Table of Contents |
Figure I.1: Head Start Program Performance Measures Conceptual Framework
This figure is a pyramid that graphically represents the Head Start Performance Measures Conceptual Framework. At the peak of the pyramid is the desired outcome, which is the development of children's school readiness. The lower three tiers represent the processes by which this outcome can be achieved. Specifically, the second tier of the pyramid includes the first two processes or objectives, which are, respectively, to enhance children's healthy growth and development and strengthen families as the primary nurturers of their children. The third tier of the pyramid incorporates the third process objective of providing children with educational, health and nutritional services. It also includes the fourth process of linking children and families to needed community services. Lastly, the base of the pyramid, and the fifth identified process, is that programs be well managed and involve parents in decision-making.(back)
Figure 1.1: Most Children Entering Head Start Have Academic Skills Below National Norms
This bar chart shows the academic skills levels of children upon entry into Head Start. The x axis on this bar chart depicts four skill areas: vocabulary, early writing, letter identification, and early math. The y axis depicts the mean standard score for all FACES 2000 children assessed in English in both fall and spring where the norm is 100, the mean standard score for the lowest quartile, and the mean standard score for the highest quartile. The mean standard score for all children was 85.3 for vocabulary, 85.1 for early writing, 92.4 for letter identification, and 87.9 for early math. The mean standard score for the lowest quartile was 67 for vocabulary, 70.8 for early writing, 82.9 for letter identification, and 69 for early math. The mean standard score for the highest quartile was 102.8 for vocabulary, 101.1 for early writing, 104 for letter identification, and 104.7 for early math.(back)
Figure 1.2: Head Start Students Show Gains in Vocabulary and Early Writing Skills During Program Year
This bar chart shows the gains made by Head Start children in vocabulary and early writing skills during the program year. The x axis on this bar chart depicts four skills areas namely Vocabulary (PPVT-III), Early Writing (WJR-Dictation), Letter Identification (WJR-Letter Word ID), and Early Math (WJR-Applied Problems). The y axis depicts the mean standard score for FACES 2000 children assessed in English both at the beginning of the program year (fall 2000) and at the end of the program year (spring 2001). The mean standard score on Vocabulary was 85.3 at the beginning of the program year and 89.1 at the end of the program year. The mean standard score on Early Writing was 85.1 at the beginning of the program year and 87.1 at the end of the program year. The mean standard score on Letter Identification was 92.4 at the beginning of the program year and 92.9 at the end of the program year. The mean standard score on Early Math was 87.9 at the beginning of the program year and 89 at the end of the program year.(back)
Figure 1.3: Vocabulary Standard Scores of Children in Fall and Spring of Head Start Year: FACES 1997 versus FACES 2000
This bar chart compares the mean standard scores on vocabulary of FACES 1997 and FACES 2000 children who received English language assessments in both Fall and Spring of the Head Start year. While the x axis shows the FACES 1997 and FACES 2000 cohorts, the y axis shows the mean standard scores in fall, in spring, and the difference score between fall and spring. For FACES 1997 children, the mean standard score increased from 83.5 in the fall to 87.7 in the spring thereby resulting in a difference score of 4.2. For FACES 2000 children, the mean standard score increased from 85.3 in the fall to 89.1 in the spring thereby resulting in a difference score of 3.8.(back)
Figure 1.4: Early Writing Standard Scores of Children in Fall and Spring of Head Start Year: FACES 1997 Versus FACES 2000
This bar chart compares the mean standard scores on early writing of FACES 1997 and FACES 2000 children who received English language assessments in both Fall and Spring of the Head Start year. The x axis shows the FACES 1997 and FACES 2000 cohorts, and the y axis shows the mean standard scores in fall, in spring, and the difference score between fall and spring. For FACES 1997 children, the mean standard score increased from 84.3 in the fall to 88.2 in the spring thereby resulting in a difference score of 3.8. For FACES 2000 children, the mean standard score increased from 85.1 in the fall to 87.1 in the spring thereby resulting in a difference score of 2.0.(back)
Figure 1.5: Letter Identification Standard Scores of Children in Fall and Spring of Head Start Year: FACES 1997 Versus FACES 2000
This bar chart compares the mean standard scores on letter identification of FACES 1997 and FACES 2000 children who received English language assessments in both Fall and Spring of the Head Start year. While the x axis shows the FACES 1997 and FACES 2000 cohorts the y axis shows the mean standard scores in fall, in spring, and the difference score between fall and spring. For FACES 1997 children, the mean standard score decreased from 91.2 in the fall to 90 in the spring thereby resulting in a difference score of -1.2. For FACES 2000 children, the mean standard score increased from 92.4 in the fall to 92.9 in the spring thereby resulting in a difference score of 0.5.(back)
Figure 1.6: Early Math Standard Scores of Children in Fall and Spring of Head Start Year: FACES 1997 Versus Faces 2000
This bar chart compares the mean standard scores on early math of FACES 1997 and FACES 2000 children who received English language assessments in both Fall and Spring of the Head Start year. While the x axis shows the FACES 1997 and FACES 2000 cohorts the y axis shows the mean standard scores in fall, in spring, and the difference score between fall and spring. For FACES 1997 children, the mean standard score increased from 85.8 in the fall to 87 in the spring thereby resulting in a difference score of 1.2. For FACES 2000 children, the mean standard score increased from 87.9 in the fall to 89.0 in the spring thereby resulting in a difference score of 1.2.(back)
Figure 1.7: Children Who Enter Head Start With Lower Skills Show Larger Gains in Comparison to National Norms
This bar chart shows that children who begin Head Start with lower skill levels show greater gains compared to national averages. The four skill areas depicted on the x axis of this bar chart are Vocabulary, Early Writing, Letter Identification, and Early Math. The y axis depicts the difference in the mean standard scores from fall to spring of the Head Start year for all children receiving assessments in English at both times, for children in the lowest quartile, and children in the highest quartile. For vocabulary, the difference in the mean standard scores from fall to spring of the Head Start year for all children was 3.8, for those in the lowest quartile it was 8.4 and for those in the highest quartile it was -0.5. For early writing, the difference in the mean standard scores from fall to spring of the Head Start year for all children was 2.0, for those in the lowest quartile it was 7.0 and for those in the highest quartile it was -2.7. For letter identification, the difference in the mean standard scores from fall to spring of the Head Start year for all children was 0.5, for those in the lowest quartile it was 1.3 and for those in the highest quartile it was -0.6. For early math, the difference in the mean standard scores from fall to spring of the Head Start year for all children was 1.2, for those in the lowest quartile it was 5.9 and for those in the highest quartile it was –3.2.(back)
Figure 1.8: English-Language Vocabulary Skills of Head Start Children: Spanish-Speaking Language-Minority, Language-Majority, and Combined Population
This bar chart compares the mean standard scores on vocabulary (PPVT-III) of language-minority children, language-majority children, and the combined population in the Fall and Spring of the Head Start year, as well as the difference between the two time points. The x axis has a set of bars for the language-minority children, the language-majority children, and the combined population and the y axis shows the mean standard PPVT-III scores in fall, in spring, and the difference score between fall and spring. For the language-minority children, the fall score was 59.7, the spring score was 66.7, and the difference between the two was 7.0 points. For language-majority children, the fall score was 85.2, the spring score was 89.1, and the difference between the two was 3.8 points. For the combined population, the fall score was 81.4, the spring score was 85.7, and the difference between the two was 4.3 points.(back)
