The experiment was a 2 (professional vs. trainee) × 2 (vertical vs. horizontal layout)
× 2 (source vs. target language) mixed design, with one between-subjects factor, Expertise, and two within-subjects factor, Layout and Language. The dependent variables were various eye movement measures, which will be explained in detail in later sections.
3.4 Materials
3.4.1 Speeches (Appendix B)
The four English speeches used in this study were all excerpts from authentic speech texts and were recorded at a rate of 100 to 120 words per minute by a native English speaker. The recording of each speech was approximately three minutes in length. All four speeches were on general topics without technical terms and were suited to the reading ability of high-school students in an English speaking country.
They all showed similar levels of reading difficulties in the Flesch Reading Ease rating (Table 4). The speech used in the practice trial was a brief introduction to Taiwan’s geography and history, about 1.5 minutes long.
Topic Word
Education reform in Australia 301 40.1 6
Table 4. Speech materials
37
3.4.2 Notes (Appendix B)
Notes taken by student interpreters in consecutive interpreting class were collected, compared, and analyzed. Common features of the notes, such as symbols, arrows, abbreviations, and Chinese characters were tabulated. These common features were used, wherever applicable, in the designed notes as shown in Appendix B. Other combinations of mathematical symbols and universal logical indicators were also used. Two experimenters listened to the speeches and took notes in the vertical style that they were both familiar with. Then the two sets of notes were compared to determine what elements should be kept in the designed notes. Two students from translation and interpretation graduate programs who had similar backgrounds with the subjects did the pilot test to examine the legibility and readability of notes. Afterwards, the notes were adequately revised, scanned into digital pictures, and manipulated by using Photoshop to create the four conditions for each speech. Certain elements in the notes were chosen by the experimenters to present in Chinese or in English to test the effect of source and target language on the subjects’ eye movements, as shown in the squared areas in Appendix B. Four note conditions from left to right are shown in the order of vertical layout with source language (VS), vertical layout with target language (VT), horizontal layout with source language (HS), and horizontal layout with target language (HT).
3.4.3 Comprehension Questions (Appendix C)
Each speech had two true or false questions in Chinese to test the subjects’
understanding of the speeches.
3.4.4 Questionnaire (Appendix D)
The questionnaire included three parts. The participant was asked to fill out the first part before the experiment and the other two parts after completing the experiment.
The first part was about the participant’s basic information, including age, sex, background, whether he had received consecutive interpreting training, had real interpreting experience outside of classroom, and the experience of teaching interpreting. The second part, designed based on the Likert scale, was about the participant’s consecutive interpreting training in the past and asked the participant to examine his personal note-taking style and preferences in terms of layout and language, and whether such style and preferences change with experience. The third part asked the participant to recall the experiment. Complemented with an interview, the participant was asked to assess the level of difficulty of the experimental
38
materials in terms of their contents and note legibility, and was also invited to share his feelings about this experiment, any difficulties encountered during the experiment and possible causes, and provide suggestions to the experimenters.
3.5 Apparatus
Stimuli, the notes, were presented on a computer monitor (1024×768 pixels) and eye movements were recorded using an Eyelink 1000 Desk Mount eye tracker, manufactured by SR Research. The sampling rate was 1000 samples per second.
Stimuli were shown in the middle of the screen one page of note at a time. Eyelink 1000 was equipped with a chin rest for participants to position their heads during the experiment. However, since this experiment required the participants to speak, the chin rest was not used. The distance between the participant and the monitor was 70 centimeters.
3.6 Procedures
The subjects first signed a consent form and filled out the first part of the questionnaire. They were then seated in a dimly-lit room in front of a computer monitor while the experimenter sat in the same room but facing another computer where he could monitor the subjects’ eye movements. Each subject was tested individually. Prior to the experiment, the subjects were tested for their dominant eye, which was the eye recorded during the experiment. In the beginning, instructions about how the experiment was going to proceed were presented on the screen. The nine-point calibration and validation procedure followed to determine the correspondence between pupil position and gaze position. The experiment started with one practice trial followed by four formal trials. The purpose of the practice trial was to familiarize the participants with the experiment procedures.
Once the experiment formally began, the participants listened to the recorded speeches while looking at the corresponding designed notes displayed on the screen.
Since they were not allowed to take notes by themselves, they could use the time to comprehend the speeches and familiarize themselves with the notes. After a whole speech was delivered and before each page of note was shown, the participants were asked to fixate their eyes on a cross which appeared on the upper left corner on the screen. If the eyes fixated in an acceptable range, the cross would vanish and the next page of note would appear. The calibration and validation process would be performed again if the participants failed to accurately fixate on the cross. After they finished interpreting one page, they pressed any button on the response box in front of them to turn to the next page. The recording device was activated automatically whenever the participants were rendering the interpretation. In the end of
39
interpreting each speech, the participants were asked to answer two comprehension questions that appeared on the screen to ensure that they comprehended the content. A “yes” and a “no” were displayed on the screen’s left and right bottom corner, each of which corresponded to the left and right buttons respectively on the response box. Feedback indicating whether the participant’s answer was correct would be presented on the screen after the button was pressed. The entire experiment took about one hour, including a short break between the first two and the last two speeches. However, there was no time constraint on how much time the participants had to complete the whole experiment. The speed was controlled by the participants depending on how fast they spoke and when they pressed the button.
One thing that is worth noting here is that the participants performed consecutive interpreting for the four speeches in rotating orders. In other words, different participants encountered different combinations of the four note conditions of the four speeches in order to minimize the possible carryover effects coming from practice and fatigue.
After the experiment ended, the participants were asked to fill out the last two parts of the questionnaire and were interviewed. The questionnaire and interview content, the qualitative data, would be used to complement the statistical analyses on the eye movement measures, the quantitative data.