UNITED STATES DEPARTMENT OF EDUCATION OFFICE OF INSPECTOR GENERAL AUDIT SERVICES Chicago/Kansas City Audit Region Control Number ED-OIG/A05N0006 September 26, 2013 Michael L. Williams Commissioner of Education Texas Education Agency 1701 North Congress Avenue Austin, TX 78701 Dear Mr. Williams: This Final Audit Report, “The Texas Education Agency’s System of Internal Control Over Statewide Test Results,” presents the results of our assessment of selected aspects of the systems of internal control over statewide test results designed and implemented by the Texas Education Agency (TEA) and three local educational agencies (LEAs). Our work in the State of Texas (Texas) was part of a nationwide audit of the systems of internal control over statewide test results put in place by the U.S. Department of Education (Department) and five State educational agencies (SEAs). The purpose of the nationwide audit is to determine whether the Department and the five SEAs have systems of internal control that prevent, detect, and require corrective action if they find indicators of inaccurate, unreliable, or incomplete statewide test results. We will issue an audit report that presents the results of the nationwide audit to the Department. In Texas, we performed our audit work at TEA, La Joya Independent School District (La Joya), Lufkin Independent School District (Lufkin), and Marion Independent School District (Marion). Our audit covered statewide test results for school years 2007–2008 through 2009–2010.1 The purpose of this report is to separately address internal control weaknesses so the Department, TEA, and La Joya can take appropriate corrective action before the next round of statewide tests. TEA and La Joya need to improve their systems of internal control designed to prevent, detect, and require corrective action if they find indicators of inaccurate, unreliable, or incomplete statewide test results. TEA could improve its system of internal control by (1) using reviews of test results and analyses of erasure data to identify LEAs and schools to monitor, (2) identifying ways in which LEAs and schools can improperly influence the State of Texas Assessments of Academic Readiness (STAAR) test results and designing mitigating controls, and 1 In Texas, classes generally run from August to May or June. The Department of Education’s mission is to promote student achievement and preparation for global competitiveness by fostering educational excellence and ensuring equal access. Final Report ED–OIG/A05N0006 Page 2 of 15 (3) documenting the corrective actions that it recommends and verifying that the LEAs implement the corrective actions. La Joya could strengthen its system of internal control by (1) properly administering statewide tests, (2) adequately documenting its reviews of potential test administration irregularities, (3) retaining records of all reviews of potential test administration irregularities, and (4) reporting all test administration irregularities to TEA. We did not identify any significant control deficiencies at Lufkin or Marion. In response to the draft of this report, both TEA and La Joya agreed with the findings and recommendations. In addition, they described the corrective actions that they have already initiated and those that they plan to initiate to address our recommendations. See Attachment 2 for the full text of the comments. Other than including TEA’s and La Joya’s comments, we did not make any changes to the report. BACKGROUND The Elementary and Secondary Education Act of 1965, as amended (ESEA), requires States to establish a set of high-quality, yearly student academic tests. The tests must measure the proficiency of students in mathematics, reading or language arts, and science. States must establish a single minimum percentage of students who are required to meet or exceed the proficient level on these tests. States use these tests to determine the yearly performance of the SEA, each LEA, and each public school in the State. Section 1111(b)(3)(C)(iii) requires each State’s tests to be valid, reliable, and consistent with relevant, nationally recognized professional and technical standards. From school years 2007–2008 through 2012–2013, TEA developed seven tests to measure SEA, LEA, and school performance: 1. The Texas Assessment of Knowledge and Skills (TAKS) is a general test, offered with or without accommodations, given to students in grades 3 through 11 and exit-level students during school years 2007–2008 through 2010–2011. TEA has been phasing out the TAKS during school years 2011–2012 through 2013–2014. 2. The TAKS-Modified is an alternative test that uses modified academic achievement standards and is given to students receiving special education services. TEA has been phasing out the TAKS-Modified during school years 2011–2012 through 2013–2014. 3. The STAAR is a test or set of tests given to students in different grades. The STAAR given to students in grades 3 through 8 assesses the same subjects as the TAKS. The STAAR given to high school students consists of 12 end–of–course tests rather than the grade-specific TAKS. The STAAR began replacing the TAKS in school year 2011–2012. Final Report ED–OIG/A05N0006 Page 3 of 15 4. The STAAR-Modified is an alternative test that uses modified academic achievement standards and is given to students receiving special education services. The STAAR- Modified began replacing the TAKS-Modified in school year 2011–2012. 5. The STAAR-Alternate is a test given to students in grades 3 through high school who have significant cognitive disabilities and are receiving special education services. 6. The STAAR-L is a version of the STAAR test given to English language learners who meet the linguistic accommodation requirements. 7. The Texas English Language Proficiency Assessment System consists of tests given to students with limited English language proficiency to assess their progress in learning the English language. Students take these tests at various times throughout the school year depending on their grade levels and the specific test being administered. TEA developed all the tests with its contractor. After TEA’s Committee of Texas Educators approved the tests, the contractor printed the test materials and distributed them to LEAs and schools. Schools administered the tests, and the LEA or schools returned completed tests and unused test materials to the contractor. The contractor scored the tests and provided the results to TEA for review and verification. After TEA verified the results, the contractor made the results available to LEAs and schools via a secure Web site. TEA uploaded required test results data to the Department’s Education Data Exchange Network (EDEN) Submission System, part of the Department’s EDFacts initiative. EDFacts is a Department initiative to make performance data available for policy, management, and budget decisions for all