Figure 1.9: Letter-Identification in English by Head Start Children: Spanish-Speaking Language-Minority, Language-Majority, and Combined Population
This bar chart compares the mean standard scores on letter identification (WJ-R LWI task) of language-minority children, language-majority children, and the combined population in the Fall and Spring of the Head Start year, as well as the difference between the two time points. The x axis has a set of bars for the language-minority children, the language-majority children, and the combined population, and the y axis shows the mean standard WJ-R LWI task scores in fall, in spring, and the difference score between fall and spring. For the language-minority children, the fall score was 89.5, the spring score was 87.5, and the difference between the two was -2.0 points. For language-majority children, the fall score was 92.4, the spring score was 92.9, and the difference between the two was 0.5 points. For the combined population, the fall score was 91.9, the spring score was 91.9, and there was no difference between the two time points.(back)
Figure 1.10: Vocabulary and Letter Identification Skills in English and in Spanish of Head Start Children From Spanish-Speaking Language-Minority Families
This bar chart shows the change in English and Spanish vocabulary and letter identification skills of Head Start children who are members of Spanish-speaking families. The four skill areas depicted on the x axis of this bar chart are Vocabulary in English, Vocabulary in Spanish, Letter Identification in English, and Letter Identification in Spanish. The y axis depicts the mean standard scores of children from Spanish-speaking Language-Minority families in the fall, in the spring, and the difference in the scores between fall and spring. For vocabulary in English, the mean standard score increased from 59.7 in the fall to 66.7 in the spring, a difference of 7.0. For vocabulary in Spanish, the mean standard score decreased from 84.9 in the fall to 84.4 in the spring, a difference of -0.6. For letter word identification in English, the mean standard score decreased from 89.5 in the fall to 87.5 in the spring, a difference of -2.0. For letter word identification in Spanish, the mean standard score decreased from 89.6 in the fall to 86.2 in the spring, a difference of -3.4.(back)
Figure 1.11: Head Start Graduates Show Further Progress Toward National Norms in Kindergarten
This bar chart shows the progress toward national norms that Head Start children make in Kindergarten. The three skill areas depicted on the x axis of this bar chart are Vocabulary (PPVT-III), Early Writing (WJR Dictation), and Early Math (WJR Applied Problems). The y axis depicts the mean standard scores of children at the end of Head Start (Spring 1998 and Spring 1999 combined) and at the end of kindergarten (Spring 1999 and Spring 2000 combined). The sample sizes vary from 1215 to 1252. For Vocabulary (PPVT-III), the mean standard score at the end of Head Start was 88.0 and 92.7 at the end of Kindergarten. For Early Writing (WJR Dictation), the mean standard score at the end of Head Start was 87.9 and 96.2 at the end of Kindergarten. For Early Math (WJR Applied Problems), the mean standard score at the end of Head Start was 86.5 and 92.3 at the end of Kindergarten.(back)
Figure 2.1: Head Start Students Show Gains in Cooperative Classroom Behavior During Program Year
This bar chart shows the gains in cooperative classroom behavior for Head Start children. The four groups of children depicted on the x axis of this bar graph are FACES 2000-01 (all children), FACES 1997-98 (all children), the lowest quartile of the FACES 2000-01 cohort, and the lowest quartile of the FACES 1997-98 cohort. The y axis depicts the mean Cooperative Behavior Scale Score for Teacher Ratings in the fall and in the spring. For the FACES 2000-01 group the mean score for cooperative behavior in the fall was 14.6 and 16.6 in the spring. For the FACES 1997-98 group the mean cooperative behavior score was 14.7 in the fall and 16.1 in the spring. For the lowest quartile of the FACES 2000-01 cohort, the mean cooperative score was 8.6 in the fall and 13.3 in the spring. For the lowest quartile of the FACES 1997-98 cohort, the mean cooperative behavior score was 9.7 in the fall and 13.2 in the spring.(back)
Figure 2.2: Head Start Students Show Significant Reductions in Hyperactive Behavior During Program Year
This bar chart shows the decrease in hyperactive behavior of children in Head Start in one program year. The four areas of behavior depicted on the x axis of this bar chart are total problem behavior, aggressive behavior, hyperactive behavior, and withdrawn behavior. The y axis depicts the mean behavior scale score for teacher ratings in the fall and the spring. For total problem behavior the mean behavior scale score for the fall was 5.6 and 5.2 in the spring. For aggressive behavior, the mean behavior scale score was 1.8 in the fall and 1.7 in the spring. For hyperactive behavior it was 1.4 in the fall and 1.2 in the spring. For withdrawn behavior it was 2.5 in the fall and 2.3 in the spring.(back)
Figure 2.3: Head Start Students Showed Similar Patterns of Problem Behavior in 2000-2001 and 1997-1998 Program Years
This bar chart illustrates the patterns of problem behaviors for the two cohorts of FACES—the children in the 2000-2001 program year (n = 2,130) and the 1997-1998 program year (n = 1,953). The x axis has a set of bars (fall and spring scores) for all children in 2000-01, all children in 1997-98, highest quartile in 2000-01, and highest quartile in 1997-98 and program year, and the y axis has the mean Problem Behavior Score—Teacher Ratings, and ranges from 0 to 16.0. The first group shown is All Children 2000-2001. The fall score was 5.6, and the spring score was 5.2. For the second group, All Children 1997-1998, the fall score was 5.5 and the spring score was 5.2. The third group is the Highest Quartile 2000-2001. The fall score was 11.9 and the spring score was 8.7. The last group shown is Highest Quartile 1997-1998. The fall score was 11.8 and the spring score was 9.2.(back)
Figure 2.4: Head Start Students Who Enter With High Levels of Problem Behavior Show Improvement During Program Year
This bar chart depicts the changes across the program year for children with high levels of problems behaviors. The four areas of problem behavior depicted on the x axis of this bar chart are total problem behavior, aggressive behavior, hyperactive behavior, and withdrawn behavior. The y axis depicts the mean problem behavior scale score--teacher rating for the highest quartile of children in the fall and for the same children in the spring. For total problem behavior, the highest quartile of children had a mean score of 11.9 in the fall and 8.7 in the spring. For aggressive behavior, the highest quartile of children had a mean score of 4.4 in the fall and 3.4 in the spring. For hyperactive behavior the mean score was 3.0 in the fall and 2.1 in the spring. For withdrawn behavior, the mean score was 5.9 in the fall and 4.0 in the spring.(back)
Figure 2.5: Parents of Head Start Children Report Reductions in Problem Behavior During Program Year
This bar chart depicts the changes in problem behavior across the program year according to parent report. The four types of problem behavior depicted on the x axis of this bar chart are total problem behavior, aggressive behavior, hyperactive behavior, and withdrawn behavior. The y axis depicts the mean problem behavior scale score for parent ratings, and ranges from 0 to 16.0. For total problem behavior, parents reported a mean score of 6.2 in the fall and 5.6 in the spring. For aggressive behavior, parents reported a mean score of 3.2 in the fall and 2.8 in the spring. For hyperactive behavior the mean score was 1.9 in the fall and 1.7 in the spring. For withdrawn behavior, the mean score was 0.6 in the fall and 0.6 in the spring.(back)
Figure 2.6: Parents Report Greater Improvement in Head Start Students Who Enter With High Levels of Problem Behavior
This bar chart depicts the improvement in problem behavior across the program year for students who entered with high levels of problem behavior (n = 694-883) according to parent report. The four types of problem behavior depicted on the x axis of this bar chart are total problem behavior, aggressive behavior, hyperactive behavior, and withdrawn behavior. The y axis depicts the mean problem behavior scale score for parent ratings, and ranges from 0 to 16.0. For total problem behavior, parents reported a mean score of 10.4 in the fall and 8.1 for the same children in the spring. For aggressive behavior, parents reported a mean score of 4.9 in the fall and 3.7 for the same children in the spring. For hyperactive behavior the mean score was 3.7 in the fall and 2.6 in the spring. For withdrawn behavior, the mean score was 1.6 in the fall and 1.0 in the spring.(back)
Figure 3.1: The Majority of Teachers Used Either Creative Curriculum or High/Scope
