DOCUMENT RESUME ED 481 219 PS 031 560 AUTHOR Whitebook, Marcy Early Education Quality: Higher Teacher Qualifications for TITLE Better Living Environments. A Review of the Literature. PUB DATE 2003-00-00 NOTE 26p.; Produced by University of California at Berkeley, Institute of Industrial Relations. Center for the Study of Child Care Employment. AVAILABLE FROM Center for the Study of Child Care Employment, Institute of Industrial Relations, University of California at Berkeley, 2521 Channing Way, Suite 5555, Berkeley, CA 94720-5555. Tel: 510-643-5091; Fax: 510-642-6432; Web site: http://www.iir.berkeley.edu/cscce. For full text: http://www.iir.berkeley.edu/cscce/pdf/teacher.pdf. PUB TYPE Information Analyses (070) EDRS PRICE EDRS Price MF01/PCO2 Plus Postage. DESCRIPTORS Bachelors Degrees; *Child Care; Child Care Quality; *Child Caregivers; Child Development; *Early Childhood Education; Educational Environment; *Educational Quality; Literature Reviews; Outcomes of Education; *Preschool Teachers; *Teacher Qualifications IDENTIFIERS *Caregiver Qualifications ABSTRACT Driven in part by research findings in early childhood development, the academic, business, and policy communities now recognize that high-quality preschool programs are an important way to help children reach their full potential. Although some current research suggests that teacher preparation at the four-year college degree level is the best way to achieve such quality, recommending such a standard for teachers in early childhood settings has raised considerable debate. This paper reviews research literature on the relationship between teacher preparation and child outcomes in ECE programs, focusing on the central question of whether teachers with at least a BA degree in ECE provide better quality preschool experiences for 3- to 5-year-olds than those with less education. The review was limited to peer-reviewed journal articles and reports, to studies examining education and care programs with classrooms serving 3- to 5-year- olds, and to studies that allowed for exploration of particular features of high-quality programs. The review concentrated on studies with samples including more settings in diverse regulatory environments, and those employing multivariate analyses exploring the relative contribution of teacher education/training to positive program quality and child development. The review notes that although the studies had some limitations in sample, measures, and analytic methods, they underscore the importance of more higher education and specialized training, and identify the role of the bachelor's degree, most often in ECE, in producing teacher behaviors consistent with high-quality programming, which in turn supports better developmental outcomes. The body of research raises several questions requiring further investigation, particularly regarding thresholds of education and training; the content, format, and quality of specialized early childhood training; variations in strategies for teachers with varying characteristics and needs; and the aspects of the adult work environment that scaffold teachers' knowledge, enabling them to engage in effective strategies with children. The review also points out that decisions about qualifications and requirements for preschool and other early childhood teachers will be driven by their feasibility as well as empirical evidence, and that the more diverse a state's preschool system, the more complex it will be to enforce common standards for professional development. (Contains 94 references.) (KB) Reproductions supplied by EDRS are the best that can be made from the original document CENTER'FOR THE STUDY OF Child Care Employment Early Education Quality: Higher Teacher Qualifications for Better Learning Environments - A Review of the Liturature Marg Whitebook' Institute of Industrial Relations University of California, Berkeley 2003 BEST COPY AVAILABLE U.S. DEPARTMENT OF EDUCATION Office of Educational Research and Improvement PERMISSION TO REPRODUGE AND EDUCATIONAL RESOURCES INFORMATION DISSEMINATE THIS MATERIAL HAS CENTER (ERIC) BEEN GRANTED BY This document has been reproduced as received from the person or organization )C originating it. 0 Minor changes have been made to improve reproduction quality. Points of view or opinions stated in this TO THE EDUCATIONAL RESOURCES document do not necessarily represent INFORMATION CENTER (ERIC) official OERI position or policy. 1 Institute of Industrial Relations Center for the Study of Child Care Employment 2521 Channing Way #5555 Berkeley, CA 94720-5555 Recognizing the importance of early learning experiences for young children's school readiness and lifelong success, states and communities throughout the country are planning or implementing initiatives to assure greater access to high-quality, publicly supported preschool services. In order to design such efforts, planners need accurate information about the particularly programmatic factors that lead to best practices and best outcomes in early education in the area of professional preparation for teachers. Increasingly, the national trend is to raise teacher qualifications, with an emphasis on college degrees in early childhood education (ECE), child development (CD) or a related field. Head Start has raised its standards to require every classroom to have an associate degree-level (AA/AS) teacher by fall 2003, and the pending reauthorization bill sets a new goal of 50 percent of Head Start teachers holding a