 4.1 (2) | -- |
| D | 4.2 (16) | 4.4 (36) | -- | 5.9 (1) | -- | -- |
| E | 6.3 (8) | 5.6 (4) | 4.2 (3) | 4.5 (7) | 3.8 (6) | 3.7 (8) |
| F | -- | -- | 2.5 (1) | 4.1 (3) | 2.7 (3) | 3.9 (1) |
| G | 6.0 (4) | 5.8 (3) | -- | -- | 4.0 (13) | 4.1 (14) |
| H | 5.5 (2) | 6.3 (3) | 4.0 (5) | 4.7 (11) | 3.4 (6) | 4.3 (7) |
| I | 6.3 (4) | 5.9 (36) | 2.8 (2) | 2.9 (9) | 2.4 (1) | 2.0 (2) |
| J | -- | -- | 2.5 (3) | 3.1 (1) | 2.0 (1) | -- |
| K | -- | -- | 4.5 (1) | 5.0 (1) | 3.7 (4) | 3.3 (11) |
| L | 5.5 (2) | 5.7 (6) | 3.1 (6) | 3.4 (7) | 3.2 (3) | -- |
| M | -- | -- | 4.3 (7) | 4.4 (10) | 3.3 (4) | 4.1 (4) |
| N | 4.8 (14) | 5.2 (14) | 2.6 (2) | 2.9 (4) | 2.6 (2) | 2.1 (1) |
| O | 4.8 (32) | 4.6 (17) | 3.9 (2) | 4.9 (3) | -- | 3.0 (1) |
| P | -- | -- | 6.3 (3) | 5.9 (3) | 3.6 (7) | 4.5 (2) |
| Q | -- | -- | 5.2 (3) | 5.7 (4) | 3.9 (6) | 4.5 (7) |
| Average | 5.3 (96) | 5.3 (130) | 3.7 (45) | 4.4 (75) | 3.3 (65) | 3.5 (65) |
| SOURCE: | Based on classroom observations of the child care settings of program children conducted when children were 14 and 24 months old. The average scores include observations received from the field from October 1997 through September 1999. The average scores for community child care centers and family child care homes may include observations of child care arrangements that families chose on their own without assistance from the Early Head Start program or after dropping out of the Early Head Start program. The numbers in parentheses represent the number of classroom or home observations conducted for each program and type of child care. |
| NOTE: | The average scores shown here represent the average quality of Early Head Start and community child care settings, determined at the classroom level, used by program families. Average scores for each program are not weighted to reflect the number of program children participating in each classroom, center, or home. |
| ITERS = FDCRS = | Infant-Toddler Environment Rating Scale |
We also examined scores on subscales of the ITERS. The programs achieved good quality, on average, in all areas, although scores were somewhat lower in three areas: learning activities, adult needs, and furnishings (Figure VIII.3). Thus, programs may want to focus on these areas in future quality enhancement efforts. Programs were particularly strong in the area of interactions: three had average scores of 7.0 on this subscale.
Observed child-teacher ratios and group sizes were good in both time periods. Over time, as the centers became fully enrolled and as more children were being observed at 24 months of age rather than 14 months, average observed group sizes and ratios tended to increase slightly, but they remained well below the thresholds set by the revised Head Start Program Performance Standards (four children per teacher and eight children per group). Average child-teacher ratios rose slightly, from 2.3 in the first year to 2.9 in the second year (Table VIII.2). Average group sizes also rose slightly, from 5.3 to 5.9 (Table VIII.3).
2. Observed Child Care Quality in Community Child Care Centers
Our analyses suggest that the quality of child care received by Early Head Start children in community child care centers varied widely, but was consistently minimal, on average (Table VIII.1).5 However, the average ITERS score for classrooms that we observed in community child care centers was 3.7 in 1997-1998 and 4.4 in 1998-1999, indicating that the quality of care in community child care settings may have improved over time.6 However, average quality remained lower than the quality of care provided in Early Head Start centers.