This bar chart shows that the majority of Head Start teachers used either Creative Curriculum or High/Scope. The x axis on this bar chart depicts three curricula, namely, Creative Curriculum, High/Scope and Other curricula. The y axis depicts the percentage of teachers using the three types of curricula. The chart shows that 39.1 percent of teachers used Creative Curriculum, 20 percent used High/Scope and 40.9 percent used Other curricula.(back)
Figure 3.2: Teachers Liked Their Respective Curricula for a Variety of Reasons
This bar chart shows that Head Start teachers liked their respective curricula for a variety of reasons. The x axis on this bar chart depicts the reasons teachers liked their curricula and the y axis depicts the percent of teachers who provided these reasons. Nearly all teachers (99.1 percent) liked their curriculum because it addressed multiple domains, 98.3 percent because it was easy to use and adapt, 96.6 percent because it involved parents, 96.4 percent because it gave room for teacher creativity, and 92.1 percent because it had adequate learning materials.(back)
Figure 3.3: Most Teachers From Northeast, Midwest, and West Used Creative Curriculum or High/Scope While Majority of Teachers From the South Used Other Curricula
This bar chart shows the distribution of curricula used by region of the country. The x axis on this bar chart depicts the distribution in use of Creative Curriculum, High/Scope and Other curricula among the Northeast, Midwest, South, and West regions. The y axis depicts the percentage of teachers using the three types of curricula in the four regions. In the Northeast, 61.4 percent of the teachers used Creative Curriculum, 17.3 used High/Scope, and 21.3 used Other curricula. In the Midwest, 42.7 percent of the teachers used Creative Curriculum, 16.7 used High/Scope, and 40.6 used Other curricula. In the South, 24.5 percent of the teachers used Creative Curriculum, 21.1 used High/Scope, and 54.4 used Other curricula. In the West, 44.5 percent of the teachers used Creative Curriculum, 24.9 used High/Scope, and 30.5 used Other curricula.(back)
Figure 3.4: The Classrooms of Teachers Who Used Creative Curriculum or High/Scope Had Significantly Higher Average ECERS-R Total Scores and ECERS-R Language Scores Than Classrooms of Teachers Who Used Other Curricula
This bar chart shows the average ECERS-R Total Scores and ECERS-R Language scores by curricula used. The x axis of the bar chart depicts the three types of curricula namely Creative Curriculum, High/Scope and Other Curricula. The y axis depicts the Average ECERS-R Total and ECERS-R Language Scores of their classrooms. The classrooms of teachers who used Creative Curriculum had an average ECERS-R Total score of 5.02 and an average ECERS-R Language score of 5.03. The classrooms of teachers who used High/Scope had an average ECERS-R Total score of 5.04 and an average ECERS-R Language score of 5.12. The classrooms of teachers who used Other curricula had an average ECERS-R Total score of 4.55 and an average ECERS-R Language score of 4.58.(back)
Figure 3.5: The Classrooms of Teachers Who Used Creative Curriculum and High/Scope Had Significantly Higher Quality Composite Scores Than Classrooms Using Other Curricula
This bar chart shows the quality composite score by curriculum used. The x axis of the bar chart depicts the three types of curricula namely Creative Curriculum, High/Scope and Other Curricula. The y axis depicts the Average Quality Composite scores of classrooms. The classrooms of teachers who used Creative Curriculum had an Average Quality Composite score of 0.27. The classrooms of teachers who used High/Scope had an Average Composite Factor score of 0.26. The classrooms of teachers who used Other curricula had an Average Composite Factor score of -0.18.(back)
Figure 3.6: Classrooms That Used Creative Curriculum and High/Scope Had Significantly Higher Average ECERS-R Total Scores and ECERS-R Language Scores Than Classrooms That Used All Other Curricula
This bar chart shows the average ECERS-R Total Scores and ECERS-R Language Scores across curriculum type. The x axis of the bar chart depicts the four types of curricula namely Creative Curriculum, High/Scope, Widely Available curricula and All Other Curricula. The y axis depicts the Average ECERS-R Total and ECERS-R Language Scores of classrooms. The classrooms of teachers who used Creative Curriculum had an average ECERS-R Total score of 5.02 and an average ECERS-R Language score of 5.03. The classrooms of teachers who used High/Scope had an average ECERS-R Total score of 5.04 and an average ECERS-R Language score of 5.12. The classrooms of teachers who used Widely Available curricula had an average ECERS-R Total score of 4.82 and an average ECERS-R Language score of 5.15. The classrooms of teachers who used All Other curricula had an average ECERS-R Total score of 4.47 and an average ECERS-R Language score of 4.4.(back)
Figure 3.7: Classrooms That Used Creative Curriculum and High/Scope Had Significantly Higher Average Quality Factor Scores Than Classrooms That Used All Other Curricula
This bar chart shows the Average Quality Factor Scores by curriculum type. The x axis of the bar chart depicts the four types of curricula namely Creative Curriculum, High/Scope, Widely Available curricula and All Other Curricula. The y axis depicts the Average Quality Composite scores of classrooms. The classrooms of teachers who used Creative Curriculum had an Average Quality Composite score of 0.27. The classrooms of teachers who used High/Scope had an Average Quality Composite score of 0.26. The classrooms of teachers who used Widely Available curricula had an Average Quality Composite score of 0.07. The classrooms of teachers who used All Other curricula had an Average Quality Composite score of -0.27.(back)
Figure 4.1: Classroom Quality in Head Start Compared With Other Preschool and Child Care Settings
This line graph illustrates how classroom quality in Head Start compares with other preschool and child care settings found in other research studies (as measured by the ECERS). The x axis consists of 9 studies of the quality of Head Start or other preschool or child care settings, and the y axis is the mean ECERS score, and ranges from 1.0 to 7.0. Across the graph is a dashed line at about 2.8 running parallel to the x axis indicating minimal quality as judged by the ECERS. For each study, the mean ECERS score is provided, as well as the variation in quality found. The first point is Head Start FACES Fall 2000 (258 classrooms), for which the mean ECERS score was 4.8. The variation ran from about 3.5 to about 6.2. The second point is Head Start FACES Fall 1997 (518 classrooms). The mean ECERS score was 4.9, and the variation went from about 4 to just under 6. The sample for the third study cited consists of 39 Head Start classrooms (Layzer et al., 1993). The mean ECERS score was 4.9, and the variation ranged from about 4 to about 5.8. The sample for the fourth point is 246 community- and school-based centers in New Jersey (Resnick et al., 2001). The mean ECERS score was 4.6, and the variation went from slightly under 3 to about 6.4. The fifth point indicates the quality of 38 school-based centers from the Layzer et al. study (1993). The mean ECERS score was 4.5, and the variation ran from about 3.5 to about 5.5. The next sample is 50 non-profit child-care centers from the Cost, Quality and Outcomes Study (1995). The mean ECERS score was 4.5, and the variation ranged from slightly over 3 to just under 6. The sample for the seventh datapoint consists of 42 child-care centers (Layzer et al., 1993). The mean ECERS score was 4.2, and the variation was from about 3 to slightly over 5. The eighth sample is 120 preschool centers (Scarr et al., 1992). The mean ECERS score was 4.0, and the variation ranged from about 2.6 (under the threshold for minimal quality) to about 5.4. The sample for the final datapoint is 227 preschool centers from the National Child Care Staffing Study (Whitebook et al., 1989). The mean ECERS score was 4.0, and the variation ran from about 2.6 (under the threshold for minimal quality) to slightly over 5. The variability was defined as +/- 1.5 standard deviations.(back)
Figure 4.2: Distribution of Classrooms on ECERS and ECERS-R, Fall 1997 and 2000
