Faced with reports of technology glitches in online testing, the state Board of Education last week waived the requirement that Georgia Milestones scores influence promotion and retention decisions in third, fifth and eighth grade.
That waiver does not mean elementary and middle school students who failed a Milestones test -- parents are supposed to learn Thursday or Friday how their child fared -- won't have to retake it. Their children could retest over the next two weeks, depending on the policy of local districts. The state school board left the retesting question in the hands of the districts.
(High schools did not experience the same level of technical glitches, and, at this point, End of Course tests will still count for 20 percent of a student’s final grade. Those scores began reaching districts today.)
There’s strong motivation for districts to retest low-scoring students. While the Milestones may not count for retention, the scores could still count for next year's College and Career Ready Performance Index, a complicated state Department of Education rating system that relies heavily on test performance.
“It is simply too early for us to make any determination about any impact on the CCRPI. As we communicated to districts, we have sought this pause in End of Grade test scores to allow time to conduct additional analyses to detect any potential impact resulting from the interruptions for the students who experienced them,” said DOE spokesman Matt Cardoza.
Some systems experienced no computer issues with the Milestones. So, while Cobb Schools will not use the Milestones for retention, the district will retest students Tuesday through Friday.
Clarke County chose not to retest students due to the computer problems. While Clarke had fewer snafus last month than last year, it still had students who lost connectivity during testing and had to wait 15 minutes or more to reconnect. In 2014-15, Clarke tested 99 percent of its students online. This year, it tested 84 percent.
While the DOE blamed most computer meltdowns on local capacity, technology directors around the state cite a design flaw with the Testing Site Manager backlog -- when students lost connectivity during testing, the system required 15 minutes or more to store completed answers. (You can find in-depth discussion of testing problems by school IT directors here. And you can see here that such problems are not unique to Georgia. ) Because students in a single classroom lost connectivity at various times, testing schedules were disrupted.
“Losing connectivity is not an overwhelming issue if a student can log back in and connect in two or three minutes,” said Tim Jarboe, Clarke County Schools Director of Assessment and Accountability. “But when you have one child taking 15 minutes and another taking maybe a half an hour, then you have a group of kids sitting there getting nervous.”
“We are not using the scores for retention and we are not conducting retests,” said Jarboe. "Because CCRPI is a public accountability document and directly uses Milestones scores, there are political implications in that decision as the state counts the highest performing scores for CCRPI calculations. But, in talking to our superintendent and principals, we all felt very strongly what’s right by students is to not do retests. If we don’t know whether the technical issues have been fixed, we just feel it is the right thing. We had a lot of students experiencing interruptions this year during testing. Why make these students go through that again? Or have that stress of worrying -- despite whatever happens with CCRPI.”
Jarboe said Clarke County Schools expects to receive test scores tomorrow from the DOE and will send them home to parents explaining “that while the scores will not be used for promotion and retention, we want them to have this information”
Some districts may still consider End of Grade scores in retention decisions, despite the state waiver. The Macon Telegraph reports: "In Bibb County, however, test results won’t completely be thrown out as a factor for whether students move on to the next grade. Superintendent Curtis Jones said that more than 71 percent of the county’s tests were done online and no glitches were reported...district educators and administrators will still consider test results when decisions are made on promotion, even if the state rule is relaxed."
This patchwork quilt of retention/retesting would seem to undermine Georgia’s quest for common standards and practices around state testing. DOE disagrees.
“The standards are common: all students learn them in the classroom. Like many things, local control allows for decision-making at the local level. We set the minimum requirements bar, and districts can set a higher bar if they want,” said Cardoza.
“This year’s online administration was a significant improvement over last year’s and the number of interrupted sessions is significantly fewer. Nonetheless, in an abundance of caution we sought the waiver for promotion and retention from the State Board of Education,” said Cardoza. “We will produce a CCRPI for the 2015-2016 school year. Just as we did for the 2014-2015 CCRPI, we will use what we learn from the analyses to make informed decisions about CCRPI calculations stemming from this year’s administration.”
Last year, DOE analyzed testing data to see how many students were undermined by technical problems. (DOE can track testing interruptions.) In Clarke County, the tally was 6 percent of math test takers. As a result, those students' scores were not factored into the district’s CCRPI grade.
“We thought that was very fair,” said Jarboe. “The state said it would do the same thing this year. DOE has been very responsive to me when I have called with issues and problems. Sometimes, the problems have not been fixed as fast as I wanted, but they have been responsive. Unfortunately, it doesn’t help when kids are sitting there who can’t get into tests.”
"Electronic, online testing is an innovation that our district fully supports as an assessment method. Testing is more efficient, students are engaged at higher levels and technology-enhanced items can better measure a student's ability to problem-solve. However, innovation usually comes with growing pains in its implementation, and that is certainly true in this case," said Jarboe.
"Therefore, it is unfortunate that state legislation and the State Board of Education have made Milestones testing a high-stakes event for students and teachers," he said. "The difficulties with its implementation have resulted in frustration and anxiety for students and teachers, which has diminished much of the innovative value of the online platform."