Overview

Although No Child Left Behind legislation does not require states to report enrollment, participation and performance data for English language learners (ELLs) with disabilities either online or in print-based reports, there is increasing interest in the academic success of this subgroup of students, estimated at about 357,325 nationwide (Zehler, Fleischman, Hopstock, Pendzick, & Stephenson, 2003). This interest is heightened by the fact that the current context of grade level standards-based reform has prioritized the educational success for all students, including those with limited English proficiency who also have disabilities.

In previous searches for state data on this subgroup of students, there have been little if any data reported. In a 2000 National Center on Educational Outcomes (NCEO) report for a Harvard Civil Rights Conference (Thurlow & Liu, 2001), only New Jersey was found to have reported information on ELLs with disabilities, though limited to participation data without performance. In this same report, a search for school year 1998-1999 data among the top five states ranked by percentage of ELLs nationwide, showed that none of these states reported data publicly at the state level. Even in a more recent update (Albus, Liu, & Thurlow, 2002), only three states reported participation data for ELLs with disabilities in a general state assessment. One of these states could also have reported performance data if the number of students tested had reached the minimum number required for maintaining confidentiality.

As the consequences associated with assessments increase and as the number of students who are English language learners with disabilities increase, it is likely that states will determine that it is important not only to keep track of their assessment data but also to publicly report those data. The purpose of this report is to examine the extent to which states report enrollment, participation rates in state administered tests, and the percent of those students who are proficient and above during 2002-2003. This was the first year that states were required to publicly report AYP data by the fall following the spring of the year during which the data were collected. This requirement caused a flurry of reporting activity in states. There are several factors that could potentially influence the data that states report. These factors also may influence our findings. The factors are:

• ELLs with disabilities are not defined the same way across states. The greatest differences in defining these students exist at the exiting and monitoring end of the service continuum. During spring 2003, federal officials proposed that students who had met a state’s English proficiency requirements could continue to be counted in the subgroup of limited English proficient students for two years after the proficiency determination. Nevertheless, the definitions of "proficient" in the upper levels still vary across states.

• ELLs with disabilities may be either over or under identified from state to state, which affects participation numbers. Although misidentification between two disability categories may also occur, this does not generally affect the numbers of ELLs identified as having a disability participating in large-scale assessments, except when looking at data for specific disability categories.

• Variability among state tests nationwide make performance comparisons of these students across states problematic. Therefore, we discourage across-state comparisons of performance. We present the data in order to highlight the educational needs of these students with respect to each state’s own standards.

• Participation rates may also vary across states due to differences in the time of year that student totals are counted for enrollment. For example, numbers enrolled at the time of testing would be more accurate than calculating participation rates using enrollment from a different point in the school year.

With these factors in mind, the information in this report is meant to focus needed attention on the participation and performance of English language learners with disabilities at the national level. It is also meant to promote discussion about specific concerns (e.g., appropriate minimum numbers of students able to be reported) within the larger conversation of how best to meet the educational needs of this diverse subgroup of students.

Method

Staff searched for data on ELLs with disabilities between September 17, 2003, and December 22, 2003. Because a few states had communicated during this time that their data would be updated on their Web sites in early January, actual collecting and updating of data continued until the end of January, 2004.

Charts were created to send to state assessment directors in order to check the accuracy of the data found on state Web sites. A copy of the e-mailed verification request is provided in Appendix A. The data chart attachment is included in Appendix B. The chart includes whether a state reported data on the enrollment, participation, or performance of ELLs with disabilities for any state tests.

These verification requests and chart attachments were sent to state assessment directors in early January in order to evaluate the accuracy of our collected data and to update any data not yet found on state Web sites. Sixty-one verification e-mails were sent out to the 50 states and additional unique states that receive special education funds (e.g., Bureau of Indian Affairs, Department of Defense, Guam, Mariana Islands, Marshall Islands, Micronesia, Puerto Rico). A total of 30 states responded to the request for verification. Of these, 29 responded that the data found were accurate, and 1 provided a link to data not previously found. One state that verified the accuracy of our data decided to post its state’s data in response, and three other states offered non-posted data to us for inclusion in our public data collection efforts. Seven of the original e-mails were undeliverable, and thus were followed by attempts to re-send verification requests with updated contact information. Two months after the first notification, a second request was sent. The final date for verifying or collecting data from states was February 29, 2004..

During the verification process, some states made comments in addition to whether the data found were accurate. These comments are also presented in the results section in a format that maintains the anonymity of the states from which the comments originated.

Participation and performance data for state tests are presented in this report by grade level ranges. The elementary, middle, and high school grades are those of 4^th, 8^th, and 10^th grade students respectively, and have been chosen to be representative of the educational levels because most states have content assessments in these grades. Where states do not have a test in one of these three grades, our practice was to use the next grade below if tested, and if not, to use the next grade above, if tested. For the complete data collected by state for each grade, see Appendix C.

