Why Do Some Texas Children in Poverty Succeed in School and What Can We Do to Foster More Success?

Transcription

1 Why Do Some Texas Children in Poverty Succeed in School and What Can We Do to Foster More Success? Jennifer L. Matjasko, Ph.D. The University of Texas at Austin This project was supported with a grant from the University of Kentucky Center for Poverty Research through the U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, grant number 5 ASPE The opinions and conclusions expressed herein are solely those of the author and should not be construed as representing the opinions or policy of the UKCPR or any agency of the Federal government. 1

2 The Project Children from low-income families start school with fewer academic competencies than their more affluent counterparts, and many of the disparities widen during elementary school (Alexander & Entwisle, 1996; Lee & Burkam, 2002; Stipek & Ryan, 1997). Moreover, the risks persist into adolescence and early adulthood (Clements, Reynolds, & Hickey, 2004; Duncan, Yeung, Brooks-Gunn, & Smith, 1998). Youth in poverty are at risk for low marks, grade retention, and drop-out (Wirt et al., 2004). In San Antonio, the annual recent proportion of high school students who drop out is 10.8% (Texas Education Agency, 2002). Yet, a considerable group of students in disadvantaged urban schools perform well, succeeding academically and socially in school and preparing themselves for adaptive trajectories in adulthood. To date, most studies of school success have compared children and youth in low-income families with those in more affluent families (but see Ackerman, Brown, & Izard, 2004; see also Clements et al., 2004). Instead of a cross-group approach, we used a multi-disciplinary withingroup approach. In this report, I will describe our progress on collecting San Antonio teacher data and provide some descriptive statistics on the academic progress of the low-income children and youth in Texas. Once data collection is complete, we will also estimate a series of multivariate models that are designed to provide some insight into why some Texas children and youth who grow up in poverty are academically successful. The findings will not only spotlight the fact that many at-risk Texas children do indeed succeed in school, but they will also inform Texas policymakers, school leaders, parents, and teachers about what they can do to improve children s opportunities for academic success. 2

3 Description of the Larger Study and the Three-City Teacher Survey I am currently a collaborator on a longitudinal study of 2,447 children, primarily African- American and Latino, from low-income neighborhoods in Chicago, Boston, and San Antonio, called Welfare, Children, and Families: A Three-City Study ( The Three- City Study was originally designed to examine the impact of welfare reform on preschoolers and adolescents starting in 1999 (about 400 of them are from Texas), and then reinterviewed them in As part of the third wave of data collection, we are getting teacher reports on child achievement and behavior as part of the Three-City Teacher Survey (TCTS). The TCTS includes two central components to create a comprehensive, multi-method view of school success: 1) interviews with the teachers of the students in San Antonio; and 2) acquisition of students school records and test scores. Both efforts are currently ongoing. We have obtained approval to collect teacher data and school records from 40 of the 46 schools and districts in Texas. We are continuing our efforts to obtain approval from the remaining six schools and districts in the San Antonio area. Half of the San Antonio children in our sample ranged in age from infancy to 4 years at wave 1 in 1999, and they have transitioned into elementary school (ages 6-10) at the time of the TCTS data collection. The other half consisted of young adolescents, 10 to 14 years of age in They have made the transition to high school or they have graduated or dropped out (ages 16-20) by the time of the TCTS. The San Antonio sample provides a unique opportunity to examine the specific risk and protective factors associated with school success within one of the largest states in the country. The sample also contains a relatively high proportion of Hispanic children and adolescents, many of whom face both poverty and language barriers to educational success. 3

4 Measures For the purposes of this report, I utilized a subset of the teacher data from the larger Three-City Teacher Survey. The preliminary data contained in this report are derived from Texas teachers responding to the survey as of May Below are the measures that were used in the descriptive analyses. Child Characteristics and Achievement Child grade level. Teachers were asked to indicate in what grade the focal child was in during the current academic year. Responses ranged from 2 nd to 12 th grade. Child grades. Teachers were asked, In general, what grades does the child receive in school? Responses ranged from mostly A s to mostly failing. Child Problem Behaviors and Disabilities Disciplinary Actions. Teachers were asked if they have taken any disciplinary action regarding the focal child during the prior two months. Responses ranged from never to several times a week. Disabilities. To assess child disabilities, teachers were asked the following set of questions: because of an ongoing physical or mental health problem or disability, is the child limited in any way in school? and If so, has child been diagnosed with any of the following disorders: learning-disabled, behavior-disordered, mentally handicapped, physically handicapped, or other. About 10% of the teachers responded that the focal child had a disability. 4

