6
Comparability of Estimates of Immigrant School-Age Children

In this chapter we discuss the definitions of immigrant students from the two allowable data sources and compare those definitions. In discussing the American Community Survey (ACS), we apply several of the analytical techniques used in Chapter 2 when considering the English language learner (ELL) estimates. We also assess the strengths and weakness of estimates based on state administrative data.

As noted in Chapter 1, Title III of the Elementary and Secondary School Act requires the U.S. Department of Education (DoEd) to allocate funds to all 50 states, the District of Columbia, and Puerto Rico on the basis of a formula that incorporates the population of immigrant children and youth in each state. Specifically, the legislation states that 20 percent of the funds are to be based on the population of “recently immigrated children and youth (relative to national counts of these populations).” Section 3301(6) defines an “eligible immigrant student” as an individual who (A) is aged 3 through 21; (B) was not born in any state; and (C) has not been attending one or more schools in any one or more states for more than 3 full academic years. In this definition, language spoken by an individual is not a criterion for classification as an immigrant. This absence in the statutory definition may affect the allocation of the 20 percent of the funds to the extent that the English speaking ability of immigrant populations might vary by state because the source of the immigrant population varies by state.

As with the counts of eligible limited English proficiency (LEP) children and youth, the data on immigrant students can and have been derived from both the ACS and administratively reported state counts, and both data series have been used in the allocation formula. Prior to fiscal 2005, the DoEd allocated the immigration-related portion of Title III funds to the states on the basis of the state-reported counts of the number of immigrant children and youth; since then, the department



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 133
6 Comparability of Estimates of Immigrant School-Age Children In this chapter we discuss the definitions of immigrant students from the two allowable data sources and compare those definitions. In discussing the American Community Survey (ACS), we apply several of the analytical techniques used in Chapter 2 when considering the English language learner (ELL) estimates. We also assess the strengths and weakness of estimates based on state administrative data. As noted in Chapter 1, Title III of the Elementary and Secondary School Act requires the U.S. Department of Education (DoEd) to allocate funds to all 50 states, the District of Columbia, and Puerto Rico on the basis of a formula that incorpo - rates the population of immigrant children and youth in each state. Specifically, the legislation states that 20 percent of the funds are to be based on the population of “recently immigrated children and youth (relative to national counts of these popu - lations).” Section 3301(6) defines an “eligible immigrant student” as an individual who (A) is aged 3 through 21; (B) was not born in any state; and (C) has not been attending one or more schools in any one or more states for more than 3 full aca - demic years. In this definition, language spoken by an individual is not a criterion for classification as an immigrant. This absence in the statutory definition may affect the allocation of the 20 percent of the funds to the extent that the English speaking ability of immigrant populations might vary by state because the source of the im - migrant population varies by state. As with the counts of eligible limited English proficiency (LEP) children and youth, the data on immigrant students can and have been derived from both the ACS and administratively reported state counts, and both data series have been used in the allocation formula. Prior to fiscal 2005, the DoEd allocated the immigration- related portion of Title III funds to the states on the basis of the state-reported counts of the number of immigrant children and youth; since then, the department 133

OCR for page 133
134 ALLOCATING FEDERAL FUNDS has used ACS data for the number of immigrant children and youth to determine the state allocations. ACS DATA AND ESTIMATES Definition of Immigrant Children and Youth Like the ACS data that the department uses to determine students with LEP, the data on immigrant status are based on self-reports. Three ACS questions are used to identify recent immigrants: (1) whether each household member was born in the United States, (2) whether he or she is a citizen, and (3) for those not born in the United States, when the person entered the country—see Box 6-1. Household mem- bers between the ages of 3 and 21 are classified as recent immigrants if they are not U.S. citizens at birth1 and entered the country less than 3 years prior to the survey. Evaluation of the Survey Questions Like the questions on language spoken and English speaking ability, the ACS questions that define an immigrant child or youth were adopted from the long form of the decennial census at the time the ACS was developed. They have also been a part of the Current Population Survey for some time. They play a critical role in the Census Bureau’s annual population estimates program as the basis for the net international migration estimate. Owing to their importance, the objectivity and collectability of these questions has been the subject of several analyses over the years, culminating in a major 2006 Census Bureau ACS test of the new and modified item content (Harris et al., 2007). Although much of the research has focused on missing content, such as parental nativity and date of naturalization, the “year of arrival” question has been the subject of some evaluation because of the concern that the current question allows report - ing of only one entry to the United States even when the respondents have entered multiple times, and the interpretation of “coming to live” in the United States may be too broad. Redstone and Massey (2003) identified problems with the year of entry question as a source of underestimation of the number of years that have elapsed since a person’s arrival. Most likely, the inconsistencies were the result of multiple entries into the United States by persons who may have provided the year of a recent entry rather than their first entry (Schmidley and Robinson, 2003). The content test report suggested that there was confusion among both respon - dents and ACS field representatives about the kind of information that the entry question was seeking and about how to report multiple arrivals. This confusion was evident in the content test itself, when a test group that was asked further probing questions about year of arrival was not able to provide accurate answers to the ques - 1 The category includes respondents who indicate they are U.S. citizens by naturalization.

OCR for page 133
135 COMPARABILITY OF ESTIMATES OF IMMIGRANT SCHOOL-AGE CHILDREN BOX 6-1 ACS Questions on Birth, Citizenship, and Year of Entry into the United States 1. Where was this person born? In the United States Print Name of State Outside the United States Print Name of Foreign Country, or Puerto Rico or Guam etc 2. Is this person a citizen of the United States? Yes, born in the United States, Yes, born in Puerto Rico, Guam, the U.S. Virgin Islands, or Northern Marianas Yes, born abroad of U.S citizen parent or parents Yes, U.S. citizen by naturalization Print Year of Naturalization No, not a U.S. citizen 3. When did this person come to live in the United States? Print year of Arrival tions. Based on the results of the content test, in 2008 the ACS added a question on year of naturalization. Unfortunately, it is not known how these issues with the precise timing of the date of entry affect the precision of the estimate of immigrant children and youth for purposes of Title III allocations. If a significant number of children and youth who had originally arrived 3 years ago or earlier reported a subsequent arrival because of confusion over the meaning of the question, there would be tendency for the count of recent immigrant children and youth from the ACS to be an overestimate. Effect of Nonresponse on Data Quality We next considered the possible effect of item nonresponse on the ACS esti - mates. The allocation (imputation) rates (described in Chapter 2) for the “place of birth” item were 7.0 percent in 2008, which is considered moderate, and the alloca - tion rates for the “year of entry” items were also moderate, at 10.4 percent. However, the amount of imputation required has trended upwards from year to year for each of the immigrant-identifying questions: see Table 6-1. The implications of nonresponse for the accuracy of estimates are not evident. Its effects depend, first of all, on the accuracy of the assumptions underlying the

