Study Description 
Summary of Findings From All Pioneer Partnerships  
Conclusions  
References  
Appendix A: Child Care Data From All Pioneer Partnerships  

Smart Start, the North Carolina Early Childhood Initiative, was established in 1993 as a partnership between state government and local leaders, service providers, and families to better serve young children and their families, ensuring that all children enter school healthy and prepared to succeed. Smart Start’s innovative approach requires local community partnerships to plan how best to meet their own community’s needs, improve and expand previous programs for children and families, and design and implement new programs. Although each partnership decides how best to meet the needs of children and families, they are all working to improve the quality of early childhood education, including center-based care. 

Are we providing high quality child care in North Carolina? Has the quality of center-based child care in pioneer Smart Start partnerships changed over time? These questions were addressed in a Smart Start evaluation study described in an earlier report, Effects of Smart Start on the Quality of Preschool Child Care (April 1997). Findings from this study suggest that child care quality has improved over time in these pioneer partnerships. From 1994 to 1996 the percentage of classes rated as providing good or better quality care increased from 14% to 25%. 

The purpose of this report is to present more detailed information about the child care centers than was included in the earlier report. The April 1997 report focused mainly on overall child care quality. This report summarizes more detailed information about the characteristics, services provided, teacher education and training, quality, and other aspects of the participating child care centers. Data were gathered in 1994 and 1996 through classroom observations and interviews with child care directors. Information from both 1994 and 1996 child care samples is included in this report. The more detailed information presented in this report should be useful to partnerships in monitoring the progress of their child care initiatives and planning new initiatives to meet the child care community’s needs. Child care directors and providers may find the information helpful in understanding the characteristics of child care centers in their community and planning strategies for improving child care services. 
 

Study Description   

In the fall and winter of 1994-95, researchers visited 184 child care centers in the first 12 Smart Start partnerships (22% of the 831 licensed centers in those counties). In 1996-97, 188 child care centers from the same counties were visited (19% of the 995 licensed centers). Ninety-one (91) centers were visited in both 1994 and 1996. Of the centers invited to participate in the study, 75% agreed to do so in 1994; 64% in 1996. Some child care directors in the 1996 sample said they did not want to participate because they had participated in too much research recently or they were dissatisfied with local Smart Start decision-making. Although the participation rate dropped from 1994 to 1996, both years’ participation rates were equal to or higher than participation rates in two recent child care observation studies with large samples (the Cost, Quality, and Outcomes Study in North Carolina and the National Institute of Child Health and Human Development Study of Early Child Care). 

In each year of data collection, data were obtained from two samples of child care centers: a partnership-nominated sample and a random sample. The nominated sample consisted of child care centers that the 12 partnerships noted were involved in local Smart Start child care quality improvement efforts. These centers were invited to participate in 1994 and again in 1996. The nominated sample was included to study directly the effect of Smart Start on child care in centers that were confirmed to be participating. The second sample of centers was randomly selected from the 1994 and 1996 lists of licensed child care centers in the partnerships, regardless of a center's participation in Smart Start. The random sample was included to measure the overall quality of care in each partnership community and to provide a comparison with the nominated sample. This process resulted in the selection of some centers both randomly and by nomination, a more frequent occurrence in small counties with fewer child care centers. These two samples were not significantly different on any child care variable in either 1994 or 1996, so they are combined in all further analyses presented here. 

At each center, data collectors completed the Early Childhood Environment Rating Scale (ECERS, Harms & Clifford, 1980) in one randomly selected preschool classroom. The ECERS is a well-established measure of child care quality that assesses seven general areas: personal care routines, furnishings and display for children, language-reasoning experiences, fine and gross motor activities, creative activities, social development, and adult needs. Scores on each of 37 items can range from 1 to 7 with the overall mean score obtained by averaging all items typically used as a global measure of the developmental appropriateness or quality of the classroom. An overall score from 1 to 3 is considered poor; scores from 3 to 5 are considered medium; and scores of 5 or greater are considered good. Data collectors also interviewed center directors to obtain information about center characteristics and services, including a checklist of 14 different Smart Start improvement activities the center or center staff might have participated in during the past year. The child care providers in the observed classrooms were asked to provide basic demographic information about themselves. 

