WEBVTT

00:00:00.040 --> 00:00:04.930 
>> What I think is important to mention is
these two caveats which are huge, right.

00:00:04.930 --> 00:00:10.140 
In other words, in many cases we don't know
exactly in the studies that they analyze.

00:00:10.140 --> 00:00:12.640 
They didn't know exactly
what was going on in terms

00:00:12.640 --> 00:00:16.330 
of the instruction because it's so malleable.

00:00:16.330 --> 00:00:18.590 
I mean you say, "Well, I use
a communicative approach."

00:00:18.590 --> 00:00:20.640 
Well, what does that mean, what do you do?

00:00:20.640 --> 00:00:23.740 
Or we tend to focus on form.

00:00:23.740 --> 00:00:24.440 
What does that mean?

00:00:24.440 --> 00:00:26.430 
It changes from one instructor to the next.

00:00:26.430 --> 00:00:31.220 
There are many different schools of thought in
terms of how you interpret, how you implement,

00:00:31.220 --> 00:00:36.220 
how you deploy all those
different structures, methods.

00:00:36.220 --> 00:00:37.810 
And also test effects.

00:00:37.810 --> 00:00:43.910 
And by in large you can see that the majority
of the tests were mostly discrete, focused,

00:00:43.910 --> 00:00:50.320 
fill in the blanks, multiple-choice, et
cetera, which are not necessarily the type

00:00:50.320 --> 00:00:58.480 
of more open-ended type of test, which we
could argue is perhaps a more realistic version

00:00:58.480 --> 00:01:02.270 
of language or much more demanding type of test,

00:01:02.270 --> 00:01:06.140 
because now all the different
contextual factors come into play.

00:01:06.140 --> 00:01:07.680 
In many cases for example, it's online.

00:01:07.680 --> 00:01:10.170 
Think of an oral proficiency interview.

00:01:10.170 --> 00:01:13.060 
And I'm not talking about any
particular proficiency interview.

00:01:13.060 --> 00:01:14.080 
I'm just talking in general.

00:01:14.080 --> 00:01:18.110 
Oral proficiency interviews
where the demands are online,

00:01:18.110 --> 00:01:19.880 
meaning that you cannot just ask for time.

00:01:19.880 --> 00:01:23.950 
"Okay, give me a minute and I'll get back to
you," that answer, or let me think for 2 minutes

00:01:23.950 --> 00:01:27.130 
about this-- something you can
do with more discrete tests.

00:01:27.130 --> 00:01:30.150
So those are two important
caveats we need to keep in mind.