OCR for page 133
136 ALLOCATING FEDERAL FUNDS TABLE 6-1 Allocation Rates for Nonresponse on Immigrant Items in the ACS, 2005-2008 (in percentage*) Item 2008 2007 2006 2005 Place of Birth 7.0 5.5 4.8 4.7 Citizenship 2.5 1.8 1.6 1.6 Year of Entry: 10.4 8.5 7.9 6.9 Total Population Not Born in U.S. *The item allocation rates for year 2005 are for housing units only. The item allocation rates for 2006 to 2008 include housing units and group quarters populations. SOURCE: From the Census Bureau Quality Measures Page, http://www.census.gov/acs/www/UseData/ sse/ita/ita_def.htm [May 2010]. procedures used to impute the missing values, and, consequently, on whether they impute recent immigration at approximately the correct rate. Furthermore, even if estimates of the total number of recent immigrants are inaccurate, Title III al - locations to states would only be affected if the errors are disproportionate across states. For 2008, the range in the imputation rates for “place of birth” and “year of entry” across states are fairly narrow with interquartile ranges (the area between the 25th and 75th percentiles) of only 1.4 percent and 3.2 percent, respectively. Given the moderate overall item nonresponse rates and the fairly limited range of rates among the states, the effects of nonresponse on the allocations are not likely to be substantial. The ACS Estimates The 1-year and 3-year estimates of immigrant children and youth for 2005- 2008 were derived from special tabulations provided by the Census Bureau: they are shown in Table 6-2. Along with the number of immigrant children and youth aged 3-21 years, the table shows the corresponding standard errors and coefficients of variation. The panel’s conclusions regarding the characteristics of the 1- and 3-year esti - mates for ELL estimates (see Chapter 2) also apply to the ACS estimates of immi - grant children. Because each 3-year estimate is based on three times the sample size of the 1-year estimates, the standard errors of the former are substantially lower. The shares of the states that are based on 1-year estimates do not fluctuate a great deal. The variation is further dampened when shares are based on 3-year estimates because consecutive 3-year estimates include 2 overlapping years (in this case, 2006 and 2007) and so are more stable than 1-year estimates: see Table 6-3. However, the 1-year estimates respond more quickly to changes in economic and social character - istics than the 3-year estimates. The percentage share of each state’s estimate of immigrant children and youth is shown in Table 6-4.

OCR for page 133
TABLE 6-2 Number of Immigrant Children and Youth Aged 3-21, by State ACS 2005 ACS 2006 ACS 2007 ACS 2008 State Estimate SE CV Estimate SE CV Estimate SE CV Estimate SE CV Alabama 7,710 862 0.11 7,335 829 0.11 9,815 1,404 0.14 5,405 920 0.17 Alaska 965 431 0.45 1,765 487 0.28 2,555 728 0.28 1,095 433 0.40 Arizona 35,660 2,855 0.08 32,565 2,852 0.09 37,565 3,158 0.08 22,250 2,429 0.11 Arkansas 4,680 935 0.20 5,330 838 0.16 5,785 1,044 0.18 4,510 914 0.20 California 251,275 9,185 0.04 214,095 6,483 0.03 208,295 6,405 0.03 179,500 6,834 0.04 Colorado 16,835 1,897 0.11 13,405 1,539 0.11 13,560 1,558 0.11 12,780 1,417 0.11 Connecticut 10,670 1,395 0.13 11,275 1,573 0.14 11,095 1,550 0.14 10,165 1,073 0.11 Delaware 2,495 491 0.20 1,980 706 0.36 1,515 512 0.34 1,100 451 0.41 District of Columbia 1,285 454 0.35 4,065 810 0.20 1,445 456 0.32 1,635 362 0.22 Florida 93,535 4,263 0.05 88,770 4,553 0.05 83,245 3,785 0.05 69,790 5,551 0.08 Georgia 36,945 3,031 0.08 31,160 2,172 0.07 26,295 2,172 0.08 28,460 2,670 0.09 Hawaii 6,645 1,512 0.23 8,495 1,273 0.15 5,110 858 0.17 7,810 2,288 0.29 Idaho 5,010 1,343 0.27 3,550 837 0.24 4,130 679 0.16 4,425 732 0.17 Illinois 35,965 2,710 0.08 35,225 2,771 0.08 44,240 2,900 0.07 32,535 2,505 0.08 Indiana 11,985 1,368 0.11 11,160 1,436 0.13 10,665 1,015 0.10 9,160 1,172 0.13 Iowa 4,150 733 0.18 4,685 704 0.15 4,580 812 0.18 5,540 818 0.15 Kansas 6,035 913 0.15 7,030 1,236 0.18 6,815 940 0.14 6,330 988 0.16 Kentucky 5,275 961 0.18 5,600 679 0.12 6,355 1,202 0.19 5,710 961 0.17 Louisiana 3,185 602 0.19 4,980 798 0.16 5,940 1,113 0.19 4,910 910 0.19 Maine 995 497 0.50 1,715 469 0.27 1,075 393 0.37 1,520 436 0.29 Maryland 26,765 2,621 0.10 23,940 1,858 0.08 22,260 1,974 0.09 19,500 1,392 0.07 Massachusetts 23,935 1,876 0.08 21,920 1,855 0.08 24,380 2,174 0.09 26,290 2,327 0.09 Michigan 20,640 2,328 0.11 17,700 2,054 0.12 21,810 2,030 0.09 18,945 1,904 0.10 Minnesota 14,420 1,562 0.11 17,285 2,044 0.12 15,100 1,536 0.10 10,980 1,165 0.11 Mississippi 2,695 703 0.26 2,925 663 0.23 3,300 534 0.16 3,420 896 0.26 Missouri 7,315 1,244 0.17 7,370 1,123 0.15 8,090 1,062 0.13 8,940 1,031 0.12 137 continued