This figure shows the distribution of classrooms on the ECERS and ECERS-R for the Fall 1997 (n=518) and Fall 2000 (n=258) cohorts of FACES. The x axis consists of the ECERS scale ratings, and ranges from 1 to 7, with the following anchors: 1 is “inadequate,” 3 is “minimal,” 5 is “good,” and 7 is “excellent.” The y axis is the percent of classrooms, and ranges from 0 to 100. No classrooms from either cohort received a score of 1, “inadequate.” No classrooms from the Fall 1997 sample received a score of 2, however 1.9 percent of the Fall 2000 classrooms received a score of 2. The next category is 3, “minimal,” and 0.8 percent of Fall 1997 classrooms and 5.7 percent of Fall 2000 classrooms received this rating. The next ranking is 4, which 26.8 percent of Fall 1997 classrooms and 22.8 percent of Fall 2000 classrooms received. Most classrooms fell into the 5 or “good” range: 53.7 percent of Fall 1997 classrooms and 48.1 percent of Fall 2000 classrooms received this score. The next score is 6, and 18.1 percent of Fall 1997 classrooms and 20.9 percent of Fall 2000 classrooms received this score. Finally, 0.6 percent of Fall 1997 classrooms and 0.7 percent of Fall 2000 classrooms were rated 7, “excellent.” ECERS scale categories were coded as follows: 0-1.49=”1,” 1.5-2.49=”2,” 2.50-3.49=”3,” 3.5-4.49=”4,” 4.5-5.49=”5,” 5.5-6.49=”6,” 6.5-7=”7.”(back)
Figure 4.3: Teacher Beliefs Are Correlated With Classroom Quality, Fall 2000
This bar chart shows the simple correlations (using the Pearson Product-Moment Correlation coefficient) between teacher beliefs and 15 measures of classroom quality. The y axis depicts the 15 classroom quality measures while the x axis indicates the size of the correlations, from –0.10 to +0.35. Significance is shown using asterisks for the three levels: one asterisk indicated significant at .05, two asterisks indicate significance at .01, and three asterisks indicate significance at .001. A negative correlation, which was not statistically significant, was found between teacher beliefs and child-adult ratio (-0.06). Positive correlations were found for the remaining quality indicators, with non-significant correlations between teacher beliefs and Assessment Profile Scheduling (0.05) and between teacher beliefs and Assessment Profile Individualizing Composite (0.09). A positive and significant correlation at the .05 level was found between teacher beliefs and Assessment Profile Learning Environment (0.13). Positive and significant correlations at the .01 level were found between teacher beliefs and Assessment Profile Individualizing Observed (0.15), between teacher beliefs and ECERS-R Personal Care (0.17), and between teacher beliefs and ECERS-R Furnishings (0.18). Positive and significant correlations at the .001 level were found between teacher beliefs and ECERS-R Social (0.24), between teacher beliefs and ECERS-R Motor Skills (0.25), between teacher beliefs and Quality Factor Score (0.27), between teacher beliefs and Arnett Lead Teacher Total (0.28), between teacher beliefs and ECERS-R Creative (0.29), between teacher beliefs and ECERS-R Language (0.29), between teacher beliefs and ECERS-R Program Structure (0.31), and between teacher beliefs and ECERS-R Total Score (0.32).(back)
Figure 4.4: A Model of Center- and Classroom-Level Factors Explaining Quality in Head Start
This chart depicts the factors related to the quality within the classroom that is being tested by the data. The first level is the characteristics of the Head Start program, which is bi-directionally related to the primary curriculum that is being used. The characteristics of the Head Start program are also linked to classroom quality that is defined by the classroom process, quality of language/literacy environment, teacher-child interactions, and classroom structural quality. The program level is also linked to teacher backgrounds at the classroom level. The diagram also depicts the relationship between teacher backgrounds, teacher education and experience, and teacher attitudes and knowledge. Teacher education and experience show both direct and indirect relationships to the quality of the classroom. Teacher background and teacher education and experience are a bi-directional relationship.(back)
Figure 5.1: Analytical Model of Multi-Level Factors Predicting to Classroom Quality and Children’s Achievement and Gains in the Head Start Year
This diagram represents a model of the relationship between the levels within a Head Start program, namely, program level, classroom level, and child level. The model demonstrates the hypothesized relationship between the program level with teacher characteristics and classroom quality at the classroom level. Within the classroom level there is a relationship between teacher characteristics and classroom quality as well as full-day versus part-day classes. The classroom level then influences the child level in the following manner, full-day versus part-day class and classroom quality are directly related to children’s level of achievement and fall-spring gains. Child and family characteristics also directly influence children’s level of achievement and gains.(back)
Figure 5.2: Children in Head Start Programs Using High/Scope Curriculum Show Greater Gains in Letter Recognition Skills
This line graph depicts the mean standard score of the Woodcock Johnson (WJR) Letter-Word Identification Task of children in the Fall of 2000 and the Spring of 2001 as used in three different types of curriculum, High/Scope, Creative Curriculum, and Other. Children in the High/Scope curriculum had a mean standard score on the WJR Letter-Word Identification Task of 92.2 in the Fall of 2000 and 93.8 in the Spring of 2001. Children in the Creative Curriculum had a mean standard score on the WJR Letter-Word Identification Task of 91.9 in the Fall of 2000 and 91.6 in the Spring of 2001. For the Other curriculum group children had a mean standard score on the WJR Letter-Word Identification Task of 91.7 in the Fall of 2000 and 90.7 in the Spring of 2001.(back)
Figure 5.3: Children in Head Start Programs With Higher Teacher Salaries Show Larger Gains in Cooperative Classroom Behavior
This line graph illustrates the fall to spring changes in cooperative classroom behavior for children in classrooms with varying levels of teacher salaries. The x axis shows the time of teacher ratings, and the y axis shows the mean score on the Cooperative Classroom Behavior Rating Scale. The mean score on the Cooperative Classroom Behavior Rating Scale for children with teachers with salaries in the highest quartile had the highest gains, from 14.6 in Fall 2000 to 17.7 in Spring 2001. Children with teachers in the middle quartiles had a mean score of 14.6 in Fall 2000 and 16.3 in Spring 2001. Children with teachers whose salaries were in the lowest quartile had a mean score of 14.8 in Fall 2000, and a score of 16.3 in Spring 2001.(back)
Figure 5.4: Children in Head Start Programs With Higher Teacher Salaries Show Larger Declines in Hyperactive Behavior
This line graph illustrates the fall to spring changes in hyperactive behavior for children in classrooms with varying levels of teacher salaries. The x axis depicts the time of the teacher ratings, and the y axis shows the mean score on the Hyperactive Rating Scale. The mean score on the Hyperactive Rating Scale for children with teachers with salaries in the highest quartile had the largest decline, from 1.4 in Fall 2000 to 1.0 in Spring 2001. Children with teachers in the middle quartiles had a mean score of 1.4 in Fall 2000 and 1.3 in Spring 2001. Children with teachers whose salaries were in the lowest quartile had no change, with a mean score of 1.3 in Fall 2000 and Spring 2001.(back)
Figure 5.5: Children in Head Start Classes Taught by Teachers With Bachelors’ or Associates’ Degrees Show Gains in Early Writing Skills
This line graph depicts the change from Fall to Spring on the Woodcock-Johnson-Revised (WJ-R) Dictation Task for children in Head Start classes taught by teachers with Bachelors’ degrees, Associates’ degrees, and teachers with less than an Associates’ degree. The x axis indicates the time of assessments, and the y axis shows the mean standard score on the WJ-R Dictation task. Children with teachers with a Bachelors’ degree had a mean standard score on the WJ-R Dictation task of 86.3 in Fall 2000 and a mean standard score of 88.8 in Spring 2001. Children with teachers with an Associates’ degree had a mean standard score on the WJ-R Dictation task of 84.5 in Fall 2000 and a mean standard score of 87.1 in Spring 2001. Children with teachers with less than an Associates’ degree had a mean standard score on the WJ-R Dictation task of 83.9 in Fall 2000 and a mean standard score of 85.5 in Spring 2001.(back)
Figure 5.6: Head Start Children in Full-Day Classes Show Larger Gains in Letter Recognition Skills Than Those in Part-Day Classes
This line graph depicts the mean standard score of the Woodcock Johnson (WJR) Letter-Word Identification Task of children in the Fall of 2000 and the Spring of 2001 for children participating in full-day and part-day Head Start classes. The x axis indicates the time of assessments, and the y axis shows the mean standard score on the WJ-R Letter-Word Identification task. Children in full-day classes had a gain in their mean standard scores, with a mean standard score on the WJR Letter-Word Identification Task of 91.7 in the Fall of 2000 and 92.8 in the Spring of 2001. However, children in part-day classes scores’ declined, with a mean standard score on the WJR Letter-Word Identification Task of 92.0 in the Fall of 2000 and 91.1 in the Spring of 2001.(back)
