WEBVTT 00:00:00.130 --> 00:00:04.910 >> But what are some of the disadvantages with fill in the blank? 00:00:04.910 --> 00:00:06.710 >> They're still not producing a lot. 00:00:06.710 --> 00:00:07.430 >> Okay, yes. 00:00:07.430 --> 00:00:09.850 >> They're just producing a word. 00:00:09.850 --> 00:00:11.470 >> Yeah, yeah, yeah. 00:00:11.470 --> 00:00:15.940 >> From a test-taker's perspective what can be very frustrating and if you now know 00:00:15.940 --> 00:00:18.890 that two words and you're pretty much screwed or-- 00:00:18.890 --> 00:00:19.260 >> Yeah. 00:00:19.260 --> 00:00:19.770 >> Yeah, yeah. 00:00:19.770 --> 00:00:20.330 [ Laughter ] 00:00:20.330 --> 00:00:23.980 >> So they can be frustrating in their own way. 00:00:23.980 --> 00:00:27.560 You're trying to-- it's almost like a puzzle, right, like a jigsaw puzzle or something. 00:00:27.560 --> 00:00:30.830 You're trying to fit things into these blanks and-- 00:00:30.830 --> 00:00:33.400 >> And they sit there forever. 00:00:33.400 --> 00:00:34.270 >> Other things? 00:00:34.270 --> 00:00:36.650 >> And possible answers can be various. 00:00:36.650 --> 00:00:37.840 >> Right, right. 00:00:37.840 --> 00:00:42.110 So we think they'll be easier to score and probably than an extended response they are 00:00:42.110 --> 00:00:46.450 but when you actually get the answers you realize there are a lot more possibilities 00:00:46.450 --> 00:00:48.140 than you realize when you wrote that test. 00:00:48.140 --> 00:00:54.040 So a good answer key and piloting it is very, very important for the scoring certainly. 00:00:54.040 --> 00:00:57.600 Any other things? 00:00:57.600 --> 00:01:04.060 >> Like in our example in the example test, it can decontextualize the language 00:01:04.060 --> 00:01:07.730 or it can contextualize it, it's just a matter of how it's designed. 00:01:07.730 --> 00:01:11.200 >> And how it's designed, yeah, yeah, that's a good point. 00:01:11.200 --> 00:01:16.360 >> Another disadvantage actually was an advantage we talked about also and the fact 00:01:16.360 --> 00:01:19.150 that it could assess a lot of different things. 00:01:19.150 --> 00:01:23.580 So if you are really looking at reading and you're using a fill in the blank, 00:01:23.580 --> 00:01:27.370 there might be some blanks that are actually testing grammar 00:01:27.370 --> 00:01:30.530 or something other than reading comprehension. 00:01:30.530 --> 00:01:35.690 And so sometimes it can be difficult to interpret why a student chose the correct answer 00:01:35.690 --> 00:01:38.910 or an incorrect answer even though we think we're testing reading 00:01:38.910 --> 00:01:42.930 or we think we're testing grammar, maybe its vocabulary, or something like that. 00:01:42.930 --> 00:01:45.920 So while it can be an advantage, it can also make it more difficult 00:01:45.920 --> 00:01:48.370 to interpret sometimes the results, okay. 00:01:48.370 --> 00:01:54.420 So we just talked about two kinds of indirect assessment and there are plenty of others. 00:01:54.420 --> 00:01:57.120 But just to summarize some of the things that came 00:01:57.120 --> 00:02:00.240 up that they can be problematic 'cause they're not very communicative. 00:02:00.240 --> 00:02:03.940 You guys talked about meaningfulness and production. 00:02:03.940 --> 00:02:07.430 They aren't all that authentic to what students are actually gonna have to do 00:02:07.430 --> 00:02:10.310 with language probably outside of the classroom. 00:02:10.310 --> 00:02:14.900 And some people would say that they're maybe not very motivating 00:02:14.900 --> 00:02:19.580 because students aren't using language or they're not getting to create things 00:02:19.580 --> 00:02:26.000 that also might not make them very meaningful to the test taker and they can be hard to write. 00:02:26.000 --> 00:02:31.760 >> However, we still use them and we see them in a lot of places for a lot of reasons. 00:02:31.760 --> 00:02:35.590 One is they seem easier to grade and keep in mind 00:02:35.590 --> 00:02:41.100 that some people are teaching language classes with-- in China in a foreign language class, 00:02:41.100 --> 00:02:43.960 how many students would you say you have? 00:02:43.960 --> 0 students. 00:02:44.970 >> I have 50 students. 00:02:44.970 --> 0 students, right. 00:02:46.440 >> 50 students, right. 00:02:46.440 --> 0 students [laughter] it's sort of unmanageable. 00:02:50.500 So imagine doing an oral interview for 50 students [laughter] it's sort of unmanageable. 00:02:50.500 --> 00:02:55.750 And so there are contexts where or if you are teaching six classes in a semester 00:02:55.750 --> 00:02:58.500 where you need something that's gonna be a little bit easier to grade 00:02:58.500 --> 00:03:02.440 and still get some information about students' abilities. 00:03:02.440 --> 00:03:08.080 It's low in terms of production so we're not finding information about how students are able 00:03:08.080 --> 00:03:11.880 to use the language but for beginning learners, they may not be ready to use the language yet 00:03:11.880 --> 00:03:14.010 or comfortable using the language. 00:03:14.010 --> 00:03:16.680 And so if there are things like listening and reading, sometimes it's helpful 00:03:16.680 --> 00:03:22.120 to use an indirect method to find out if they're able to comprehend without asking them 00:03:22.120 --> 00:03:22.120 to say something or to write something.