I have chosen to take the current thread on April testing and isolate Eric’s set of excellent questions, presented by him this morning, and make a new thread out of it. So it is a new thread from an old one. We need to answer Eric’s questions, and a lively discussion on this could lead to some good assessment design, which we need.
The original post is here:
https://benslavic.com/blog/april-test-prep/
One of Eric’s questions has to do with assessing listening. In DPS we do that in classes via Quick Quizzes and via translation of content. On the exit test, we do that by videotaping two teachers, the two with the best French accents (Paul and Sabrina) as they talk to each other. The image is projected into the classroom and the kids answer the multiple choice questions in the listening section of the hard copy test.
Anyway, here is Eric’s comment from this morning. Please don’t be stingy with your comments. This has never been properly discussed, to my knowledge, in any academic circles. We haven’t even known what the questions were, but Eric seems to have found some of the key ones.
So here is Eric’s post and Sean’s comments under it. It’s a good start to answering some important questions:
[Quoting me in this first paragraph]:
“Diana had to write them, to turn the pedagogical focus of all 100 WL teachers in our district to CI. Without having control over that change in how we test over the past five years, I doubt that Diana would have been able to change the number of CI teachers in DPS from about 5 six years ago to at least 80 now. Since the exit tests are CI based, any teacher still teaching grammar is screwed.”
Yesss! This is exactly what I have been saying! We all need help in writing CI-friendly, proficiency-based common assessments! That has been a catalyst for change in Denver and could be for teachers everywhere! It will give a lot of aspiring and current CI teachers the freedom to try out and teach with CI. So, I really wish the PLC members of this listserv would help to create a template for this CI exam. Like it or not, testing can spark change.
I have heard what people have said that listening is best tested with translation, but how do you do that except with dictation? And then when I suggested dictation, I was told that puts focus on form, rather than what we want: focus on meaning. So, I suggested reading a story 3xs and having kids write in L1 the story elements. I also want to know where we are supposed to get our listening/reading passages from or else, what are we to use as our guide in writing them? I figured the more high-frequency words included in the passage, the more accurately it represents real-life language. Somebody please critique my ideas. I need some constructive feedback. Don’t worry at all about being direct. That is what I want. If I were ever to propose an idea that was not CI/acquisition friendly, I would WANT people to call it what it is and I would never be offended!
I think testing is ONLY helpful if used to give feedback to the teacher to validate or modify instruction.
“Targeting vocabulary is not something that I believe in, even when time limitations require it in the minds of some in our district. Diana made that exact statement to me yesterday in a discussion with Annick, that we must address the frequency list since we have but a fraction of the time first language acquirers have).”
To clarify, Ben, you aren’t talking about whether non-targeted CI vs. targeted CI is better, but rather whether one set of vocabulary should be targeted vs. it not being important what vocabulary we target.
I thought it was you who has made this same statement to me before: We don’t have the instruction time to use non-targeted CI. For that reason, you were all about targeting input and all about optimizing the repetitions.
I understand that what you are saying here is that targeted CI is the way to go in a FL classroom, but “what” vocabulary is targeted shouldn’t matter. I agree with the latter, since the more we try to target a pre-prescribed set of words, the more constraint we put on interest and the more constraint we put on conversation in general. And conversation is not pre-planned as you often remind us with that quote from the French 🙂
“the best CI is non-targeted, but the exit tests require that we target most frequently occurring words, which is unnatural.”
Haven’t we been told that the Zen part of this work is that if we don’t shelter our grammar, we will automatically and naturally include those high-frequency words? I know I’ve heard before that we just need to speak naturally with the target structures and we’ll be including the high-frequency words. Except that that is not completely true, since when I went over the high-frequency list, I realized a few words I tend not to use very often. Additionally, I think it takes a deliberate effort to include some of the high-frequency verbs. In order to best include the high-frequency verbs, seems they have to be part of the target structure. At least I am finding that the TPRS readers can fill in some of the “verb gaps,” those high-frequency verbs I haven’t yet included in my CI.