K–12 educational programs. EDFacts centralizes K–12 performance data supplied by SEAs with other data, such as financial grant information, within the Department to enable better analysis and use in policy development, planning, and management. EDFacts includes statewide test data on student proficiency, participation rates, and graduation rates at the SEA, LEA, and school levels. As part of our LEA and school selection methodology, we used the EDFacts data to identify 935 LEAs in Texas that had at least 1 school with total enrollment of more than 200 students during school years 2007–2008, 2008–2009, and 2009–2010. These 935 LEAs had more than 4.3 million students attending 6,105 schools that each had more than 200 students. For each grade tested in the subjects of math and reading at each of the schools, we calculated a risk score. We calculated the risk scores to evaluate how anomalous increases and decreases in student proficiency for a grade and subject were from one year to the next in relation to the change for that grade and subject across the State.2 A school’s composite risk score was the average of the school’s five highest risk scores across all years, grades, and subjects. We selected the three LEAs with the highest composite risk scores (see Scope and Methodology for a further 2 We used the risk score to select LEAs and schools for internal control reviews, not to determine whether cheating occurred at a particular LEA or school. Final Report ED–OIG/A05N0006 Page 4 of 15 description of our selection methodology). We then selected two schools from each of the three LEAs with the largest annual test score fluctuations. Tables 1 and 2 show the number of schools and students for the three LEAs and six schools that we selected for review. Table 1. Schools and Students in the LEAs Reviewed Number of LEA Number of Schools Students La Joya 30 26,401 Lufkin 7 5,878 Marion 3 1,053 Totals 40 33,332 Table 2. Students in the Schools Reviewed Number of LEA School Name Students La Joya La Joya High School 2,186 La Joya Sam Fordyce Elementary School 517 Lufkin Brandon Elementary School 449 Lufkin Lufkin High School 2,257 Marion Marion High School 409 Marion Marion Middle School 361 Total 6,179 AUDIT RESULTS TEA could improve its system of internal control over preventing, detecting, and taking corrective actions if it finds indicators of inaccurate, unreliable, or incomplete statewide test results. Specifically, TEA could strengthen its risk assessment and monitoring processes by (1) using reviews of test results and analyses of erasure data to identify LEAs and schools to monitor, (2) identifying ways that LEAs and schools can improperly influence STAAR test results and designing mitigating controls, and (3) documenting the corrective actions that it recommends and verifying that the LEAs implement the corrective actions. TEA performed some internal control activities related to the administration of statewide tests. The control activities included providing LEA and school personnel with guidance manuals and training documents that emphasized test security and proper handling of test administration irregularities, as well as requirements for reporting improper conduct. TEA also required campus coordinators and principals, LEA coordinators, and test administrators to sign a statement acknowledging these guidelines. In addition, TEA generally established adequate control over input, processing, and reporting of statewide test results. Final Report ED–OIG/A05N0006 Page 5 of 15 We also identified weaknesses in La Joya’s system of internal control over preventing, detecting, and taking corrective actions if it finds potential test administration irregularities that could call into question the validity of the statewide test results. La Joya could strengthen its system of internal control by (1) properly administering statewide tests, (2) adequately documenting its reviews of potential test administration irregularities, (3) retaining records of all reviews of potential test administration irregularities, and (4) reporting all test administration irregularities to TEA. We did not identify any significant control deficiencies at Lufkin or Marion. The two LEAs generally were following the guidance that TEA provided. In response to the draft of this report, both TEA and La Joya agreed with the findings and recommendations. In addition, they described the corrective actions that they have already initiated and those that they plan to initiate to address our recommendations. See Attachment 2 for the full text of the comments. FINDING NO. 1 – TEA Could Strengthen Its Risk Assessment and Monitoring Processes TEA could strengthen its risk assessment and monitoring processes by (1) using reviews of test results and analyses of erasure data to identify LEAs and schools to monitor, (2) identifying ways that LEAs and schools can improperly influence STAAR test results and designing mitigating controls, and (3) documenting the corrective actions that it recommends and verifying that the LEAs implement the corrective actions. TEA Did Not Use Analyses of Test Results and Erasures to Identify LEAs or Schools to Monitor From school years 2007–2008 through 2010–2011, TEA conducted onsite monitoring visits to schools during test administration. After school year 2010–2011, TEA relied on LEAs to conduct their own onsite monitoring using guidelines that TEA provided to them. The LEAs were supposed to report test administration irregularities to TEA so that TEA could consider additional follow-up activities. However, TEA could identify more potential test administration irregularities if it used the results of forensic analyses, rather than relying solely on LEA reporting. TEA had access to test results data and erasure analyses from its testing contractor, but it did not use these resources to proactively identify LEAs and schools with possible test administration irregularities. TEA received test results data from its statewide testing contractor and used the data to identify test values outside an expected range, such as excessive absences during testing. However, TEA did not analyze year-to-year test score changes to identify possible test administration irregularities and determine potential follow-up activities. As of school year 2012–2013, TEA was still reviewing the practice of using test score analyses to identify possible test administration irregularities and had not implemented the practice. In addition, since school year 2005–2006, TEA’s statewide testing contracts have contained provisions for collecting and summarizing erasure data for all schools and students. TEA required its statewide testing contractor to summarize the data by listing the total number of erasures, wrong-to-right erasures, right-to-wrong erasures, and wrong-to-wrong erasures by subject and student. In addition, when an LEA reported a test administration