TaKS pro­jec­tion rat­ings at­tacked on their ac­cu­racy

Austin American-Statesman - - METRO & STATE - By Holly K. Hacker and Jef­frey Weiss

DAL­LAS — TAKS re­port cards sent this spring to Texas par­ents in­cluded a new piece of in­for­ma­tion: pre­dic­tions of whether kids are on track to be ready for col­lege.

For some par­ents, the pro­jec­tions brought the dis­turb­ing news that their chil­dren might lose their “com­mended” rat­ings. But the state didn’t tell par­ents that some of those warn­ings could be false.

School dis­tricts were left in the same bind: They re­ceived pro­jec­tions that could be off by as much as 20 per­cent­age points.

Some crit­ics see the com­mended pre­dic­tions as one more rea­son to scrap the Texas Pro­jec­tion Mea­sure al­to­gether.

“If you don’t have con­fi­dence that the mea­sures you’re get­ting from the state are ac­cu­rate, then even the le­git­i­mate suc­cesses are sub­ject to ques­tion. And that’s not fair to the ed­u­ca­tors and the par­ents who work

so hard,” Rep. Scott Hochberg, D-Hous­ton, said.

The Texas Pro­jec­tion Mea­sure was im­ple­mented last year to al­low schools to count as pass­ing any stu­dent who failed a TAKS test but was pro­jected to pass fu­ture tests. Do­ing so helped more than 2,000 schools earn higher school per­for­mance rat­ings.

Pro­jec­tions about com­mended rat­ings were added this year to give par­ents and ed­u­ca­tors an early look at how stu­dents might per­form on harder tests sched­uled to start in three years. A com­mended score on the TAKS in­di­cates mas­tery of a sub­ject.

Hochberg, chair­man of a House sub­com­mit­tee on ed­u­ca­tion, raised doubts about the ac­cu­racy of the pre­dicted TAKS pass­ing rates dur­ing a hear­ing last month. That hear­ing did not touch on the com­mended pro­jec­tions. But The Dal­las Morn­ing News ex­am­ined those com­mended pre­dic­tions and found that many show a trou­bling de­cline.

On Thurs­day, Ed­u­ca­tion

‘You would not want to draw the con­clu­sion that we be­lieve per­for­mance in the state of Texas is go­ing to plunge.’

TEA’s deputy as­so­ci­ate com­mis­sioner for as­sess­ment

Com­mis­sioner Robert Scott told the state Board of Ed­u­ca­tion that the Texas Pro­jec­tion Mea­sure is “sta­tis­ti­cally ac­cu­rate, valid and rea­son­able.” But he has said that he’ll con­sider scrap­ping it or scal­ing back its use.

Other top agency of­fi­cials say the pre­dic­tions are in­evitably im­per­fect and shouldn’t cause alarm.

“You would not want to draw the con­clu­sion that we be­lieve per­for­mance in the state of Texas is go­ing to plunge,” Glo­ria Zyskowski, TEA’s deputy as­so­ci­ate com­mis­sioner for as­sess­ment, said. Nor should par­ents panic if their chil­dren are pre­dicted to fall from “com­mended” sta­tus, she said.

“I would just in­ter­pret it with a grain of salt,” she said.

Glo­ria Zyskowski

Zyskowski said any sta­tis­ti­cal mea­sure used to pre­dict fu­ture test scores won’t be to­tally ac­cu­rate, es­pe­cially for chil­dren al­ready achiev­ing at high lev­els. And of­fi­cials say they se­lected the best model avail­able with feed­back from test­ing ex­perts, teach­ers and oth­ers.

State of­fi­cials knew about lim­i­ta­tions of the Texas Pro­jec­tion Mea­sure be­fore the data was dis­trib­uted to par­ents and school dis­tricts.

“Pre­vi­ous analy­ses in­di­cate that un­der­pro­ject­ing com­mended per­for­mance by 10 per­cent to 20 per­cent is not un­usual,” TEA of­fi­cials wrote in re­sponse to ques­tions from the News. That in­for­ma­tion, how­ever, was not in­cluded on the re­port card sent to par­ents or dis­closed to school dis­tricts.

One ed­u­ca­tion pol­i­cy­maker said the pre­dic­tions are too in­ac­cu­rate to help fam­i­lies or ed­u­ca­tors.

“It’s send­ing mixed sig­nals to schools, to par­ents and to kids them­selves,” said Daria Hall, di­rec­tor of K-12 pol­icy for the Ed­u­ca­tion Trust, a Washington-based non­profit or­ga­ni­za­tion that ad­vo­cates for poor and mi­nor­ity stu­dents. Hall served on a fed­eral panel that re­viewed Texas’ new model, which the U.S. Depart­ment of Ed­u­ca­tion ap­proved in Jan­uary 2009.

“These data, at their core, ought to be pro­vid­ing use­ful and quite clear in­for­ma­tion to par­ents, teach­ers and stu­dents about where they are and where they need to go,” Hall said.

The Texas Pro­jec­tion Mea­sure pre­dicts stu­dent scores on fu­ture Texas As­sess­ment of Knowl­edge and Skills tests based on the stu­dents’ cur­rent scores and av­er­age scores at their schools. Yet some pre­dicted com­mended rates de­part rad­i­cally from re­cent trends.

For ex­am­ple, 46 per­cent of Texas stu­dents who took the eighth-grade read­ing TAKS this spring scored com­mended. Next spring, only 27 per­cent of stu­dents are ex­pected to score com­mended on the same test — an un­prece­dented 19-point drop.

But be­cause the sys­tem may be un­der­pro­ject­ing by up to 20 per­cent­age points, it’s pos­si­ble the com­mended rate could ac­tu­ally go up by 1 point.

De­spite the lack of pre­ci­sion, Zyskowski said the pro­jec­tions are a valu­able tool when as­sess­ing stu­dent progress. “It’s a pow­er­ful piece of in­for­ma­tion, but it’s just one piece of in­for­ma­tion about the stu­dent,” she said.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.