OCR for page 133
TABLE 6-2 Continued 138 ACS 2005 ACS 2006 ACS 2007 ACS 2008 State Estimate SE CV Estimate SE CV Estimate SE CV Estimate SE CV Montana 465 170 0.37 1,240 366 0.30 1,415 462 0.33 800 281 0.35 Nebraska 4,130 904 0.22 4,820 759 0.16 2,790 560 0.20 3,885 797 0.21 Nevada 9,445 1,158 0.12 11,900 1,337 0.11 11,425 1,517 0.13 9,620 1,477 0.15 New Hampshire 1,155 368 0.32 2,635 779 0.30 1,640 439 0.27 1,395 381 0.27 New Jersey 38,670 2,645 0.07 38,475 2,245 0.06 34,525 2,470 0.07 37,725 2,451 0.06 New Mexico 5,720 1,063 0.19 9,465 1,439 0.15 3,920 759 0.19 2,340 500 0.21 New York 83,310 3,942 0.05 95,185 4,360 0.05 79,390 3,826 0.05 84,055 3,810 0.05 North Carolina 27,890 3,614 0.13 27,175 2,541 0.09 24,495 2,165 0.09 21,660 1,887 0.09 North Dakota 415 212 0.51 1,805 501 0.28 1,265 343 0.27 1,105 426 0.39 Ohio 13,525 1,331 0.10 11,720 1,531 0.13 12,625 1,638 0.13 16,370 1,932 0.12 Oklahoma 5,935 1,021 0.17 6,515 1,025 0.16 7,950 1,123 0.14 5,305 902 0.17 Oregon 10,925 1,638 0.15 12,480 1,255 0.10 9,450 1,388 0.15 9,860 1,303 0.13 Pennsylvania 16,150 1,430 0.09 18,285 1,856 0.10 21,255 2,165 0.10 19,565 1,498 0.08 Rhode Island 4,610 1,169 0.25 3,465 712 0.21 3,410 798 0.23 3,205 636 0.20 South Carolina 11,865 1,452 0.12 7,005 903 0.13 9,715 1,200 0.12 5,695 845 0.15 South Dakota 1,835 876 0.48 715 241 0.34 665 185 0.28 180 176 0.98 Tennessee 9,800 1,260 0.13 10,845 1,152 0.11 11,885 1,333 0.11 10,150 1,670 0.16 Texas 130,990 5,851 0.04 122,375 5,277 0.04 110,375 5,088 0.05 95,575 4,515 0.05 Utah 7,410 1,179 0.16 7,950 1,048 0.13 9,420 1,115 0.12 8,630 1,569 0.18 Vermont 645 178 0.28 880 394 0.45 805 204 0.25 970 283 0.29 Virginia 25,835 2,306 0.09 26,545 2,072 0.08 23,800 1,673 0.07 22,240 1,691 0.08 Washington 24,375 2,018 0.08 28,775 3,064 0.11 31,535 2,641 0.08 24,160 2,083 0.09 West Virginia 200 108 0.54 945 252 0.27 1,335 318 0.24 2,095 573 0.27 Wisconsin 8,805 1,057 0.12 9,110 1,029 0.11 7,525 890 0.12 6,300 1,010 0.16 Wyoming 1,085 422 0.39 1,290 362 0.28 680 281 0.41 165 92 0.56 United States 1,082,255 17,490 0.02 1,046,930 14,440 0.01 1,008,330 13777 0.01 895,760 15,175 0.02 NOTES: SE = standard error; CV = coefficients of variation. SOURCE: U.S. Census Bureau Special Tabulations.

OCR for page 133
139 COMPARABILITY OF ESTIMATES OF IMMIGRANT SCHOOL-AGE CHILDREN TABLE 6-3 Average Number of Immigrant Children and Youth Aged 3-21, by State ACS 2005-2007 ACS 2006-2008 State Estimate SE CV Estimate SE CV Alabama 8,680 575 0.07 7,295 607 0.08 Alaska 2,080 483 0.23 1,800 303 0.17 Arizona 35,815 1,995 0.06 30,470 1,584 0.05 Arkansas 5,355 486 0.09 5,300 557 0.11 California 225,860 4,538 0.02 198,565 3,896 0.02 Colorado 14,790 942 0.06 13,305 787 0.06 Connecticut 11,480 938 0.08 11,040 794 0.07 Delaware 1,950 367 0.19 1,455 353 0.24 District of Columbia 2,360 312 0.13 2,165 260 0.12 Florida 89,035 2,463 0.03 80,605 2,367 0.03 Georgia 31,355 1,458 0.05 27,390 1,319 0.05 Hawaii 6,540 713 0.11 7,065 913 0.13 Idaho 4,715 793 0.17 4,195 527 0.13 Illinois 39,335 1,603 0.04 36,725 1,497 0.04 Indiana 11,545 733 0.06 9,980 605 0.06 Iowa 4,635 404 0.09 4,935 459 0.09 Kansas 6,740 547 0.08 6,340 623 0.10 Kentucky 6,370 584 0.09 5,785 505 0.09 Louisiana 4,820 674 0.14 5,230 578 0.11 Maine 1,345 293 0.22 1,650 353 0.21 Maryland 24,730 1,313 0.05 22,530 1,268 0.06 Massachusetts 24,605 1,118 0.05 24,085 1,312 0.05 Michigan 21,035 1,208 0.06 19,370 907 0.05 Minnesota 16,315 956 0.06 14,895 1,005 0.07 Mississippi 3,425 472 0.14 3,315 426 0.13 Missouri 8,280 696 0.08 8,230 678 0.08 Montana 1,060 190 0.18 1,305 228 0.17 Nebraska 4,075 477 0.12 3,785 422 0.11 Nevada 10,925 789 0.07 11,055 931 0.08 New Hampshire 1,995 281 0.14 1,860 319 0.17 New Jersey 36,740 1,357 0.04 36,920 1,386 0.04 New Mexico 6,105 588 0.10 5,265 557 0.11 New York 88,090 2,421 0.03 8,5340 2,635 0.03 North Carolina 25,720 1,382 0.05 23,915 1,160 0.05 North Dakota 1,455 270 0.19 1,300 200 0.15 Ohio 13,595 1,006 0.07 13,660 1,013 0.07 Oklahoma 7,345 622 0.08 6,920 687 0.10 Oregon 11,435 921 0.08 10,615 945 0.09 Pennsylvania 19,500 1,315 0.07 19,725 1,095 0.06 Rhode Island 3,920 564 0.14 3,350 483 0.14 South Carolina 9,950 810 0.08 7,910 713 0.09 South Dakota 1,155 300 0.26 540 135 0.25