Figure 5.7: Children in Full-Day Head Start Classes Show Greater Gains in Early Writing Skills
This line graph depicts the mean standard scores on the Woodcock Johnson (WJR) Dictation Test for children in either full-day or part-day classes in the Fall of 2000 and the Spring of 2001. The x axis indicates the time of assessments, and the y axis shows the mean standard score on the WJ-R Dictation task. Children in full-day classes had a mean standard score on the WJR Dictation Test of 84.9 in the Fall of 2000 and 88.3 in the Spring of 2001. Children in part-day classes had a mean standard score on the WJR Dictation Test of 85.0 in the Fall of 2000 and 86.1 in the Spring of 2001.(back)
Figure 5.8: Head Start Children Whose Parents Read to Them More Often Show Larger Vocabulary Gains
This line graph depicts the change from Fall to Spring on the mean standard score of the Peabody Picture Vocabulary Test-III (PPVT-III) for children whose parents read to them every day, three to six times per week, once or twice per week, or not at all. The x axis indicates the time of assessments, and the y axis shows the mean standard score on the PPVT-III. Children whose parents read to them daily had a mean standard score on the PPVT-III of 87.9 in Fall 2000 that increased to 92.5 in Spring 2001. Children whose parents read to them three to six times per week had a mean standard score on the PPVT-III of 86.1 in Fall 2000 that increased to 89.4 in Spring 2001. Children whose parents read to them once or twice per week had a mean standard score on the PPVT-III of 80.8 in Fall 2000 that increased to 84.5 in Spring 2001. Finally, children whose parents did not read to them at all had a mean standard score on the PPVT-III of 82.6 in Fall 2000 that increased slightly to 84.7 in Spring 2001.(back)
Figure 5.9: Children in Head Start Classes With Higher, Moderate, and Lower Quality Language Activities Show Parallel Gains in Vocabulary Knowledge
This line graph depicts the mean standard score on the Peabody Picture Vocabulary Test (PPVT-III) for children in the highest quartile on ECERS Language Scale, children in the middle quartile on the ECERS Language Scale and children in the lowest quartile on the ECERS Language Scale during the Fall 2000 and Spring 2001. The x axis indicates the time of assessments, and the y axis shows the mean standard score on the PPVT-III. Children in the highest quartile on ECERS Language Scale had a mean standard score on the PPVT-III of 84.4 in the Fall of 2000 and 88.6 in the Spring of 2001. Children in the middle quartile of the ECERS Language Scale had a mean standard score on the PPVT-III of 80.5 in the Fall of 2000 and 84.8 in the Spring of 2001. Children in the lowest quartile of the ECERS Language Scale had a mean standard score on the PPVT-III of 79.5 in the Fall of 2000 and 84.1 in the Spring of 2001.(back)
Figure 5.10: Children in Head Start Classes With Higher Child-Adult Ratios Show Larger Gains in Letter Recognition
This line graph depicts the mean standard score on the Woodcock Johnson (WJ-R) Word Identification Task for children in the highest quartile of Child-Adult Ratios, children in the middle quartile of Child-Adult Ratios, and children in the lowest quartile of Child-Adult Ratios during the Fall of 2000 and Spring of 2001. The x axis indicates the time of assessments, and the y axis shows the mean standard score on the WJ-R Letter-Word Identification task. Children in the highest quartile of Child-Adult Ratios had a mean standard score on the WJ-R Letter-Word Identification Task of 92.1 in the Fall of 2000 and 93.3 in the Sspring of 2001. Children in the middle quartile of Child-Adult Ratios had a mean standard score on the WJ-R Letter-Word Identification Task of 91.7 in the Fall of 2000 and 91.1 in the Spring of 2001. Children in the lowest quartile of Child-Adult Ratios had a mean standardized score on the WJ-R Letter-Word Identification Task of 92.0 in the Fall of 2000 and 91.1 in the Spring of 2001.(back)
Figure 7.1: Correlations and Standardized Multiple Regression Coefficients Between Reading Scale Scores at End of Kindergarten Year and FACES Assessment Scale Scores at End of Head Start Year
This flow diagram shows the correlations and standardized multiple regression coefficients between Reading scale scores at the end of the Kindergarten year and FACES assessment scale scores at the end of the Head Start year. The bivariate correlation coefficient for WJR Letter-Word ID at the end of Head Start and Reading Scale Score at the end of Kindergarten is .55, for WJR Applied Problems at the end of Head Start and Reading Scale Score at the end of Kindergarten is .52, for WJR Dictation at the end of Head Start and Reading Scale Score at the end of kindergarten is .48, for One-to-One Counting at the end of Head Start and Reading Scale Score at the end of kindergarten is .40, and Draw-A-Design at the end of Head Start and Reading Scale Score at the end of Kindergarten is .33. When the assessment scores are combined in a multiple linear regression predicting Reading scale scores at the end of Kindergarten, the multiple regression coefficient is .68. The beta values in the multiple regression are 0.32 for WJR Letter-Word ID, 0.16 for WJR Applied Problems, 0.14 for WJR Dictation, 0.12 for One-to-One Counting, and 0.10 for Draw-A-Design. The Head Start data are from Spring 1998 while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 718 to 786 for the bivariate correlations and 623 for the multiple regression coefficients.(back)
Figure 7.2: Correlations and Standardized Multiple Regression Coefficients Between General Knowledge Scale Scores at End of Kindergarten Year and FACES Assessment Scale Scores at End of Head Start Year
This flow diagram shows the correlations and standardized multiple regression coefficients between General Knowledge scale scores at the end of the Kindergarten year and FACES assessment scale scores at the end of the Head Start year. The bivariate correlation coefficient for PPVT-III Vocabulary at the end of Head Start and General Knowledge Scale at the end of Kindergarten is .77, for WJR Applied Problems at the end of Head Start and General Knowledge Scale at the end of Kindergarten is .62, for Book Knowledge at the end of Head Start and General Knowledge Scale at the end of Kindergarten is .52, for WJR Dictation at the end of Head Start and General Knowledge Scale at the end of Kindergarten is .46, and for Draw-A-Design at the end of Head Start and General Knowledge Scale at the end of Kindergarten is .30. When the assessment scores are combined in a multiple linear regression predicting General Knowledge scale scores at the end of kindergarten, the multiple regression coefficient is .81. The beta values in the multiple regression are .62 for the PPVT-III, .09 for WJR Applied Problems, .06 for Book Knowledge, .11 for WJR Dictation, and .06 for Draw-A-Design. The Head Start data are from Spring 1998 and the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 720 to 787 for bivariate correlations and 624 for multiple regression coefficients.(back)
Figure 7.3: Correlations and Standardized Multiple Regression Coefficients Between Reading Scale Scores at End of Kindergarten Year and FACES Assessment Scale Scores at End of Head Start Year—Norm-Referenced Tests Only
This flow diagram shows the correlations and standardized multiple regression coefficients between reading scale scores at the end of the kindergarten year and FACES assessment scale scores at the end of the Head Start year for the norm-referenced tests. The bivariate correlation coefficient for WJR Letter-Word ID at the end of Head Start and Reading Scale Score at the end of kindergarten is .55, for WJR Applied Problems at the end of Head Start and Reading Scale Score at the end of kindergarten is .52, for WJR Dictation at the end of Head Start and Reading Scale Score at the end of kindergarten is .48, for PPVT-III vocabulary at the end of Head Start and the Reading Scale Score at the end of kindergarten is .42. When the assessment scores are combined in a multiple linear regression predicting Reading scale scores at the end of Kindergarten, the multiple regression coefficient is .66. The beta values in the multiple regression are 0.33 for WJR Letter-Word ID, 0.24 for WJR Applied Problems, 0.19 for WJR Dictation, and .05 for the PPVT-III. The Head Start data are from Spring 1998 while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 718 to 737 for the bivariate correlations and 640 for the multiple regression coefficients.(back)
Figure 7.4: Correlations and Standardized Multiple Regression Coefficients Between General Knowledge Scale Scores at End of Kindergarten Year and FACES Assessments Scale Scores at End of Head Start Year—Norm-Referenced Tests Only
