Hi Ben,
This is Kelly Williams, a member of your PLC. I need your help. Starting in 2015 in the state of IL, a minimum of 30% of our teacher evaluation will be based on student growth. This June our WL teachers will meet to create pre-tests and post-tests for all 4 levels of French and Spanish for all 4 skills – listening, reading, writing, and speaking. These tests will be used to judge our growth for the evaluations. I read your post on the PLC about the CI exit tests that you give at DPS. You gave an example of the listening test. I realize that your tests are the intellectual property of DPS and cannot be shared. But would you be able to tell me the format of each test to give us something to go on? We are also struggling to figure out how to feasibly give all of our students speaking tests.
Could you tell me how the speaking pre-tests/post-tests are set up at DPS? Half of our teachers are CI and half are more traditional. We CI teachers do not want to end up with grammar tests. Anything you could tell me would be greatly appreciated.
Sincerely,
Kelly Williams
My response:
Yes this is an incredibly important thing to talk about before the grammar teachers taint the instrument you end up with in their favor with 1950’s types of questions. You really need to hear Diana Noonan on this topic since she is the originator of every detail, and there are a lot of details, in our DPS WL testing program. I will ask her to comment here.
I hope we all appreciate what a big deal this is and I would welcome comments from anybody that has anything to say on this topic since it is going to affect us all in coming years as the research slowly chokes off the grammar approach to testing. Five years ago the grammar teachers would have had their way on the test design you are facing but since it seems that your school is split there is hope for a test that actually refects current and not outdated research (actually there was never any research to support the grammar/translation method, there never was…).
Here’s what we do in DPS:
1. LISTENING. Paul and Sabrina, who have the two best French accents, filmed a series of short dialogues. Each was laboriously written by our writing team last summer. So the kids on the listening test, which takes about a half hour and consists of seven or eight dialogues with two or three questions about each one, have the multiple choice answers in front of them in the test booklet and look up at the screen and listen to the dialogue and answer the questions.
2. SPEAKING. Three strips of images just like a comic strip, with three or four panels in each one, is given to the kids. These are the speaking prompts. Each student comes up to the teacher’s desk and picks one of the strips and is given a few seconds to prepare a short description of what they see, then they are invited to start talking. This can be done by the doorway to the classroom while the other kids watch a video or something.
Now the other parts of the test, writing and reading, you didn’t ask about so I won’t address. The writing part is still a problem in the district because each teacher (we have 100 WL teachers in DPS) will grade their students writing in very different ways. So we have certainly not perfected the test yet. We used to get subs for every teacher twice a year to meet and grade in a very stringent process but were denied those funds this year.
One big thing is the fine rubrics we have created for speaking and writing. They are incredible, really, so easy to follow to determine where the student is on the ACTFL proficiency scale (NL, NM, etc.) – each year those rubrics have gotten more refined.
OK now I will try to get Diana to address this. She will also be at iFLT this summer, as she and Carol Gaab are the organizers, and she is the Keynote Speaker at NTPRS in Chicago this year as well. She really is the best one to ask about all the details.
Just one other point. We weigh the speaking and writing at 15% each, and the listening and reading at 35% each. I fought against this but that is what we came up with. In spite of the fact that we now know that kid with just a few hundred hours of exposure to the language can’t really speak for real in any authentic way, the output skills got 30% of the grade. Oh well.