OCR for page 133
140 ALLOCATING FEDERAL FUNDS TABLE 6-3 Continued ACS 2005-2007 ACS 2006-2008 State Estimate SE CV Estimate SE CV Tennessee 11,460 905 0.08 10,660 759 0.07 Texas 122,765 3,237 0.03 109,105 2,996 0.03 Utah 8,535 828 0.10 8,685 811 0.09 Vermont 745 138 0.19 800 160 0.20 Virginia 25,690 1,291 0.05 23,380 1,129 0.05 Washington 27,935 1,260 0.05 27,800 1,528 0.05 West Virginia 1,045 217 0.21 1,520 263 0.17 Wisconsin 8,615 673 0.08 7,340 518 0.07 Wyoming 1,005 223 0.22 750 164 0.22 United States 1,064,075 8853 0.01 977,220 9,638 0.01 NOTES: SE = standard error; CV = coefficients of variation. SOURCE: U.S. Census Bureau Special Tabulations. The corresponding standard errors are shown in Table 6-5. The percentage shares are the basis for the portion of the Title III allocations based on immigrants. The nine states with the largest shares—Arizona, California, Florida, Georgia, Il - linois, New Jersey, New York, Texas, and Washington—together account for more than 70 percent of immigrant youth. We also calculated the ratio of immigrant children and youth aged 5-18 to all children and youth within each state of that age who are enrolled in public school. This ratio is computed by simply dividing the ACS estimate of immigrant children and youth enrolled in public school by the ACS estimate of all children and youth in public school: see Table 6-6. Tables 6-3 through 6-6 are based on special tabulations of ACS data provided by the U.S. Census Bureau. Sensitivity of the Estimates to Variation in Subpopulations We analyzed how the percentage share of states in the 3-year ACS estimates for 2006-2008 were affected by modifying the statutory criteria (3-21 years old, foreign born, and entered the United States after a particular year) to assess sensitivity of allocations to these criteria. We first limited the count to those aged 5-18 years old, an age range more similar to those reported to the states by local education authori- ties. We also examined the effect of limiting the count to either only those enrolled in school or only those enrolled in public schools. Table 6-7 compares allocations with various combinations of these modifica - tions to those under the base (statutory) criteria. Variations in age criteria did not influence the allocation of states very much (mean absolute difference, MAD, of 0.09%). For those aged 3-21, restricting the estimates to enrolled children and youth

OCR for page 133
141 COMPARABILITY OF ESTIMATES OF IMMIGRANT SCHOOL-AGE CHILDREN TABLE 6-4 Share of Immigrant Children and Youth Aged 3-21, by State (in percentage) ACS ACS ACS ACS ACS ACS State 2005 2006 2007 2008 2005-2007 2006-2008 Alabama 0.71 0.70 0.97 0.60 0.82 0.75 Alaska 0.09 0.17 0.25 0.12 0.20 0.18 Arizona 3.29 3.11 3.73 2.48 3.37 3.12 Arkansas 0.43 0.51 0.57 0.50 0.50 0.54 California 23.22 20.45 20.66 20.04 21.23 20.32 Colorado 1.56 1.28 1.34 1.43 1.39 1.36 Connecticut 0.99 1.08 1.10 1.13 1.08 1.13 Delaware 0.23 0.19 0.15 0.12 0.18 0.15 District of Columbia 0.12 0.39 0.14 0.18 0.22 0.22 Florida 8.64 8.48 8.26 7.79 8.37 8.25 Georgia 3.41 2.98 2.61 3.18 2.95 2.80 Hawaii 0.61 0.81 0.51 0.87 0.61 0.72 Idaho 0.46 0.34 0.41 0.49 0.44 0.43 Illinois 3.32 3.36 4.39 3.63 3.70 3.76 Indiana 1.11 1.07 1.06 1.02 1.08 1.02 Iowa 0.38 0.45 0.45 0.62 0.44 0.51 Kansas 0.56 0.67 0.68 0.71 0.63 0.65 Kentucky 0.49 0.53 0.63 0.64 0.60 0.59 Louisiana 0.29 0.48 0.59 0.55 0.45 0.54 Maine 0.09 0.16 0.11 0.17 0.13 0.17 Maryland 2.47 2.29 2.21 2.18 2.32 2.31 Massachusetts 2.21 2.09 2.42 2.93 2.31 2.46 Michigan 1.91 1.69 2.16 2.11 1.98 1.98 Minnesota 1.33 1.65 1.50 1.23 1.53 1.52 Mississippi 0.25 0.28 0.33 0.38 0.32 0.34 Missouri 0.68 0.70 0.80 1.00 0.78 0.84 Montana 0.04 0.12 0.14 0.09 0.10 0.13 Nebraska 0.38 0.46 0.28 0.43 0.38 0.39 Nevada 0.87 1.14 1.13 1.07 1.03 1.13 New Hampshire 0.11 0.25 0.16 0.16 0.19 0.19 New Jersey 3.57 3.68 3.42 4.21 3.45 3.78 New Mexico 0.53 0.90 0.39 0.26 0.57 0.54 New York 7.70 9.09 7.87 9.38 8.28 8.73 North Carolina 2.58 2.60 2.43 2.42 2.42 2.45 North Dakota 0.04 0.17 0.13 0.12 0.14 0.13 Ohio 1.25 1.12 1.25 1.83 1.28 1.40 Oklahoma 0.55 0.62 0.79 0.59 0.69 0.71 Oregon 1.01 1.19 0.94 1.10 1.07 1.09 Pennsylvania 1.49 1.75 2.11 2.18 1.83 2.02 Rhode Island 0.43 0.33 0.34 0.36 0.37 0.34 South Carolina 1.10 0.67 0.96 0.64 0.94 0.81 South Dakota 0.17 0.07 0.07 0.02 0.11 0.06 Tennessee 0.91 1.04 1.18 1.13 1.08 1.09 Texas 12.10 11.69 10.95 10.67 11.54 11.16 Utah 0.68 0.76 0.93 0.96 0.80 0.89 Vermont 0.06 0.08 0.08 0.11 0.07 0.08 Virginia 2.39 2.54 2.36 2.48 2.41 2.39 Washington 2.25 2.75 3.13 2.70 2.63 2.84 West Virginia 0.02 0.09 0.13 0.23 0.10 0.16 Wisconsin 0.81 0.87 0.75 0.70 0.81 0.75 Wyoming 0.10 0.12 0.07 0.02 0.09 0.08 SOURCE: U.S. Census Bureau Special Tabulations.