![[FIGURE VIII.3 EARLY HEAD START CENTERS AVERAGE ITERS SUBSCALE SCORES, 1998-99]](images/pways_fig8_3.gif)
[D] |
| Early Head Start Centers | Community Child Care Centers | Family Child Care Homes | ||||
|---|---|---|---|---|---|---|
| 10/97-9/98 | 10/98-9/99 | 10/97-9/98 | 10/98-9/99 | 10/97-9/98 | 10/98-9/99 | |
| A | -- | -- | -- | 5.3 (1) | 4.5 (7) | 2.5 (3) |
| B | -- | -- | 3.9 (6) | 3.7 (7) | -- | 4.5 (2) |
| C | 1.9 (14) | 2.2 (11) | 6.8 (1) | 2.3 (3) | 2.2 (2) | -- |
| D | 3.3 (16) | 2.9 (36) | -- | 2.7 (1) | -- | -- |
| E | 2.5 (8) | 2.8 (4) | 2.8 (3) | 4.4 (7) | 2.8 (6) | 3.6 (8) |
| F | -- | -- | 3.6 (1) | 3.1 (3) | 6.8 (3) | 4.0 (1) |
| G | 2.7 (4) | 3.5 (3) | -- | -- | 3.8 (13) | 4.5 (14) |
| H | 1.3 (2) | 3.2 (3) | 4.7 (5) | 4.5 (11) | 2.0 (6) | 2.0 (7) |
| I | 2.8 (4) | 2.9 (36) | 3.0 (2) | 6.3 (9) | 4.0 (1) | 1.7 (2) |
| J | -- | -- | 4.0 (3) | -- | 1.0 (1) | -- |
| K | -- | -- | 4.1 (1) | 3.5 (1) | 3.6 (4) | 3.1 (11) |
| L | 1.6 (2) | 2.6 (6) | 4.2 (6) | 5.9 (7) | 1.6 (3) | -- |
| M | -- | -- | 5.3 (7) | 4.4 (10) | 5.6 (4) | 6.1 (4) |
| N | 2.3 (14) | 2.3 (14) | 6.0 (2) | 8.2 (4) | 1.0 (2) | -- |
| O | 2.7 (32) | 3.6 (17) | 3.8 (2) | 3.1 (3) | -- | 9.5 (1) |
| P | -- | -- | 6.0 (3) | 4.6 (3) | 3.1 (7) | 5.5 (2) |
| Q | -- | -- | 4.0 (3) | 4.3 (4) | 4.6 (6) | 3.3 (7) |
| Average | 2.3 (96) | 2.9 (130) | 4.4 (45) | 4.4 (74) | 3.3 (65) | 4.2 (65) |
| SOURCE: | Based on classroom observations of the child care settings of program children conducted when children were 14 and 24 months old. The average ratios include observations received from the field from October 1997 through September 1999 for all programs with at least three observations for a particular type of child care setting (Early Head Start centers, community child care centers, or family child care homes). The average ratios for community child care centers and family child care homes may include observations of child care arrangements that families chose on their own without assistance from the Early Head Start program or after dropping out of the Early Head Start program. The numbers in parentheses represent the number of classroom or home observations conducted for each program and type of child care. |
| NOTE: | The average ratios shown here are the average teacher-child ratios in Early Head Start and community child care settings, determined at the classroom level, used by program families. Average ratios for each program are not weighted to reflect the number of program children participating in each classroom, center, or home. |
| Early Head Start Centers | Community Child Care Centers | Family Child Care Homes | ||||
|---|---|---|---|---|---|---|
| 10/97-9/98 | 10/98-9/99 | 10/97-9/98 | 10/98-9/99 | 10/97-9/98 | 10/98-9/99 | |
| A | -- | -- | -- | 9.0 (1) | 4.6 (7) | 3.0 (5) |
| B | -- | -- | 5.5 (6) | 8.9 (7) | -- | 6.7 (2) |
| C | 5.9 (14) | 7.1 (11) | 13.7 (1) | 6.7 (3) | 5.9 (2) | -- |
| D | 8.0 (16) | 6.3 (36) | -- | 13.7 (1) | -- | -- |
| E | 5.9 (8) | 6.1 (4) | 5.8 (3) | 10.9 (7) | 2.9 (6) | 4.1 (8) |
| F | -- | -- | 8.7 (1) | 8.1 (3) | 6.8 (3) | 4.0 (1) |
| G | 5.2 (4) | 5.2 (3) | -- | -- | 5.0 (13) | 5.3 (14) |
| H | 2.5 (2) | 6.0 (3) | 7.9 (5) | 7.6 (11) | 3.1 (6) | 3.2 (7) |
| I | 5.5 (4) | 4.8 (36) | 3.9 (2) | 8.7 (9) | 4.0 (1) | 2.9 (2) |