Results

States Reporting Enrollment

Before presenting the results for those states that reported data for participation or performance of English language learners with disabilities, we want to first show those states that, at a minimum, had reported enrollment information for these students on their state Web sites. Enrollment information by grade is important because it provides the basis for figuring the participation rates for state assessments by grade. Table 1 shows the six states (Delaware, Iowa, Kentucky, Maine, Ohio, and Pennsylvania) that reported at least some data on state enrollment totals of ELLs with disabilities. Three of these states provided data for selected grades and three states provided overall enrollment totals. Dashes in the table show where no data were reported or provided.

Table 1.  States That Reported Enrollment of English Language Learners with Disabilities

Note: Dashes indicate no data reported or provided.

Of these six states, Kentucky reported the disability categories of students who were receiving ESL services (English as a Second Language). These totals were not reported by grade, but rather reported as state totals for each disability category. The numbers reported for Kentucky are presented in Table 2.

Table 2. Special Education Enrollment Reported by Students Receiving Services in English as a Second Language.  

Kentucky

 Dashes indicate no student in category.

^aMMD= Mild Mental Disability, FMD= Functional Mental Disability, HI = Hearing Impairment, S/L= Speech or Language Impairment, VI= Visual Impairment, EBD= Emotional/Behavioral Disorder, OI = Orthopedic Impairment, OHI= Other Health Impaired , SLD= Specific Learning Disability, D/B= Deaf-Blind, MD= Multiple Disabilities, AUT= Autistic, TBI= Traumatic Brain Injury, DD= Developmental Delay.

States Reporting Participation and Performance

Table 3 shows those states that reported participation or performance data for ELLs with disabilities by type of state test, grouped as "General" or "Other" State Assessments. General state assessments include assessments given to all students in the state for measuring reading, mathematics, science, social studies, writing, and other subjects. We include in this category the reporting of assessment data for other language versions of the general tests, such as Texas’ Spanish version of the TAKS assessment, and other language assessments, such as California’s SABE/2. Although a state may be listed as reporting data under "General State Assessment," this does not mean that the state reports data on all of its general state assessments. In the table, "Other State Assessments" includes alternate assessments for students receiving special education services, and state assessments designed for English language learners to measure language proficiency. Two states (California and Texas) provided data across all three types of state tests, general, special education alternate, and language proficiency assessments.

Table 3. States Reporting Assessment Data for ELLs with Disabilities By Type of Assessment

 ¹ Indicates non-identical version of general state test in another language.

² Indicates version of general state test in another language.

 Table 4 also presents this same information, but for those states that did not have the data publicly posted online. These states, which submitted the data directly to NCEO, were Alaska, Pennsylvania, and Wyoming. Alaska provided participation and performance information on its general state assessment and its special education alternate test. Pennsylvania and Wyoming provided these data for their general state assessments.

The following maps (Figures 1-3) show the regular states, and unique states that reported data for English language learners with disabilities by type of information reported (e.g., participation or performance) and type of assessment reported.

Table 4. States That Provided Assessment Data for ELLs with Disabilities to NCEO Though Not Reported on Web Site

Figure 1. States That Reported Participation and Performance Data for ELLs with Disabilities on at Least One General Assessment

Figure 2. States That Reported Participation and Performance Data for ELLs with Disabilities on State Special Education Alternate

Figure 3. States That Reported Participation and Performance Data for ELLs with Disabilities on a Language Proficiency Test

Several states also reported additional information in unique ways. These are described for California, Colorado, and Delaware.A

California: For the California alternate assessment for special education students, the state reports by English language learner status with additional disaggregation of test performance by type of language service provided. The state reported data by general ELD (students receiving English learner services only), ELD with SDAIE (English learners receiving specially designed academic instruction in English with a minimum of two academic subjects taught), and ELD with SDAIE with native language support.

Colorado: For the Spanish version of the Colorado Reading and Writing tests, the state reported additional information by disability category and accommodations. More specifically, for the Lectura and Escritura tests for 3^rd and 4^th grades the state reported the number tested and performance by grade, for each disability category. The reports also provided the number of students who used a certain accommodation and how those students performed. However, it is not clear whether student performance was reported for students who may have used more than one accommodation on the test, or whether they were reported once by a primary accommodation if they had used more than one. Also, because the accommodation data do not separate ELLs from ELLs with disabilities in the reporting of accommodated performance, it is not clear how these data were reported for students who may have used accommodations for both language proficiency and disability related needs.

Delaware: Delaware’s Web site allows users to generate reports for selected criteria in an interactive format. Although it is possible to generate reports for any general state assessments for ELLs with disabilities, the numbers in the state are so small by grade tested that often the numbers are too small to report for privacy reasons. We note that of all grades and tests reported in 2002-03, only 8^th grade Science and Social Studies had the required minimum number (i.e.,15 for Delaware) of students to be able to show performance data.

Participation and Performance for General State Assessments

Participation and performance data for the general state assessments are presented by content area (i.e., reading, mathematics, science, social studies, and writing) including participation data, participation rates if available, and performance.

