Resource Center

Testimony by Albert Cortez, Ph.D.

on behalf of the Intercultural Development Research Association
Senate Education Committee – October 12, 1999

Mr. Chairman, members of the committee, Senator Barrientos, I appear before you today representing the Intercultural Development Research Association (IDRA), a research, training and technical assistance organization based in San Antonio that is dedicated to the improvement of educational opportunities for all students.

Under a contract with the then Texas Department of Community Affairs, IDRA was the first organization to conduct extensive research to estimate the number of pupils who were completing high school in Texas. At that time, we developed a formula for calculating the attrition rate – which the percent of students lost from enrollment. We take the number of students enrolled in the ninth grade and compare it to the number of students enrolled in the 12th grade four years later. Recognizing that some new students enter Texas high schools between the 10th and 12th grades, we incorporate a statistical adjustment for that in-migration of pupils.

We began using this approach 14 years ago because the Texas Education Agency (TEA) did not have the capacity to allow for tracking individual pupils through the Texas school system. We chose to continue conducting the analysis in the same way to enable us to track any changes that may have occurred in attrition over several years. We contend that our attrition estimate more closely tracks the actual longitudinal dropout rate than does the current state dropout estimates produced by TEA, and it can be used as one of several methods for cross-checking dropout reports.

Among the attachments provided to the committee are tables from IDRA’s latest attrition analyses, just completed this month. You can see on Table 1 that the attrition rates for the state as a whole and for subgroups of students – particularly Hispanic students and African American students, continue to be higher than 40 percent. We have seen this trend over several years.

Despite ongoing state concerns, the dropout rate is not getting better. What has been occurring, however, is that the state education agency, through its authority to create certain definitions, is reporting a decrease in the state dropout estimates. They do this not by actually decreasing the number of students receiving a high school diploma but by manipulating the dropout numbers by changing the state definitions of what constitutes a dropout.

Over the years, IDRA has seen the state create new categories of pupils who are not counted in the calculation of state and school district dropout statistics. Students who drop out of the regular program and either acquire a GED or are reported as being enrolled in a GED program are not counted as dropouts. Students who complete all their required course work but do not pass the Texas Assessment of Academic Skills (TAAS) are denied a high school diploma but are not counted as dropouts (giving districts no incentive to see that those pupils really finish high school). Students who are reported as “having returned to their home country” are not counted as dropouts. Students who say they are transferring, but for whom no transcripts are requested by the receiving schools are also not counted as dropouts.

This “dropout definitions game” has led many people to question the accuracy of the Texas dropout data that is reported by the state. In turn, the integrity of the entire state accountability system is thrown into question when districts with excessive dropout rates are given “acceptable” and even “recognized” ratings.

But we are not here today to debate statistical approaches for calculating dropouts. IDRA’s approach to calculating attrition is sound, and we will continue to employ it and share it with the public. We have provided information to the committee on IDRA’s attrition analyses and have also shared our comparisons of IDRA attrition estimates and TEA’s dropout estimates. We are clear that IDRA’s numbers are a closer approximation of the numbers of Texas pupils who are being lost by the public school system. We will continue our approach as a way of providing you and the public an independent measures of the dropout problem in Texas and a way of cross-checking the rates that are generated by this state.

We at IDRA also want to offer some assessments regarding the proposed new method for calculating Texas school dropout rates that tracks individual pupils all the way through the Texas school system and assigns leaver codes to those who leave the system prior to graduation. TEA’s new school leaver reporting system has the potential to – for the first time – provide data on the number of individual pupils who actually receive a Texas high school diploma and those who do not – which is the question that is at the center of the controversy about the accuracy of dropout reporting in this state. Using the state’s new school leaver system we have a new opportunity to provide more precise student-, district-, and state-level reporting on high school graduation and – more importantly – to account for the array of pupils who do not earn a regular diploma for a number of different reasons.