OCR for page 133
142 ALLOCATING FEDERAL FUNDS TABLE 6-5 Standard Errors of Shares of Immigrant Children and Youth Aged 3-21, by State (in percentage) ACS ACS State ACS 2005 ACS 2006 ACS 2007 ACS 2008 2005-2007 2006-2008 Alabama 0.08 0.08 0.14 0.10 0.05 0.06 Alaska 0.04 0.05 0.07 0.05 0.05 0.03 Arizona 0.26 0.27 0.31 0.27 0.19 0.16 Arkansas 0.09 0.08 0.10 0.10 0.05 0.06 California 0.76 0.55 0.57 0.68 0.39 0.34 Colorado 0.17 0.15 0.15 0.16 0.09 0.08 Connecticut 0.13 0.15 0.15 0.12 0.09 0.08 Delaware 0.05 0.07 0.05 0.05 0.03 0.04 District of Columbia 0.04 0.08 0.05 0.04 0.03 0.03 Florida 0.37 0.42 0.36 0.61 0.22 0.23 Georgia 0.27 0.20 0.21 0.29 0.13 0.13 Hawaii 0.14 0.12 0.08 0.25 0.07 0.09 Idaho 0.12 0.08 0.07 0.08 0.07 0.05 Illinois 0.24 0.26 0.28 0.27 0.15 0.15 Indiana 0.13 0.14 0.10 0.13 0.07 0.06 Iowa 0.07 0.07 0.08 0.09 0.04 0.05 Kansas 0.08 0.12 0.09 0.11 0.05 0.06 Kentucky 0.09 0.06 0.12 0.11 0.05 0.05 Louisiana 0.06 0.08 0.11 0.10 0.06 0.06 Maine 0.05 0.04 0.04 0.05 0.03 0.04 Maryland 0.24 0.17 0.19 0.15 0.12 0.13 Massachusetts 0.17 0.17 0.21 0.25 0.10 0.13 Michigan 0.21 0.19 0.20 0.21 0.11 0.09 Minnesota 0.14 0.19 0.15 0.13 0.09 0.10 Mississippi 0.06 0.06 0.05 0.10 0.04 0.04 Missouri 0.11 0.11 0.10 0.11 0.07 0.07 Montana 0.02 0.03 0.05 0.03 0.02 0.02 Nebraska 0.08 0.07 0.06 0.09 0.04 0.04 Nevada 0.11 0.13 0.15 0.16 0.07 0.09 New Hampshire 0.03 0.07 0.04 0.04 0.03 0.03 New Jersey 0.24 0.21 0.24 0.26 0.12 0.14 New Mexico 0.10 0.14 0.08 0.06 0.06 0.06 New York 0.34 0.40 0.36 0.39 0.22 0.26 North Carolina 0.33 0.24 0.21 0.21 0.13 0.12 North Dakota 0.02 0.05 0.03 0.05 0.03 0.02 Ohio 0.12 0.15 0.16 0.21 0.09 0.10 Oklahoma 0.09 0.10 0.11 0.10 0.06 0.07 Oregon 0.15 0.12 0.14 0.14 0.09 0.10 Pennsylvania 0.13 0.18 0.21 0.16 0.12 0.11 Rhode Island 0.11 0.07 0.08 0.07 0.05 0.05 South Carolina 0.13 0.09 0.12 0.09 0.08 0.07 South Dakota 0.08 0.02 0.02 0.02 0.03 0.01 Tennessee 0.12 0.11 0.13 0.19 0.08 0.08 Texas 0.50 0.48 0.48 0.47 0.29 0.29 Utah 0.11 0.10 0.11 0.17 0.08 0.08 Vermont 0.02 0.04 0.02 0.03 0.01 0.02 Virginia 0.21 0.19 0.16 0.18 0.12 0.11 Washington 0.18 0.29 0.26 0.23 0.12 0.15 West Virginia 0.01 0.02 0.03 0.06 0.02 0.03 Wisconsin 0.10 0.10 0.09 0.11 0.06 0.05 Wyoming 0.04 0.03 0.03 0.01 0.02 0.02 SOURCE: U.S. Census Bureau Special Tabulations.

OCR for page 133
143 COMPARABILITY OF ESTIMATES OF IMMIGRANT SCHOOL-AGE CHILDREN TABLE 6-6 Ratio of Immigrant Children Aged 5-18 Enrolled in Public School to All Children Aged 5-18 Enrolled in Public School (in percentage) ACS ACS State ACS 2005 ACS 2006 ACS 2007 ACS 2008 2005-2007 2006-2008 Alabama 0.32 0.29 0.38 0.28 0.36 0.35 Alaska 0.20 0.88 1.01 0.58 0.87 0.88 Arizona 1.56 1.40 1.46 0.81 1.51 1.22 Arkansas 0.43 0.49 0.59 0.09 0.52 0.42 California 1.81 1.45 1.45 1.27 1.57 1.39 Colorado 0.90 0.86 0.72 0.67 0.86 0.74 Connecticut 0.86 1.04 0.98 0.77 0.97 0.95 Delaware 0.85 1.08 0.63 0.50 0.82 0.71 District of Columbia 0.90 1.93 1.19 0.40 1.28 1.08 Florida 1.77 1.58 1.52 1.29 1.64 1.46 Georgia 0.91 0.68 0.59 0.76 0.74 0.68 Hawaii 2.68 3.11 1.53 2.31 2.37 2.29 Idaho 1.18 0.77 0.62 0.87 0.94 0.80 Illinois 0.78 0.69 0.90 0.68 0.80 0.75 Indiana 0.48 0.43 0.36 0.35 0.41 0.36 Iowa 0.43 0.39 0.37 0.49 0.38 0.44 Kansas 0.84 0.58 0.53 0.50 0.67 0.54 Kentucky 0.42 0.32 0.38 0.43 0.39 0.38 Louisiana 0.21 0.16 0.32 0.30 0.23 0.23 Maine 0.13 0.50 0.17 0.25 0.27 0.40 Maryland 1.76 1.40 1.36 1.18 1.50 1.36 Massachusetts 1.21 0.95 1.21 1.09 1.14 1.08 Michigan 0.62 0.57 0.67 0.62 0.62 0.61 Minnesota 0.98 1.15 0.93 0.65 1.04 0.96 Mississippi 0.09 0.18 0.09 0.35 0.16 0.26 Missouri 0.38 0.38 0.31 0.40 0.38 0.38 Montana 0.12 0.52 0.21 0.21 0.34 0.45 Nebraska 0.35 0.66 0.47 0.65 0.54 0.57 Nevada 0.87 0.89 0.96 1.20 0.92 1.03 New Hampshire 0.31 0.55 0.18 0.19 0.36 0.28 New Jersey 1.22 1.26 1.22 1.52 1.23 1.37 New Mexico 0.74 1.29 0.72 0.19 0.89 0.72 New York 1.42 1.45 1.26 1.35 1.37 1.34 North Carolina 0.99 0.65 0.74 0.57 0.78 0.68 North Dakota 0.12 0.56 0.36 0.63 0.40 0.45 Ohio 0.44 0.30 0.28 0.37 0.36 0.32 Oklahoma 0.28 0.38 0.60 0.28 0.43 0.45 Oregon 0.77 0.79 0.61 0.58 0.76 0.65 Pennsylvania 0.46 0.44 0.59 0.41 0.49 0.47 Rhode Island 1.00 1.31 1.41 0.79 1.29 1.20 South Carolina 0.87 0.31 0.52 0.29 0.61 0.43 South Dakota 0.99 0.31 0.15 0.00 0.52 0.15 Tennessee 0.36 0.36 0.41 0.41 0.40 0.42 Texas 1.54 1.26 1.20 1.01 1.35 1.17 Utah 0.56 0.54 0.55 0.65 0.59 0.59 Vermont 0.09 0.66 0.28 0.42 0.28 0.37 Virginia 0.99 0.86 0.99 0.80 0.94 0.84 Washington 1.12 1.48 1.65 1.10 1.41 1.41 West Virginia 0.04 0.14 0.09 0.47 0.08 0.23 Wisconsin 0.47 0.50 0.28 0.32 0.42 0.36 Wyoming 0.15 0.69 0.20 0.13 0.40 0.36 SOURCE: U.S. Census Bureau Special Tabulations.

