How Do Schools and Teachers Affect Immigrant Students Science Performance? Mido Chang, Virginia Tech, USA Kusum Singh, Virginia Tech, USA Youngji Y. Sung, Virginia Tech, USA Sunha Kim, Virginia Tech, USA Abstract: Science education for immigrant students, particularly English language learners (ELL), cannot be accomplished successfully unless teachers and schools bridge the gap between cultural and linguistic differences (Cho & McDonnough, 2009; Lee, 2005). Several studies report that immigrant students who do not have an American cultural and linguistic understanding display significantly lower science performance than native-born English-speaking students in schools (Crosnoe, Lopez-Gonzalez, & Muller, 2004; Pong & Hao, 2007; Schnepf, 2004). The main objective of this project is to examine the effects of teachers certification in science and school demographic environments (minority student proportion, ELL student proportion, and the proportion of students who are eligible for free lunch) on the science performance of eight grade students, with focused attention to ELL students. The study employed a three-level Hierarchical Linear Modeling (HLM) to a US national data, the Early Childhood Longitudinal Survey (ECLS-K). The study found that while teacher certification in science and the ELL student proportion in a school did not have a significant effect on students science performance, the minority student proportion and the proportion of students who are eligible for free lunch had a negative effect on the average performance of students. Interestingly, ELL students displayed comparatively higher science performance in schools with high minority population where native-born Englishspeaking students had low science performances. The findings of the study from national databases will lay the foundation for further research regarding science outcomes of immigrant students. Keywords: immigrant students, science performance, teacher science degree, school environments Introduction Despite the widespread belief that science can be learned using a universal language, research has shown that science subjects include a large amount of content-specific vocabulary and background knowledge rooted in American and European cultures (Lee, 2005). Science education for immigrant students, particularly ELL students, cannot be accomplished successfully unless teachers and schools bridge the gap between cultural and linguistic differences (Cho & McDonnough, 2009; Lee, 2005). Several studies report
that immigrant students who do not have an American cultural and linguistic understanding display significantly lower science performance than native-born Englishspeaking students in schools (Chang & Kim, 2009; Crosnoe, Lopez-Gonzalez, & Muller, 2004; Pong & Hao, 2007; Schnepf, 2004). The performance gap is particularly large for immigrant students with limited English proficiency (LEP) (Baldi, Jin, Green, & Herget, 2007; Chang, 2008; Chang, Singh, & Filer, 2009; Haile & Nguyen, 2007; Kim & Chang, 2008, 2010; Muller, Stage, & Kinzie, 2001; Sung & Chang, 2008). The main objective of this project is to examine the effects of teacher certification in science and school demographic environments on the science performance of students in middle school. The study particularly paid attention to the science proficiency of immigrant students to suggest educational policies that are responsive to the needs of immigrant students for their school success. The study used the eighth grade data from the Early Childhood Longitudinal Survey Kindergarten Cohort (ECLS-K), a US nationally representative data. As a main statistical tool, the study employed a three-level Hierarchical Linear Modeling (HLM) to consider the nested structure of students, teachers, and schools. By benefiting from advanced features of the HLM analysis, this study explored how students get influenced from teachers educational preparation in science, and how dynamics between students and teachers are influenced by school environments. The overarching research questions of this study are as follows: 1. Does a teacher s certificate in science show an effect on the science performance of students in middle school? If it does, is the effect significantly different for immigrant students? 2. Do school environmental factors (proportions of racial minorities, English language learners, and students who are eligible for free lunch) have effects on the average science performance of middle schools? If they do, are the effects different for native-born and ELL student groups? Literature Review Immigrant Students and School Performance Over the past ten years from 1996 to 2006, the growth rate of ELL students was 57.17% in the total PK-12 enrollment, having 3.66% of a growth rate of total enrollment (NCELA, 2009). Approximately 10 million students in the US, aged from 5 to 17, speak other languages than English at home (NCES, 2005). Unfortunately, those large number of ELL students have displayed significantly lower school performance, including science. The study on ELL students science performance in middle school and significant factors associated with their school performance is particularly important to reduce the science performance gap and to ensure the academic success in the later schooling (National Research Council, 2009).