Nine states reported data for ELLs with disabilities on at least one content area for a general state assessment (see Table 5). Of these, all nine states reported on reading and mathematics, three states reported on science, one reported on social studies, and eight states reported on writing. Only six states reported on general state assessments in English. The other three states (California, Colorado and Texas) reported on Spanish language tests across content areas.

Table 5.  States with Participation and Performance Data on General State Assessments by Content Area and by Language of Assessment

Note: This table includes data from three states that provided data to us; the data were not publicly available online (Alaska, Pennsylvania, & Wyoming).

Numbers Tested on General State Assessments

Table 6 presents the numbers of students tested for five content areas by the elementary grade level (i.e., grade 4). Nine states had data. For reading and mathematics, California had the most ELLs with disabilities tested (N=610 on both). In the other states, numbers ranged from 374 to below 15 tested for reading and mathematics. Delaware had too few students to report for any of the content areas at the elementary level. For science, Ohio and Texas reported between 148 and 229 students. Only one state, Ohio, reported numbers tested for social studies. Eight states reported numbers tested for writing; these numbers were comparable within states to those reported for reading and mathematics. Because Minnesota reported on writing numbers in four separate categories (i.e., descriptive, narrative, problem solution, and clarification), we decided to report the average of the numbers tested for the participation number. An asterisk in the table indicates there were too few students for a state to report data publicly.

We note here also a reminder that data reported for elementary, middle, and high school grade levels are those of 4^th, 8^th, and 10^th grade students respectively. Where states did not have a test in one of these three grades, our practice was to use the next grade below if tested, and if not, to use the next grade above, if tested.

Table 6. Numbers of English Language Learners with Disabilities Tested in Content Areas at Elementary Grade Level¹

 
* Indicates too few to report according to state policy.
¹Elementary grade level, as noted in the methods section, means 4^th grade tests. If a state did not have a 4^th grade test, a 3^rd grade or 5^th grade test was chosen as representative for elementary grades.

 At the middle school level (see Table 7), more states were unable to report numbers tested because there were too few students tested to report publicly. In those states where data were reported, the numbers in some states were similar to those at the elementary school level (e.g., Alaska, Minnesota, Ohio), but in others these were different (e.g., California’s numbers were much lower). 

 Table 7. Numbers of English Language Learners with Disabilities Tested in Content Areas at Middle School Level¹

 
* Indicates too few to report according to state policy.
¹ Middle School level, as noted in the methods section, means 8^th grade tests. If a state did not have an 8^th grade test, a 7^th grade or 9^th grade test was chosen as representative for middle school grades.

At the high school level (see Table 8), five states had enough students to report for reading and mathematics. Three states could have reported science participation, and two states social studies participation, but the numbers for both of these subjects were too small to report according to state policy. The numbers reported by state for writing were comparable to those reported for reading and mathematics.

Table 8. Numbers of English Language Learners with Disabilities Tested in Content Areas at High School Level¹

 
* Indicates too few to report according to state policy.
¹ High School level, as noted in the methods section, means 10^th grade tests. If a state did not have a 10^th grade test, a 9^th grade or 11^th grade test was chosen as representative for high school grades.

Participation Rates on General State Assessments

Of the states that reported data on the number of ELLs with disabilities tested on general state assessments, Alaska, Ohio, and Pennsylvania were three that also reported on the number of ELLs with disabilities enrolled by grade. This made it possible for participation rates to be calculated for the assessment. Recall, however, that Alaska and Pennsylvania provided data to us after our requests; these were not data that were publicly reported.

Figure 4 shows participation rates at the elementary level in the three states for the content areas where data were available. Rates were high, with between 97% and 99% assessed at the elementary level across content areas. Figure 5 presents participation rates for middle school grades in the three states. Percentages again were quite high (93% to 107%). Ohio showed a percentage over 100%. The state explained that this was probably due to students being enrolled in one grade but taking the test in another. For example, the 9^th grade test, which is required for graduation, can be taken beginning in grade 8 through grade 12, although the majority of students participate as 9^th graders. (J. Dannemiller, personal communication, May 19, 2004).

High school participation rates are shown in Figure 6. Only Alaska and Pennsylvania had data at this level, and they had data only for reading, mathematics, and writing. Participation percentages at the high school level were slightly lower, from 80% to 91%.

Figure 4. Elementary Participation Rates Across Content Areas for ELLs with Disabilities

Figure 5. Middle School Participation Rates Across Content Areas for ELLs with Disabilities

Figure 6. High School Participation Rates Across Content Areas for ELLs with Disabilities
 

Performance on General State Assessments

Performance figures are presented for each subject area, showing states and school level of data. Recall that these figures should not be used for across-state comparisons of performance. Student populations and assessments used to assess skills vary considerably from state to state. We remind the reader that data reported for elementary, middle, and high school grade levels are those of 4^th, 8^th, and 10^th grade students respectively. Where states did not have a test in one of these three grades, our practice was to use the next grade below if tested, and if not, to use the next grade above, if tested.

