Available at

3 Data and methodology 
This chapter is divided into four sections. Section 3.1 gives information about the 
interpreter trainees and explains how the corpus used in the present study was 
compiled. The transcription guidelines used for my data are explained and 
illustrated in Section 3.2. Section 3.3 gives information about the corpus used, 
such as source speech topic, source speech length, or the total number of running 
words in each source and target text. Finally, the construction of the database on 
which the analyses are based is explained in Section 3.4. 
3.1 Corpus collection 
In order to analyse the differences between a given source speech in English and 
its interpretations into French, I had to collect relevant data. That is why I recorded 
and transcribed the performances of eight master’s students of the UCLouvain 
(Belgium), named Student 1 to Student 8, who were all in the process of getting 
their master’s degree in conference interpreting. I recorded four of their 
simultaneous interpretations in November and December 2016, across four 
different sessions of their simultaneous interpreting course. They filled in a 
consent form (see Appendix 2) before the interpreting tasks and knew they were 
being recorded. I have not collected metadata related to the students’ gender, age, 
or language proficiency because I knew from the start that these factors would not 
be investigated in the present case study. However, it is important to say that the 
students were all native speakers of French. During these four interpreting 
sessions, their teacher, a non-native speaker of English (NNS), was reading out 
texts that were transcripts of oral speeches, and the students were alone in the 
interpreting booths. In other words, they did not have anyone to help them out 
when they were facing a difficulty. The four speeches were rather long, between 
30 minutes and 1 hour, and I therefore decided to study the first twelve to fourteen 
minutes of each speech, starting at the very beginning and stopping when the 
lecturer was finishing the idea that was uttered around the minute thirteen. Every 

---

Data and methodology 
 
page 40 
performance was then transcribed according to the EPTIC guidelines. More 
information about the transcription process and about these guidelines can be 
found in Section 3.2. 
The eight students did not attend all four classes, which is why I have collected 
and transcribed twenty-two outputs in total. Table 1 shows the number of outputs 
per student. Student 1 and Student 2 attended the four sessions while Student 8 
only took part in one of them. It will be interesting to investigate if there is some 
kind of improvement across time (see research question 2) for students who 
attended at least three out of the four classes (Students 1, 2, 4, 6 and 7).
Table 1 – Distribution across speeches and students 
Speech 1 
(3rd 
November)
Speech 2 
(10th 
November)
Speech 3 
(17th 
November)
Speech 4 
(8th 
December)
Total 
Student 1
X 
X 
X 
X 
4 
Student 2 
X 
X 
X 
X 
4 
Student 3 
X 
X 
2 
Student 4 
X 
X 
X 
3 
Student 5 
X 
X 
2 
Student 6 
X 
X 
X 
3 
Student 7 
X 
X 
X 
3 
Student 8 
X 
1 
Total 
5 
5 
6 
6 
22 
In the present dissertation, the first speech will be referred as IN01, which stands 
for “input number 1” and the same principle applies to the other source speeches. 
The first student will be referred as STU01, which stands for “student number 1” 
and the same principle also applies to the other students. 

---

Data and methodology 
 
page 41