We cannot display this gallery

Traduction Interobserver Agreement

In the case of questionnaires in which several counsellors fill out the same instrument for each person studied (for example. B a control list of behaviour/symptoms), it is possible to assess the consistency of counselors` observations in the same group of reviewers. This consistency is called inter-rated reliability or inter-rated tuning and can be estimated on the basis of Kappa statistics. [33] Suppose two physicians independently evaluated the same group of patients after the operation with their mobility (for example. B 0 – requires help from 2 people; 1 – requires 1-person help; 2 – independent), kappa () can be calculated as follows: If items close to the end are to be used, should multiple choices, Likert, true/false scales or other nearby formats be used? How many response options should be available? If a Likert scale is to be adopted, which scale anchors should be used to indicate the degree of compliance (z.B. strongly agree, agree, not, not agree, highly graduated), the frequency of an event (for example. B almost never, from time to time, sometimes, often, almost always) or other different options? In order to use participants` responses for future statistical analyses, researchers should keep in mind that the elements must be modulated to create sufficient variability among the respondents considered. [6.7] Where is the observed proportion of the observations in which the two councillors agree, and Pe is the expected portion of the observations in which the two councillors coincide by chance. As a result, the share of the agreement between the two advisers, after the share of the agreement was taken into account at random. – ranges from 0 to 1, No.

0 indicating all coincidences and No. 1 constituting a perfect match between the two advisors. other than the de minimis agreement, 0.41 0 that did not agree, – 0.01 x 0.20 as bad deal, – n 0.41 x 0.60 as fair agreement, – 0.61 – 0.80 as good agreement, – 0.81 – 0.92 as very good agreement and – 0.93 – 1 as excellent agreement. [34.35] If more than two advisors are used, an extension of Cohens` statistics is available to calculate inter-rated reliability across multiple tranches. [36] Today, the field of translation studies (TS) confirmed the need for experimental and empirical evidence of the evaluation and quality of translation tests (Akbari – Segers, 2017b; Anckaert, Eyckmans, Segers, 2008). Although educational and professional institutions have implemented “translation skills certifications” (Eyckmans, Anckaert, Segers, 2013) on the basis of different testing administrations, the reliability and validity of these test administrations remains at stake. In other words, translation evaluation is somehow linked to codes of conduct (written rules that express how a researcher should behave in a specific situation and profession) and not empirical experimental research around the world (ibid.). The term “translation evaluation” refers to the translation product (for example.

B the target text), the translation process (i.e. how the translator transmits the content of a basic language to the target language), the translation service (for example. B billing, customer, claims, compliance agreement) and hence translator competence. However, a translator`s translation product, process, service and competency cannot be evaluated in the same way and require different types of evaluation approaches. Translation evaluation is largely characterized by criteria-based evaluation (Schmitt, 2005). Based on educational and professional contexts, evaluation grids are used to make translation evaluation more objective, valid and reliable (ibid.). Although the use of assessment grids is caused by the student`s desire to consider different dimensions of translation proficiency, it must be argued that they fail to reduce the “subjectivity of translation” (Anckaert et al., 2008).