In reading, at the elementary level, between 7% and 64% of ELLs with disabilities were considered proficient or above on reading/English language arts (see Figure 7). Yet only one of these states had data showing 50% or more of its ELLs with disabilities at proficient or above. At the middle school level it was only slightly better. There were 5% to 81% of ELLs with disabilities at proficient or above, with just two states with 50% or more at the proficient level or above. At the high school level, the three states reporting data showed between 0% and 38% of ELLs with disabilities as proficient or above. None of these states had its ELLs with disabilities reaching this 50% proficiency level. States without data plotted indicate no data were reported for a test in that grade level, except in the case of Wyoming, which did have tests in all three levels but no percentage of students at proficient or above. An asterisk (i.e., Delaware) indicates numbers too small for a state to report publicly according to its policy.

 Figure 7. Performance on General Reading/English Language Arts Assessments for ELLs with Disabilities

In mathematics, at the elementary level, between 14% and 64% of ELLs with disabilities were considered proficient or above across the six states that had data (see Figure 8). Two of these states had 50% or more of ELLs with disabilities performing at proficient or above. At the middle school level, 4% to 48% of these students were performing at proficient or above. Of these states, only one state had 50% or more of its ELLs with disabilities performing at proficient or above. At the high school level, the three states reporting data showed between 0% and 13% as proficient or above. None of these states had 50% or more of its ELLs with disabilities performing at proficient or above. States without data plotted indicate no data were reported for a test in that grade level, except in the case of Wyoming, which had tests in all three levels but no percentage of students at proficient or above.

 Figure 8. Performance on General Mathematics Assessments for ELLs with Disabilities

In science, three states reported data (see Figure 9). These were Delaware, Ohio and Texas. Between 15% and 50% of the English language learners with disabilities were reported as  performing at proficient or above at the elementary level. At the middle school level, only two states had data. These states, Delaware and Ohio, showed between 7% and 36% performing at proficient or above. No states at this grade level had at least 50% of its ELLs with disabilities performing at proficient or above on a general science assessment. There were no data reported at the high school level. Delaware had enough students for this test in the middle school grade to report data, but not for other school levels.

Figure 9. Performance on General Science Assessments for ELLs with Disabilities

Two states, Delaware and Ohio, reported performance on social studies (see Figure 10). Delaware reported 6% as proficient or above at the middle school level. Ohio reported 50% proficient or above at the elementary and 43% proficient or above at the middle school levels. No data were reported at the elementary or high school level for Delaware because there were too few students to publicly report. These data show that only one state had at least 50% proficient at the elementary level, and no states reported 50% or more of its ELLs with disabilities at proficient or above in middle or high school levels. Ohio did not report scores for a high school test.

Figure 10. Performance on General Social Studies Assessments for ELLs with Disabilities

For writing, it is noticeable that of the 8 states that reported performance on a state writing test, the majority reported performance below 50% proficient across all education levels. One state, Texas, reported performance above 50% for ELLs with disabilities at the elementary level (see Figure 11).  However, none of the states with middle school data (4 states) or high school data (3 states) had at least 50% of their English language learners with disabilities performing at proficient or above on a general writing assessment. Delaware, again, had too few students to report publicly.

 Figure 11. Performance on General Writing Assessments for ELLs with Disabilities

State Alternate Assessments

Table 9 shows that four states reported data on ELLs with disabilities on the state alternate assessments developed for students receiving special education services. All four states, Alaska, California, Delaware, and Maryland, reported on reading and mathematics. No state separated its reporting for English and other language (e.g., Spanish) versions of alternate assessments.

Table 9.  States That Reported Participation and Performance Data for ELLs with Disabilities on a Special Education Alternate by Content Area

 ¹Reading includes English Language Arts, if there are no distinct reading and writing tests.

²Reports Performance, but state reports show a symbol to indicate too few students to report actual data for confidentiality reasons.

³Reports on Skills for Healthy Living.

Numbers Tested on State Alternate Assessments

Four states, Alaska, California, Delaware and Maryland, reported data on state alternate reading and mathematics assessments. Across grade levels, California reported the highest numbers of ELLs with disabilities tested with 129 at the elementary level, 164 at the middle school level, and 94 at the high school level. Alaska, Delaware, and Maryland had too few students in this subgroup to report. We again note that data reported for elementary, middle, and high school grade levels are those of 4^th, 8^th, and 10^th grade students respectively. Where states did not have a test in one of these three grades, our practice was to use the next grade below if tested, and if not, to use the next grade above, if tested.

Table 10. Numbers of ELLs with Disabilities Assessed on Alternate Assessments by Content Area Across Grade Levels

* Indicates too few to report according to state policy.

Only two states, Alaska and Delaware, reported data on alternate assessments for other content areas (see Table 9). However, the numbers of students tested for each state were too small to report.