5 Teacher, Classroom, and School Characteristics Teacher Education Level. Teachers were asked to indicate the highest degree or level of education that they completed. Responses ranged from some college to doctoral degree or equivalent. Years of Experience. Teachers indicated the number of years that they taught in their current district. Responses ranged from 1-6 years. Teacher Ethnicity. Teachers were asked to identify one of the following racial groups that best described them: Asian or Pacific Islander, Black or African-American, White, American Indian or Alaskan Native, Hispanic, or Other. Classroom Size. Teachers indicated the number of students that were in the child s class. Responses ranged from 1 to 50 students. School Size. Teachers also provided the total number of students within the child s school. Responses ranged from 50 to 6,000 students. Analyses For the following descriptive analyses, both oneway analysis of variance and chi-square tests were performed in order to ascertain whether there were significant between-group differences in child grades based on child, teacher, classroom, and school characteristics. Given that we have obtained data on only one-half of the Texas teachers, all of the following results are preliminary and are reported for descriptive purposes only. TCTS: San Antonio Children and Youth Child Characteristics and Achievement To date, we have preliminary teacher data on just over one-half of our San Antonio children and youth (n=210). Approximately three-quarters of them are in elementary school 5

6 (grades 2-6) while the rest are in high school or a combined grade classroom. Of the teachers who have completed the survey, 43% of them reported that the children are making mostly A s or B s; 31% reported that the children are making mostly C s; and, 19% reported that the children are making mostly D s or are failing. Of those who are reporting school failure, a majority of the children (77%) are in the second, third, or fourth grades. Child Problem Behaviors and Disabilities In terms of disciplinary actions, 33 teachers reported that they have taken some action regarding the child once a week or more while about half of the teachers reported that they have never taken any action regarding the child. There was not a significant relationship between grades and the number of disciplinary actions that the teachers reported (χ 2 = 41.85; p=0.198). A small percentage of teachers reported that the child had school limitations due to a physical or mental disabilities (n=21 or 10% of children). Eleven teachers reported that the child had a learning disability, and one teacher reported a behavior disorder, while the remainder of the teachers reported other child disabilities (e.g., asthma, dyslexic hyperactive, seizures). Children with physical or mental disabilities were more likely to be failing school compared to children who did not (χ 2 = 26.93; p<.05). Teacher, Classroom, and School Characteristics In terms of teacher characteristics, most reported holding a bachelor s degree (44%) and 28% of the teachers held a master s degree. There was not a significant relationship between teacher-reported child grades and the teacher s own education level (χ 2 = 28.95; p=0.415). The amount of teaching experience may be related to teacher efficacy. Hence, we asked the teachers how long they had taught in their districts. A majority reported teaching in their districts for six years (n=110 or 64%). Only six of the teachers reported teaching in their districts for a year; 6

7 thirty teachers taught for 2-3 years, while 22 teachers taught for 4-5 years. There was not a significant relationship between the number of years that the teachers taught in the district and child grades (χ 2 = 40.78; p=0.761). Finally, about half of the teachers were Caucasian (n=99 or 47%), 37% were Hispanic (n=77), and 11% were African American (n=24), while the other 10 identified themselves as either Asian or Native American. There was not a significant relationship between teacher ethnicity and child achievement (χ 2 = 43.38; p=0.662). In terms of classroom and school size, classroom size ranged from 1-50 students, with an average size of There was not a significant relationship between classroom size and child grades (F=0.44; p=0.90). School size ranged from 50-6,000 students among Texas schools. The average school contained approximately 828 students. As with classroom size, there was not a significant relationship between school size and child grades (F=1.18; p=0.32). Future Data Analysis Upon the completion of TCTS data collection, my collaborators and I plan to conduct a series of multivariate analyses that will be based on what is referred to as an accumulation of inputs model, which has been articulated most clearly in the work of NICHD-ECCRN and Duncan (2003) and Blau (1999). This model, based on Figure 1, suggests that child i's development at time t is an additive function of all family, school, and child inputs to development up to that point in time. This model is represented in equation one below: (1) Child Outcome i3 = B 1E Family i1-3 + B 2E School i1-3 + B 3E Child i1-3 + ε i1-3 The challenge of considering the effects of family, school, and child characteristics on children s academic and social development using this approach is that children are not randomly assigned to their family or school contexts. This raises significant concerns of omitted variable bias (Johnston & DiNardo, 1997). To reduce this sort of bias in the estimates, we will 7