OCR for page 133
150 ALLOCATING FEDERAL FUNDS COMPARISON OF ACS AND STATE ESTIMATES There are important differences in the source, methodology, and results between the estimates of recent immigrant students that come from the ACS and the counts provided by the local school districts through the states. The differences in source and methodology were summarized in the GAO report: see Table 6-9. Not surpris- ingly, the very different sources and methods used in the two allowable sources result in very different estimates of the recent immigrant student population by state. One way to depict the differences is to compute the ratio of the state student immigration counts with the ACS estimates of recent immigrant students. This com- parison is shown in Table 6-10. In 17 states, the state count was higher than the ACS estimate, while it was lower in the remaining 34 states and the District of Columbia. These relationships are illustrated graphically in Figure 6-2, which shows the ACS and state-reported counts for each state, together with the regression line through the origin (lower dashed line) and the line of equality (upper dashed line). Dotted lines around each state abbreviation represent 95 percent confidence inter - vals. Almost every state is below the line of equality, illustrating the generally lower TABLE 6-9 Key Features of ACS and State-Collected Data on Immigrant Children and Youth ACS Dataa State-Collected Datab Feature Measures Number of foreign-born persons aged 3 Number of (public and private school) Provided to 21 who arrived in the United States students in grades K-12 identified as within the 3 years prior to the survey recent immigrants How It Is Self-report (sample of population) States make determinations based on Measured student records or other information. Some states told us that they are not able to directly ask students questions related to their immigration status. Timing Annual average of monthly sample Varies Purpose To comply with Immigration Nationality To comply with the ESEA requirement Act and Public Health Service Act to assess progress of all limited requirements; to provide data to set and English proficient children, including evaluate immigration policies and laws. immigrant children and youth, to attain English proficiency (Department Work with Census to make sure Education collects this number from of ) Education’s appropriate questions are included. the states in the Consolidated State Role in Data Can propose new questions, if necessary. Performance Reports. Collection aThis column refers to data obtained by the U.S. Department of Education from ACS, but ACS collects additional data. bSome states may have data available for children prior to kindergarten. SOURCE: U.S. Government Accountability Office (2006b, p. 23).

OCR for page 133
151 COMPARABILITY OF ESTIMATES OF IMMIGRANT SCHOOL-AGE CHILDREN TABLE 6-10 Comparison of State Student Immigrant Counts and American Community Survey Estimates of Recent Immigrant Students ACS 3-Year Average, State 2007-2008 Ratio of ACS Estimate State 2006-2008 Estimate School Year Count to State Count Alabama 7,295 4,142 1.761 Alaska 1,800 880 2.045 Arizona 30,470 15,503 1.965 Arkansas 5,300 4,187 1.266 California 198,565 241,024 0.824 Colorado 13,305 12,940 1.028 Connecticut 11,040 13,571 0.813 Delaware 1,455 1,164 1.250 District of Columbia 2,165 993 2.180 Florida 80,605 142,333 0.566 Georgia 27,390 33,891 0.808 Hawaii 7,065 3,032 2.330 Idaho 4,195 3,188 1.316 Illinois 36,725 43,274 0.849 Indiana 9,980 11,763 0.848 Iowa 4,935 4,122 1.197 Kansas 6,340 11,206 0.566 Kentucky 5,785 7426 0.779 Louisiana 5,230 2,583 2.025 Maine 1,650 431 3.828 Maryland 22,530 16,617 1.356 Massachusetts 24,085 20,458 1.177 Michigan 19,370 11,052 1.753 Minnesota 14,895 15,985 0.932 Mississippi 3,315 6,007 0.552 Missouri 8,230 442 18.620 Montana 1,305 170 7.676 Nebraska 3,785 3,609 1.049 Nevada 11,055 14,694 0.752 New Hampshire 1,860 1,769 1.051 New Jersey 36,920 36,614 1.008 New Mexico 5,265 11,606 0.454 New York 85,340 98,797 0.864 North Carolina 23,915 23,365 1.024 North Dakota 1,300 497 2.616 Ohio 13,660 11,309 1.208 Oklahoma 6,920 4,954 1.397 Oregon 10,615 2,397 4.428 Pennsylvania 19,725 11,387 1.732 Rhode Island 3,350 2,903 1.154 South Carolina 7,910 6,415 1.233 South Dakota 540 197 2.741 continued

OCR for page 133
152 ALLOCATING FEDERAL FUNDS TABLE 6-10 Continued ACS 3-Year Average, State 2007-2008 Ratio of ACS Estimate State 2006-2008 Estimate School Year Count to State Count Tennessee 10,660 1,5815 0.674 Texas 109,105 93,627 1.165 Utah 8,685 7,935 1.095 Vermont 800 556 1.439 Virginia 23,380 29,284 0.798 Washington 27,800 15,142 1.836 West Virginia 1,520 1,599 0.951 Wisconsin 7,340 5,882 1.248 Wyoming 750 391 1.918 United States 977,220 1,029,128 0.950 SOURCE: ACS estimates from U.S. Census Bureau Special Tabulation. State counts from U.S. Depart - ment of Education’s EDEN Database. Immigration Rate from ACS 2006-2008 Immigration Rate from State Data FIGURE 6-2 Immigrant ratio from state counts (2007-2008 academic year) and ACS 3-year estimates (2006-2008). NOTES: The vertical dashed lines represent 95 percent confidence intervals for ACS esti - Figure 6-2 mates. The upper diagonal line is the line of equality; the lower diagonal line is proportional regression (regression through origin). R01924 bitmapped, axis labels replaced

