Tuesday, June 17, 2014

my pronunciation final exam: a marriage of complex and simple

I'm bizarrely proud of the pronunciation exam I've designed. I inflict it on my students tomorrow (Wednesday). The exam is in six parts—twice as many sections as the diagnostic exam and the midterm. The first four parts have to do with speaking; the last two have to do with listening (you can't teach proper pronunciation to students who can't hear the differences between and among sounds). Speaking is valued at 60% of the exam grade; listening is the remaining 40%.

Speaking Section 1 comes right from the earlier tests, and is, in fact, an exact copy of those previous sections. In this part of the test, students must read aloud from a paragraph of movie dialogue. I selected the dialogue because it represents spoken American English, but spoken English that isn't too slang-ridden or too laced with dialect. The sentences in this paragraph are long, which means students need to be mindful of rhythm. The paragraph is also long enough to cover a wide range of phonemes, and because it's a paragraph, these are phonemes-in-context, which is also crucial for exposing student pronunciation problems. Students will read the paragraph into KakaoTalk on their cell phones, then send me their sound files.

Section 2 is more focused and less context-dependent. In this part of the test, students must read a series of sentences containing an obstacle course of difficult phonemes. There's still some discursive context; the sounds of English aren't presented in total isolation, and I'm not sure that testing students on totally isolated sounds is all that useful. That said, the real point of this section is to test the students' intonation. They've been told to stress the important words in a sentence, and they've learned a bit about using intonation to express doubt, curiosity, excitement, and other states of mind. As above, students will send me KakaoTalk sound files.

Section 3 features tongue twisters. This is the most focused trial yet, and the least contextualized. In this section, I've borrowed and/or created tongue twisters that will challenge the students to produce difficult sounds repeatedly (think: "this, that, these, those" as a way to practice the voiced th). As with the two previous sections, students must send me KakaoTalk sound files.

Section 4 introduces the random naturalness and freewheeling creativity of actual conversation into the mix. In this part of the test, students will sit down for a one-minute conversation with the teacher about anything at all. I added this section because Korean students are infamously good at memorization and other rote activities. Spontaneity trips them up, and my kids often tend to forget what they've learned, in terms of proper pronunciation, whenever they find themselves in a free-talk environment. This section may very well catch them at their worst. I've advised them to be mindful and to speak slowly and clearly, which is much more important to me than speaking rapidly and fluently. No KakaoTalk this time: I'll be the one recording students on my own cell phone.

Listening Section 1 is a do-over of the minimal pairs work done in both the diagnostic and midterm exams. Despite my hatred of multiple-choice questions, I made this and the next section multiple choice, as the listening problems are purely about raw discernment. In Listening Section 1, students will hear an utterance like "sit," after which they'll have to choose between "sit" and "seat." This section will be done together as a class: the students will listen and mark their answers while I call out various words and phrases.

Listening Section 2, the final section, is all about syllable stress. Students will hear a sentence twice; they'll be given about 20 seconds to figure out the syllable-stress pattern for that sentence. Riffing off the notation from one of the several pronunciation resources I used during this course, I'll be using "o" to represent weak stress and "O" to represent strong stress. Here's an example of a typical utterance and its stress pattern:

"Where are you going?" (oooOo)—five syllables, fourth one stressed

The test questions for this section will be, as mentioned above, multiple choice, and will look something like this:

[teacher reads utterance: "Where are you going?"]
a. oOooo
b. oooOoo
c. oooOo

Note that I've included a head-fake in the above problem: answer (b) is incorrect because it has too many syllables, so students will have to be able to count syllables in order to get the stress pattern right.

That's basically it. I warned my students that grading would be extremely strict. For the speaking sections, students will be scored on a three-point scale, with a 3 representing absolute, natural perfection, and a 1 representing absolute incomprehensibility. I suspect most of my students will get 2's, which means most will average a 66.7% for the spoken section. How well the kids do on the listening section is anyone's guess.

The exam has been deliberately designed to be brutal. This was a necessary corrective to my students' earlier performance: most of them had gotten A's on the midterm, and most were averaging an A for the class. Since, alas, I have to make my kids conform to a school-sanctioned grade curve, I have little choice but to blast them.

There's one student in this class who worries me. He's the only one who can't speak English at all. He has done remarkably well with memorized English, but any sort of spontaneous conversation is totally beyond him because his listening comprehension is near zero. I feel bad for him: the rest of the class is easily intermediate level, on average; some students might even be considered advanced. The student in question has an A or a high B right now, so I'm not worried that he'll fail the course, but I do worry about how many pegs he's going to be knocked down after my final exam grabs him by the scruff of the neck and worries him violently.

I titled this post "a marriage of complex and simple." I think I've given you an idea of the exam's complexity (fine: it's not that complex); the simple aspect of the exam is in how easy it will be to grade. As I mentioned before, to get a 3, student output needs to be absolutely perfect. This means that, the moment I hear any mistake at all, a student automatically drops to a 2. I reserve 1's only for those kids whose words I simply can't understand, and I doubt that that applies to anyone in the class.

Here's hoping my kids do well tomorrow. I admit I'm morbidly curious as to how this test, my proud achievement, will rearrange the grade-curve landscape.



Charles said...

Why wouldn't you give "absolute incomprehensibility" a 0 instead of a 1? Are they getting that point just for taking the test?

Kevin Kim said...

Something like that, yeah. A 33.3% is a pretty lame F, after all.

Kevin Kim said...

Also, for grading purposes, introducing a zero would mean adopting a four-point scale, which would make it harder for me to judge performance. The more fine-grained a grading scale is, the more a teacher finds himself facing "borderline" dilemmas. "Is this student a 6 or a 7?" Or worse: "Is this student a 6.4 or a 6.5?" Admittedly, a four-point scale isn't all that fine-grained, but in contrast to a three-point scale, it entails extra work.

Charles said...

You could just be a bastard and give them a 0, 1, or 2...

Kevin Kim said...

Ha ha—true enough.