irregularity, TEA Final Report ED–OIG/A05N0006 Page 6 of 15 had the option to require the testing contractor to report any class in the subject school with an average erasure rate greater than three standard deviations above the statewide mean in wrong- to-right erasures for each of the subjects within each grade tested. TEA had the option to use these analyses to obtain additional information on the specific test administration irregularity. TEA agreed that reviews of test results and analyses of erasure data could help identify LEAs and schools to monitor. However, TEA informed us that its statewide testing contractor cautioned in a report to the Texas Technical Advisory Committee (TTAC) that erasure analyses alone provided no insight into the reason behind excessive erasures and stressed that these statistical analyses serve only to identify an extreme number of light marks or erasures. TEA’s statewide testing contractor further reported that these statistical analyses should serve only as a screening device. As a result, TEA chose to couple the data from the erasure analyses with information that LEAs provided through the incident reporting process and to use this data as supporting evidence of a potential test administration irregularity. TEA added that it considered two new methods for identifying possible test administration irregularities: residual analysis and pass rate analysis. TEA presented the results from the pilot analyses for both statistical methods to the TTAC in March 2012 and January 2013, respectively. The TTAC expressed concern about using data derived from these methods as sole identifiers of test administration irregularities, noting that TEA could use the same data to identify effective teaching. The TTAC also agreed with the contractor’s recommendation that the use of residual and pass rate analyses should always take place within a larger review process. The TTAC stated that these two analyses should be supplemented with the collection of additional evidence, such as locally maintained seating charts, reports of test administration irregularities, test participation data, erasure analyses, and records of test security and administration training for campuses. The TTAC concluded that statistical measures alone should not be used to trigger a review. Because of its limited resources and the caution expressed by the TTAC, TEA concluded that it was not feasible or appropriate to monitor schools based on outcomes of proposed statistical analyses at this time. TEA added that it will continue to discuss with its statewide testing contractor and TTAC members the use of statistical measures to determine appropriate and fair methods by which TEA can use these data. We agree that TEA should not use forensic analyses as the sole evidence of test administration irregularities. There are a variety of possible explanations for anomalous wrong-to-right erasure counts and changes in proficiency, and sanctions should not be based solely on forensic analysis. However, TEA could use the results of these analyses to identify LEAs and schools for possible monitoring. Improved monitoring procedures will help TEA ensure that LEAs and schools are properly administering statewide tests. TEA Did Not Ensure That LEAs Tested All Qualified 10th Grade Students and Has Not Assessed How LEAs or Schools Could Influence Outcomes of New State Tests TEA did not track students’ grade-level reclassifications or assess the potential impact of these reclassifications on the accuracy, reliability, and completeness of test results and determinations of adequate yearly progress (AYP). For school years 2007–2008 through 2012–2013, TEA used the results of the prior year’s TAKS given to qualified 10th grade students and graduation rates to determine whether a high school made AYP. High school students took the TAKS only if Final Report ED–OIG/A05N0006 Page 7 of 15 their LEA classified them as 10th grade students at the time of the test. In Texas, LEAs are responsible for assigning grade-level classifications to students. Therefore, an LEA could avoid classifying students as 10th grade students so that lower performing students would never take the 10th grade TAKS. The three LEAs that we reviewed reclassified students to the next grade level only at the beginning of each school year. The three LEAs did not classify a high school student as a 10th grade student until the student earned the number of credits that each LEA’s policy required. However, some students might not have earned enough credits by the beginning of their second year of high school for the LEA to classify them as 10th grade students. Because the LEAs would classify them as 9th grade students for their entire second year of high school, these students would not be given the 10th grade TAKS at any time during their second year of high school. By the beginning of their third year of high school, the students might have earned enough credits for the LEA to classify them as 11th grade students. The students never would have been classified as 10th grade students at any time during high school and never would have been given the 10th grade TAKS that TEA used to determine AYP. Section 1111(b)(2)(A) of the ESEA requires States to implement a statewide accountability system that ensures LEAs and public schools make AYP.3 The accountability system must be based on academic standards and annual academic tests that demonstrate AYP towards academic achievement standards. TEA’s accountability system requires LEAs and public schools to meet AYP in three measures: reading/language arts, mathematics, and either graduation rate for LEAs and high schools or attendance rate for elementary and middle schools. In addition, section 1111(b)(3)(C) of the ESEA requires States to use such annual academic tests for purposes for which such tests are valid and reliable. States must administer the tests not less than once in grades 10 through 12 and provide for the participation of all students, including limited English proficient students. TEA mitigated the risk that an LEA could avoid testing students by never classifying them as 10th grade students when it instituted the STAAR end-of-course tests. Starting with school year 2011–2012, high school students take 12, end–of–course tests instead of the grade-specific TAKS. Students take STAAR end-of-course tests whenever they complete specific courses, regardless of their grade levels. TEA includes the students’ results on the STAAR end-of-course tests when determining whether high schools are making AYP. Although TEA eliminated the potential impact of grade-level reclassifications on test results for students who were 9th grade students beginning with school year 2011–2012, it had not assessed the possibility that LEAs and schools might influence STAAR test results in other ways. For example, a school potentially could improve its test results in a given year