OCR for page 133
153 COMPARABILITY OF ESTIMATES OF IMMIGRANT SCHOOL-AGE CHILDREN ACS estimates. States are scattered above and below the regression line, indicating the deviations from proportionality of ACS and state-reported estimates. However, only in some cases are these deviations from proportionality statistically significant (confidence interval does not cross regression line). TABLE 6-11 Comparison of Volatility in ACS Estimates of Youth Aged 5-18 and Enrolled in Public School and State Counts of Recent Immigrants (in percentage) Sum of Absolute Difference in Mean Absolute Relative Share of States Difference in Share of States ACS 2006 to 2007 11.68 31.66 All Large 3.83 8.98 Medium 5.32 21.85 Small 2.05 42.43 Minimum 0.48 73.09 ACS 2007 to 2008 15.97 38.08 All Large 8.17 21.12 Medium 4.78 24.22 Small 2.75 53.58 Minimum 0.27 73.27 ACS 2005-2007 to 2006-2008 6.16 14.65 All Large 3.53 6.34 Medium 1.71 8.22 Small 0.73 20.90 Minimum 0.20 34.01 State 2006-2007 to 2007-2008 9.76 26.62 All Large 3.60 6.62 Medium 4.63 19.05 Small 0.93 20.11 Minimum 0.59 99.37 State 2007-2008 to 2008-2009 35.45 44.35 All Large 22.92 33.61 Medium 10.28 47.39 Small 2.02 41.39 Minimum 0.23 57.20

OCR for page 133
154 ALLOCATING FEDERAL FUNDS Volatility of State and ACS Estimates As we did for the analysis of counts of ELL students (in Chapter 2), we report our assessment of changes in shares between consecutive years using several differ - ent measures of immigrant children, which are summarized in Table 6-11. Both the absolute and relative year-to-year changes in shares that are based on the 1-year ACS estimates are much greater than those that are based on the 3-year estimates. The 1-year estimates consequently are more volatile than the 3-year estimates, although they are also more responsive to year-to-year changes. In relative terms, the volatility of the ACS estimates increases monotonically from “large” to “small” states, as would be expected with diminished sample size, although in absolute terms the largest part of the share (and hence money moved) is in the “large” states. The relative changes in share between consecutive 1-year ACS estimates are quite large, in most cases averaging more than 10 percent for all but the “large” states. This observation, with the earlier analysis of reliability of changes—which showed that interyear changes in these estimates are largely noise—together sug - gest that the volatility of the 1-year estimates outweighs the value of their greater responsiveness. The interyear changes in shares based on state-provided data are surprisingly large, even though they are based on administrative data and therefore not subject to sampling error. This is especially notable from 2007-2008 to 2008-2009, when the shifts in share across states were even larger than those from single-year ACS estimates. (There seemed to be a substantial, but smaller, shift in single-year ACS estimates at about the same time.) This finding suggests that there might be substantial instability in the way child immigration counts are collected and reported by the states. CONCLUSION 6-1 Due to greater stability and insensitivity to poorly es- timated changes, the American Community Survey(ACS) 3-year estimates of immigrant children are statistically preferable to ACS 1-year estimates and more plausible at present than the state-provided counts. ACS Data and LEA Reports To better understand the relationship between ACS estimates and LEA-provided counts (through the states) of immigrant children, we performed an analysis of consistency between these measures across school districts within each state. The methods (correlations adjusted for sampling error and a hierarchical model), data (2006-2008 ACS), and inclusion criteria (districts with at least 20,000 population) are very similar to those presented in the parallel analysis of ELL rates in Chapter 5 and are not repeated here. Table 6-12 presents summary statistics by state. Rates of immigrant status by school district vary substantially, often dramatically, within each state. ACS esti - mates are almost always lower than LEA-provided estimates, but the ratio varies greatly from state to state. This table shows that the state immigration rates were

OCR for page 133
TABLE 6-12 Rates of Immigrant Children by Eligible School District State Counts State Immigrant Student Rates School Number of Overall Mean 20th 80th ACS Overall Ratio of Unadjusted State Enrollment LEAs Rate of LEAs Percentile Percentile Rate ACS/State Correlation Alaska 77,679 4 1.0 0.5 0.0 1.3 1.0 0.97 0.16 Arizona 834,896 64 1.7 1.5 0.1 3.0 1.4 0.82 0.50 Arkansas 217,450 28 1.6 1.1 0.1 1.7 0.6 0.38 0.60 California 5,491,668 411 4.1 3.9 2.0 5.6 1.4 0.33 0.51 Colorado 640,769 33 1.9 2.1 0.3 3.0 0.8 0.42 0.42 Connecticut 378,744 56 3.2 2.5 1.1 3.7 1.2 0.38 0.75 Delaware 102,396 13 1.1 0.9 0.5 1.1 0.7 0.62 0.09 District of Columbia 57,877 1 1.3 1.3 1.3 1.3 1.1 0.82 NA Florida 2,619,362 54 5.4 3.6 1.2 5.8 1.5 0.27 0.51 Georgia 1,412,950 91 2.3 1.2 0.3 1.8 0.7 0.32 0.66 Hawaii 179,897 1 1.7 1.7 1.7 1.7 2.3 1.36 NA Indiana 701,769 86 1.4 1.3 0.3 2.2 0.5 0.33 0.59 Iowa 210,815 27 1.3 1.0 0.1 1.9 0.6 0.46 0.39 Kansas 250,297 23 4.3 2.7 0.4 2.5 0.8 0.19 –0.05 Kentucky 371,406 38 1.9 0.8 0.1 0.7 0.6 0.30 0.45 Louisiana 566,824 43 0.4 0.3 0.1 0.4 0.3 0.58 0.40 Maine 26,010 6 1.5 1.4 0.0 3.3 1.8 1.18 0.92 Maryland 843,426 23 2.0 1.0 0.2 1.3 1.4 0.69 0.68 Massachusetts 590,965 97 3.2 2.4 0.2 4.2 1.4 0.45 0.64 Michigan 445,557 46 1.9 1.9 0.3 3.0 1.0 0.53 0.60 Mississippi 288,725 42 1.5 1.4 0.4 2.4 0.3 0.22 0.26 Missouri 94,131 9 0.4 0.5 0.2 0.7 0.8 1.74 –0.12 Montana 34,686 8 0.4 0.6 0.1 0.8 0.9 2.16 –0.63 Nebraska 169,074 15 1.7 1.4 0.4 2.5 0.8 0.47 0.27 155 continued