The potential of having an accurate report on Texas school dropouts, however, can be severely compromised if the state continues to minimize the magnitude of the state dropout problem by lowering the counts through manipulations of the leaver codes in any new dropout calculation procedure. IDRA raises the concern because this kind of manipulation of the numbers has already occurred in the dropout calculations incorporated into TEA’s first report that used this new data.

Table 2 that we provided you is labeled “PEIMS Leaver Codes Used in 1996-97 and/ 1997-98.” This was pulled from the TEA report entitled 1996-97 and 1997-98 Returning and Non-Returning Students in Grades 7-12, released on May 14, 1999, that for the first time made use of the state’s new lever reporting system.

In this table, you will notice that the state now provides for reporting of student status for cohorts of students using the new school leaver codes that are part of a new student status reporting system developed by TEA. The agency should be commended for creating this new student leaver code tracking system. On that form you will note that there is an opportunity for reporting the status of students who have been enrolled in Texas public schools and specifying whether they: (1) graduated from high school, which is given a code of “1”or (2) are a “leaver.” The school leaver codes include 36 options ranging from “deceased” (assigned a code of “03”) to “withdrew for reasons unknown” (assigned a code of “99”).

However, the new procedure that uses the school leaver codes already has incorporated flaws into the formula used by the state to calculate “leavers” and “completers.” The TEA document referenced provides summary information about the sub-categories for reported “leavers” and whether those sub-categories of leavers were included in or excluded from 1997-98 dropout calculations for individual schools and districts.

IDRA is concerned that the dropout calculating procedures used in this categorizes too many pupils as non-dropouts who did not graduate and receive a diploma, young people who should be considered dropouts. These include, for example, all GED students – whether enrolled in a GED program or completed a GED program – because in order to participate in such programs a person must withdraw from the regular high school program.

Leaver categories that are not included in dropout calculations that IDRA finds questionable are listed below.

A second major flaw in the new school leaver-based dropout counts, is the former state education commissioner’s decision to exclude all pupils for whom no leaver data were provided from the calculation of school district dropout counts in the report. The commissioner explained that because this was the first year that the leaver codes were being used and the inclusion of the missing data in dropout counts would make many districts have significantly higher dropout counts, he believed that schools should be given one more year to account for all pupils before such unreported pupils are included in dropout calculations. IDRA is deeply concerned that, unless such missing data is included in the dropout counts, schools will have another loophole to underreport the numbers of dropouts in the Texas student accounting system.

A final observation regarding the school leaver data concerns the large numbers of pupils reported as withdrawn for “reasons unknown,” which accounted for 18.1 percent or about one of every five of all school leavers reported in 1997-98. While it is conceivable that a small percentage of pupils may leave Texas schools giving little information regarding why they are leaving school, lack of data on so many pupils and why they left provides very little support for developing an intervention system that will effectively re-enroll these students or prevent them from dropping out of Texas schools.

IDRA proposes that creating some state specified limit on the percentage of pupils that may be reported in this category (to include sanctions for exceeding those percentages) would provide schools a greater incentive to confer with prospective leavers and thus provide critical information that might be needed to prevent our young people from eventually dropping out of school.

IDRA recommendations related to use of the school leaver code-based dropout reporting system include the following.

1. TEA should be supported in developing and refining a valid, reliable and accountable school leaver reporting system for all schools in Texas.
2. These school leaver-based data should replace the current longitudinal dropout calculation procedure used to calculate school and district dropout rates used in assigning ratings to schools and school districts in the state school accountability rating system.
3. The state longitudinal dropout rate should be based on the number of pupils from a cohort of pupils (e.g, in grades seven through 12) who actually do not receive a high school diploma (in Texas or elsewhere). Specifically, graduation and dropout rates should be calculated based on the number of students who were originally enrolled in each cohort (seventh grade and ninth grade) divided by the number of students who receive a high school diploma. The remaining pupils should be identified as dropouts. Breakouts of sub-groups of students by ethnic group should also be generated and included in the accountability system.
4. There should be a specific accounting that includes all pupils in the cohorts who did not receive a high school diploma, including a breakout of the numbers of pupils included in all sub-categories of school leaver codes provided in the new state school leaver reporting system.