by either preventing low-performing students from taking these courses during that year or encouraging high- achieving students from multiple grades to take these courses during that year. As Texas implements the STAAR end-of-course tests and phases out the 10th grade TAKS, the AYP calculations will be modified to include the STAAR end-of-course tests. In addition, 3 AYP is a measurement used to determine how States, LEAs, and public schools are performing academically according to students’ results on standardized tests. Final Report ED–OIG/A05N0006 Page 8 of 15 TEA’s transition plan for school years 2010–2011 through 2013–2014 includes long-term goals and targets for graduation rates. TEA also should take this opportunity to determine how best to prevent or detect potential manipulation of test results, which could impact whether schools make AYP. TEA Neither Documented Corrective Actions That It Recommended for La Joya Nor Verified That La Joya Implemented These Corrective Actions For school years 2007–2008 through 2009–2010, TEA did not document its recommended corrective actions to address 3 of 12 statewide test administration irregularities that La Joya reported through incident reports. These irregularities included 13 students not taking a statewide test, students completing answer documents that La Joya assigned to other students, and a student receiving an incorrect test score because the campus testing coordinator coded the answer document incorrectly. TEA informed us that it provided La Joya with recommendations to resolve the three test administration irregularities. TEA made the recommendations when La Joya called TEA to report the three irregularities and seek guidance. However, TEA did not record these recommendations in its online incident report system. Instead, TEA relied on La Joya to record the recommendations that TEA made to La Joya in the online incident reports that La Joya later submitted to TEA. If TEA relies on LEAs to record its recommendations, rather than recording the recommendations itself, LEAs might incorrectly record TEA's recommendations. TEA would have no record of the actual recommendations that it made. In addition, TEA did not maintain records showing that it verified La Joya implemented the corrective actions. Instead, TEA documented only that it notified La Joya that it was closing its records of the incidents with no further action. If TEA does not maintain these records, then it cannot ensure that the LEA implemented the corrective actions. TEA stated that it will start documenting all communications and retaining support for recommendations that it provides to LEAs regarding corrective actions needed to address test administration irregularities. Federal Regulations and Guidance Require Monitoring for Compliance With Federal Requirements According to 34 Code of Federal Regulations § 80.40(a), grantees are responsible for monitoring grant- and subgrant-supported activities to ensure compliance with applicable Federal requirements and the achievement of performance goals. According to “Key Policy Letters from the Secretary or Deputy Secretary,” June 24, 2011, States are urged to make test security a high priority. States should review and, if necessary, strengthen efforts to protect student achievement and accountability data, ensure the quality of those data, and enforce test security. Recommendations We recommend that the Assistant Secretary for Elementary and Secondary Education require TEA to strengthen its risk assessment and monitoring processes by— 1.1 Using the test results and erasure analyses completed by its testing contractor more effectively to identify LEAs and schools for possible monitoring. Final Report ED–OIG/A05N0006 Page 9 of 15 1.2 Identifying ways that LEAs and schools can improperly influence STAAR test results and design mitigating controls. This could include reviewing enrollment and participation trends in end-of-course tests. 1.3 Documenting the corrective actions that it recommends LEAs take in response to self– reported testing irregularities and verifying that the LEAs implement the corrective actions. TEA Comments TEA agreed with the finding and recommendations and stated that it already implemented, or has plans to implement, corrective actions for all issues identified in the finding. TEA added that it will implement the new policies and controls to coordinate with its upcoming test administrations. However, it is likely the activities will be ongoing for the next several years. Legislation passed during the most recent Texas legislative session will result in components being added to the accountability system and changes being made to the State assessment system. FINDING NO. 2 –La Joya Could Strengthen Its System of Internal Control Over Administering Tests, Retaining Records, and Reporting Test Administration Irregularities La Joya did not (1) properly administer statewide tests, (2) adequately document its reviews of potential test administration irregularities, (3) provide complete records of all reviews of potential test administration irregularities, or (4) report all test administration irregularities to TEA in a timely manner. TEA needs to identify why the test administration irregularities occurred and ensure that La Joya implements procedures to prevent them from happening in the future. La Joya Did Not Properly Administer Statewide Tests For school years 2007–2008 through 2009–2010, of the 65 total test administration irregularities, La Joya identified 25 in which it did not properly administer statewide tests. Of these 25, La Joya: Did not administer the correct statewide test to students 16 times. Each irregularity affected at least one student and in one case affected an entire classroom. La Joya placed students in incorrect testing rooms, provided the wrong grade-level tests, provided tests in the wrong language, provided versions of tests with accommodations that students were not approved to receive, and provided versions of tests that did not allow students to have the accommodations that they were approved to receive according to their Individualized Education Programs. Provided answer documents to the wrong students four times. In each instance, the irregularity affected 2 to 3 students. Students with identical names received each other’s answer documents, students who took a test in April used versions of the answer documents from the March test, and an 11th grade student was provided an answer document intended for a 10th grade student. Final Report ED–OIG/A05N0006 Page 10 of 15 Coded answer documents with incorrect information in the “Test Taken Info” field or did not correct inaccurate information on pre–coded student labels on answer documents five times. In each instance, the irregularity affected one to three students. According to pages 13–14 of its “Test Security Supplement 2007–2008,” TEA requires campus testing