Summary of Child Care Findings From All Pioneer Smart Start Partnerships  
 

	Observed Quality  In 1994, only 14% of the child care centers provided "good" quality care. In 1996, 25% of the centers provided "good" quality care. This increase in observed quality of care was also evident in the 91 child care centers that were visited in both 1994 and 1996.
	Licensing  The percentage of centers licensed at the higher AA level was greater in the 1996 sample than in the 1994 sample. Of the 91 child care centers that were visited in both years, the percent of AA-licensed centers rose from 37% to 52%. The increase in AA-licensed centers is additional evidence that the quality of child care is improving in these Smart Start counties.
	NAEYC Accreditation  The percent of centers voluntarily accredited by the National Association for the Education of Young Children (NAEYC) as providing high quality care remained the same across both the 1994 and 1996 samples (6%). This suggests that although child care quality is improving, many centers do not yet meet the highest standards of quality.
	Group Size and Teacher-Child Ratios  Children’s care is more likely to be developmentally appropriate if there are fewer children in the class (i.e., smaller group sizes) and more teachers per children (i.e., better teacher-child ratios). Group sizes for infants and preschoolers were slightly smaller in the 1996 sample than in the 1994 sample. The median class size for infants in 1994 (8) and 1996 (7) met the infant class size recommended by NAEYC (8). Teacher-child ratios were the same in the 1994 and 1996 samples.
	Teacher Education  Teachers with more education generally provide more appropriate care for children (Whitebook, Howes, & Phillips, 1989). More child care teachers in the 1996 sample had some college or community college coursework than did teachers in the 1994 sample (59% and 48%, respectively), and fewer teachers in 1996 had a high school education or less. This improvement in teacher education was also evident in the sample of 91 centers visited both years.
	Compensation  Although teachers in the 1996 sample were better educated, they were not generally better compensated. Teachers in 1996 earned an average of $6.00 an hour, compared to $5.77 earned by teachers in 1994. The percent of centers offering retirement benefits did not change over time. The percent of centers offering health insurance was slightly higher in 1996 than in 1994.
	Teacher Turnover  Keeping teacher turnover low is key to providing high quality care. Having warm relationships with consistent caregivers fosters children’s development. When caregivers change frequently, they cannot get to know each child and his or her unique learning styles. The average (mean) turnover rate for lead teachers across all centers was approximately the same in both the 1994 and 1996 samples (29% and 32%, respectively). Notably, some child care centers did not have any teachers leave during the previous year. In 1994, 34% of the centers had no lead teacher turnover in the previous year. In 1996, 39% had no lead teacher turnover. In the child care centers that experienced turnover, the average turnover rate for lead teachers was 47% in 1994 and 54% in 1996. These turnover rates are similar to those reported in a 1990 national survey of child care centers (Kisker, Hofferth, Phillips, & Farquhar, 1991) and are much higher than the 10% turnover rate among public school teachers.
	Poor Children Served  More centers in the 1996 sample than in 1994 served children who received government subsidies, providing opportunities for more children from low-income families to participate in preschool programs. The increased number of centers serving children who receive government subsidies was also seen in the sample of 91 centers. The median percent of subsidized children per center was similar in both the 1994 and 1996 samples (38% vs. 40%).
	Children with Disabilities  More centers in 1996 than in 1994 served children with disabilities, suggesting that there are more opportunities in the community for young children with disabilities to be served in settings with typically developing children. Directors reported more resources and supports available for staff who serve children with disabilities, particularly training and resource materials about children with disabilities.
	Screening Services  More child care centers in the 1996 sample provided vision, hearing, dental, and speech/language screenings to children compared to the 1994 sample. Sixty-seven percent (67%) of the centers provided some type of developmental screening in 1994. In 1996, 79% of the centers provided some type of screening. The increase in center-based screening was also evident in the sample of 91 centers visited in both years. The increased number of child care center screening programs may be due to local Smart Start efforts to improve local screening efforts to identify and treat children’s problems as early as possible.
	Participation in Smart Start  The percent of centers in the sample participating in one or more Smart Start quality improvement efforts remained approximately the same in 1996 as in 1994. However, more centers in the 1996 sample received on-site technical assistance and more centers received funds to move to a higher licensing level. More center directors in 1996 than in 1994 also reported using a Smart Start-sponsored lending library. The increased participation in on-site technical assistance and use of lending libraries were also evident in the sample of centers visited in both 1994 and 1996.

