1
00:00:00,000 --> 00:00:18,720
Okay, so first I will investigate the output produced by learners and the instructor in

2
00:00:18,720 --> 00:00:25,880
a synchronous online clean learning environment provided by the online clear degree course

3
00:00:25,880 --> 00:00:33,760
in applied computer science implemented by the Information Science and Technology Institute

4
00:00:33,760 --> 00:00:41,260
of Urbino University. I will also examine whether clean chat classrooms are actually

5
00:00:41,260 --> 00:00:53,120
a setting of equal or unequal power discourse. So I actually analyzed the online first year

6
00:00:53,120 --> 00:00:59,640
computer architecture course taught entirely in English at Urbino University. And in this

7
00:00:59,640 --> 00:01:06,600
learning environment, learners are provided with clear designed teaching materials. The

8
00:01:06,600 --> 00:01:14,360
participants of the study were just one non-native speaker instructor whose L1 was Italian and

9
00:01:14,360 --> 00:01:22,000
30 non-native speaker first year university students featuring different L1s and aged

10
00:01:22,000 --> 00:01:31,920
between 30 and 55. And I carried out my analysis on chat postings actually featuring the interactions

11
00:01:31,920 --> 00:01:38,600
between learners and the instructor. Chat postings were just the written transcripts

12
00:01:38,600 --> 00:01:49,760
of 22 weekly chat classrooms which were collected over kind of a two semester period. So I carried

13
00:01:49,760 --> 00:01:56,880
out both a quantitative and a qualitative analysis of language. In terms of language

14
00:01:56,880 --> 00:02:03,600
complexity, well the language was analyzed in terms of word number, words per sentence,

15
00:02:03,600 --> 00:02:12,120
passive sentences, academic vocabulary, and subject specific terminology. As regards qualitative

16
00:02:12,120 --> 00:02:19,840
analysis, well the analysis was carried out on the interactions implemented by the participants

17
00:02:19,840 --> 00:02:29,640
in chat classrooms. So as the chart shows, the amount of words produced by the instructor

18
00:02:29,640 --> 00:02:36,760
in chats was much higher than the amount of words produced by learners, which was kind

19
00:02:36,760 --> 00:02:44,400
of a result that we didn't really expect because we thought that chat classrooms were

20
00:02:44,400 --> 00:02:54,240
kind of equalizers in participation. In terms of distribution of talk between the participants,

21
00:02:54,240 --> 00:03:04,040
the instructor's talk amounted to 90% of the total output, quite a striking result. But

22
00:03:04,040 --> 00:03:13,000
in terms of words per sentence, well the instructor produced more words per sentence

23
00:03:13,000 --> 00:03:22,640
in every chat, more than of course learners, just in 50% of all chats. In general, learners

24
00:03:22,640 --> 00:03:30,680
output ranged between 7 and 23 words per sentence, while the instructor's output ranged between

25
00:03:31,120 --> 00:03:40,480
8 and 19 words per sentence, okay? So most of the times, learners' output equaled or

26
00:03:40,480 --> 00:03:47,480
sometimes even outnumbered the instructor's output. So here we started to see that actually

27
00:03:47,480 --> 00:03:56,880
learners were able to produce quite a complex amount of language. In terms of passive sentences,

28
00:03:56,880 --> 00:04:04,360
well learners almost never used passive structures, while the instructor used passive structures

29
00:04:04,360 --> 00:04:12,840
quite regularly, okay? This was kind of an important difference. We also used the word

30
00:04:12,840 --> 00:04:22,520
frequency text profiling to compare chat language with two word frequency lists, okay? The first

31
00:04:22,520 --> 00:04:30,080
one, which is the 2000 most frequent word list featuring general non-academic English

32
00:04:30,080 --> 00:04:38,440
and another word list featuring academic English, and it is a word list devised by Paul Nation.