bachelor's degree (BA/BS) by 2008. The accompanying chart (Barnett, 2003b) indicates whew the states have currently set their minimum post-secondary degree standards for teachers in state-financed prekindergarten (pre-K) programs. These range from 24 credit hours in California, a Child Development Associate (CDA) certificate in 11 states, and a BA degree in 20 states and the District of Columbia, to a master's (MA) degree (after five years of employment) in New York. Thus far, however, states have had mixed success in meeting their own standards, particularly in privately operated sectors of state-funded pre-K systems (Be llm, Burton, Whitebook, Broatch & Young, 2002). Several previous research reviews address the relationship of teacher background and quality (Barnett, 2003a; Bowman, Donovan & Burns, 2001; Howes & Brown, 2000), and have all come to the conclusion that the presence of BA-level teachers with speciali7ed training in early childhood education leads to better outcomes for young children. As we review the research on teacher qualifications and preschool program quality, we must recognize that in most states, recommending in a BA in ECE or a similar standard would result in a significant raising of standards for teachers early childhood settings, and such a suggestion can therefore trigger considerable debate. In some cases, the underlying cost implications of raising standards drive this debate, and in some communities, concerns about the existing child care workforce's ability to meet a higher also prompt questions and/or issues of linguistic and cultural diversity in the workforce standard about raising qualifications. Other concerns include the higher education system's capacity to respond to a demand for more college-educated early childhood teachers, and skepticism about how this long-underpaid field will manage to match higher standards with sufficient teacher compensation (Be llm & Whitebook, 2003). A great deal appears to be at stake for young children and their families as states and communities grapple with these issues, and the discussion about preschool teacher qualifications is in sharp contrast to trends in K-12 education. The chair of the National Research Council's Committee on Integrating the Science of Early Childhood Development, Dr. Jack Shonkoff of Brandeis University, posed a question to Congress in 2002 about the gap between our current understanding of child development and public policies related to the early education workforce: 'Thanks to Mire lla Almaraz, Dan Be llm, Erika Cruz, Michelle Munn, Laura Sakai and Marci Young for their assistance in the preparation of the review. 1 Table 1: Minimum Post-Secondary Degree Requirements For Preschool Teachers, By States Child Care' State Financed Pre-K Kindergarten State BA' None BA' ALABAMA CDA' None ALASKA BA None CDA BA ARIZONA BA' None BA ARKANSAS 6 credits 40 credits' BA CALIFORNIA 3 None CDA COLORADO BA None BA' CDA CONNECTICUT CDA BA' CDA DELAWARE BA' CDA BA DISTRICT OF COLOMBIA None None FLORIDA BA AA' None BA GEORGIA CDA CDA HAWAII BA None N/A BA IDAHO CDA or CCP BA ', BA ILUNOIS None N/A BA INDIANA None BA' IOWA None CDA BA KANSAS BA BA' None CDA KENTUCKY BA' None BA' LOUISIANA BA, None MAINE BA BA' None BA' MARYLAND 3 credits° 3 credits' MASSACHUSETI'S BA' None AA MICHIGAN BA CDA CDA MINNESOTA BA None BA N/A MISSISSIPPI BA' None CDA MISSOURI None N/A MONTANA BA BA' None NEBRASKA BA BA' None BA' NEVADA 12 credits' CDA' BA NEW HAMPSHIRE BA' CDA BA NEW JERSEY None None BA NEW MEXICO None BA NEW YORK BA AA' None BA' NORTH CAROLINA None N/A NORTH DAKOTA BA AA" BA' None OHIO BA' None OKLAHOMA BA WA None OREGON BA None BA PENNSYLVANIA BA BA, BA' BA' RHODE ISLAND BA' None SOUTH CAROUNA BA None N/A SOUTH DAKOTA BA BA' None BA TENNESSEE BA' None TEXAS BA None BA' N/A UTAH BA' 12 credits° BA VERMONT BA, None CDA VIRGINIA AA' None BA WASHINGTON None WEST VIRGINIA BA BA BA' None BA' WISCONSIN None N/A WYOMING BA .. .. ._ . re n erga re en a ; n; CCP re- gree; BA ache or's Degree; CDA Child Development Associate AA Associates no post-secondary degree requirements. Certified Childcare Professional. N/A state does not provide finances for pre-k; None 2 year many states require professional training or ongoing development. 3 1 with courses or certification in early childhood. 2 in topics related to early childhood education or child vocational child care course or 6 credits in early childhood education. 4 24 credits in Head Start requirements used because all state pre-k funds supplement Head Start program. 6 development. 5 in early childhood education, 6 of which may be non-credit early childhood education and 16 credits more in general education. 7 courses. 8 update June, 2003. Source: Bamett, W. S. (2003b). Better teachers, better preschools: Student achievement linked to teacher qualifications. NIEER Policy Facts. 