Available at

 
 
IN01 
IN02 
IN03 
IN04 
Total % 
STU01 
Close 
rendition 
50.00% 
(37/74) 
59.78% 
(55/92) 
54.88% 
(45/82) 
57.83% 
(48/83) 
55.62% 
Divergent 
rendition 
16.22% 
(12/74) 
17.39% 
(16/92) 
24.39% 
(20/82) 
19.28% 
(16/83) 
19.32% 
STU02 
Close 
rendition 
52.70% 
(39/74) 
69.56% 
(64/92) 
54.88% 
(45/82) 
59.04% 
(49/83) 
59.04% 
Divergent 
rendition 
10.81% 
(8/74) 
16.67% 
(15/92) 
19.51% 
(16/82) 
15.66% 
(13/83) 
15.57% 
STU04 
Close 
rendition 
41.89%
(31/74) 
/ 
57.32% 
(47/82) 
48.19% 
(40/83) 
49.13% 
Divergent 
rendition 
17.57% 
(13/74) 
/ 
19.51% 
(16/82) 
26.51% 
(22/83) 
21.19% 
STU06 
Close 
rendition 
/ 
46.74% 
(43/92) 
41.46% 
(34/82) 
43.37% 
(36/83) 
43.86% 
Divergent 
rendition 
/ 
18.48% 
(17/92) 
31.71% 
(26/82) 
24.10% 
(20/83) 
24.76% 
STU07 
Close 
rendition 
58.11% 
(43/74) 
57.61% 
(53/92) 
/ 
53.01% 
(44/83) 
56.24% 
Divergent 
rendition 
9.46% 
(7/74) 
15.22% 
(14/92) 
/ 
15.66% 
(13/83) 
13.45% 
It appears that when the student’s delivery rate is much lower than the 
corresponding IN delivery rate, then the student correctly translates less than half 
of the occurrences. This is the case for STU06’s three interpretations and STU04’s 
interpretation of IN01. However, there are not enough examples to confirm that 
assumption and it would need deeper investigation. Furthermore, as said in 
Section 4.1.1, the delivery rate gradually increases and thereby increases the 
difficulty of the interpreting tasks to be performed by the students. IN03 and IN04 
respectively have a delivery rate of 108.14 w/m and 109.9 w/m and the proportion 
of wrongly translated items is the most significant for these two source speeches. 
There is thus a direct link between a faster delivery rate and the proportion of 
divergent renditions. 
As for lexical density, we can say that a high lexical density (LD) score seems to 
have a direct link with the proportion of wrongly translated items as IN03 has the 
highest LD score (57.83%) and the highest percentage of divergent renditions 

---

Results and discussion 
 
page 88 
(23.78%). However, the four IN lexical densities are between 52 and 58%, which 
means that all source speeches are rather dense, not just IN03. Another 
observation is the potential link between STU lexical density scores and STU 
divergent renditions. Indeed, there are four students (STU02, STU04, STU06, 
STU07) who support this idea as the IN for which their LD score is the lowest is 
also the IN for which their percentage of divergent renditions is the highest. 
Nevertheless, all STU lexical density scores are between 50% and 55% so the 
difference between LD scores of a same student is not significant at all. Finally, 
there is no potential direct link between close/divergent renditions and expansion 
rates. 
In conclusion, there is no marked improvement in the quality of the trainees’ 
interpretations considering their renditions of four linguistic items. In addition, 
some observations and assumptions have been made concerning potential links 
between some of the global features (lexical density and delivery rate) and 
close/divergent renditions. These observations though need deeper investigation.
It is very important to keep in mind that this is only one way to assess the quality of 
an interpretation and does not mean that they did not improve across time at all. It 
would be very interesting to investigate non-content related aspects of interpreting 
such as fluency (disfluency markers such as repetitions, self-corrections, 
hesitations or silent pauses), intonation, etc. 

---

Results and discussion 
 
page 89