Participation Rates on State Alternate Assessments

Of the states that reported data on state alternate assessments for ELLs with disabilities, only Alaska reported the numbers enrolled by grade so that participation rates could be calculated. Figure 12 shows these rates by school level and reveal that the levels of participation vary considerably between the elementary and middle school levels. The participation rates for reading and mathematics are both at 100% at the elementary level, and 40% at the middle school level. There were no participation rate data reported at the high school level in Alaska.

Figure 12. Elementary and Middle School Participation Rates Across Content Areas for Alaska for ELLs with Disabilities

Performance on State Alternate Tests

The next figures show the performance data reported for state alternate assessments. We note that California reported data for its alternate in two groups, those in instructional level 1 defined as students between the ages of 7 and 16 (grades 2-11) "with profound disabilities and functioning developmentally at or below 24 months," (California Department of Education, 2004) and a second group in levels 2-5 with less severe disabilities. The data set in these figures represent the second group. Figures in this section do not include the data for the "other" content area reported for Alaska. These additional data are provided in Appendix C. 

Performance for reading by school level on alternate assessments is presented in Figure 13.  Four states had a place designated for reporting performance data. In Alaska and Delaware, the numbers of students were too few to publicly report performance data. California reported data for all three school levels. At these three levels, a range of 35% to 46% of the students was at proficient or above. Maryland reported data at the elementary level only, with 100% of its ELLs with disabilities at proficient or above.

 Figure 13. Reading Performance on Alternate Assessments for ELLs with Disabilities

The pattern of Mathematics performance on state alternate assessments is similar to that of reading (see Figure 14). Alaska and Delaware, again, had too few numbers to publicly report performance. In California, 59% of elementary level ELLs with disabilities were proficient or above, with slightly fewer proficient or above in the middle (28%) and high school (31%) grades than in reading. In Maryland, 72% of those students at the elementary level were proficient or above.

 Figure 14. Mathematics Performance on Alternate Assessments for ELLs with Disabilities

State Alternative Assessments

One state reported data on an alternative assessment designed for students with disabilities (see Table 11). This state, Texas, reported on reading, mathematics and writing. It is reported separately here from other state alternate assessments because Texas has not considered this test to be the state’s alternate assessment developed for the small percentage of students with significant cognitive disabilities (Thurlow, Wiley, & Bielinski, 2003).

Table 11.  States That Reported Participation or Performance Data for ELLs with Disabilities on a Special Education Alternative Assessment by Content Area

 ¹Reading includes English Language Arts, if there are no distinct reading and writing tests.

Numbers Tested on State Alternative Assessment

The numbers tested on the Texas alternative test (SDAA), are noticeably higher than the number of ELLs with disabilities on other state assessments, with between 3,200 and 3,800 tested at the elementary grade and middle school grade levels for reading and mathematics (see Table 12). For writing, the range increases to 4,800 to 5,700 students tested in the elementary and middle school levels. There were no data reported at the high school level. Participation rates for the alternative tests were not available, nor calculable.

As noted previously, data reported for elementary, middle, and high school grade levels are those of 4^th, 8^th, and 10^th grade students respectively. Where states did not have a test in one of these three grades, our practice was to use the next grade below if tested, and if not, to use the next grade above, if tested.

Table 12. Numbers of ELLs with Disabilities Tested on Texas Alternative Tests in Reading, Mathematics, and Writing by Grade Level¹

 ¹ All numbers for high school level were too few to report publicly.

Performance on State Alternative Tests

Performance on the SDAA assessment in Texas showed a relatively high percentage of students achieving proficient status as determined by the ARD committees. At the elementary level, 87% of students tested in reading, 89% tested in mathematics, and 83% tested in writing were proficient (see Figure 15). At the middle school level, the numbers were slightly lower, with 81% of students in reading considered proficient, 66% in mathematics, and 67% in writing.

 Figure 15. Performance on Alternative Tests in Texas for ELLs with Disabilities

Language Proficiency Tests

Only two states, California and Texas, reported data for ELLs with disabilities on assessments designed to measure language proficiency of ELLs (see Table 13). Because these tests are a measure of English, states do not have alternate forms in other languages (In the past, some states, such as Colorado, reported the number of students not able to take their Spanish version tests because of limited proficiency in Spanish). States did not report performance results for Spanish proficiency tests.

Table 13.  States That Reported Participation or Performance for ELLs with Disabilities on a Language Proficiency Test by Content Area

 The reported participation and performance data on English language proficiency tests for California and Texas are presented in Table 14 and Figure 16. Texas reports scores for the Reading Proficiency Test in English which combines performance on reading and writing. Therefore, there is only one percentage per grade for this test for students performing at "intermediate" or "advanced" levels. Texas did not report the number of these students enrolled, so the percent tested cannot be reported here. California reports on the California English Language Development Test (CELDT) which includes reading, writing, and listening/speaking. Although the report includes the mean scale scores for these separate areas, the overall proficiency levels are reported as a composite of all of these skills.

