RE: Exemptions to STAR and other standardized testing

I agree essentially with everything Linda wrote. She is correct to point out that "authentic" does not have to be "subjective", although it frequently is. She is also right to point out the power of a well-written authentic assessment. 

Where we possibly (slightly, I hope) disagree, is where and when such assessment should be used. I am all for in-class use -- among many others -- of authentic assessment. Exactly for the reasons Linda mentions. But using them for external high-stakes decision making is extremely difficult and extremely expensive, as Vermont found out in recent years. So while standardized way of measuring "authentic" tasks can be done in principle, and intensively training enough people to achieve reasonable consistency in scoring is possible in theory, neither almost ever happens in reality. That's why instead of keeping finding explanations of why external authentic assessment didn't work yet another time, I prefer to leave them in the classroom, where they can be adjusted by the common sense of teachers. 

To sharpen the distinction, I should point out that variety of standardized norm and criterion referenced tests are frequently attacked on the grounds of "repeatability" - that the scores of the same student taking the same test multiple times vary. Yet such variance is almost never more than 10% of the score, and almost always less than half that. Compare it with rubric-based holistic scoring, where getting 10% agreement between graders within 1 rubric-level (typically 20%-25% grade-points span!) is considered exemplary, is rarely possible to achieve, and never maintained over time. And it costs a fortune.

So authentic assessment definitely needs to have a significant place -- but in the classroom.