| J | -- | -- | 10.7 (3) | -- | 1.0 (1) | -- |
| K | -- | -- | 11.3 (1) | 8.8 (1) | 6.7 (4) | 3.2 (11) |
| L | 2.8 (2) | 4.2 (6) | 8.1 (6) | 10.7 (7) | 1.7 (3) | -- |
| M | -- | -- | 8.1 (7) | 6.9 (10) | 6.5 (4) | 6.9 (4) |
| N | 5.2 (14) | 5.9 (14) | 11.5 (2) | 13.8 (4) | 1.0 (2) | -- |
| O | 6.7 (32) | 7.2 (17) | 7.5 (2) | 7.9 (3) | -- | 9.5 (1) |
| P | -- | -- | 12.3 (3) | 8.0 (3) | 3.5 (7) | 5.8 (2) |
| Q | -- | -- | 9.8 (3) | 6.3 (4) | 6.5 (6) | 5.0 (7) |
| Average | 5.3 (96) | 5.9 (130) | 8.9 (45) | 9.1 (74) | 4.2 (65) | 5.0 (64) |
| SOURCE: | Based on classroom observations of the child care settings of program children conducted when children were 14 and 24 months old. The average group sizes include observations received from the field from October 1997 through September 1999. The average group sizes for community child care centers and family child care homes may include observations of child care arrangements that families chose on their own without assistance from the Early Head Start program or after dropping out of the Early Head Start program. The numbers in parentheses represent the number of classroom or home observations conducted for each program and type of child care. |
| NOTE: | The average group sizes shown here are the average group sizes in Early Head Start and community child care settings, determined at the classroom level, used by program families. Average group sizes for each program are not weighted to reflect the number of program children participating in each classroom, center, or home. |
The average child-teacher ratios in classrooms in community child care centers, 4.4 in both the first and second years, exceeded the maximum ratio of four children per teacher specified in the revised Head Start Program Performance Standards (Table VIII.2). Similarly, the average group size in community child care centers, 8.9 in 1997-1998 and 9.1 in 1998-1999, exceeded the maximum group size of eight children specified in the standards (Table VIII.3).
3. Observed Child Care Quality in Family Child Care Settings
Observational data suggest that the observed quality of child care that Early Head Start children received in family child care settings was consistently minimal, but ratios and group sizes were good.7 The average FDCRS score for the family child care settings was 3.3 in 1997-1998 and 3.5 in 1998-1999 (Table VIII.1), both in the minimal quality range. The average child-caregiver ratio in the family child care settings that we were able to observe was 3.3 in the first year and 4.2 in the second (Table VIII.2). The average group size in the family child care settings that we were able to observe was 4.2 children in the first year and 5.0 in the second observation period (Table VIII.3).
D. INPUTS TO THE QUALITY OF CHILD DEVELOPMENT HOME VISITS
The inputs to the quality of child development home visits that we rated indicate that overall, the quality of child development home visits improved substantially between the fall 1997 and fall 1999 site visits. By fall 1999, the quality of child development home visits in 11 research programs that served some or all families in a home-based option was rated as good or high, up from 9 programs in fall 1997 (Figure VIII.4).