This flow diagram shows the correlations and standardized multiple regression coefficients between general knowledge scale scores at the end of the kindergarten year and FACES assessment scale scores at the end of the Head Start year for the norm-referenced tests. The bivariate correlation coefficient for the PPVT-III Vocabulary at the end of Head Start and the General Knowledge Scale at the end of kindergarten is .77, for the WJR Applied Problems at the end of Head Start and the General Knowledge Scale at the end of kindergarten is .62, for the WJR Dictation at the end of Head Start and General Knowledge Scale at the end of kindergarten is .46, and for the WJR Letter-Word Identification at the end of Head Start and the General Knowledge Scale at the end of kindergarten is .40. When the assessment scores are combined in a multiple linear regression predicting General Knowledge scale scores at the end of kindergarten, the multiple regression coefficient is .81. The beta values in the multiple regression are .63 for the PPVT-III, .13 for WJR Applied Problems, .15 for WJR Dictation, and .02 for the WJR Letter-Word Identification. The Head Start data are from Spring 1998 and the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 720 to 748 for the bivariate correlations and 642 for the multiple regression coefficients.(back)
Figure 7.5: Correlations and Standardized Multiple Regression Coefficients Between Reading Scale Scores at End of Kindergarten Year and FACES Assessment Scale Scores at End of Head Start Year—Criterion-Referenced Tests Only
This flow diagram shows the correlations and standardized multiple regression coefficients between reading scale scores at the end of the kindergarten year and FACES assessment scale scores at the end of the Head Start year for the criterion-referenced tests. The bivariate correlation coefficient for One-to-One Counting at the end of Head Start and Reading Scale Score at the end of kindergarten is .40, for Book Knowledge at the end of Head Start and Reading Scale Score at the end of kindergarten is .39, for Color Naming at the end of Head Start and Reading Scale Score at the end of kindergarten is .36, for Draw-A-Design at the end of Head Start and Reading Scale Score at the end of kindergarten is .33, and Social Awareness at the end of Head Start and Reading Scale Score at the end of kindergarten is .23. When the assessment scores are combined in a multiple linear regression predicting Reading scale scores at the end of Kindergarten, the multiple regression coefficient is .55. The beta values in the multiple regression are 0.23 for One-to-One Counting, 0.19 for Book Knowledge, 0.16 for Color Naming, 0.20 for Draw-A-Design, and 0.08 for Social Awareness. The Head Start data are from Spring 1998 while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 772 to 786 for the bivariate correlations and 761 for the multiple regression coefficients.(back)
Figure 7.6: Correlations and Standardized Multiple Regression Coefficients Between General Knowledge Scale Scores at End of Kindergarten Year and FACES Assessment Scale Scores at End of Head Start Year—Criterion-Referenced Tests Only
This flow diagram shows the correlations and standardized multiple regression coefficients between the general knowledge scale scores at the end of the kindergarten year and FACES assessment scale scores at the end of the Head Start year for the criterion-referenced tests. The bivariate correlation coefficient for Book Knowledge at the end of Head Start and the General Knowledge Scale at the end of kindergarten is .52, for Color Naming at the end of Head Start and the General Knowledge Scale at the end of kindergarten is .38, for One-to-One Counting at the end of Head Start and the General Knowledge Scale at the end of kindergarten is .36, for Draw-A-Design at the end of Head Start and the General Knowledge Scale at the end of kindergarten is .30, and for Social Awareness at the end of Head Start and the General Knowledge Scale at the end of kindergarten is .30. When the assessment scores are combined in a multiple linear regression predicting General Knowledge scale scores at the end of kindergarten, the multiple regression coefficient is .60. The beta values in the multiple regression are .35 for Book Knowledge, .14 for Color Naming, .13 for One-to-One Counting, .15 for Draw-A-Design, and .13 for Social Awareness. The Head Start data are from Spring 1998 and the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 773 to 787 for the bivariate correlations and 762 for the multiple regression coefficients.(back)
Figure 7.7: Correlation Coefficients Between Parent and Teacher Ratings of Children’s Social Competence at the End of Head Start and Teacher Reported Ratings of Cooperative Classroom Behavior at the End of Kindergarten
This bar chart shows the correlation coefficients for 10 parent- and teacher-reported behavior ratings of children’s social competence at the end of Head Start and teacher reported ratings of cooperative classroom behaviors at the end of kindergarten. The x axis consists of the behavior ratings, and the y axis is the magnitude of the correlation coefficient, and ranges from –0.4 to 0.4. Parent reported aggression at the end of Head Start had a –0.23 correlation coefficient with cooperative classroom behavior at the end of kindergarten. A correlation coefficient of –0.14 was found for both parent reported hyperactivity and parent reported withdrawal at the end of Head Start and cooperative classroom behavior at the end of kindergarten. Parent reported total behavior problems at the end of Head Start had a –0.23 correlation coefficient with cooperative classroom behavior at the end of kindergarten. A 0.08 correlation coefficient was found between parent reported positive approach to learning at the end of Head Start and cooperative classroom behavior at the end of kindergarten. A similar but stronger pattern was found for teacher reported behaviors. Teacher reported aggression at the end of Head Start had a –0.36 correlation coefficient with cooperative classroom behavior at the end of kindergarten. A correlation coefficient of –0.25 was found for teacher reported hyperactivity, and a correlation coefficient of –0.24 was found for teacher reported withdrawal at the end of Head Start and cooperative classroom behavior at the end of kindergarten. Teacher reported total behavior problems at the end of Head Start had a –0.36 correlation coefficient with cooperative classroom behavior at the end of kindergarten. Finally, a 0.33 correlation coefficient was found between teacher reported cooperative classroom behavior at the end of Head Start and cooperative classroom behavior at the end of kindergarten.(back)
Figure 7.8: Correlations and Significant Standardized Multiple Regression Coefficients Between Teacher Ratings of Cooperative Classroom Behavior at End of Kindergarten Year and Parent and Teacher Behavior Ratings at End of Head Start Year
This flow diagram shows the relationship between cooperative classroom behavior at the end of the Kindergarten year and parent and teacher behavior ratings at the end of the Head Start year. The bivariate correlation coefficient for Teacher ratings of Aggression at the end of Head Start and Teacher ratings of Cooperative Classroom Behavior at the end of Kindergarten is -.36, for Parent ratings of Aggression at the end of Head Start and Teacher ratings of Cooperative Classroom Behavior at the end of Kindergarten is -.23, and for Teacher ratings of Withdrawal at the end of Head Start and Teacher ratings of Cooperative Classroom Behavior at the end of Kindergarten is -.24. When the behavior ratings are combined in a multiple linear regression predicting Teacher ratings of Cooperative Classroom Behavior at the end of kindergarten, the multiple regression coefficient is .42. The beta values in the multiple regression are -.25 for Teacher ratings of Aggression, -.13 for Parent ratings of Aggression, and -.11 for Teacher ratings of Withdrawal. The Head Start data are from Spring 1998 and the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 556 to 575 for bivariate correlations and 542 for multiple regression coefficients.(back)
Figure 7.9: Correlations and Significant Standardized Multiple Regression Coefficients Between Teacher Ratings of Total Problem Behavior at End of Kindergarten Year and Parent and Teacher Behavior Ratings at End of Head Start Year
This flow diagram shows the relationship between teacher rating of total problem behavior at the end of kindergarten and parent and teacher ratings at the end of the Head Start year. The bivariate correlation coefficient for Teacher ratings of Aggression at the end of Head Start and Teacher ratings of Total Problem Behavior at the end of Kindergarten is .37, for parent ratings of Withdrawal at the end of Head Start and Teacher ratings of Total Problem Behavior at the end of Kindergarten is .23, and for Parent ratings of Hyperactivity at the end of Head Start and Teacher ratings of Total Problem Behavior at the end of Kindergarten is .28. When the behavior ratings are combined in a multiple linear regression predicting Teacher ratings of Total Problem Behavior at the end of kindergarten, the multiple regression coefficient is .49. The beta vales in the multiple regression are .23 for Teacher ratings of Aggression, .12 for Parent ratings of Withdrawal, and .09 for Parent ratings of Hyperactivity. The Head Start data are from Spring 1998 and the Kindergarten Follow-Up data are from Spring 1999. The total sample ranges from 556 to 575 for bivariate correlations and 542 for multiple regression coefficients.(back)