33
00:04:38,440 --> 00:04:45,320
In terms of general non-academic English, no big differences were detected between learners

34
00:04:45,320 --> 00:04:53,000
and the instructor's output. But in terms of academic language, things were slightly

35
00:04:53,000 --> 00:04:59,640
different, okay? If you take a look at the chart, you can see that learners' output ranged

36
00:04:59,640 --> 00:05:07,720
between, in terms of course of academic language, between 2 and 14%, while the instructor's

37
00:05:07,720 --> 00:05:17,920
output ranged between 6 and 15%. But what is actually most striking is the fact that,

38
00:05:17,920 --> 00:05:23,840
just look at the chart, learners' use of academic English visibly increased during the second

39
00:05:23,840 --> 00:05:31,600
semester, okay? So these results were likely to measure learners' acquisition of academic

40
00:05:31,640 --> 00:05:38,280
language, and we know that academic language is one of the main features of CLIL.

41
00:05:38,280 --> 00:05:50,000
Now, as regards all the words not included in either list, well, learners' output ranged between

42
00:05:50,640 --> 00:06:02,720
10 and 34%, while the instructor's output ranged between 10 and 21%. But the thing was that we

43
00:06:02,720 --> 00:06:12,240
realized that learners used subject-specific terminology quite extensively and effectively

44
00:06:12,240 --> 00:06:19,840
during the whole target period, but there was no significant increase during the target period.

45
00:06:19,840 --> 00:06:26,480
But they used this kind of subject-specific terminology quite extensively. So these results

46
00:06:26,480 --> 00:06:33,520
were likely to measure learners' acquisition of subject-specific terminology, which is one of the

47
00:06:33,520 --> 00:06:43,760
key features and one of the key objectives of CLIL instruction. We also investigated whether

48
00:06:43,760 --> 00:06:52,240
chat classrooms were a setting of equal or unequal power discourse, and how power was negotiated

49
00:06:52,240 --> 00:07:01,680
in chat classrooms. We had also some research questions, such as what kind of questions does

50
00:07:01,680 --> 00:07:09,280
the instructor ask? What kind of questions do learners ask? What kind of output are learners

51
00:07:09,280 --> 00:07:17,200
expected to produce? What kind of output do learners produce? Moreover, all the questions

52
00:07:17,200 --> 00:07:24,480
asked by the participants were classified according to a taxonomy featuring content

53
00:07:24,480 --> 00:07:32,240
and procedural questions, open and closed questions, questions for facts, explanations,

54
00:07:32,240 --> 00:07:38,560
reasons, and opinions. Moreover, we tried to investigate the role of questioning and answering

55
00:07:38,560 --> 00:07:47,280
in the negotiation of power in chat classrooms. Well, to analyze chat practice, we used half-frame,

56
00:07:48,080 --> 00:07:55,520
and the main purpose of chats was for the participants to discuss computer architecture

57
00:07:55,520 --> 00:08:05,280
related topics. Now, the analysis showed that in terms of turn-taking, well, turn-taking was

58
00:08:05,280 --> 00:08:13,680
definitely managed by the instructor by access to the opening move, and it was still the instructor

59
00:08:13,680 --> 00:08:21,120
who actually filtered students' questions. In terms of roles and rights of the participants,

60
00:08:21,120 --> 00:08:29,280
learners were simply expected to ask content-related questions. And in terms of act sequencing,

61
00:08:29,280 --> 00:08:36,800
well, the instructor actually managed openings, transitions, closings, as well as asking and

62
00:08:36,800 --> 00:08:45,920
answering questions. But the analysis of chat postings revealed that in every chat,

63
00:08:46,480 --> 00:08:52,880
the instructor produced slightly more chat postings than learners, which makes perfect

64
00:08:52,960 --> 00:09:00,720
sense, since, as we have just mentioned, it was the instructor who managed openings, transitions,

65
00:09:00,720 --> 00:09:10,240
rights, and closings. So the analysis revealed also that questioning was the main interactional

66
00:09:10,240 --> 00:09:18,320
pattern. So the way content was conveyed was deeply affected by this interactional pattern.

