IB Proponent Just Makes Stuff Up

In Southern California's Press-Enterprise Newspaper, reporter Dayna Straehley pens an article headlined: Test scores prove International Baccalaureate program's mettle.

Since it began a program proposed for Eastvale's next school, test scores at an elementary school in Walnut rose 41 percent, the principal told Corona-Norco school officials Wednesday.

"We're still a Title 1 school," serving a number of economically disadvantaged students, C.J. Morris Elementary School Principal Susan Arzola told officials visiting from the Corona-Norco Unified School District...

Morris' Academic Performance Index rose from about 600 before the program to 847, Arzola said.

Now I'm surprised by this headline because the only academic study I've seen on the International Baccalaureate program shows that the IB-MYP leads to a mixed result on standardized tests. But this is the IB-PYP, so it is worth investigating.

First, we must determine when C.J. Morris Elementary School became an IB school so that we can compare these claims. According to the IBO, Morris Elementary became and IB school in July 2006.

The next step is to compare Morris Elementary's Academic Performance Index (API) before and after that date. The State of California helpfully posts this info on the web.

2007-2008 API for Morris Elementary: 846
2005-2006 API for Morris Elementary: 836

Morris' API is 846, not 847 as Ms. Arzola claimed, but that's small potatoes. Morris Elementary's API score went up 10 points. I don't think it is appropriate to calculate a percentage increase like Ms. Arzola did in the article, but Morris Elementary's API score increased 1% since it became an IB school, not 41%. And Morris Elementary's pre-IB API of 836 is not about 600 as Ms. Arzola claims.

[The average API in Walnut Valley USD elementary schools went up 1 point over this period with a standard deviation of 12. This means that Morris Elementary outperformed its peer group by 0.75 standard deviations. Although professional statisticians would prefer to see a variation of around 2 standard deviations before declaring an observation to be statistically significant, we use a much, much lower threshold. Back in May 2008, we agreed to define variations greater than 1 standard deviation to be significant. So even with our much lower threshold, Morris Elementary's improvement is still insignificant.]

We know that Ms. Arzola's statement is simply inaccurate. Maybe, though, Ms. Arzola is employing some sort of Clintonian version of the truth. We just have to figure out some way to make Ms. Arzola's statement technically true even though it is blatently misleading.

Again, the State of California is helpful. We have Morris Elementary's API scores going back to 1999. Morris Elementary's API score was not "around 600" during that period.

I have seen this stretching of truth beyond the point of dishonesty time and again with respect to the IB Program. If the IB Program is so good, why can't its supporters tell the truth? What values do students in the IB Program learn when they see the behaviors of the Program's adult advocates?

===

I want to make my position clear on this matter:

The best way to determine whether the IB Program improves student performance is to create two separate groups to study, one with students preparing to go through the IB Program and one with students going through another academic program. The two groups should perform similarly on standardized tests at the beginning of the study. Several years later, we should compare the performance of the very same students to see if one group outperforms the other.

I have seen such a study as I alluded earlier. When the sample was adjusted to match the demographics of the school studied, standardized science and history test scores for kids who went through the IB-MYP program in Virginia were marginally lower than peers in the same school district who went through traditional education programs. IB students performed marginally better in English and math. Nothing was statistically significant (except the size of the school district's checking account).

Evaluating changes in a single school like was done in this Morris Elementary example is an easier task. But there are many factors that could influence results and lead to faulty conclusions. Demographics of a school could change. [Recently in Pearland ISD, there was virtually 100% turnover at Jamison Middle School due to rezoning.] Sometimes, one age cohort performs significantly differently than other cohorts at the same school for no apparent reason. So even though this kind of analysis seen at Walnut Valley is easy, we must be aware of its limitations.