OCR for page 133
TABLE 6-12 Continued 156 State Counts State Immigrant Student Rates School Number of Overall Mean 20th 80th ACS Overall Ratio of Unadjusted State Enrollment LEAs Rate of LEAs Percentile Percentile Rate ACS/State Correlation Nevada 419,488 8 3.5 1.8 0.8 2.8 1.1 0.30 0.59 North Carolina 1,373,592 89 1.7 1.2 0.5 1.8 0.7 0.42 0.57 North Dakota 41,118 5 1.2 1.2 0.1 2.6 0.8 0.69 0.27 Oklahoma 274,584 23 1.5 1.0 0.3 1.8 0.7 0.46 0.33 Oregon 421,039 43 0.5 0.5 0.1 0.8 0.7 1.45 0.48 Pennsylvania 794,662 101 1.3 0.9 0.2 1.4 0.8 0.63 0.34 Rhode Island 112,975 19 2.2 1.1 0.2 1.5 1.4 0.63 0.72 South Carolina 632,088 49 1.0 0.8 0.2 1.3 0.5 0.47 0.54 Tennessee 806,634 61 2.0 0.9 0.1 1.5 0.5 0.23 0.30 Texas 3,616,573 183 2.4 1.7 0.6 2.5 1.3 0.56 0.60 Virginia 1,075,882 66 2.7 1.6 0.1 1.7 0.9 0.34 0.76 Washington 639,676 60 2.1 1.7 0.5 3.3 1.4 0.70 0.69 West Virginia 212,029 26 0.7 0.5 0.1 0.7 0.3 0.40 0.13 Wisconsin 413,481 49 1.1 0.8 0.3 1.2 0.5 0.43 0.36 NOTE: See text for discussion.

OCR for page 133
TABLE 6-13 Relationship at the School District Level Between ACS Estimates and State-Provided Estimates of the Rate of Immigrant Children Among Public School Enrollees, in Eligible Districts as Described in Text, for States with at Least 10 Eligible Districts Model with Intercept, State Data Rate No-Intercept Model Ratio of Intercept State Adjusted State Number Sigma Coefficient Coefficient Sigma Correlation Coefficient Sigma of LEAs Estimates Arizona 0.0071 0.1461 0.0035 0.6680 0.3562 0.0065 64 1.86 Arkansas 0.0028 0.1081 0.0004 0.9802 0.2268 0.0001 28 0.22 California 0.0033 0.1507 0.0036 0.7168 0.2217 0.0038 411 1.06 Colorado 0.0031 0.1506 0.0030 0.7115 0.2573 0.0032 33 1.05 Connecticut –0.0015 0.3513 0.0011 0.9862 0.3003 0.0010 56 0.91 Delaware 0.0034 0.0030 0.0000 0.3748 0.2115 0.0001 13 1.17 Florida 0.0023 0.1699 0.0039 0.8306 0.2076 0.0038 54 0.98 Georgia 0.0007 0.2040 0.0016 0.8647 0.2328 0.0016 91 0.97 Indiana 0.0019 0.1100 0.0008 0.8801 0.1969 0.0011 86 1.43 Iowa 0.0015 0.1848 0.0007 0.9543 0.2636 0.0006 27 0.75 Kansas 0.0058 –0.0015 0.0005 0.1722 0.0271 0.0047 23 8.73 Kentucky 0.0021 0.1749 0.0001 0.9993 0.2432 0.0001 38 1.08 Louisiana 0.0003 0.2898 0.0003 0.9591 0.3554 0.0001 43 0.52 Maryland 0.0015 0.4868 0.0031 0.8727 0.5511 0.0032 23 1.01 Massachusetts 0.0031 0.2607 0.0015 0.9825 0.3320 0.0001 97 0.09 Michigan 0.0028 0.2796 0.0025 0.9454 0.3704 0.0025 46 1.00 Mississippi 0.0023 0.0936 0.0001 0.9956 0.1955 0.0001 42 1.05 Nebraska 0.0000 0.3624 0.0003 0.9965 0.3657 0.0003 15 0.85 North Carolina –0.0003 0.2857 0.0013 0.9087 0.2686 0.0013 89 1.03 Oklahoma 0.0045 0.0144 0.0016 0.0836 0.2255 0.0028 23 1.76 Oregon 0.0023 0.8148 0.0015 0.9321 1.1533 0.0001 43 0.06 Pennsylvania 0.0041 0.0407 0.0020 0.1765 0.2971 0.0029 101 1.48 157 continued

OCR for page 133
TABLE 6-13 Continued 158 Model with Intercept, State Data Rate No-Intercept Model Ratio of Intercept State Adjusted State Number Sigma Coefficient Coefficient Sigma Correlation Coefficient Sigma of LEAs Estimates Rhode Island 0.0016 0.4022 0.0002 0.9997 0.4494 0.0002 19 1.01 South Carolina 0.0023 0.1685 0.0015 0.5744 0.3782 0.0018 49 1.19 Tennessee 0.0020 0.0785 0.0012 0.6575 0.1358 0.0004 61 0.36 Texas 0.0019 0.3480 0.0031 0.8401 0.4218 0.0031 183 0.98 Virginia 0.0006 0.2743 0.0015 0.9836 0.2933 0.0014 66 0.94 Washington 0.0009 0.3521 0.0023 0.9128 0.3911 0.0027 60 1.15 West Virginia 0.0025 0.0736 0.0001 0.9938 0.2416 0.0001 26 1.12 Wisconsin 0.0024 0.0619 0.0012 0.4047 0.1859 0.0013 49 1.09

OCR for page 133
159 COMPARABILITY OF ESTIMATES OF IMMIGRANT SCHOOL-AGE CHILDREN substantially different from the ACS rates in many states. In the most populous state, California, the state reports yielded an immigrant student estimate of 4.1 percent of all students in school year 2007-2008 while the ACS estimate was 1.4 percent. In most, but not all states, the state-reported rates were higher than the ACS rates. Table 6-12 also summarizes the strength of the association between ACS and state-provided rates within each state, corrected for overfitting due to sampling error. With some exceptions, these correlations tend to be generally quite high, with half of the 30 states included showing adjusted correlations of higher than 0.90. This finding suggests that the measures are usually fairly consistent within each state, holding constant the state procedures and to some extent the immigration patterns (to the extent that they are more consistent within than between states). However, the correlations are considerably weaker in some states, perhaps providing evidence of inconsistent collection of immigration data or of varying patterns of immigration that affect consistency of reporting. We also tested the relationship at the school district level between ACS estimates and state-provided estimates of rate of immigrant children among public school enrollees. As shown in Table 6-13, the results are mixed, with some states showing a very good consistency between the ACS and state-provided numbers for immi - grants, and other states showing a very weak relationship between the series. When compared with the results of this test for the ELL estimates and counts in Chapter 5, these findings suggest that there are perhaps systemic differences between the ACS and state-provided counts at the school district level. The results suggest the possibil- ity of less consistent procedures and criteria within many states than was observed with the within-state counts of ELL students, an indication that caution should be exercised in using the state-provided counts of immigrant children.

OCR for page 133