Numbers Tested on State English Language Proficiency Tests

The numbers of ELLs with disabilities reported as taking a state language proficiency test are the largest of all types of tests in this report. Between Texas and California, the numbers of students tested ranged from 1,180 at the high school level for the Texas RPTE to 11,455 at the elementary level for California’s CELDT test. We again note that data reported for elementary, middle, and high school grade levels are those of 4^th, 8^th, and 10^th grade students respectively. Where states did not have a test in one of these three grades, our practice was to use the next grade below if tested, and if not, to use the next grade above, if tested.

Both states show the similar pattern of higher numbers of students tested in the elementary grade with fewer students tested in the middle and high school years. For Texas, the numbers of ELLs with disabilities tested were 5,280 at the elementary level, 2,389 at the middle school level, and 1,180 at the high school level (see Table 14). For California, the number of students tested was higher overall with 11,455 tested at the elementary level, 8,472 at the middle school level, and 6,048 at the high school level. The participation rates for students taking English language proficiency tests were not available, and could not be calculated for either state.

Table 14. Numbers of ELLs with Disabilities Tested on Language Proficiency Assessments by Grade Level

Performance on State English Language Proficiency Tests

Figure 16 presents the performance data for ELLs with disabilities who took the state English language proficiency tests for Texas (RPTE) and California (CELDT). Overall, all students across elementary, middle, and high school grades were below 40% proficient. Although elementary grades had the most numbers tested, the performance of these students was from 13% to 18% proficient (see Figure 16). In the middle school grades proficiency rates were from 29% to 32%. The students at the high school level had proficient rates from 35% to 38% proficient. Therefore, both states showed more students proficient in the high school grades than in the elementary or middle school grades.

Figure 16. Performance on English Language Proficiency Tests for ELLs with Disabilities

Information on Students Not Tested

Three states, California, Colorado and Texas, reported data on the number of ELLs with disabilities not taking a given test, or other information that indicates that not all students are included in a report category due to missing student information about either having disabilities or being an English language learner. The following are descriptions of the information provided by each of these three states.

California: The report for the SABE/2 for students receiving special education services mentioned that off-level test takers were not included in the state’s report. However, the report also mentioned that students who used special accommodations were included in the data. Similar information was not presented for the other state tests that reported data for ELLs with disabilities. For the CELDT test, there was exemption information reported for the Listening/Speaking component, but this was for students who had already passed this portion of the test previously and who did not need to retake that portion.

Texas: For the State Developed Alternate Assessment (SDAA), numbers were reported for students who did not have information about English proficiency status (e.g., unknown status), therefore there may have been more ELLs with disabilities tested but reported separately under this other reporting column due to incomplete student information. Information was also provided regarding students who were not tested for: "Students Absent," "Students Exempt: ARD," "Students Exempt: LEP," and "Other Students Not Tested." Because the SDAA is an alternative assessment for students receiving special education services, information on ELLs with disabilities being exempted would be clearly reported for the "Students Exempt: LEP" category. The other exemption categories are general, and may or may not include ELLs with disabilities.

Similar to the SDAA test, the Texas language proficiency test (RPTE) report included the number of students who had incomplete information about English proficiency and special education status. Again, there may have been more ELLs with disabilities tested but reportedly separately. Also, the report summarizes categories of those not tested: "Students Absent," "Students Exempt: ARD," "Other Students Not Tested," and "Not Tested: 2^nd Semester Immigrants Non-English Readers." The category "Students Exempt: ARD" clearly reports those ELLs exempted from the language proficiency test based on ARD, a special education decision. The other general "not tested" categories may or may not include ELLs with disabilities.

Colorado: This state had detailed information on students with no scores by category:

"Test deferred due to language, Taking CSAP Alt. Assessment, Eligible to take CSAP Alt. Assess. N/A, Parental refusal, Test not completed, Student withdrew before completion, Extreme frustration, Nonapproved accom/modification, and Test is invalid: incomplete."

Although some of these categories would clearly pertain to ELLs with disabilities such as those not tested because they took the CSAP Alternate, other categories of those not tested such as parental refusal may or may not include ELLs with disabilities.

Issues Raised by States

During the verification process, some states responded to our request for verification with additional comments. These comments included issues such as noting that federal and state laws do not require reporting on ELLs with disabilities, the expressed ability or inability to provide this information if requested at the state level, the fact that LEAs may analyze this information but it is not reported at that level of detail at the state level, and the issue of how a small population size of ELLs with disabilities in a state affects the usefulness of public reporting due to the need to maintain privacy of students by not reporting their performance when numbers fall below a state specified minimum. Another state offered, in response to our query, to post the data on its Web site after sending a completed chart of the data for our verification request. Although this was not the intent of the verification process, some states clearly were interested in demonstrating that they had the data when they did have them. Some state comments, with identifiers removed, are included here:

"We currently don't post the data as you are looking for, however we collect the data and post it as required by state and federal law. We collect data from testing companies and from districts by individual student ID, therefore we can provide the data you are asking for."

"If your question refers to standard public reports on the performance of students who belong to both categories (ELL and Students with Disabilities), your data is correct: [state] does not report to that level of detail. The data is provided to LEAs at an individual student level with fields that identify both categories. Therefore they could easily analyze the data to that level."