The greatest improvements in inputs to the quality of child development home visits were in the areas of supervision, emphasis on child development, and home visit planning. In other areas, most programs received high ratings in both fall 1997 and fall 1999.
The number of programs that were rated as providing good- or high-quality supervision of home visitors increased from 8 programs in fall 1997 to all 13 programs in fall 1999 (Figure VIII.5). The programs rated as providing good-quality home visitor supervision provided regular individual and group supervision that included support, teaching, and evaluation; they also provided mentoring. Supervisors paid some attention to child development, tracked the frequency of home visits carefully, and accompanied home visitors on some home visits. Programs rated as providing high-quality supervision also provided regular opportunities for home visitors to discuss their experiences with peers, and supervisors had a regular plan for accompanying home visitors on home visits.
The number of programs rated as providing good or high quality in terms of the number of completed home visits per month increased from six to eight. The relatively small improvement in the number of programs with quality rated as good or high on this dimension reflects in part the increase between 1997 and 1999 in the number of completed home visits required for a good rating (from two to three per month). Thus, the small increase in the number of programs rated as providing good or high quality understates the progress programs made in this area.
![[FIGURE VIII.4 EARLY HEAD START CHILD DEVELOPMENT HOME VISITS OVERALL RATINGS OF QUALITY INPUTS]](images/pways_fig8_4.gif)
[D] |
![[FIGURE VIII.5 EARLY HEAD START CHILD DEVELOPMENT HOME VISITS RATINGS OF QUALITY INPUTS]](images/pways_fig8_5.gif)
[D] |
The number of programs that were rated as good or high quality in terms of their emphasis on child development during home visits increased from 7 to 10. In these programs, home visitors were reported to spend at least half an hour during each home visit on child development activities either with the child or with the child and parent together.
With respect to home visit planning, the number of programs rated as good or high quality increased from 9 to 12. In programs receiving a good rating, home visits were planned based on program goals and expected outcomes, and home visitors developed plans for each visit using a curriculum or protocol to guide child development activities, which were then individualized to meet the needs of individual parents and children. In programs that received a high quality rating, home visitors also worked in partnership with parents to plan child development activities.
E. SUMMARY
Between fall 1997 and fall 1999, the 17 research programs had notable success in providing consistently good-quality care in Early Head Start centers. Although the observed quality of care in community child care settings was somewhat lower, observation data indicate that quality in community child care centers may have improved over time. In addition, programs made considerable progress in improving key inputs to the quality of child care and child development home visits between fall 1997 and fall 1999. The pathways that programs took as they worked towards improving quality are examined in the next chapter.
1Observations subsequently conducted when children were 36 months old are reported in a separate paper on child care.(back)
2The shortened version of the ITERS we used excludes three items from the adult needs category (opportunities for professional growth, adult meeting area, and provisions for parents).(back)
3The average ITERS and FDCRS scores reported here have not been weighted to reflect the number of program children participating in each classroom, center, or home.(back)
4The average for the first year has been updated since preliminary findings were presented in Leading the Way, Volume III: Program Implementation (ACYF 2000a), because more observations were received from data collectors.(back)
5The community child care centers that we observed include both those that Early Head Start assessed and monitored and those that parents selected without help from Early Head Start.(back)
6This change may indicate real improvement over time, but we are cautious in making this interpretation because response rates were low in some sites. With fewer than three observations in a number of sites, we may not have sufficient data to consider this to be a representative sample of Early Head Start children’s community child care arrangements. In addition, it is possible that higher quality scores are somewhat easier for centers to attain when serving older children.(back)
7The family child care settings that were observed include both family child care homes that Early Head Start assessed and monitored and family child care homes that parents selected without help from Early Head Start.(back)
| Table of Contents | Previous | Next |