Teacher Certification in Science Research has shown that teacher s qualification in science fields is associated with high science performance of students (Darling-Hammond, Berry, & Thoreson, 2001; Goldhaber & Brewer, 1996). A major policy issue in the U.S. is that science education suffers from a lack of certified teachers. Only 69% of science teachers in middle and high schools majored in science fields in college (Goldhaber & Brewer, 1996). Teachers without standard certification tend to teach African-American students and students from low SES (Darling-Hammond, Holtzman, Gatlin, & Heilig, 2005). Students in schools with a high percentage of ELL students are likely to be served by unqualified and substitute teachers, as compared to those in schools with a low percentage of ELL students (de Cohen, Deterding, & Clewell, 2005). School Environments The performance of immigrant students is more strongly affected by the school environment than that of non-immigrant students (Han, 2008). Conversely, schools can be an effective assimilation vehicle for immigrant students. The average income of students families in a school has been found to have a long-term effect on immigrant students science achievement, just as their personal family income is a major factor for school performance (Kyriakides & Creemers, 2008). Many immigrant families live in inner-city areas, where immigrant students are negatively influenced by native-born English-speaking peers who tend to exhibit a lower level of school engagement (Hao & Pong, 2008). Yet, when immigrant students are enrolled in schools in which ELL students are highly concentrated (usually in urban areas), they are segregated from mainstream education. Moreover, students in schools with a high percentage of ELL students are more likely to be served by unqualified and substitute teachers, as compared to those in schools with a low percentage of ELL students (de Cohen et al. 2005). These school environment factors affect students educational performance. Methods Data The study used the ECLS-K eighth-grade data which were collected from spring 2006 to spring 2007. The ECLS-K is a nationally representative cohort from kindergarten through eighth grade. The total of 21,260 kindergarteners in the fall of 1998 participated in the base year data and the total 9,725 eighth grade students participated in the various measurements until the end of spring 2007. The sampling method of the ECLS-K used a multistage probability sample design. In the primary sampling of the ECLS-K, the units were randomly selected from 90 strata of geographic areas consisting of counties. In the second stage schools were randomly selected within sampled counties. The total 1277 schools, 914 public and 363 private, participated in the data collection. At the final stage all students within the selected schools became final units (Tourangeau, et al. 2006). Variables The science performance score measured by Item Response Theory (IRT) was the dependent variable of the study. The major benefit of the IRT scale score is that it
measures students ability by separating it out from test characteristics (e.g., item difficulty and item discrimination). In other words, the IRT score measures comparatively true student ability that was not contaminated by test characteristics. As the objectives of the study indicated, this study focused on immigrant students, specifically ELL students. The study conducted analyses for two language groups of students: native-born English-speaking and ELL students. The native-born English speaking group was coded as 0 and served as the reference group in the analysis while the ELL group was coded as 1. The main predictor variables for the study are teacher certification in science and three school environment variables (minority student proportion, ELL student proportion, and the proportion of students who are eligible for free lunch). The variable of teacher certification in science is a composite score of science teacher certificates either in the fall or the spring of the eighth grade (Yes=1; No=0). The proportion of minority students in a school was a variable with five categories (1=less than 10%; 2=10% to less than 25%; 3=25% to less than 50%; 4=50% to less than 75%; and 5=75% or more). The variables for proportions of ELL students and students eligible for free lunch were actual percentages of those students. The variable of the proportion of students who are eligible for free lunch in a school was used as a proxy for the average school poverty level. The important student variables of gender (male=0; female=1), and socio-economic status measured in a continuous scale were specified in the analysis. Analysis The study ran a three-level HLM analysis applying a proper weight (the 8th grade full child weight full sample: Cwc90) to treat design effects and to have the sample representative of the US national 8 th grade student population and arrive at study conclusions with generalizability. The three models at each level of HLM analysis are specified as follows: Level 1 model: Y =π0 + π1*(sex) + π2*(ell)+ π3*(ses) + e. Level 2 model: π0 = β00 + β01*(science Teacher), π1 = β10 + r1, π2 = β20, and π3 = β30. Level 3 model: β00 = γ000 + γ001(minority) + γ 002(ELL) + γ 003(Free Lunch) + μ00, β01 = γ 010, β10 = γ 100,
β20 = γ 200 + γ 201(Minority) + γ 202(ELL) + γ 203(Free Lunch), and β30 = γ 300. Results As shown in Table 1, this paper paid attention to the science performance of ELL students comparing it with that of native-born English-speaking students. The overall performance of ELL students in science was significantly lower than that of native-born English-speaking students (γ 200 = -3.092, p<0.01) after controlling for student gender and socio-economic status. One of the main research questions of the study, the effect of a teacher certificate in science, did not show a significant effect on the science performance of students which was contrary to our expectation. Another main research question regarding school environments indicated some significant effects on student science performance. The first school demographic environmental factor, minority proportion, showed a negative effect on the average science performance of students (γ001 = -1.266, p<0.01). In other words, when a school had a high proportion of minority student population, the overall science performance of students in that school tended to be low as compared to the average performance of a school that have a zero minority student population. The second school environmental factor in the study, ELL student proportion did not indicate a significant effect. The last school environment factor, the proportion of students who are eligible for free lunch did reveal a significant, negative effect (γ003 = -0.082, p<0.01), indicating the higher the proportion of those students in a school, the lower the average science performance of the students in the school. The important results of the study are the differential effects of the main predictor variables on the science performance of ELL students. The effects of the teacher science certificate, the ELL proportion of a school, and the school proportion of students who are eligible for free lunch did not show significant differential effects for ELL students. Therefore, the teacher certificate in science and the ELL proportion of a school did not show significant associations with the science performance of ELL students. However, when ELL students go to a school that has a high proportion of students who are eligible for free lunch, they tend to indicate a low science performance level. The last important finding of the study is the effect of the school minority proportion on the science performance of ELL students. The effect was significantly positive (γ 201 = 4.087, p<0.01), with ELL students having higher science performance in a school with the high minority population as compared to native-born students in the same condition. In other words, ELL students did not get affected by the negative conditions as much as native-born English-speaking students.
