more conversations on AP and the Reply to my letter

This was something I just posted on the Cambridge List.  FYI.

****

> DD: Statistical Validity is a measurement of how well the test

> instrument is measuring whatever it is supposed to be measuring.

> Rarely done. Requires two sets of data on the same subjects and the

> correlation between them. Basically first the test, then a prediction,

> then a test of the prediction's results, then a correlation. Consider

> a test for suitability for the ability to be a military pilot. The

> test and then the follow up of how many of the first test actually got

> the wings.

> Validity is seldom tested. I have seen some scientific studies that

> did it, but rarely.

<snip>

> DD: They are trying to confuse the folk in "fly-over-land."

OK, but are they really? I was just as upset as anyone when this first happened, and the first to forward word of this to several people etc.

My husband has an ABD in Ed admin. (ABD=all but dissertation) We have discussed both in the past and currently issues relating to quality test writing, and how most teachers don't know how to write a test that could meet standards of validity and statistical repetition (or whatever--I'm out of my league with such terminology). And I don't think that's a matter of being hypercritical and saying teachers are dumb. That's not it at all. Just inexperienced

But we are well aware, if we are honest with ourselves, that we have probably all used at some point in our own personal history as teachers some test or quiz that really didn't measure what we wanted it to or certainly didn't get the results we predicted.

If we are good teachers we consider the results and try to determine what changes need to be made, whether it is in teaching or test design or both. I've been changing up my tests this year to include a small portion to translate from stories we've read as well as reading comp questions in Latin and English over a sight story (and this year they have been written by yours truly), followed by multiple choice on aspects of grammar and new forms, etc. Do I think my test is statistically sound? Would it stand up to scrutiny? I have no idea. Maybe some parts of it. Maybe more of it than I think. Maybe less.

BUT I DO NOT KNOW STATISTICS. As a liberal arts major I steered clear of stats--foolishly.

What I understand is this. When a test, like AP Vergil or Latin Lit reaches a certain critical mass--large enough numbers to make reliable data--then certain statistical characteristics fall into play. Since the Latin Lit test has choices among the authors, it may have reached critical mass but the data is unreliable because the choices vary.

That's my take on it anyway.

Look, I was talking (privately) a week ago about whether we should get lawyers and force a meeting with AP, if that was even feasible. That is, I was pretty wound up about it. I was frustrated. I was feeling like this is more interference from some national source that's interfering with what we do.

But let's be fair. Let's take off our Latin hat for a moment and consider how a teacher of Greek (ancient? modern?) might feel because they can't get enough bodies to stay in Greek because there is no AP test and the students all want that for their transcripts. Or--even if I think some of the Asian languages are going to be a fad, like Russian back in the 80s--if I were a teacher of Japanese, having captured the interest of a large number of students, only to have them desert me after a couple of years because there is no AP.

Please, we are part of the language community at large and we should not think ourselves better than other languages. Do I think Latin may be more valuable in relationship to our culture? To our language? To certain aspects of our history? Law? Medicine? I would be lying if I said otherwise. But am I going to tell the French teacher that I think Latin is more important to French? No, because she can go to Canada or France and actually communicate in the language. Jobs can be gotten purely for being bilingual in French. I am not fluent in Latin. I can read it but I'm not fluent and I know it. And when my numbers shot up for next year at school, one of the first things I did is check with my department chair to make sure my numbers didn't cut into French and cause her to have to teach something other than French. (Only overloaded Spanish had a dip in numbers, which they were glad off.) Because frankly, I'm exhausted from teaching English as well as Latin and don't wish that on anyone.

I feel like I'm part of a community with the foreign language teachers. It's like we're all serving ice cream but the students like different flavors. Does that mean that one flavor is better than the rest? Surely not.

I think Spanish is justified to have 2 tests because their numbers are HUGE *and* let's face it, our country is bilingual English/Spanish, whether we want it to be or not. (I personally don't mind, and like to read the signs in Spanish to see if I can figure out what they mean.) But no other language comes near--check out the data in the annual report.

So am I still fighting to keep the Latin Lit test? Frankly, I think it's time to accept that it's going. And I think that truly, as Trevor Packer said, they are the experts on test writing and psychometrics. And even if there is obfuscation there--even if only a little--that still leaves us with how do we justify having two exams? I'm certainly not going to tell French that our tests are more important than their tests....

The question is then why do WE want two tests so badly. 1) We like different authors too--more flavors of ice cream; 2) because of playing the number game in order to "make" classes, we need to combine classes and to do that we need to alternate test topics every other year.

In answer to the first, perhaps in a few years the Vergil test will also include other authors (not as choices), and thus we will get variety. In answer to the second, I'm going to say something that may upset people, so I'll apologize in advance. But...is there something different we can do in our teaching to make the language more accessible to a wider audience so that we can have the larger numbers and make an AP class without it needing to be a combined Latin 4/5 or even 3/4? In talking to my strugglers today (after handing back a test that many struggled with), I told them not to give up hope. I would work on the things they found difficult and together we would all keep going to the next level. It's not just a matter of "you must study more." Many don't have a clue what to do differently on their own. But TOGETHER, I think I can get them to the next level.

Our AP classes should not just be the kids that were the straight A students in Latin 1 from the beginning. They should also somehow include students that didn't think they were language learners at first, but who--with help and guidance--persevered to become readers of Latin.

And perhaps I'm just rambling.

And, as others have said, if you need an alternate class/a Latin V class, why not make it an SAT II prep class of sorts, reading whatever authors you like while fine-tuning grammatical concepts?

Yes, many of us if not most of us will need to rethink how we view the structure of our programs. But if we want to be included, we need to be part of the change.

Someone asked whether anyone else had gotten as thorough a response as I got to my letter. As far as I know, the answer is no. Why? Probably because I took the time to consider what AP thinks is important, what their big picture is, and where we might fit into that picture. I still think convincing them to invest in foreign languages in middle school may not only boost our numbers but also create the change in closing the achievement gap that they are after.

Of course, we have to do our part in producing more teachers. How many of you did something for National Latin Teacher Recruitment Week this past March? Are we going to be able to replace the retiring teachers? Can we fill the positions opening up at middle schools???

I don't know what the multimillion dollar investment into the Vergil test is all about, but I want to know more and be a part of it. I want to learn more about psychometrics now. Instead of this being a word that we come to hate (this is how National Board Certification for Latin was shut down for us as well), I want to learn how to use it to our advantage.

The cheese has been moved. We can go after it, or we can stay where we are wondering why it's no longer here. And if we do that, we'll starve.

I know many of you will disagree with all of this, but we can make it work if we view it in a different light. We just have to figure out what angle the sun's shining in at and get ourselves in that light.

ginnyL, using really weird imagery today...(I blame lack of sleep)