"[State] does not disaggregate the LEP group by those with disabilities; however, our data system allows us to do so, if requested.  We have never received such a request."

"Currently we are not capable of reporting by this category, but we have asked our testing company to provide additional disaggregated data at the state, … and district level that will provide this information. We should have these data electronically in early summer, but the method of disseminating the information has not been determined."

"I am not sure that we disaggregated data even at the state level, although I think we could at this level. I have not seen the full context within which this data will be presented, but it seems very important to point out the small numbers of students with limited English proficiency in the [state] public school system. In most schools and districts, the data element would be left blank because of small numbers." 

Summary

Overall, most states are not reporting participation and performance data on English language learners with disabilities. With the growing population of these students, information on their participation and performance is that more important. We recognize that there are a few states that are reporting state totals of these students and at least one by specific categories of disability. Yet, still very few are reporting these students’ enrollment, number tested, and performance on state assessments by grade.

Unique states were included because they receive special education funds. While they have not been required to disaggregate the performance of ELLs with disabilities for either IDEA or NCLB, it was possible that they would do so. It is important to recognize at the same time that certain issues related to reporting on ELLs with disabilities are different depending on the context because the official language of a territory may be Spanish, or offical bilingualism such as English and Spanish. Although communication with such territories as Puerto Rico was attempted regarding the potential of such students disaggregated in their reports, the queries did not receive a response.

Overall performance across assessment types for ELLs with disabilities was fairly low, with only a few states showing 50% or more of these students performing at proficient or above. General assessments showed that only two states, of the nine states with reading data and eight states with math data, had 50% or more students performing at proficient or above for at least one grade. Across all of the general content area assessments, elementary school students were more likely scoring at 50% or above than middle or high school students, most likely because of increased difficulty in academic content in later grades. For state alternates, one state out of four with data had 50% or more students at proficient or above in reading and mathematics. The Texas alternative assessment showed over 50% of all grade levels across all content areas as meeting the ARD goals as set by special education Admission, Review and Dismissal committees. Finally, language proficiency test performance showed no states with 50% proficient or above. Across the two states, performance ranged from under 20% for elementary, to just under 40% for high school level students. We note that student populations that take state language proficiency tests change due to the function of the test in assessing students’ abilities to an acceptable cut-off point. Thus, students scoring higher on the proficiency test are more likely to be reclassified as language proficient or monitor status. This may result in lower overall percentages of students scoring higher on English language proficiency tests over time.

Another observation, focused on the representation of content areas in the reported data, shows that about half the number of states that reported data on general tests also reported on other content such as science and social studies. States reporting on alternates and alternative tests were more limited to the content areas of reading and mathematics. This was most likely due to states’ responding to requirements of content test development along legislative timelines. Language proficiency assessments were not as affected by this, due to the inherent focus of the tests on reading and writing as language skills, though these tests may in fact incorporate other content topics such as science or social studies in the texts used to measure language proficiency.

States vary in the number of ELLs in their student populations. States with very few ELLs have an even smaller number who have disabilities. This raises both the issue of student confidentiality in the reporting of data and the question of what minimum number is significant enough to report. Although larger states do not face this issue, they face their own unique challenges of tracking potential thousands of these students across several types of state assessments.

Of the states that reported numbers of ELLs with disabilities tested on state assessments, those states with larger populations of these students reported data on special education alternate tests, alternative tests, and language proficiency assessments rather than general state assessments. Further, the data that were reported for general state assessments, were only from the Spanish versions. In contrast, states with smaller populations of these students reported their participation and performance data on general state assessments in English, and not on state alternate assessments (except for Massachusetts) or assessments of English language proficiency. This may be due to the fact that states with smaller populations of these students may have even fewer students taking the alternate, resulting in states not being able to report the performance of students in order to maintain confidentiality. This rationale would not apply similarly to explain why fewer states reported data for these students on language proficiency tests.

Although we use the term participation rates to describe levels of participation in testing, we acknowledge that these may vary based on timing of student counts. Variability in how and when students are counted and reported for enrollment by grade affects the ultimate calculation of participation rates, as do individual state practices of deciding to include the numbers of students tested out of their current grade level in grade level totals. Even though this is the reality of the numbers we are dealing with, we can observe from the data that although some states have very small numbers of ELLs with disabilities reported for participation in state assessments, the participation rates are fairly high for those that reported them. In contrast, states in this study with larger numbers of these students did not have information on participation rates available.

Issues raised by states during the verification process shared a common theme that states with smaller numbers of ELLs have much smaller populations of ELLs with disabilities. Further, the students’ rights to confidentiality should not be compromised. These are understandable arguments for why states with very small populations of these students do not report these data. Yet, it is worth noting that at least a few states did report on the participation of students, even though the numbers were too low to report student performance. This reflects the variability among states in how they have set minimum numbers for reporting. Some states have a minimum of 5 to 10 students, whereas others may be 25 to 30, depending on the purpose of reporting (e.g., general testing or accountability). Because this was not a primary focus on this report, we suggest that further research address states’ rationales for setting the minimum number of students acceptable for reporting for these two purposes.

