State releases La. PARCC results

Louisiana students performed “consistently” with previous years, according to spring standardized test results released Monday by the state.

At most grade levels, about 30 to 40 percent of the students who took the Partnership for Assessment of Readiness for College and Careers (PARCC) questions scored “mastery” or above, which State Superintendent John White said shows readiness for community college and universities.

These results were typical in comparison to previous years’ data, White said in a media call Monday.

“These results are not surprising,” White said. “They are roughly consistent with how Louisiana students have performed on the LEAP and the NAEP (National Assessment of Educational Progress).”

The preliminary data show a range for students scoring at least mastery in English — from 33 percent of fifth-graders to 40 percent of eighth-graders at mastery or above in English. Scores were slightly lower for math, ranging from 22 percent of seventh-graders to 37 percent of third-graders scoring mastery or above.

The number of students scoring “basic” on the tests seems to be is shrinking as compared to previous years’ standardized test scores as more students are scoring either mastery or approaching basic.

These results are based on “cut scores” that go before the Louisiana Board of Elementary and Secondary Education on Tuesday. BESE is to decide score ranges representing a student’s achievement level, like scales used on Advanced Placement and the ACT. There are five levels ranging from unsatisfactory to advanced on the 850-point PARCC scale.

“They are the exact same cut scores as every other state in the country that takes the PARCC test,” White said.

After BESE sets the cut scores, DOE staff will convert the raw data to the PARCC scale and set off a domino effect for the state’s accountability and choice systems.

Once the scores are scaled, the department will generate and release reports to schools and districts, which helps determine school performance scores and letter grades that the state hopes to issue by Oct. 30. The department then would issue individual student reports for districts, teachers and families “detailing scores and skills for every student” in November.

Many have voiced that the department has taken too long to release scores, but the release could be delayed further should BESE not approve any cut scores Wednesday. A vote to delay cut scores would delay the entire accountability system.

Louisiana’s contracted vendor and publisher Data Recognition Corp. scored individual test questions in June, providing “raw data,” or the percent of questions answered correctly out of the total number of questions. Then the DOE’s “double-checked” that each question was recorded properly and such. He said there were no “major” discrepancies, if any, between DRC’s scoring and the department’s.

“Every test was scored exactly the same way,” White said.

Some districts have requested and received raw data, including a 32-person coalition that filed a formal public records request for such data on Oct. 1.

On Monday White cautioned that raw data would not provide a complete picture or comparable data. He said the PARCC tests more than one standard in one moment in time as a general pop quiz might.

“PARCC is looking for multiple standards over the course of a school year,” he said.

Students answer multi-step problems that can test several standards in one problem.

“Because the test asks students to do so much, we don’t grade it on percent right,” White said.

Some district officials have said they intend to compare the raw data with the converted scores after BESE sets the cut scores.

Mike Deshotels, retired educator and author of The Louisiana Educator blog, said the raw data and conversion table he received as part of the group’s public records request show a lowering in cut scores. He said he found that for eighth-grade students to score “basic,” which is to pass, in English would need to answer 37 percent correct, and the percentages drop for the lower grades and for math. The lowest is 22-percent correct needed for eighth-graders to score basic in math if the scaled cut score is set at 725, Deshotel said.

Deshotels called this “sugar coating a bitter pill” rather than letting the scores reflect how “poorly designed” PARCC is.

But White said “cut scores have never been lowered” and that the scoring process for standardized tests is not based only on percentage correct.

These results from 2015 and 2016 will combine to form a “baseline” measurement of Louisiana performance on new standards, in comparison with other states. Louisiana is the first state to fully report its scores out of the 11 states and Washington, D.C., that took PARCC this spring.

From this baseline, BESE has committed to creating a steady transition toward 2025, when an A‐rated school in Louisiana will have an average performance of “mastery” rather than “basic,” as is the case today. This means that each year between 2017 and 2025, the state’s accountability system will increasingly reward “mastery” results more and “basic” results less, according to the state.