Table 1. HLM Analyses for Science Achievement Using Three-Level Model Baseline Model Teacher & School Model Fixed Component Coefficient SE Coefficient SE Initial Score Intercept 84.092** 0.270 84.858** 0.311 Intercept -1.266** 0.249 Minority Proportion 0.039 0.039 ELL Proportion -0.082** 0.014 Free Lunch Eligible Proportion Science Degree 0.573 0.480 Gender -2.738** 0.475 ELL Intercept Intercept -3.092** 0.843 Minority Proportion 4.087** 0.661 ELL Proportion -0.105 0.063 Free Lunch Eligible Proportion -0.042 0.034 SES 6.601** 0.352 Random Component 2 2 Variance df Variance df Level 3 67.152 ** 4481.31 2330 34.661** 3461.25 1708 Level 2 38.469** 3385.01 2161 127.665** 4887.36 3586 Level 1 151.371 81.449 Deviance 66671.04 4 52341.62 14 Reliability of Gender 0.438 Reliability of Intercept 0.396 0.400 p < 0.05, ** p < 0.01 Discussion The primary goal of this study was to provide a sound empirical basis for policy development on the effects of teacher certification in science and school demographical
environments on the science performance of immigrant students. The study presented empirical data on the effects of those factors on the science performance of ELL students in middle school. In order for study findings to have high generalizability, the study employed an advance statistical tool, a three-level HLM to a nationally representative database of ECLS-K with proper weight adjustment. The study found that the science performance of ELL students was significantly lower than that of native-born English-speaking students, and this result supports previous research findings that immigrant students lag behind native-born English-speaking students (Baldi, Jin, Green, & Herget, 2007; Chang & Kim, 2009; Chang, 2008; Chang, Singh, & Filer, 2009; Haile & Nguyen, 2007; Muller, Stage, & Kinzie, 2001; Sung & Chang, 2008). Against the general expectations, teacher certification in science and the ELL student proportion in a school did not have a significant effect on students science performance in the study. In that regard these results did not match with prior research findings. According to Darling-Hammond, Berry, and Thoreson (2001) and Goldhaber and Brewer (1996), teacher preparation or qualification in science subject areas are important conditions for effective science teaching. The study did not find a significant interaction of poor quality of teachers and the large population of ELL students (de Cohen et al. 2005). However, the minority student proportion and the proportion of students who are eligible for free lunch had a negative effect on the average performance of students as noted in several studies (Kyriakides & Creemers, 2008). One interesting, important finding of the study is that ELL students displayed comparatively higher science performance in schools with high minority population whereas native-born English-speaking students had low science performance. This condition should be further studied in conjunction with other school factors such as school programs for ELL students. The findings of the study from national databases will lay the foundation for further research regarding science outcomes of immigrant students. Though the study is based on survey questionnaires, thus rendering any causal inferences tentative at best, this study is a contribution to our knowledge of the effects of school environments on the educational performance of the ELL students. Although school environment affects a student s academic science performance, it is not the only factor. This study further points to the need for more future studies.