8 implement a two-pronged approach. First, we will include an extensive set of control variables that may be linked to children s development and their family and school contexts. Three waves of data from the Three-City Study offer a more extensive set of control variables than are typically available in a data set of this size. Even after controlling for these measured characteristics, however, there still may be unmeasured variables that bias our results. In the absence of data from randomized experiments, the only way to address concerns related to omitted variables is to do so statistically. Thus our second approach will be to estimate lagged regression models, of the form shown in Equation 2 below. (2) Child Outcome i3 = B 1 Family i1-3 + B 2 School i1-3 + B 3 Child i1-3 + B 4 Child Outcome i1-2 + ε i1-3 When predicting child outcomes at wave three, we will include a variable that represents an average of children s scores on this same outcome across wave one and wave two of the survey. This approach allows us to control for unmeasured heterogeneity that may be related to assessments of children s development at wave 1 and wave 2 of the survey (Cain, 1975). In addition to the child, classroom, and school characteristics presented above, we will also include measures on family characteristics and processes, as well as classroom and school processes in our multivariate models. Contribution to Theory or Practice of Education Public, political, and academic concerns are widespread regarding low achievement, school disengagement, and behavior problems, particularly among disadvantaged students in Texas, which ranks 23 rd on spending per student and high school completion rates remain relatively low. Facing concentrations of poverty, unequal funding, and extensive new requirements of the No Child Left Behind Act, Texas schools are working to identify best 8

9 practices and methods to engage and challenge low-income children. At the same time, developmental researchers seek to enhance understanding of the family practices that lead to healthy trajectories for children. I believe that our study will contribute new theoretical models and methods of research, spanning multiple environments and processes that promote children s school success in a state that ranks 45 th out of 50 states in educational attainment. In addition, we will identify the combinations of family and school environments that are best suited for children with various skills and strengths. Findings will benefit students, educators, policymakers, and scholars in Texas and across the country. 9

10 Figure 1: Family and School Influences on Children and Youth FAMILY ENVIRONMENT Family Economic Resources Family Structure Family Climate FAMILY PROCESSES Parenting Practices Parent Academic Practices Parent Characteristics SCHOOL SUCCESS Academic Competence Social Competence Teacher Characteristics SCHOOL ENVIRONMENT School Economic Resources School and Classroom Structure School and Classroom Climate SCHOOL PROCESSES Teacher Practices Child Characteristics 10

11 References Ackerman, B. P., Brown, E. D., & Izard, C. E. (2004). The Relations Between Persistent Poverty and Contextual Risk and Children's Behavior in Elementary School. Developmental Psychology, 40(3), Alexander, K. L., & Entwisle, D. R. (1996). School and Children at Risk. In A. Booth & J. F. Dunn (Eds.), Family-School Links: How Do They Affect Educational Outcomes? (pp ). Mahwah, New Jersey: Lawrence Erlbaum Associates. Blau, D. (1999). The effect of income on child development. The Review of Economics and Statistics, 81, Cain, G. G. (1975). Regression and selection models to improve nonexperimental comparisons. In C. A. Bernett & A. A. Lumsdiane (Eds.), Evaluation and experiment (pp ). New York:: Academic Press. Clements, M. A., Reynolds, A., & Hickey, E. (2004). Site-level predictors of children's school and social competence in the Chicago Child-Parent Centers. Early Childhood Research Quarterly, 19(2), Duncan, G. J., Yeung, W., Brooks-Gunn, J., & Smith, J. R. (1998). How much does childhood poverty affect the life chances of children? American Sociological Review, 63(3), Johnston, J., & DiNardo, J. (1997). Econometric Methods: Fourth Edition. New York:: The MCGraw- Hill Companies, Inc. Lee, V. E., & Burkam, D. T. (2002). Inequality at the starting gate: Social background differences in achievement as children begin school. Washington, DC: Economic Policy Institute. NICHD-ECCRN. (2002). The relation of first classroom environment to structural classroom features and teacher and student behavior. The Elementary School Journal, 102, NICHD-ECCRN. (2004). Multiple Pathways to Early Academic Achievement. Harvard Educational Review, 74, NICHD-ECCRN, & Duncan, G. J. (2003). Modeling the impacts of child care quality on children's preschool cognitive development. Child Development, 74, Stipek, D. J., & Ryan, R. H. (1997). Economically disadvantaged preschoolers: Ready to learn but further to go. Developmental Psychology, 33(4), Texas Education Agency. (2002). Secondary School Completion and Dropouts in Texas Public Schools, Austin, TX: Department of Accountability Reporting and Research, Texas Education Agency. Wirt, J., Choy, S., Rooney, P., Provasnik, S., Sen, A., & Tobin, R. (2004). The Condition of Education (NCES ). Washington, DC: U.S. Department of Education, National Center for Education Statistics. 11