coordinators to “ensure that students are being administered the appropriate tests and have the correct corresponding answer document.” On pages 95–100 of its “2008 District and Campus Coordinator Manual,” TEA explains how LEAs need to properly attach pre–coded student labels to answer documents and correct inaccurate information displayed on the labels. Pages 109–113 of the manual explain how LEAs need to properly fill out the “Test Taken Info” field, because failure to do so might result in a student receiving an incorrect score. Without identifying the underlying causes of these instances and developing mitigating controls, La Joya does not have reasonable assurance that tests will be administered properly in the future. Additionally, because it did not always properly administer the statewide tests, La Joya cannot provide assurances about the validity of student test results. La Joya Did Not Adequately Document Its Reviews of Potential Test Administration Irregularities For school years 2007–2008 through 2009–2010, La Joya did not adequately document its review of 17 of the 65 potential test administration irregularities. The potential irregularities included incidents of students caught using or possessing cell phones or other electronic devices during test administration, a student distracting other students, a student taking a restroom break and never returning, students caught cheating, and students receiving test materials with broken seals. Although La Joya has its own form to document the required information, the documentation that it maintained for each incident did not clearly describe (1) what happened, (2) how the incident occurred, (3) the actions La Joya took to resolve the incident, or (4) La Joya’s determination of whether the incident constituted a deviation from documented testing practices that should be reported to TEA. According to pages 25–26 of TEA’s “Test Security Supplement, 2007–2008,” TEA requires LEAs to review each potential test administration irregularity and collect statements and reports that clearly outline the sequence of events, explain exactly what happened and how it occurred, include information about how the problem was resolved or remedied, and document the LEA’s determination in the matter. Without complete documentation of the potential test administration irregularity review process, La Joya cannot show that it reviewed and resolved all potential test administration irregularities fairly, appropriately, and consistently. La Joya Did Not Provide Records of All Reviews of Potential Test Administration Irregularities La Joya was unable to provide documentation for all of its reviews of potential statewide testing irregularities during school year 2007–2008. It could locate documentation for only nine of its reviews. As a result, we could not determine the total number of reviews that La Joya conducted during that school year. Final Report ED–OIG/A05N0006 Page 11 of 15 Without records for all of its reviews of potential test administration irregularities, La Joya cannot show that it thoroughly reviewed all of them and took corrective actions to address internal control deficiencies identified as a result of these reviews. La Joya could not tell us the total number of reviews of potential test administration irregularities that it conducted during school year 2007–2008. According to page 75 of its “2008 District and Campus Coordinator Manual,” “Beginning with the spring 2008 TAKS administration, districts will be required to maintain test security materials for five years.” La Joya Did Not Report All Test Administration Irregularities to TEA or Did Not Report the Irregularities Timely For school years 2007–2008 through 2009–2010, of the 28 incidents of test administration irregularities from the two schools we visited, La Joya did not report 16 to TEA. La Joya also provided us with records for 37 incidents that occurred during school year 2009–2010 at the remaining La Joya schools, but it never reported the incidents to TEA. The irregularities included giving students the wrong answer documents, misplacing testing booklets and answer documents, and leaving students unmonitored during a testing session. Additionally, La Joya gave students the wrong statewide tests, resulting in students (1) receiving accommodation that they were not approved to receive, (2) not receiving accommodations that they were approved to receive according to their Individualized Education Programs, or (3) taking tests that did not include all required subject areas. In addition, for school years 2009–2010 through 2011–2012, La Joya reported 13 statewide test administration irregularities to TEA more than 6 months after the irregularities occurred. In 1 of the 13 cases, La Joya reported an irregularity to TEA more than 2 years after the irregularity occurred. To test for timely reporting, we used the date that the irregularity occurred instead of the date that the district test coordinator became aware of the incident because we could not determine when the district coordinator learned of the incident or actually received a complaint. La Joya did not explain why it took more than 6 months to report the test administration irregularities. See Table 3 for a breakdown of the number of test administration irregularities at La Joya. According to pages 21–24 of its “2008 District and Campus Coordinator Manual” and pages 19– 23 of its “Test Security Supplement, 2007–2008,” TEA requires LEAs to report all incidents resulting in deviations from documented testing procedures, including eligibility errors, improper accounting for secure test materials, monitoring errors, and procedural errors. According to page 23 of its “2009 District and Campus Coordinator Manual,” and page 30 of its “Test Security Supplement, 2009–2010,” TEA requires LEAs to submit all required documentation for testing irregularities, including statements from individuals involved and an incident report, “within 10 working days of the district testing coordinator being made aware of the incident.” Because TEA was not aware of all test administration irregularities at La Joya, it did not have a complete understanding of La Joya’s system of internal control over statewide testing. If La Joya had notified TEA of all the test administration irregularities, TEA might have decided to take corrective actions, such as assessing penalties or performing additional monitoring. Final Report ED–OIG/A05N0006 Page 12 of 15 Table 3. Test Administration Irregularities Number of La Joya Incidents Reviewed Number of School Years Incidents for 2007–2008 School Which TEA Through Years 2007– Provided 2009–2010: School Year 2008 Records: Two Schools 2009–2010: Through School Years That We All Other 2009–2010: 2010–2011 and Category Visited Schools(a) Total 2011–2012 Total Improper Test Administration 17 8(b) 25 (c) 25 Inadequate Documentation 4 13 17 (c) 17 Did Not Report to TEA 16 37 53 (c) 53 Reported to TEA Untimely 1 0 1 12 13 Reported to TEA 11 0 11 15 26 Timely Total Number of Records of Incidents 28 37 65(d) 27 92 (a) La Joya supplied unreported incidents from other schools only for school year 2009–2010. (b) There were seven separate incidents but one incident had two different improper test administration irregularities associated with it. (c) We did not review any records provided by TEA for incidents in this category. (d) We did not review the reported incidents for all other La Joya schools. La Joya could not determine the total number of incidents. Recommendations We recommend that the Assistant Secretary for Elementary and Secondary Education require TEA to— 2.1 Require La Joya to identify why it did not administer the correct tests to all students and ensure that La Joya implements procedures to prevent it from administering incorrect tests to students in the future. 