In conclusion, because this report is only able to include the data that states are currently reporting for this sub-population, only a small fraction of the total estimated number of English language learners with disabilities nationwide is actually presented. Regardless of how states decide to report data for these students, it is important that they use these data in planning educational reforms. English language learners who have one or more disabilities are among those students with the greatest challenges in achieving proficiency and realizing high standards, and should be receiving specific attention from educators. As it is crucial to understand how these students as a group are participating and performing on assessments, more states will need to report these data in order for a more accurate picture to take shape. This information is especially needed in this era of standards-based reform, to better ensure that students are receiving the instruction and services needed to attain grade-level standards.

References

Albus, D., Liu, K., & Thurlow, M. (2002, April). Large-scale assessment participation and performance of limited English students with disabilities: What public data reports tell us. Round table presentation at the Annual Meeting of the American Educational Research Association, New Orleans, LA.

California Department of Education. (2004, January 29). Text from January 29^th CAPA training of trainers. [Webcast script]. California Department of Education. Retrieved May 17, 2004, from http://www.cde.ca.gov/sp/se/sr/webcst0129.asp

Thurlow, M., Wiley, H. I., & Bielinski, J.  (2003). Going public: What 2000-2001 reports tell us about the performance of students with disabilities (Technical Report 35). Minneapolis, MN: University of Minnesota, National Center on Educational Outcomes.

Thurlow, M., & Liu, K. (2001). State and district assessments as an avenue to equity and excellence for English language learners with disabilities (LEP Projects Report 2). Minneapolis, MN: University of Minnesota, National Center on Educational Outcomes.

Zehler, A., Fleischman, H., Hopstock, P., Pendzick, M., & Stephenson, T. (2003). Descriptive study of services to LEP students and LEP students with disabilities (Special Topic Report #4). Arlington, VA: Development Associates, Inc.

Appendix A

Verification Letter E-mailed to States

Dear                    

The National Center on Educational Outcomes will produce a report on how states report results for English language learners with disabilities. Our goal is to determine whether and the extent to which each state reports disaggregated test results for limited English proficient students/English language learners (ELLs) who are receiving special education services.

Please take a moment to verify the accuracy of the attached table reflecting your state’s online data for English language learners with disabilities for 2002-03.

•     If ALL of the information is accurate, please send an email verifying that this is so.

•     If any information is inaccurate, please provide us with the Web address containing the data. The data in the table is for 2002-2003.

Thank you,

Deb Albus,

 Research Fellow, LEP/ELL issues
National Center on Educational Outcomes

Appendix B

Verification Attachment

 Data on English Language Learners (ELLs) with Disabilities

 The National Center on Educational Outcomes will produce a report on how states report results for English language learners with disabilities. Our goal is to determine whether and the extent to which each state reports disaggregated test results for students with limited English proficiency who are receiving special education services.

 PLEASE VERIFY THE ACCURACY OF THIS TABLE:

•     If ALL of the information is accurate, please send an email verifying that this is so.

•     If any information is inaccurate, please provide us with the Web address containing the data. The data in the table is for 2002-2003.

Please address communications to Deb Albus at albus001@umn.edu

sample

--- indicates no information on ELLs with disabilities

Appendix C

Complete Participation and Performance Data by State and Assessment Type

The data presented in Tables C1-C12 are to give an idea of how these students are doing within each state, and to compare what information is and is not provided among them. Note that numbers too small for reporting are indicated by an * with a note on the minimum number required for each state in order to report participation or performance data (e.g., some states do report participation for small numbers but not performance). Dashes indicate no data. For Ohio the participation rates over 100% are due to students being enrolled in one grade but taking the test in another (personal communication, 2004). A key for state test acronyms are provided at the end of this appendix.

Table C1. Regular State Reading Tests

 NOTE: For OHIO the participation rates over 100% were explained as being due to students being enrolled in one grade but taking the test in another.

 NOTE: For OHIO the participation rates over 100% were explained as being due to students being enrolled in one grade but taking the test in another (personal communication, 2004).

 NOTE: For OHIO the participation rates over 100% were explained as being due to students being enrolled in one grade but taking the test in another (personal communication, 2004).

 NOTE: For OHIO the participation rates over 100% were explained as being due to students being enrolled in one grade but taking the test in another (personal communication, 2004).

Table C5. Regular Social Studies Tests

 NOTE: For OHIO the participation rates over 100% were explained as being due to students being enrolled in one grade but taking the test in another (personal communication, 2004).

Table C6. Alternate Reading Tests

 Table C7. Alternate Mathematics Tests

 Table C9. Alternative Reading/Language Arts Tests

 Table C10. Alternative Mathematics Tests

 Table C12. English Language Proficiency Tests

 *Advanced Category Includes Early Advanced Level

State Testing Acronym Key