How closely should virtual schools’ feet be held to the proficiency fire? Should we provide them more time before we make comparisons between them and traditional schools? Regardless, what are the reasons virtual schools perform more poorly than the general student population? All good questions. I appreciate Rob Darrow’s detailed response to my earlier post and I attempt to reply, adding more data, below. Rob’s comments are in bold. My replies follow each of his points.
First, it is difficult to judge any school with one set of data and make conclusions.
I totally agree. Multiple choice tests are just one way to demonstrate proficiency. Unfortunately, they reinforce knowledge and comprehension over higher order thinking skills, but as they say, this is all we have at the moment.
Second, research regarding Algebra in California shows that many students who do not pass Algebra by 8th grade have difficulty showing proficiency in Algebra in 9th-11th grades.
You make a good point.
Third, you compare Rocketship Schools, which are K-5 schools, with online high schools (9-12), which is not a valid comparison.
I included Rocketship only to demonstrate that blended online learning models have great potential to improve student learning. I excluded their data from my analysis, though.
Fourth, the majority of charter school students are “at risk.”
Assuming that the majority of virtual school students are “at risk”, let’s compare apples with apples. This week, we built a spreadsheet that included 203 continuation high schools and independent study programs in California. How do they compare to California’s virtual schools?
Our spreadsheet includes 28 independent study programs and 175 continuation high schools ranging in size from 20 to 800 students. Average size: 100 students. Median: 87.
ELA proficiency levels ranged from 0% to 80%. The median proficiency level was 17%, meaning that half the continuation high schools and independent study programs had no more than 17% of their students score proficient on the California Standards Test (CST). Compared to the virtual school median of 48%, alternative education programs scored much more poorly.
94 of the 203 alternative schools (46%) had less than 10% of their students score proficient on the ELA California Standards Test (CST). None of the virtual schools scored below 10% proficiency. Only two of the 203 alternative schools met this year’s NCLB goal of 66.7% proficiency versus three virtual schools (7%).
147 of the 203 alternative schools have Algebra I data. Algebra proficiency levels ranged from 0% to 46%. However, discounting the highest performing alternative school, the next highest proficiency level was 25%. The median proficiency level for the 147 schools is 8%, meaning that half of the alternative schools had no more than 8% of their students score proficient on the Algebra I test. The virtual school median score was 17%. 58 of the 147 alternative schools (39%) had NO proficient students in Algebra. Only two (6%) of the virtual schools had no students score proficient.
Assuming that Rob is right that virtual school students are similar to their alternative school counterparts, virtual school students performed much better in both English-language arts and algebra than their alternative school counterparts.
Still, one should also look at statewide averages for the CST. Overall, 54% of California’s students scored proficient in English-language arts and 50% of students were proficient in math. Subgroup proficiency is an interesting comparison. 23% of English Learners (EL) scored proficient in ELA while 37% of EL students were proficient in math. So, overall, English Learners performed poorly in ELA compared to virtual school students while they outperformed virtual students in math. (OK, let’s say that again. English Learners performed better in mathematics than virtual school students.)
Fifth, it is difficult to judge charter schools and online charter schools when they are just starting up.
Fair point. However, does this mean that parents should think twice about having their children take online courses from schools that have just started up? Should they hold off while virtual schools work out the bugs? I don’t think so.
Still, let’s look at the data from virtual schools that have been in operation for more than two years. 19 of the virtual schools have been in operation for at least three years and the median start up date for this group is 2006. Six of the 19 schools are in year two of Program Improvement, meaning that they haven’t met targeted goals for four years. Their ELA proficiency levels range from 26% to 71% with the median proficiency level being 53% (in line with the California average for all students). “Start-up” virtual schools, those that began during or after the 2009/10 school year, had proficiency levels ranging from 13% to 76%. Their median level was 39%, so Rob could have a point that it takes a few years for a virtual school to find its legs.
Mathematics scores for start-up online schools ranged from 5% to 19% with the median being 11%. By comparison, established virtual schools ranged from 4% to 85% of their students scoring proficient in mathematics. The median proficiency level for this group was 8%, so again Rob could be right that start-up virtual schools have some bugs to work out.
I’m not sure that we should excuse them for their poor performance, though.
Sixth, students move in and out of online charter schools on a weekly basis.
As this is also true of alternative schools, I think comparing the two is fair.
Seventh, it is important to look at how many students were tested.
There were 20 virtual schools with more than 200 students, ranging from 215 to 4000 students. All but two have been active more more than two years. The median ELA proficiency level for this group is 47% (48% for the original group) while the median mathematics proficiency level was 22% (17% for the original group). The median Algebra proficiency level was 8% (8% for the original group). So, you could say that larger virtual schools performed better in mathematics than the group as a whole, while they scored about the same in English-language arts and algebra.
Finally, it takes from 3-5 years for any good system to go from good to great.
Agreed. It also takes time and competition for online course curriculum to shed the instructional paradigm of “lecture, worksheet, test” in favor of teaching methods and activities that reflect a greater emphasis on higher order thinking skills. We all want online learning to succeed. In the meantime, we should hold virtual schools to the same accountability standards as all schools.