2 4 BEST COPY AVAILABLE How can the recently enacted No Child Left Behind Act emphasize the need for stronger performance standards and financial incentives to attract bright and highly motivated teachers, while we simultaneously tolerate large percentages of inadequately trained and poorly compensated providers of early child care and education who have an important influence on the foundations of school readiness? (Testimony to the U.S. Senate Committee on Health, Education, Labor and Pensions, February 12, 2002, p.3.) This paper will review what the scientific research literature tells us about the relationship between teacher preparation and child outcomes in early childhood education. This review covers familiar ground for those who have read previous assessments of this literature, but focuses on the remaining unanswered questions that emerge from this body of knowledge, and attempts to identify questions for future research, the answers to which will increase the precision with which we can predict optimal outcomes for preschoolers. The central question we pose here is whether teachers with a BA degree in early childhood education (or higher) provide better-quality preschool experiences that lead to better outcomes for three- to five-year-olds. There are a number of limitations to the research, however; relatively few studies have dealt directly with this question or posed it exactly this way, and some have dealt more generally with the issue of college-level education for early childhood teachers. Nevertheless, our review of the literature indicates a reasonably strong basis for answering our central question about specialized BA-level preparation. We do not intend to imply in our central question, however, that even, teacher in a given early education program would have a BA degree; rather, a standard might be set at one BA-level teacher per classroom, or for a certain number of children (e.g., 20), with this teacher working with one or more assistant teachers. While there has been less discussion thus far of qualifications for assistant teachers, there has been some effort to set standards at the associate (AA) degree level. Methods We restricted our literature review to articles published in peer-reviewed journals and reports issued by agencies that subject their reports to peer advising. Most of the studies reviewed here have been published in journals, with a few released as freestanding reports. For some of the large-scale studies, such as the Cost, Quality and Child Outcomes Study, the National Child Care Staffing Study, the Florida Quality Improvement Study, and the Then and Now Study, both peer-reviewed articles and fmal reports were consulted. Our initial scan of articles included those that focused on programs in home and center settings serving all ages of children prior to elementary school. Because our immediate concern revolves around the appropriate staffing of publicly financed preschool programs for three- and four-year-olds, we restricted our review to studies that could best address that issue. Thus, we initially narrowed this review to studies that examined programs with classrooms serving three- to five-year-olds, either exclusively or within a larger multi-age program, and thai allowed for exploration of the particular features of high-quality programs for the later preschool years. Consequently, studies focused exclusively on teacher background and infant/toddler development 5 3 are not discussed in detail in this review (Burchinal, Roberts, Nabors & Bryant, 1996; Burchinal, Roberts, Riggins, Zeisel, Neebe & Bryant, 2000; NICHD, 1996). We also limited this review to center-based programs, although a number of studies explore the role of family child care provider training and education as it relates to child care quality and child outcomes (Clarke-Stewart, Vandell, Burchinal, O'Brien, & McCartney, 2002; DeBord & Sawyers, 1996; Kontos & Riessen, 1993, Kontos, Howes & Galinksy, 1996). These, too, deserve exploration, particularly as some states move toward establishing publicly-supported preschool programs in home-based settings (Be Ilm & Whitebook, 2003). It is beyond the scope of this review to attempt a comparison of the literature for both center-based and home-based services, given the diversity across these sectors in terms of regulation, program structure and size, age groups of children served, and provider training, education and motivation. Having decided to focus on studies related to center-based care for three- and four-year- olds, we next categorized the studies according to research design. Some studies examine only teacher behaviors or overall program quality, and do not include child outcomes; conversely, studies looking at child outcomes may include no specific information about teacher behavior. Thus, we divided studies from our initial scan into two categories: observational studies of program quality (with and without child outcome data) and studies that explored issues of teacher preparation but did not include direct observations of teacher behavior. The review focuses primarily on the observational studies. Another distinction among the observational studies involves how they collect and analyze teacher background data. Many studies employ linear measures of education and training, either in designing variables or in conducting analyses. They look at training and education as continuous variables, and address only whether more education or training makes a difference, but do not explore the particular role of the BA or of higher degrees in contrast to other levels of education. They may find, for example, that more education and/or training is better than less, but do not establish clear cut-off points (e.g., that an AA degree is less effective than a BA). Further, many studies confound formal education and other child-related training. In part, this is because many in the field who have a BA, for example, have also completed high levels of other training in early childhood education, and this training, when analyzed as a separate variable from college education, may in fact reflect