Eric continues:
Good rubrics may be the most valid way to assess proficiency, but the result depends on the person grading. If you want reliable measures, then you have to spend time and money on training people to use the rubric and then have multiple people grade the same test. Ideally, the testers & scorers would not be the student’s teacher.
I’ve discussed, as well as Chris and others have commented, on how poor the NYS SLP exam is in measuring proficiency. And yet, I know that at least one of the TPRS vs. Traditional comparative studies shows TPRS to be superior, but uses the NYS SLP to measure that.
http://www.tprstories.com/ijflt/IJFLTSummer-09.pdf
The listening and reading on the NYS SLP is crap, because it doesn’t measure comprehension, but rather whether a student can pick out 1-2 words and the multiple choice answers are unrelated, so there’s an obvious choice.
I do like some of the NYS SLP rubrics. The writing rubric has 4 dimensions: satisfies the task, vocabulary, grammatical accuracy, word count.
I really like the way the speaking is rated: the teacher counts the student utterances that are unique, appropriate, and comprehensible. I don’t like the role-play situations, but on each task, the student only has to make 4 utterances (an utterance can be a 1-word answer). 1 extra point is given for “quality” if all 4 utterances contain evidence of fluency, complexity, and accuracy. In other words, 80% of the speaking test just counts quantity and is comprehensibility-based.
In MA we need 2 of these student-growth measures, but from what I understand, they can test whichever skills I want, not all 4. And we were told that as long as the Principal and then the Superintendent of the District approves the measures, then they are accepted. I also believe it is okay to do the testing in different academic years, e.g. pre-test at the end of 7th grade and the post-test at the end of 8th grade I don’t know how true all that is, because this system seems to change day by day and no one really knows.