Figure 7.10: Correlation Coefficients Between Parent and Teacher Ratings of Children’s Social Competence at the End of Head Start and Reading Scale Scores at End of Kindergarten
This bar chart shows the correlation coefficients for 10 parent- and teacher-reported behavior ratings of children’s social competence at the end of Head Start and Reading Scale scores at the end of Kindergarten. The x axis consists of the behavior ratings, and the y axis is the magnitude of the correlation coefficient, and ranges from –0.25 to 0.3. Parent reported aggression at the end of Head Start had a –0.13 correlation coefficient with Reading Scale scores at the end of Kindergarten. A correlation coefficient of –0.18 was found for parent reported hyperactivity at the end of Head Start and Reading Scale scores at the end of Kindergarten. Parent reported withdrawal at the end of Head Start had a -.011 correlation coefficient with Reading Scale scores at the end of Kindergarten. Parent reported total behavior problems at the end of Head Start had a –0.19 correlation coefficient with Reading Scale scores at the end of Kindergarten. A 0.05 correlation coefficient was found between parent reported positive approach to learning at the end of Head Start and Reading Scale scores at the end of Kindergarten. A similar pattern was found for teacher reported behaviors. Teacher reported aggression at the end of Head Start had a –0.1 correlation coefficient with Reading Scale scores at the end of Kindergarten. A correlation coefficient of –0.16 was found for teacher reported hyperactivity at the end of Head Start and Reading Scale scores at the end of Kindergarten. A correlation coefficient of –0.19 was found for both teacher reported withdrawal and teacher reported total behavior problems at the end of Head Start and Reading Scale scores at the end of Kindergarten. Finally, a 0.24 correlation coefficient was found between teacher reported cooperative classroom behavior at the end of Head Start and Reading Scale scores at the end of Kindergarten.(back)
Figure 7.11: Correlations and Significant Standardized Multiple Regression Coefficients Between Reading Scale Scores at End of Kindergarten Year and Parent and Teacher Behavior Ratings at End of Head Start Year
This flow diagram shows the relationship between Reading Scale scores at the end of the Kindergarten year and parent and teacher behavior ratings at the end of the Head Start year. The bivariate correlation coefficient for Teacher ratings of Cooperative Classroom Behavior at the end of Head Start and Reading Scale score at the end of Kindergarten is .24, for Parent ratings of Hyperactivity at the end of Head Start and Reading Scale score at the end of Kindergarten is -.18, and for Teacher ratings of Withdrawal at the end of Head Start and Reading Scale score at the end of Kindergarten is -.19. When the behavior ratings are combined in a multiple linear regression predicting Reading scale scores at the end of kindergarten, the multiple regression coefficient is .28. The beta values in the multiple regression are .19 for Teacher ratings of Cooperative Classroom Behavior, -.10 for Parent ratings of Hyperactivity, and -.10 for Teacher ratings of Withdrawal. The Head Start data are from Spring 1998 while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 741 to 771 for bivariate correlations and 719 for multiple regression coefficients.(back)
Figure 7.12: Correlation Coefficients Between Parent and Teacher Ratings of Children’s Social Competencies at the End of Head Start and General Knowledge Scale Scores at End of Kindergarten
This bar chart shows the correlation coefficients for 10 parent- and teacher-reported behavior ratings of children’s social competencies at the end of Head Start and General Knowledge scale scores at the end of Kindergarten. The x axis consists of the behavior ratings, and the y axis is the magnitude of the correlation coefficient, and ranges from –0.25 to 0.3. Parent reported aggression at the end of Head Start had a –0.09 correlation coefficient with General Knowledge scale scores at the end of Kindergarten. A correlation coefficient of –0.18 was found for parent reported hyperactivity at the end of Head Start and General Knowledge scale scores at the end of Kindergarten. Parent reported withdrawal at the end of Head Start had a -.012 correlation coefficient with General Knowledge scale scores at the end of Kindergarten. Parent reported total behavior problems at the end of Head Start had a –0.17 correlation coefficient with General Knowledge scale scores at the end of Kindergarten. A 0.03 correlation coefficient was found between parent reported positive approach to learning at the end of Head Start and General Knowledge scale scores at the end of Kindergarten. A similar pattern was found for teacher reported behaviors. Teacher reported aggression at the end of Head Start had a –0.06 correlation coefficient with General Knowledge scale scores at the end of Kindergarten. A correlation coefficient of –0.11 was found for teacher reported hyperactivity at the end of Head Start and General Knowledge scale scores at the end of Kindergarten. A correlation coefficient of –0.13 was found for both teacher reported withdrawal and teacher reported total behavior problems at the end of Head Start and General Knowledge scale scores at the end of Kindergarten. Finally, a 0.23 correlation coefficient was found between teacher reported cooperative classroom behavior at the end of Head Start and General Knowledge scale scores at the end of Kindergarten.(back)
Figure 7.13: Correlations and Significant Standardized Multiple Regression Coefficients Between General Knowledge Scale Scores at End of Kindergarten and Parent and Teacher Behavior Ratings at End of Head Start Year
This flow diagram shows the relationship between General Knowledge scale scores at the end of the Kindergarten year and parent and teacher behavior ratings at the end of the Head Start year. The bivariate correlation coefficient for Teacher ratings of Cooperative Classroom Behavior at the end of Head Start and General Knowledge scale score at the end of Kindergarten is .23, for Parent ratings of Hyperactivity at the end of Head Start and General Knowledge scale score at the end of Kindergarten is -.18, and for Teacher ratings of Aggression at the end of Head Start and General Knowledge scale score at the end of Kindergarten is -.06, which is not statistically significant. When the behavior ratings are combined in a multiple linear regression predicting General Knowledge scale scores at the end of kindergarten, the multiple regression coefficient is .27. The beta values in the multiple regression are .26 for Teacher ratings of Cooperative Classroom Behavior, -.12 for Parent ratings of Hyperactivity, and .11 for Teacher ratings of Aggression. The Head Start data are from Spring 1998 while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 742 to 772 for bivariate correlations and 719 for multiple regression coefficients.(back)
Figure 7.14: Correlations and Standardized Multiple Regression Coefficients Between Reading Scale Scores at End of Kindergarten Year and FACES Assessment Scale Gain Scores During Head Start Year
This flow diagram shows the relationship between Reading scale scores at the end of the Kindergarten year and FACES assessment scale gain scores during the Head Start year. The bivariate correlation coefficient for WJR Letter-Word ID gain score during Head Start and Reading scale score at the end of Kindergarten is .39, for One-to-One Counting gain score during Head Start and Reading scale score at the end of Kindergarten is .31, for WJR Dictation gain score during Head Start and Reading scale score at the end of Kindergarten is .31, for Draw-A-Design gain score during Head Start and Reading scale score at the end of Kindergarten is .24, for WJR Applied Problems gain score during Head Start and Reading scale score at the end of Kindergarten is .29 and PPVT-III Vocabulary gain score during Head Start and Reading scale score at the end of Kindergarten is .19. When the assessment gain scores are combined in a multiple linear regression predicting Reading scale scores at the end of kindergarten, the multiple regression coefficient is .56. The beta values in the multiple regression are .27 for WJR Letter-Word ID gain score, .18 for One-to-One Counting gain score, .15 for WJR Dictation gain score, .14 for Draw-A-Design gain score, .10 for WJR Applied Problems gain score, and .09 for PPVT-III Vocabulary gain score. The Head Start gain score data are from Fall 1997 to Spring 1998, while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 580 to 697 for bivariate correlations, and 441 for multiple regression coefficients.(back)