2.2 Require La Joya to ensure that it properly documents all reviews of potential test administration irregularities. 2.3 Require La Joya to retain records of its reviews of potential test administration irregularities for the minimum number of years that TEA requires. 2.4 Ensure that La Joya timely reports to TEA all test administration irregularities that TEA requires it to report. Final Report ED–OIG/A05N0006 Page 13 of 15 La Joya Comments La Joya agreed with the finding and recommendations and stated that it already implemented corrective actions for all issues identified in the finding. SCOPE AND METHODOLOGY To achieve our objective, we— 1. Reviewed and gained an understanding of Federal law, regulations, and guidance applicable to the audit objective. 2. Reviewed prior (1) Single Audits for TEA and the three LEAs and (2) Department program monitoring reports to identify areas of potential internal control weaknesses related to our audit objective. 3. Reviewed written policies and procedures, contracts for scoring tests, and test administration records at TEA, three LEAs, and six schools. 4. Interviewed officials at TEA, three LEAs, and six schools. 5. Gained an understanding and assessed the adequacy of TEA’s and the three LEAs’ systems of internal control over statewide test results in the following areas: data collection, monitoring, and guidance. Sampling Methodology We judgmentally selected three LEAs from a universe of 935 LEAs. The 935 LEAs in our universe consisted of all LEAs in Texas that had at least 1 school with more than 200 students during school years 2007–2008, 2008–2009, and 2009–2010. Next, we calculated for each school a risk score for each grade tested in the subjects of math and reading. We calculated a risk score to determine how anomalous each increase or decrease in the percentage of students who scored at a proficient level for a grade or subject was from one year to the next in relation to the change for that grade or subject across the State. A school’s composite risk score was the average of the school’s five highest risk scores across all years, grades, and subjects. We also considered the number of schools in the LEA, income level, and annual fluctuations in whether schools met AYP requirements. After selecting La Joya, Lufkin, and Marion, we selected the two schools at each LEA that had significant annual test score fluctuations. To determine whether La Joya completely and accurately documented and reported potential test administration irregularities, we reviewed documentation for 65 incidents of potential test administration irregularities for school years 2007–2008, 2008–2009, and 2009–2010. The documentation included information such as the date of each incident, the date each incident was reported to TEA, the individuals involved, explanations, resolutions, whether each incident was Final Report ED–OIG/A05N0006 Page 14 of 15 reported to TEA, and whether the resolution was satisfactory. These 65 incidents included the following: 28 incidents that occurred at the 2 La Joya schools in our sample during school years 2007–2008 through 2009–2010. La Joya reported 12 of these incidents to TEA and did not report 16. 37 unreported incidents that occurred at all other La Joya schools during school year 2009–2010. We did not review the reported incidents for all the other La Joya schools. To determine whether La Joya timely reported incidents of test administration irregularities to TEA, we reviewed documentation that La Joya provided us for 65 incidents of potential test administration irregularities that occurred during school years 2007–2008, 2008–2009, and 2009–2010. We also reviewed documentation that TEA provided us for 27 incidents that occurred at La Joya during school years 2010–2011 and 2011–2012. Data Reliability To achieve our objectives, we relied on data from EDFacts. EDFacts includes data fields for student proficiency on statewide tests, participation rates, and graduation rates at the SEA, LEA, and school levels. We used this data to select LEAs and schools to visit as part of this audit. We looked for patterns of changes in student proficiency and used enrollment data to limit the universe of schools to those with more than 200 students during school years 2007–2008, 2008–2009, and 2009–2010. To determine whether the EDFacts proficiency scores (for selected grades and subjects at selected schools) were accurate and complete, we reconciled the scores that we obtained from EDFacts with scores we calculated using data that TEA provided. We did not find any discrepancies between the scores recorded in EDFacts and the scores that we calculated using TEA’s data. Therefore, we determined that the data from EDFacts were sufficiently reliable for our intended use. We conducted this audit at TEA’s offices in Austin, Texas; La Joya’s offices in La Joya, Texas; Lufkin’s offices in Lufkin, Texas; Marion’s offices in Marion, Texas; and at our offices from August 2012 through May 2013. We discussed the results of our audit with TEA officials on May 29, 2013. We provided TEA officials with a draft of this report on July 22, 2013. We conducted this performance audit in accordance with generally accepted government auditing standards (December 2011 revision). Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions on our audit objective. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions on our audit objective. Final Report ED–OIG/A05N0006 Page 15 of 15 ADMINISTRATIVE MATTERS Statements that managerial practices need improvements, as well as other conclusions and recommendations in this report, represent the opinions of the Office of Inspector General. Determinations of corrective action to be taken will be made by the appropriate Department officials. This report incorporates the comments you provided in response to the draft report. If you have any additional comments or information that you believe may have a bearing on the resolution of this audit, you should send them directly to the following Department official, who will consider them before taking final action on this audit: Deborah S. Delisle Assistant Secretary Office of Elementary and Secondary Education 400 Maryland Avenue, SW Washington, D.C. 20202 It is the policy of the Department to expedite the resolution of audits by initiating timely action on the findings and recommendations contained therein. Therefore, receipt of any additional comments within 30 days would be appreciated. In accordance with the Freedom of Information Act (5 U.S.C. § 552), reports issued by the Office of Inspector General are available to members of the press and general public to the extent information contained therein is not subject to exemptions in the Act. Sincerely, /s/ Gary D. Whitman Regional Inspector General for Audit ATTACHMENT 1 Acronyms, Abbreviations, and Short Forms Used in this Report AYP Adequate Yearly Progress Department U.S. Department of Education ESEA Elementary and Secondary Education Act of 1965, as amended La Joya La Joya Independent School District LEA Local Educational Agency Lufkin Lufkin Independent School District Marion Marion Independent School District SEA State Educational Agency STAAR State of Texas Assessments of Academic Readiness TAKS Texas Assessment of Knowledge and Skills TEA Texas Education Agency Texas State of Texas TTAC Texas Technical Advisory Committee ATTACHMENT 2 August 26, 2013 Mr. Gary D. Whitman Regional Inspector General for Audit United States Department of Education Office of Inspector General Citigroup Center 500 West Madison Street, Suite 1414 Chicago, IL 60661 RE: Draft Audit Report; Texas Education Agency’s System of Internal Control over Statewide Test Results; Control Number ED-OIG/A05N0006 Dear Mr. Whitman, This letter is in response to the Draft Audit Report regarding TEA identified above. As noted below, the Texas Education Agency (TEA) agrees with the findings and recommendations set forth in the draft report. Response to Recommendations for Finding No. 1 – TEA Could Strengthen Its Risk Assessment and Monitoring Processes With respect to specific recommendations, TEA has the following comments and anticipated completion dates: Recommendation 1.1 – Using the test results and erasure analyses completed by its testing contractor more effectively to identify LEAs and schools for possible monitoring. TEA agrees with the recommendation, and will use information from annually conducted erasure analyses to supplement information collected by other divisions within the agency to identify campuses where monitoring or other follow-up might be warranted. Additionally, in light of the state’s new assessment and accountability systems, the agency will review and analyze its existing data validation analyses and internal controls in the fall of 2013 to determine whether they are appropriate for the new systems or whether other types of analyses and internal controls are needed. It is likely these activities will be ongoing for the next several years as additional components are added to the accountability system and as changes are made to the state assessment system as a result of legislation passed during the most recent Texas legislative session. Recommendation 1.2 – Identifying ways that LEAs and schools can improperly influence STAAR test results and design mitigating controls. TEA agrees with the recommendation and will continue to (a) identify ways that LEAs and schools can improperly influence STAAR test results; and (b) design mitigating controls. In light of the new accountability system implemented for the first time in August 2013, the agency will review and analyze, in fall 2013, its existing policies and internal controls to determine whether they are appropriate for the new accountability system or whether additional policies and internal controls are needed. While any additional policies or controls may be implemented as soon as winter 2013 or spring 2014, it is likely these activities will be ongoing as other components are added to the accountability system over the next several years, and as changes are made to the state assessment system as a result of legislation passed during the most recent Texas legislative session. Recommendation 1.3 – Documenting the corrective actions that it recommends LEAs take in response to self-reported testing irregularities and verifying that the LEAs implement the corrective actions. TEA agrees with the recommendation and has developed a process to document corrective actions recommended to districts in response to self-reported serious irregularities and to verify that LEAs have implemented them accordingly. Serious testing irregularities that are reported to TEA on the telephone or through email are documented in the database as a “Staff Report.” Districts follow up by submitting online incident reports that document their investigation of the incident. This online form will contain a field for districts to specify corrective action that they will take in response to the testing irregularity. Beginning in fall 2014 after incident reports from the spring 2014 administration have been processed by staff in the Student Assessment Division, TEA plans to audit a random sample of districts to verify that they have implemented corrective actions in response to reported irregularity events. The number of districts selected for follow-up and the nature of the follow-up activities will be dependent on available agency resources. Results will be used to reinforce actions taken by the district and to improve training on test administration and the maintenance of test security. Response to Recommendations for Finding No. 2 – La Joya Could Strengthen Its System of Internal Control Over Administering Tests, Retaining Records, and Reporting Test Administration Irregularities With respect to specific recommendations, La Joya ISD has the following comments and anticipated completion dates: Recommendation 2.1 – Require La Joya to identify why it did not administer the correct tests to all students and ensure that La Joya implements procedures to prevent it from administering incorrect tests to students in the future. La Joya ISD has investigated the errors committed by staff to identify the underlying causes.that included failure to follow procedures and misunderstanding of procedures. In response, La Joya ISD has developed more stringent controls to ensure the proper administration of statewide tests as described below. For example, in the case where 13 students were not administered the TELPAS test, we found that the campus testing coordinator (CTC) did not understand that when an ARD committee decides that a student should not be administered any of the components of the TELPAS test, the CTC is still required to register the student for the online TELPAS test and enter coding into the online system indicating that the student was not required to take the