by no means did we expect the trainees at the end of the experiment to have turned into totally competent interpreters capable of producing high-quality interpretations of a text with all its natural delicacies and complexities (lexical, syntactic, stylistic, etc.).
On the other hand, as Kim (2006, p. 255) says “interpretation ability is closely related to the development of language ability, communication ability and cognitive ability.” Among these three, the focus of the present study was on the second and in particular the third one, that is, cognitive ability.
Therefore, care had to be taken to ensure that the texts chosen for this purpose would not pose any linguistic difficulties at all. What we were seeking to find out was whether the trainees would be able to simultaneously interpret a text which would be easy for them to translate had they not been forced to work under the pressure of simultaneous interpreting. In other words, it is their ability to cope with the difficulties posed by the task of simultaneous interpreting, as opposed to their general translation ability, which is of interest to the researcher. Bearing this in mind, the audio texts were selected from among some English podcasts designed for pre-intermediate learners of English, hence simple enough grammar-wise as well as lexicon-wise for beginner interpreters to cope with. Each text was roughly five minutes long. These two audio files are recorded on a compact disk accompanying this dissertation.
188.8.131.52 Multiple Intelligences Test
Translation, being a most complicated phenomenon, is affected by a number of interdependent factors, which makes it not only possible but also interesting to be examined from different perspectives.
The researcher’s observations as an interpreter trainer in undergraduate programs convince him that certain extra-linguistic factors may affect the trainees’ ability to turn into a successful interpreter. Although the researcher is on the belief that translating, and by extension interpreting, are teachable and learnable, these factors are likely to make some trainees more apt to take on interpreting and others less apt to do so. One such factor, ostensibly, is multiple intelligences.
Therefore, as a side contribution along with the main purpose of this study, it was attempted to probe into the possible relation between the trainees’ multiple intelligences and their SI performance improvement rate during the experiment period.
To investigate this possible relation, we needed to have a yardstick to measure the participants’ multiple intelligences. It was thus decided that a test of multiple intelligences based on Howard Gardner’s (1983) theory of MI be administered to the subjects in the experimental group and then be compared and correlated with the improvement they had made within the time period. This test covers the eight intelligences: Linguistic, Logical-Mathematical, Visual-Spatial, Musical, Bodily-Kinesthetic, Interpersonal, Intrapersonal, and Naturalist. Each section of the test contains ten items amounting to a total of eighty items. Each question is answered by a ‘yes’ or a ‘no’. The test was retrieved from the following website:http://www.nedprod.com/Niall_stuff/intelligence_test.html.
It is provided in the appendices (see appendix 3).
184.108.40.206 Personality Type Test
Another factor, which seemed an appropriate option for investigation, was Personality Type. Therefore it was deemed useful to try to find out the possible correlation between personality types (i.e. extroversion vs. introversion) and the improvement the experimental subjects had made.
It was decided that Carl Jung’s test of personality type be used in this case; it is a test comprising seventy two multiple-choice items to be answered with no time limitation. This test was retrieved from the following website: http://www.humanmetrics.com/cgi-win/JTypes2.asp. It is provided in the appendices (see appendix 2)
3.4 Data Collection and Analysis
The following is a short account of how the data used for the purpose of this study was collected and analyzed.
3.4.1 General English Test
The test of General English (the two modules of IELTS) was administered to all subjects, both in the experimental group and control group, in a classroom setting at the beginning of the experiment period. The test papers were then scored by the researcher using the corresponding answer key (provided at the end of appendix 1). The participants with extreme scores (say lower than 4 or higher than 8) were left out of the study and the remaining scores were tabulated.
3.4.2 SI Pretest and Posttest
One of the most commonly criticized aspects of an experimental research design is that the researcher’s intervention can cause the setting of the experiment to be different from that in a natural situation, and by doing so it may undermine the reliability of research findings. This notwithstanding, experimental studies do continue to exist and to inform the discipline. Researchers try to counter the effects of such a shortcoming through controlling the undesired variables and making the experimental setting as similar to that of real-life situations as possible.
As regards to the simultaneous interpreting tests in this study, the participants were provided sufficient, clear instructions in advance as to how to carry the task of interpreting; this was particularly crucial before the pretest, for very few of the subjects were expected to have tried such a task previously. The audio texts for pretest and posttest were played on a player and participants listened to them through some good quality headphones. At the same time they were asked to interpret the audio texts into a voice recorder that recorded their voice for analysis to be conducted afterwards. It was important that the setting and environment be suitable for the task of interpreting. Therefore, the participants were taken into a separate room where it was possible for them to reach the level of concentration required. In an ideal situation the room needs to be soundproof. Unfortunately it was not possible to provide such an environment due to the limitations faced. However, maximum care was taken to ensure that the environment had the least distractions possible. These interpreting sessions are recorded on the compact disk which accompanies this dissertation.
The recorded interpreting sessions were then rated according to how well the trainees had managed to cope with the challenges posed by this specific mode of comprehension-production. In order to avoid subjectivity, along with the researcher himself, two other raters, both practicing interpreters, were asked and paid to rate the SI sessions. The scores were given on a scale of zero to twenty and the raters were advised to adopt a holistic approach (see section 1.9), i.e. try to assess the overall quality of the SI session taking into account all different aspects of the task-e.g. stress management, speed, compensation strategies, appropriate coordination, naturalness of expression, etc. The scores given to each SI session were then multiplied by five so that the scores could be presented on a percentile basis to be compared with multiple intelligences scores and personality type scores (which were also on a scale of zero to one hundred).
Therefore each interpreting session was given three scores, the average of which was to be calculated and considered as the final score for a given participant. It also proved necessary to ensure inter-rater reliability for the ratings. To do so, a certain statistical procedure (inter-rater reliability formula) was undertaken. Having made sure about the reliability of the ratings, we moved on to apply a statistical procedure (t-test) to check whether the pretest and posttest scores of the trainees in the two groups differed significantly or not. The statistical procedures utilized for the analysis of the data collected as well as the rationale behind the use of each will be explained in more detail in the next chapter.
3.4.3 Multiple Intelligences and Personality Type Tests
Towards the end of the experiment period and before the posttest, the two tests of multiple intelligences and personality type were administered to the subjects in the experimental group. The tests were retrieved from the Internet (http://www.nedprod.com/Niall_stuff/intelligence_test.html and http://www.humanmetrics.com/cgi-win/JTypes2.asp) and were sent to the participants via e-mail. They did the tests in their own good time and sent back the answered tests. The answers for each subject were entered into the questionnaires on the websites and the results were obtained.
To answer questions 2 and 3 put forth in chapter one, the trainees’ rate of improvement within the course of the experiment (the difference between their scores on the SI pretest and posttest) was correlated with their scores on the other two tests (multiple intelligences and personality type) to shed light on the possible relations.
4.1 Chapter Overview
The present chapter is concerned with presenting as well as discussing the results obtained, through the procedures already explained in the previous chapter, prior to the commencement and within the course of the experiment.
In order to be able to analyze the data, we needed to tabulate the data obtained so that light would be shed on different aspects of the object of study. The following sections will provide relevant tabulated results under each heading so as to make possible conclusions in each case. The statistical procedures and formulas utilized will also be discussed at length.
4.2 GE Test Scores
Out of the initial one hundred and two participants who agreed to take part in the experiment, the General English test scores for seven