differences in education as well. Some studies also lack specificity regarding what constitutes child-related training; while some include only college-level training, others include informal workshops or high school and vocational school training. Finally, the content and amount of training is often left unspecified. This issue will be addressed in the discussion. Thus, we further categorized the observational studies into two groups: those that specially weigh in on the question of the role of the bachelor's degree, and those that more generally examine the question of teacher education and training. In selecting studies for consideration in this review, we also examined the size and complexity of their samples and their analytic rigor. Some studies have small, insufficiently diverse samples with respect to locale or background characteristics of the subjects, which makes it difficult to generalize fmdings. Such studies may include so few teachers, for example, that is difficult to look at subgroup differences, or there may be little variation among subjects with respect to ethnicity or professional background. In such cases, it is impossible to determine the influence of caregiver socioeconomic status on levels of education. Further, many studies focus on samples selected from 6 4 demographic or one locale, and thus the results are not necessarily applicable to states with different regulatory features. Some studies over-represent certain subgroups (e.g., nonprofit child care centers or low- income families), which also limits the ability to generalize results. Some studies look at relationships among teacher characteristics and behaviors or child outcomes, but do not use multivariate statistics to understand predictors or effect sizes of various variables. In a few studies, observers were not completely blind to who had participated in various educational or training programs. To the extent possible, we concentrate on those with samples that include more settings in diverse regulatory environments, and those employing multivariate analyses that explore the relative contribution of teacher education/training to positive program quality and child development. There are few experimental designs or longitudinal studies in this literature. It is important to note that the studies reviewed here rely on the natural variation in teacher background that occurs in early care and education programs, and that the samples were drawn to capture the variation of programs in a particular community rather than to test various program designs. Some studies do not include information about pre-existing motivational or other differences among subjects, such as sufficient detail about how program structure, work environment or background climate (i.e., the educational and training background of one's co-workers) impact teacher behavior and child outcomes. None of the studies look at the interactions among formal education, child-related training, and the linguistic and cultural match between young children and their teachers. Also, this review encompasses early care and education programs broadly defmed, rather than looking only at publicly-funded preschool programs and variations among them based on teacher background and other features. We therefore cannot determine whether there are specific features of such programs, such as length of day or curriculum content, that may interact with teacher background so as to impact program quality and child outcomes differently than in other types of early care and education services. In many ways, state-funded preschool shares as many characteristics with early elementary school as it does with child care in terms of structure and goals, and may require different lines of investigation to answer questions around appropriate teacher preparation (Barnett, 2003a). Observations of Teacher Behavior and Program Quality: The Contribution of Education and Training More is better Initiated by the federal government in the 1970s, the National Day Care Study (NDCS) (Ruopp, Travers, Glantz & Coelen, 1979) sought to guide the construction of national child care standards by identifying the key provisions of child care quality in center-based, full-day programs that best predict good outcomes for children, and the costs associated with them. The findings earmarked specialized child-related training, along with group size, as the most potent regulatable predictors of child care quality. Among the 3,167 child care centers in seven states that were contacted by phone for the study, 70 received onsite visits. The study found that in classrooms whose lead teachers had child-related education, teachers interacted with children more, and children showed more cooperation and greater task persistence. 