References Baldi, S., Jin, Y., Green, P. J., & Herget, D. (2007). Highlights from PISA 2006: Performance of U.S. 15-Year-Old Students in Science and Mathematics Literacy in an International Context. NCES 2008-2016: National Center for Education Statistics. Chang, M. (2008). Teacher instructional practices and language minority students: A longitudinal model. Journal of Educational Research, 102(2), 83-97. Chang, M., & Kim, S. (2009). Computer access and computer use for science performance of ethnic and linguistic minority student. Journal of Educational Computing Research, 40(4), 457-489. Chang, M., Singh, K., & Filer, K. (2009). Language factors associated with achievement grouping in math classrooms: A cross-sectional and longitudinal study. School Effectiveness and School Improvement, 20(1), 27-45. Cho, S., & McDonnough, J. T. (2009). Meeting the needs of high school science teachers in English language learner instruction. Journal of Science Teacher Education, 20(4), 385-402. de Cohen, C., Deterding, N., & Clewell, B. C. (2005). Who's Left Behind? Immigrant Children in High and Low LEP Schools, Program for Evaluation and Equity Research, Washington, DC: Urban Institute. Crosnoe, R., Lopez-Gonzalez, L., & Muller, C. (2004). Immigration from Mexico into the math/science pipeline in American education. Social Science Quarterly, 85(5), 1208-1226. Darling-Hammond, L., Berry, B., & Thoreson, A. (2001). Does teacher certification matter? Evaluating the evidence. Educational Evaluation and Policy Analysis, 23(1), 57-77. Darling-Hammond, L., Holtzman, D. J., Gatlin, S. J., & Heilig, J. V. (2005). Does teacher preparation matter? Evidence about teacher certification, teach for America, and teacher effectiveness. Education Policy Analysis Archives, 13(42), 1 48. Goldhaber, D. D., & Brewer, D. J. (1996). Evaluating the effect of teacher degree level on educational performance. Rockville, MD: Westat, Inc. Haile, G. A., & Nguyen, A. N. (2007). Determinants of academic attainment in the United States: A quantile regression analysis of test scores. Education Economics, 16(1), 29-57. Han, W.J. (2008). The academic trajectories of children of immigrants and their school environments. Developmental Psychology, 44(6), 1572-1590. Hao, L., & Pong, S.-L. (2008). The role of school in the upward mobility of disadvantaged immigrants' children. The ANNALS of the American Academy of Political and Social Science, 620(1), 62-89. Kim, S., & Chang, M. (2010). Does computer use promote the mathematical proficiency of ELL students? Journal of Educational Computing Research, 42(3), 285-305. Kim, S., & Chang, M. (2008). Computer use in diverse learning contexts in ELS 2002: Gender, race and learner-centeredness effects and implications. Journal on School Educational Technology, 3(4), 55-65.
Kyriakides, L., & Creemers, B. P. M. (2008). A longitudinal study on the stability over time of school and teacher effects on student outcomes. Oxford Review of Education, 34(5), 521-545. Lee, O. (2005). Science education with English language learners: Synthesis and research agenda. Review of Educational Research, 75(4), 491-530. Muller, P. A., Stage, F. K., & Kinzie, J. (2001). Science achievement growth trajectories: Understanding factors related to gender and racial-ethnic differences in precollege science achievement. American Educational Research Journal, 38(4), 981-1012. National Center for Education Statistics. (2005). Highlights from the 2003 international adult literacy and lifeskills survey. Washington, D.C.: United States Department of Education. National Clearinghouse for English Language Acquisition & Language Instruction Educational Programs. (n.d.). The growing numbers of limited English proficient students. Retrieved August 19, 2009, from http://www.ncela.gwu.edu/files/uploads/4/growinglep_0506.pdf National Research Council (2009). Mathematics Learning in Early Childhood: Paths Toward Excellence and Equity. Washington, DC: National Academies Press. Pong, S.-L., & Hao, L. (2007). Neighborhood and school factors in the school performance of immigrants' children. International Migration Review, 41(1), 206-241. Schnepf, S. V. (2004). How different are immigrants? A cross-country and cross-survey analysis of educational achievement IZA Discussion Paper No. 1398: Institute for the Study of Labor. Sung, Y., & Chang, M. (2008). Center-based care for language minority students. Educational Research and Evaluation, 14(5), 445-463. Tourangeau, K., Nord, C., Leˆ, T., Pollack, J.M., Atkins-Burnett, S., & Hausken, E.G. (2006). Early Childhood Longitudinal Study, Kindergarten Class of 1998 99 (ECLS-K): Combined user s manual for the ECLS-K fifth-grade data files and electronic codebooks. Washington, DC: National Center for Education Statistics & U.S. Department of Education Institute of Education Sciences.