RE: Exemptions to STAR and other standardized testing

JS Kelly wrote:

"the same test is administered year after year; that this is one of the reasons there is so much "secrecy" and security surrounding the delivery and collection of the test booklets. Apparently, it is too "expensive" to come up with all of the norm-referencing & etc for original questions year after year."

The norm-referenced test (NRT) component of STAR (SAT9 until this year, CAT6 starting next year) is indeed "secret" as its composition does not change during the years in which it is administered. It is a huge effort to develop such nationwide norming, and the items stay around for a period of 5-10 years until the next edition of the test. Typically a commercial testing company owns those tests. The benefit here is the meaningful ability to compare cohorts among locations and over time. Arguments about this or that community being "different from the national norm" are irrelevant to the main purpose of this test, which is to place students relative to each other - NRT makes no claims about "why". By construction half of the students are below 50th percentile and half above, although this can vary very slightly over the life of the test if the population achievement (as a whole) moves overall since last norming. NRT does not label any score as "success" or "failure" - those terms have no meaning in this context.

The Calif. Standard Test (CST) component of STAR is not normed, but rather tests how students are doing on the Calif. standards. Here 100% of students can, in principle, succeed (or fail). We do associate labels such as "below basic" or "proficient" with this test, as Calif. believes the Standards to be meaningful benchmarks that essentially all children need, and can achieve. The test items are refreshed at 25% rate every year, and in principle cover all the Calif. standards over time. I personally would like to see the 25% (or most of them) that are taken out be published, as I find them or rather high quality and I think they would help to remove any anxiety about the test. The state and testing company doesn't like it though - they feel they may need to reuse some items one day, if a batch of new items happens to fail on a field test in a given year; so they publish only very few items.

Clearly expense has something to do with it. We spend about $60 milions per year on STAR, which is a rather small (~0.1%) part of the education spending in Calif. Adding to this the actual administration costs at the districts, probably around 1%, still leaves us with a rather small fraction the state spends on quality control. But we want to keep the costs low - people already disingenuously claim that this money would be better spent on other enrichment/enhancement. I wish 1% budget difference spent elsewhere would make a difference in education quality - it already increased tens of percents over the last decade. 

Overall I feel Calif. is doing a very decent job of testing. But some people just don't like testing. Period.