7 When preschoolers attended centers with a higher proportion of trained caregivers, they exhibited greater cognitive test score gains than those cared for by less-trained teachers. These general fmdings, commonly interpreted as "some training is better than none, and more child-related education is better than less," advanced the discussion of the importance of specialized preparation for child care quality at the time, but did not address questions related to college degrees (only information about years of education were collected about staff) or the level and intensity of effective training (only whether staff had taken courses, seminars or in-service training programs related to child care in the last year). An early study by Berk (1985) is often cited as evidence that overall education is related to positive teacher behaviors, although its conclusions are based on a small, homogeneous sample (37 teachers, all Caucasian, in 12 centers in one Midwestern city, caring mainly for middle-class three- to five-year-olds). Still, the Berk study contributed to the understanding of how the behavior of a particular group of teachers is influenced by more formal education. Unfortunately, it distinguishes only between staff with two or more years of college education,' and those with only a high school diploma. Using the Prescott, Hone and Kritchevsky (1972) observational system, teachers with at least two years of college were found to engage in more responsive encouragement of children, and teacher behaviors involving development of verbal skills were almost three times more frequent among the college-educated teachers than among those with high school only. While none of the "high-school-only" teachers had received child-related training, the college group was divided between those with child-related preparation and those with none. No differences in behaviors were found in this study based on specialized early childhood education. In a small observational study of 24 centers from eight counties in one state, Dunn (1993) found a strong relationship between the number of years of education and teacher behavior, although no distinctions were made among levels of education beyond high school. Program quality was measured with the Early Childhood Environment Rating Scale (ECERS) (Harms & Clifford, 1980),3 and through audiotaped and coded observations of lead teachers interacting with children. Slightly more than one-half of the teachers had completed some college education, most of them with a child-related major. Higher ECERS scores were highly correlated with more education and fewer years of experience. Because teachers with higher levels of education were more likely to have formal or informal training specifically related to children, and were also more likely to have college- level work in a child-related area, it is not possible to tease apart the relative roles of formal 2 All but two of the 29 teachers in the college cohort had completed a BA degree. 3 The Early Childhood Environment Rating Scale (ECERS) provides information about appropriate caregiving and activities that occur within a particular preschool classroom. The ECERS comprehensively assesses the day-to-day quality of care. The 37-item scale is organized under seven categories: Personal care routines; furnishings and displays for children; language-reasoning experience; fine and gross motor activities; creative activities; social development; and adult needs. Individual items are rated from a low of 1 to a high of 7. A rating of 3 on these scales indicates "minimally acceptable quality," while 5 indicates "good" quality and 7 indicates "excellent quality." The ECERS authors also developed similar measures for infant/toddler and home-based settings, the Infant Toddler Environment Rating Scale (ITERS) and the Family Day Care Environment Rating Scale (FDCRS). A revised ECERS scale, the ECERS-R (Harms, Clifford & Cryer, 1998), addressing certain concerns and omissions noted by researchers and practitioners, is now used in lieu of the original ECERS. A forthcoming article (Sakai, Whitebook, Wishard & Howes, in press), based on observations of the same classrooms using both measures, concludes that the two measures provide comparable measures of quality. The ECERS-R made some modifications in the subscales, but uses the same scoring system. Studies employing different versions of the ECERS can therefore be compared with confidence. 6 education per se and specifically child-related training. The study also explored the impact of teacher background on child outcomes, using teacher self-reports. Teachers were also asked to rate a randomly selected boy and girl in their classroom, ranging in age from 30 to 60 months, on selected characteristics of social and cognitive development and play. Specialized training at the college level was a positive predictor of children's cognitive development, based on the intellectual subscale of the Classroom Behavior Inventory (Schafer et al., 1987), along with higher ECERS scores and fewer years of experience in the field. When family and center selection effects were controlled, however, the findings were restricted only to children of high socioeconomic status, since they were the only children whose teachers had a child-related major. Honig and Hirallal (1998) looked at the behaviors of 81 teachers working with three- to five- year-old children in 24 centers in one mid-sized urban community. Approximately two-thirds of the teachers were Caucasian and slighdy more than one-quarter were African American. Teachers were categorized as low or high in education, experience and training, based on the following distinctions: high school through an AA degree (low), vs. a BA degree or higher (high); one to four training courses (low), vs. five or more courses (high); and one to three years of experience (low), vs. four or more years (high). Teachers were observed using the ABC Scale (Adult Behaviors in Caregiving) (Honig & Lally, 1973), which permitted examination of clusters of positive and negative behaviors. Stepwise hierarchical regression was used to determine the proportion of variance that each of the independent variables contributed to each of