test due to a decision by the student’s ARD committee. The clarification of this procedure has been emphasized during training. In 2010, La Joya ISD analyzed testing errors committed by staff and initiated procedures intended to eliminate those errors. The most common type of errors involved the implementation of special education IEPs. Therefore, written local procedures designed to prevent the administration of incorrect tests to students were developed and continue to be revised as needed. Local procedures addressing the testing of students receiving special education services were developed in May of 2010 and revised in March of 2011. Other local procedures intended to eliminate errors include procedures for the use of locally developed checklists, determining the eligibility for students to take retests, establishing a secure storage location, distributing and collecting secure materials, the verification of PEIMS coding on answer documents, the retention of testing records, the retention of Alternate testing student records, TELPAS validity and reliability procedures, a 30-minute duty-free lunch for teachers, the regrouping of students during State tests, the verification of test results, and the processing of testing incidents and irregularities. Campus principals and campus testing coordinators (CTCs) are trained annually on these procedures as well as the procedures in the State’s testing manuals. CTCs participate in several training sessions throughout the year including updates to emphasize certain procedures in order to eliminate errors. Curriculum and instruction staff members at the Central Office are trained to monitor State test administrations at each campus. The Central Office test monitors are trained to use written checklists to document their monitoring activities. Central Office test monitors are trained to notify the campus testing coordinator and the principal immediately when they discover a concern so that the concern is addressed as soon as possible. La Joya ISD provides frequent training for campus testing coordinators and encourages campus testing coordinators to contact the district testing coordinator for additional support. The participation of campus testing coordinators in training is monitored to ensure that they attend the training and make-up sessions are offered for those who do not attend the regularly scheduled training sessions. For the 2013-2014 school year, La Joya ISD has increased the number of professional staff in the central office testing department so that it may provide additional support to the campuses as needed. With more stringent controls in place, La Joya ISD has a reasonable assurance that tests will be administered properly in the future. Recommendation 2.2 – Require La Joya to ensure that it properly documents all reviews of potential test administration irregularities. In April 2013, La Joya ISD established written local procedures to supplement State procedures for the processing of irregularities and potential irregularities. Campus testing coordinators (CTCs) will contact the district testing coordinator (DTC) as early as possible after discovering a testing irregularity or potential testing irregularity. CTCs will investigate to determine whether an irregularity or student cheating occurred. CTCs will submit statements within 2 days of discovering a testing irregularity or potential irregularity. Statements will include how any testing irregularities or potential irregularities were resolved and a determination by the CTC of whether a testing irregularity or a student attempt to cheat occurred. CTCs will submit the required online reports to TEA within 6 working days of reporting it to the DTC. The written local procedures were created in part to ensure the proper documentation of investigations of potential test administration irregularities. These procedures will ensure that La Joya ISD will review and resolve all potential test administration irregularities fairly, appropriately, and consistently. Recommendation 2.3 – Require La Joya to retain records of its reviews of potential test administration irregularities for the minimum number of years that TEA requires. During 2007-2008 and 2008-2009, La Joya ISD retained testing records by submitting them to a central office of Admissions and Records which followed the record retention rules of the Texas State Library and Archives Commission. Testing records for those years were destroyed by the Office of Admissions and Records in accordance with the rules from the Texas State Library not realizing that there were other rules in the State’s testing manuals governing the retention of testing records. In December 2010, La Joya ISD established a written local procedure for the retention of testing records requiring that the records be kept at the campus with copies in the central office testing office. The procedure supplemented the State’s requirements for record retention making it clear that the records were to be kept in the campus testing coordinator’s office with copies in the district testing coordinator’s office. On August 19, 2013, La Joya ISD revised its written local procedure for the retention of testing records to include statements about irregularities and potential test administration irregularities. Recommendation 2.4 – Ensure that La Joya timely reports to TEA all test administration irregularities that TEA requires it to report. In April 2013, La Joya ISD established written local procedures for the processing of irregularities and potential irregularities that changed how the district submitted reports to TEA through the online system. The new procedures have decentralized the processing of reports so that campus testing coordinators have greater control over the submission of reports with oversight from the central office. Through TEA’s online reporting system, the district testing coordinator reviews all reports submitted by campuses. The new procedures did result in significant improvements in the timeliness of submissions during the Spring 2013 test administrations. These procedures and the results will be studied to ensure compliance with reporting requirements. The Agency appreciates the insights the report provides regarding its system of internal control over statewide test results, and would like to thank the USDE audit team for the courtesy extended to Agency and district staff during the audit process. Sincerely, /s/ Michael Williams Commissioner of Education cc: Criss Cloudt, Associate Commissioner, Texas Education Agency Gloria Zyskowski, Director, Texas Education Agency Aida Benavides, Superintendent, La Joya Independent School District
The Texas Education Agency's System of Internal Control Over Statewide Test Results
Published by the Department of Education, Office of Inspector General on 2013-09-26.
Below is a raw (and likely hideous) rendition of the original report. (PDF)