Figure 7.15: Correlations and Standardized Multiple Regression Coefficients Between General Knowledge Scale Scores at End of Kindergarten Year and FACES Assessment Scale Gain Scores During Head Start Year
This flow diagram shows the relationship between General Knowledge scale scores at the end of the Kindergarten year and FACES assessment scale gain scores during the Head Start year. The bivariate correlation coefficient for PPVT-III Vocabulary gain score during Head Start and General Knowledge scale score at the end of Kindergarten is .31, for Book Knowledge gain score during Head Start and General Knowledge scale score at the end of Kindergarten is .33, for Draw-A-Design gain score during Head Start and General Knowledge scale score at the end of Kindergarten is .27, for WJR Dictation gain score during Head Start and General Knowledge scale score at the end of Kindergarten is .26, for One-to-One Counting gain score during Head Start and General Knowledge scale score at the end of Kindergarten is .23, and for WJR Applied Problems gain score during Head Start and General Knowledge scale score at the end of Kindergarten is .30. When the assessment gain scores are combined in a multiple linear regression predicting General Knowledge scale scores at the end of kindergarten, the multiple regression coefficient is .53. The beta values in the multiple regression are .24 for PPVT-III Vocabulary gain score, .19 for Book Knowledge gain score, .18 for Draw-A-Design gain score, .13 for WJR Dictation gain score, .12 for One-to-One Counting gain score, and .09 for WJR Applied Problems gain score. The Head Start gain score data are from Fall 1997 to Spring 1998, while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 580 to 697 for bivariate correlations and 441 for multiple regression coefficients.(back)
Figure 7.16: Correlations and Standardized Multiple Regression Coefficients Between Teacher Ratings of Cooperative Classroom Behavior at End of Kindergarten Year and Parent and Teacher Behavior Ratings Gain Scores During Head Start Year
This flow diagram shows the relationship between teacher ratings of cooperative classroom behavior at the end of the Kindergarten year and parent and teacher behavior ratings change scores during the Head Start year. The bivariate correlation coefficient for the change score for parent ratings of Aggression during Head Start and Teacher ratings of Cooperative Classroom Behavior at the end of Kindergarten is -.19, for the change score for Teacher ratings of Withdrawal during Head Start and Teacher ratings of Cooperative Classroom Behavior at the end of Kindergarten is -.16, for the change score for Parent ratings of Withdrawal during Head Start and Teacher ratings of Cooperative Classroom Behavior at the end of Kindergarten is -.20, and the change score for Parent ratings of Hyperactivity during Head Start and Teacher ratings of Cooperative Classroom Behavior at the end of Kindergarten is -.14. When the behavior ratings change scores are combined in a multiple linear regression predicting Teacher ratings of Cooperative Classroom Behavior at the end of kindergarten, the multiple regression coefficient is .33. The beta values in the multiple regression are -.13 for the change score for Parent ratings of Aggression, -.13 for the change score for Teacher ratings of Withdrawal, -.13 for the change score for Parent ratings of Withdrawal, and -.10 for the change score for Parent ratings of Hyperactivity. The Head Start change score data are from Fall 1997 and Spring 1998, while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 474 to 510 for bivariate correlations, and 432 for multiple regression.(back)
Figure 7.17: Correlations and Standardized Multiple Regression Coefficients Between Teacher Ratings of Total Problem Behavior at End of Kindergarten Year and Parent and Teacher Behavior Ratings Gain Scores During Head Start Year
This flow diagram shows the relationship between teacher ratings of total problem behavior at the end of the Kindergarten year and parent teacher behavior ratings change scores during the Head Start year. The bivariate correlation coefficient for the change score for Parent ratings of Withdrawal during Head Start and Teacher ratings of Total Problem Behavior at the end of Kindergarten is .27, for the change score for Parent ratings of Hyperactivity during Head Start and Teacher ratings of Total Problem Behavior at the end of Kindergarten is .25, and for the change score for Teacher ratings of Hyperactivity during Head Start and Teacher rating of Total Problem Behavior at the end of Kindergarten is .16. When the behavior ratings change scores are combined in a multiple linear regression predicting Teacher ratings of Total Problem Behavior at the end of Kindergarten, the multiple regression coefficient is .39. The beta values in the multiple regression are .21 for the change score for Parent ratings of Withdrawal, .18 for the change score for Parent ratings of Hyperactivity, and .10 for the change score for Teacher ratings of Hyperactivity. The Head Start change score data are from Fall 1997 and Spring 1998, while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 474 to 510 for bivariate correlations and 432 for multiple regression.(back)
Figure 7.18: Correlations and Odds-Ratio Estimates Between Teachers’ Decisions at End of Kindergarten Year to Assign Child to Repeat Kindergarten and FACES Assessment Scale Scores at End of Head Start Year
This flow diagram shows the relationship between teachers’ decisions at the end of the Kindergarten year to assign a child to repeat Kindergarten and FACES Assessment Scale Scores at the end of the Head Start year. The bivariate correlation coefficient for Book Knowledge at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of Kindergarten is -.31, for One-to-One Counting at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of Kindergarten is -.23, for WJR Letter-Word ID at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten is -.23, for Color Naming at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of Kindergarten is -.21, for Social Awareness at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of Kindergarten is -.14, and for Draw-A-Design at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of Kindergarten is -.12. When the assessment scores are combined in a multiple logistic regression predicting Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of kindergarten, the R-Squared value for the multiple logistic regression is .24 and the concordance rate is 82 percent. The multiple logistic regression odds ratio estimates are .50 for Book Knowledge, .77 for One-to-One Counting, .96 for WJR Letter-Word ID, .97 for Color Naming, .98 for Social Awareness, and .98 for Draw-A-Design. The Head Start data are from Spring 1998 while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 551 to 583 for bivariate correlations, and 474 for multiple logistic regression odds ratio.(back)
Figure 7.19: Correlations and Odds-Ratio Estimates Between Teachers’ Decisions at End of Kindergarten Year to Assign Child to Repeat Kindergarten and Parent and Teacher Behavior Ratings at End of Head Start Year
This flow diagram shows the relationship between teachers’ decisions at the end of the Kindergarten year to assign a child to repeat Kindergarten and the parent and teacher behavior ratings at the end of the Head Start year. The bivariate correlation coefficient for Teacher ratings of Withdrawal at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of Kindergarten is .26, for Parent ratings of Withdrawal and Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of Kindergarten is .14, for Teacher ratings of Hyperactivity at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten is .19, and for Parent ratings of Hyperactivity at the end of Head Start and Teacher’s Decision to Assign Child to Repeat Kindergarten is .11. When the behavior ratings are combined in a multiple logistic regression predicting Teacher’s Decision to Assign Child to Repeat Kindergarten at the end of kindergarten, the R-Squared value for the multiple logistic regression is .11 and the concordance rate is 72 percent. The multiple logistic regression odds ratio estimates are 1.31 for Teacher ratings of Withdrawal, 1.25 for Parent ratings of Withdrawal, 1.23 for Teacher ratings of Hyperactivity, and 1.11 for Parents ratings of Hyperactivity. The Head Start data are from Spring 1998, while the Kindergarten Follow-Up are from Spring 1999. The total sample ranges from 551 to 583 for bivariate correlations, and 474 for multiple logistic regression odds ratio.(back)
| Table of Contents |