Articles tagged with Atlanta Journal Constitution

NSBA’s Center for Public Education reviews AJC’s investigation on school cheating

The following was also posted on the National School Boards Association’s Center for Public Education’s blog, The EDifier.

Last week I showed that the data from the Atlanta Journal Constitution’s (AJC) investigation on cheating, while certainly accurate, was not evidence of a widespread epidemic. Yes, there certainly was evidence that cheating was likely taking place in some districts. While this is inexcusable, it’s important to also remember that those instances of cheating involved few enough students that they wouldn’t have any impact on the overall state or in many cases even district results. In this case, one rotten apple wouldn’t spoil the whole barrel.

Despite the fact that AJC found that 24,000 out of 13 million students  from across the country( which is about two tenths of one percent) attended grades where cheating was likely taking place, AJC chose to declare that cheating is common practice in our nation’s schools by running an opinion piece by Robert Schaeffer of the anti-testing organization FairTest. The piece stated, “Experts may debate the methodology, but there is no question that cheating on standardized exams is widespread.”

Experts may debate the methodology, but Schaeffer is flat-out wrong to say there is no question that cheating is widespread. Whatever your personal measure of “widespread,” declaring that cheating is widespread when the data shows less than a percent of students may be involved is quite a stretch.

Schaeffer makes the same stretch when he says that his organization has confirmed cheating in 33 states and D.C. over the past three years. I don’t think a handful of unethical and misguided educators in a state means that the whole state should be labeled as cheaters.

Here’s another statistic to remember. There are hundreds, if not thousands, of schools in most states. If your child attended a school in one of those 33 states or D.C., the chances are slim their test scores were manipulated in any way. So the public should continue to have confidence in the results of the state assessments. While we must eliminate cheating, don’t let one rotten apple spoil the whole barrel of good information we can gather from state assessments.

Jim Hull|April 5th, 2012|Categories: Assessment, Center for Public Education, NSBA Opinions and Analysis, Student Achievement|Tags: , |

Is cheating prevalent in our public schools?

The following was also posted on the National School Boards Assocation’s Center for Public Education blog, The Edifier.

Articles this past weekend by the Atlanta Journal Constitution (AJC) and the Associated Press  strongly suggest the answer is yes.  AJC attempted to answer this question by analyzing state standardized test scores from all 50 states to identify districts and schools that had statistically unusual fluctuations in their year to year test scores which is an indicator that cheating may be taking place. Although the unusual fluctuations do not prove there was cheating it does point to the strong possibility that cheating is in fact taking place. As a matter of fact, the newspaper used a similar methodology in 2009 which helped uncovered extensive cheating in Atlanta public schools.   

But is cheating as prevalent across our public schools as the articles strong imply? The answer is quite simply no. When you actually look at the data from AJC you see that about 200 out of the nearly 15,000 school districts (which includes charter school districts and other special districts) analyzed by AJC had suspicious test scores like those found in Atlanta. This represents just 1.3 percent of all school districts nationwide. Keep in mind, even within these districts most schools showed no signs of cheating.

In fact, when AJC calculated how many individual students were likely to be directly impacted by cheating, just a tiny fraction (less than 1 percent) of the 13 million students examined were enrolled in the grade level within the schools where cheating likely took place.

Of course any cheating at the school or district level is not acceptable but the data does show that the AJC’s assertion that the results “…suggest a broad betrayal of schoolchildren across the nation” is not only overblown but outright wrong.  In fact their data shows that cheating is limited to a small proportion of districts and even smaller proportion of students nationwide. So parents and the general public should be confident that teachers and administrators in close to 99 percent of districts act in an ethical and professional manner.

Does this mean cheating shouldn’t be a concern? Of course not.  More work needs to be done by state and district leaders to ensure that the integrity of the test results are not compromised and that struggling students are appropriately identified so they receive the support and resources they need to actually improve their performance.  In the case of the districts identified in the article they need to look further into the data to determine if in fact cheating is going on in their schools. Or whether the fluctuations are due to highly effective instruction in high scoring grades and ineffective instruction in following low scoring grades. Either way, districts need to know why there are such fluctuations so they can either eliminate any cheating or focus on improving instruction in grades where test scores drop.

While statistically large fluctuations in scores indicates a strong possibility of cheating, as anyone who has seen the movie Stand and Deliver, great teaching can lead to unpredictable increases in student achievement. And those teachers should not be considered guilty until proven innocent. But it also doesn’t mean that such indicators of cheating should be ignored either.  Either way, the actual data shows that cheating is limited to a small number of schools nationwide contrary to what the AJC and Associated Press articles imply.

Jim Hull|March 29th, 2012|Categories: Center for Public Education, NSBA Opinions and Analysis, Reports|Tags: , , , |
Page 1 of 11