the domains of teacher behavior. With respect to teacher facilitation of language, social, and emotional development, early childhood education/child development training accounted for most of the variance in predicting teacher behavior, with education making a smaller but significant contribution. With respect to concept development, only ECE/CD training contributed, while formal education was the only variable to significantly account for the variance leading to negative socio-emotional inputs. When all subscales were combined, ECE training accounted for 62 percent of the variance in teacher scores, and formal education increased the variance by 10 percent. Neither years of experience nor stability in one's child care position increased teacher enrichments of children's learning or socio-emotional development in any way, and a high degree of formal education did not ensure positive teacher support for young children if training and experience were low. While this study attempted to tease apart the differential contributions of education, training and experience to teacher behavior, the lack of specificity around early childhood education courses, and whether or not they were delivered at the college level, makes it difficult to draw conclusions from the findings. If those courses were part of an integrated college program, or were delivered at the college level, then the categories of education and training, as delmed, are overlapping. Nonetheless, the study advanced the discussion by suggesting the role of specialized training in child development in assessing how formal education impacts teacher behavior, and added to the evidence that experience working with young children, absent education and training, does not predict appropriate caregiving with young children (Helburn, 1995; Ruopp et aL, 1979; Whitebook, Howes & Phillips, 1990). A more recent investigation of early education programs in Massachusetts by Marshall and colleagues (2001) examined quality in 90 full-day, year-round, center-based classrooms serving children 2.9 to 5 years of age, randomly selected to proportionately represent the types of care in the state. The revised version of the Early Childhood Environment Rating Scale (ECERS-R) (Harms, Clifford & Cryer, 1998) was used to assess classroom quality, and information about program features, including teacher background, was gathered through interviews. More than one-half of the classrooms in the sample did not meet the benchmark of good care as defined by the ECERS-R. Low- and moderate-income families were less likely to have access to quality preschool care and 7 education. Only 10 percent of classroom staff at centers serving predominantly low-income families had a two-year college degree or more, compared to 28 percent at centers serving low- to moderate- income families, and 62 percent at centers serving moderate- to high-income families. As with the Dunn and Berk studies, better program quality was associated with better-educated teachers, but because distinctions were made only between those with AA degrees or higher and those with less than an AA degree, it is not possible to determine the particular contribution of the BA degree over and above a two-year degree. The contribution of the BA degree and specialized early childhood training in center-based care Eight studies specifically explore the relative contributions of a bachelor's degree and specialized early childhood training to teacher behavior and program quality, and among them are the large- scale investigations of center-based early education conducted over the last 15 years.4 These are listed below in chronological order, along with the particular citations used for this review: Bermuda College Training Program Study (Arnett, 1989) O o National Child Care Staffmg Study (Howes, Phillips & Whitebook, 1992; Whitebook et al., 1990) Cost, Quality and Child Outcomes Study (Blau, 2000; Helburn, 1995; Howes, 1995; Howes, O 1997; Phillipsen, Burchinal, Howes & Cryer, 1997) Florida Quality Improvement Study (Howes, 1997; Howes et al., 1998) o Three-State Study (Massachusetts, Georgia and Virginia) (Phillips, Mekos, Scarr, McCartney o & Abbott-Shim, 2000; Scarr, Eisenberg & Deater-Deckard, 1994) Then and Now: Changes in Child Care Staffing (Whitebook, Sakai, Gerber & Howes, 2001; O Whitebook & Sakai, in press) Head Start FACES Study (Zill et al., 2001) O New Jersey studies (Barnett, Tarr, Lamy & Frede, 1999, 2001). O 4 One major investigation not included here, because it focuses on all types of child care, including home-based care, is the longitudinal NICFID Early Child Care Research Network Study. This study examines influences on child development over time by periodically assessing child outcomes; and family and child care characteristics, of 800 children from birth to 54 months. In a 2002 report, the effects of teacher education (across all types of child care) on child care quality, and the effects of child care quality on child outcomes, were tested using structural equation modeling. Reseatchers found that teachers' educational attainment predicted teacher behavior, which in turn predicted children's social and cognitive outcomes, but these effects were mediated by maternal education and parenting behavior, as well as family economic circumstances, and were less robust than maternal and family influences. Forthcoming papers on these data may provide additional information about the influence of teacher education and behavior. 8 1 0