67
00:09:19,280 --> 00:09:27,360
Overall, the instructor asked mainly closed-content questions, which means that learners

68
00:09:27,360 --> 00:09:35,680
were simply expected to produce one-word, one-number answers. To be even more specific,

69
00:09:35,680 --> 00:09:43,280
the instructor asked 29 closed-content questions with an average of two questions per chat,

70
00:09:44,160 --> 00:09:51,440
while the students asked overall 86 content questions, which are, of course, open by default,

71
00:09:52,240 --> 00:10:00,400
with an average of four questions per chat. And all the questions asked by the students fell

72
00:10:00,400 --> 00:10:07,040
into two main categories, questions regarding vocabulary meaning and real content questions.

73
00:10:07,680 --> 00:10:15,120
And actually, the instructor answered learners' questions with rather extended monologues,

74
00:10:15,120 --> 00:10:21,760
which accounts for the huge amount of words that the instructor produced, as we saw previously.

75
00:10:23,040 --> 00:10:29,680
Well, in terms of procedural questions, the instructor asked 12 procedural questions

76
00:10:29,680 --> 00:10:38,320
overall. And these questions concerned mainly exam planning, clean materials,

77
00:10:38,320 --> 00:10:48,160
and self-evaluation tests. Students asked just 15 procedural questions, always dealing with exams

78
00:10:48,160 --> 00:11:00,160
mainly and projects. So we can claim without unbuffer that questions influence the quality

79
00:11:00,160 --> 00:11:09,360
and quantity of students' contributions to classroom talk. In terms of content questions,

80
00:11:09,360 --> 00:11:18,480
learners asked 73% of them. And learners asked mainly for explanations and facts,

81
00:11:18,480 --> 00:11:26,800
while the instructor asked mainly for facts and, to a much lesser extent, for students' hypotheses.

82
00:11:29,040 --> 00:11:35,280
Now, the analysis of the role of questioning and answering in the negotiation of power

83
00:11:35,280 --> 00:11:40,880
revealed that the instructor was still conceived as the main source of knowledge,

84
00:11:41,520 --> 00:11:48,720
even though he sometimes tried to switch from instructor-led to students-led discussions.

85
00:11:49,600 --> 00:11:56,800
However, due to this large preference for students over teacher-initiated questions,

86
00:11:56,800 --> 00:12:05,920
we can claim that chat classrooms featured, to a certain extent, a setting of equal power discourse.

87
00:12:08,880 --> 00:12:16,240
Now, we also tried to answer the following question. What is the role of the triadic dialogue

88
00:12:16,240 --> 00:12:24,640
in online chat classrooms? Well, the analysis revealed that the triadic dialogue was often used,

89
00:12:24,640 --> 00:12:30,880
but especially when the interaction was teacher-initiated, such as in the example

90
00:12:30,880 --> 00:12:40,240
I've provided you with, okay? And as regards the second question, is the opening move used

91
00:12:40,240 --> 00:12:49,680
by the instructor to control topic management? Well, the answer is yes, but only if interaction

92
00:12:49,760 --> 00:12:58,720
is teacher-initiated, okay? But it's definitely not a positive answer if the interaction is

93
00:12:58,720 --> 00:13:06,400
student-initiated. And by now, we know that the instructor-student-initiated question ratio is

94
00:13:06,400 --> 00:13:16,960
23% to 73%. In conclusion, we can claim that questioning and answering were the main

95
00:13:16,960 --> 00:13:24,480
interactional patterns which deeply affected the way content was conveyed and the way

96
00:13:24,480 --> 00:13:33,200
CLIAO discourse was constructed. Second, that CLIAO chat classrooms featured, definitely,

97
00:13:33,200 --> 00:13:42,000
a certain degree of conversational symmetry. Third, that there was definitely more focus on content

98
00:13:42,000 --> 00:13:49,360
than on form. Fourth, that learners' output featured, definitely, a certain degree of

99
00:13:49,360 --> 00:13:55,440
complexity in terms of sentence length, academic English, and subject-specific terminology.

100
00:13:56,080 --> 00:14:02,320
And last but not least, in terms of future developments, the construction of knowledge

101
00:14:02,320 --> 00:14:10,640
by means of collaborative tasks and student-led discussions is advocated for. Thank you.

102
00:14:12,000 --> 00:14:15,760
Thank you.