  

Conclusions  

More child care centers in the pioneer partnerships were providing care of higher quality for young children in 1996 compared to 1994. Child care centers in the 1996 sample were also more likely to employ better educated teachers, provide developmental screenings to children, and to enroll children with disabilities as well as children from low income families. These findings suggest that Smart Start partnerships have been successful in improving the quality and quantity of child care services for preschoolers. 

However, most child care centers in this study provided care of an average or even mediocre quality that generally does not create the responsive learning environment needed to maximize children’s development and help ensure that they enter school prepared to succeed. Teacher compensation and turnover rates—two factors important in providing high quality care—also did not change between 1994 and 1996. These data suggest that child care for many preschoolers in North Carolina is still not of high quality. 

North Carolina demonstrated its commitment to young children by creating Smart Start in 1993. Since then, the first set of partnerships have worked hard to improve the quality of child care as a way of ensuring that all children enter school prepared to succeed. The data from this report suggest that these partnerships have been successful in improving the quality of child care and should be encouraged to continue their child care quality improvement efforts. The Smart Start evaluation team will continue to monitor the quality of child care in these partnerships by visiting preschool classrooms and gathering data again from child care centers in the fall of 1998. 

References 

Child Care Data From All Pioneer Partnerships  
1994 and 1996  

Type of License 

Child Care Center Characteristics 

^{1The median is the middlemost score in a distribution below which half the scores fall. When the data contain at least 1 extreme score, as do these, the median is more appropriate to report than the mean (arithmetic average) because it is less influenced by the extreme score(s).} 

^{*The number of respondents (N) is low for some variables because some of the centers did not enroll infants or toddlers, and some centers did not have assistant teachers.} 

Services Provided by Centers 

Children Served 

Director Education 

Lead Teacher Education 

Training Activities for Center Staff 

Participation in T.E.A.C.H. (Teacher Education and Compensation Helps) 

^{1The median is the middlemost score in a distribution below which half the scores fall. When the data contain at least 1 extreme score, as do these, the median is more appropriate to report than the mean (arithmetic average) because it is less influenced by the extreme score(s).}  

Classroom Information: Group Size and Ratios 

^{1The median is the middlemost score in a distribution below which half the scores fall. When the data contain at least 1 extreme score, as do these, the median is more appropriate to report than the mean (arithmetic average) because it is less influenced by the extreme score(s).} 

^{*The number of respondents (N) is low for some variables because some of the centers visited did not enroll infants or toddlers.} 

Classroom Information: Observed Quality 

Click here for Chart 
   

Availability of Resources and Supports for Serving Children with Special Needs 

Variable	1994 Sample N = 165	1996 Sample N = 185-186
Centers reporting the availability of training focusing on children with disabilities	106 64%	142 77%
Centers reporting the availability of on-site consultation from specialists	117 71%	140 75%
Centers reporting the availability of resource materials	99 60%	140 75%
Centers reporting the availability of financial incentives	36 22%	53 28%

 

Family Involvement 
 

Variable	1994 Sample N = 165	1996 Sample N = 187
Centers that have an advisory group or board of directors	98 59%	117 63%
Median1 percent of parent representatives on advisory group or board of directors  (range)	29% (0%-100%)	23% (0%-100%)

  ^{1The median is the middlemost score in a distribution below which half the scores fall. When the data contain at least 1 extreme score, as do these, the median is more appropriate to report than the mean (arithmetic average) because it is less influenced by the extreme score(s).} 

  Smart Start Participation