Critics raise questions, professor says JCPS survey lacks crucial question


Posted on January 30, 2011 at 12:34 PM

Updated Sunday, Jan 30 at 1:28 PM

Louisville, Ky. (WHAS11) - Critics of the embattled student assignment plan are dismissing a survey of Jefferson County Public School parents, saying both that the questions were skewed and that a consultant hired by the district is emphasizing only the findings which support the school board's intended aim.

And, a University of Louisville political science professor said, despite sound methodology and sample size, the survey lacks precision because it fails to ask the direct question whether parents prioritize diversity over student achievement.

The survey was presented to Jefferson County Public School board of education on Thursday night by Gary Orfield, the head of the Civil Rights Project at UCLA.  It showed that while about 90 percent of parents support the school district's diversity goals, only about 55 percent would want their child's school assignment based on fulfilling those goals.

Flanked by parents and a retired teacher dissatisfied with the student assignment plan, Teddy Gordon, the attorney whose lawsuit prompted the U.S. Supreme Court to order  JCPS to scrap its old student assignment plan, says the district is still more interested in its diversity agenda than educational achievement, and that the study was rigged.

"The questions were skewed to get the desired result accomplished," Gordon said.

Gordon mocked Orfield's characterization that "a substantial proportion, 43% of people believe that the decades of integrated scools have improved the greater Louisville community.  That means," Gordon concluded, "that 57% believe it has not improved."

"We've been asking these questions for years and years," said Shawn Herbig, President of IQS Research, the Louisville firm that conducted the poll of JCPS parents, adding that previous polls had produced "reliable and actionable results."

Herbig said the methodology of the poll is "very transparent and straightforward."

"If we can focus more on the science, maybe we can tone down some of the opinion," Herbig said, "It's still going to be a complex issue but the science is what the science is."

Herbig's firm collected the data.  Orfield and a UCLA team compiled the report.

"Unfortunately, what gets lost in that is the facts and the reality of what the research is actually saying," Herbig said.

Belinda Abernathy, who is homeschooling her kindergartener to avoid a crosstown bus trip, said she does not believe the conclusions reached by the JCPS consultant.

"You can ask anybody anything to try to get the answer that you want," Abernathy argued, "So I don't think anybody should put a lot of stock in his numbers."

Herbig denied that the survey was engineered to produce a favorable result.

"The simplest answer is no," Herbig said, "I don't walk away with any concerns whatsover that they had a slant either in their initial agenda or the way the questions were formulated."

At WHAS11's request, the national polling firm, Survey USA, reviewed the methodology and questions in the JCPS poll.
"This is an exceptionally detailed, meticulous piece of research," commented Jay Leve, Survey USA CEO, "The questions that were asked are fair, straightforward, not one-sided."

Both Leve and University of Louisville Political Science Professor Laurie Rhodebeck, however, said the poll did not adequately explain why two pools of respondents were questioned, from "area A" and "area B."

"The report fails to explain why the samples are divided into households of students K-2 grades and 3-12 grades and households in 'area A' and 'area B.'" Rhodebeck wrote in her review of the survey, also requested by WHAS11, "Comparisons are made throughout the report based on a four-fold comparison of K-2 A, K-2 B, 3-12 A, and 3-12 B.  It would be helpful to include the information that A and B are distinguished by the criteria of median household income, the educational attainment of people aged 25 or older, and percentage of minority students residing in the area."

And Rhodebeck said that a crucial question is missing from the survey.

"The problem I see." Rhodebeck wrote, "is that the parents were not asked to choose between supporting diversity and enjoying the convenience of a neighborhood school.  Some inferences can be drawn by examining the percentages of parents who said that a particular factor was important to them – e.g. educational program, student diversity, geographic location."

Deborah Stallworth, an outspoken critic of the student assignment plan, suggested that the poll deliberately avoided that comparison.

"57% of the people in this community said, 'no, I don't believe all this busing improved academic achievement.'  The survey says that," Stallworth explained, "but we want to pick and choose what we hear from the survey."

The survey is replete with contradictions.

90% of parents believe that diverse schools have important educational benefits for their children yet only 55 percent said they would send their child to a school
outside of their neighborhood if that would help the district achieve diversity.
"Was it arbitary?" Abernathy asked, "Was it random?  How did he make his selections?"

"Every household that had a student in Jefferson County Public Schools had an equal chance of being selected for the study," explained Herbig.

More than 1800 JCPS parents were polled -- a random draw from a list of about 50,000 households supplied by the school district.

"Just because the number seems small in relation to the universal population, it is still representative of that population," said Josh Holt, an IQS Research analyst.

In fact, Rhodebeck's review said the sample sizes in the JCPS study – 1852 for parents, 1095 for students – "are comparable to the sample sizes of nationwide surveys.  These local samples represent subsets of populations that are far smaller than the populations represented in national surveys.  In that sense, the local samples are actually better than “adequate”; they represent a substantial percentage of the larger population from which they were drawn."

"Also to the credit of the sampling technique are the inclusion of students from all the high schools in the district, with samples in proportion to the enrollment of the schools, and the ability to contact parents in households that lack a 'land line.'" Rhodebeck explained.

Only current JCPS parents were polled, meaning that dissatisfied parents who pulled their child out of school are not representted.

"Just statistically alone we know that when we started this that we had about 100,000 students," Gordon said, "and now we have 92,000.  We've lost ten percent.  So why don't you ask the ones that are supporting the real estate values of Oldham County, private schools, the public schools and home schools whether or not they are or are not satisfied?  If you take that sample, it will be 100 percent against the student assignement plan."

"The only way you could answer how close those two sets of opinions are," responded Herbig, "is to do a completely separate study and poll those other people."

Survey USA's Leve, said "this study appears to have been conducted with great care and to high standards," yet asked because 80 percent of the parents surveyed were mothers, if any effort should have been made to gender-balance those polled.

"It is unclear from my quick reading if moms think differently than do dads on integration," Leve wrote, "It is also unclear how many of the families in the study do not have a 'dad,' and therefore mom is the only parent. If moms and dads feel the same way on the questions asked, the distinction is not important."