A task segment framework to study keylogged translation processes

Abstract

The Task Segment Framework (TSF) is a tool to analyze full typing flows of translation tasks as keylogged with timestamps recorded for keydown, keyup, mouse clicks and moves, and actions performed in other applications. The TSF assumes that intentional pauses flag stretches where subjects concentrate on unrecorded cognitive processes such as planning and assessment. The interspersed typing stretches are task segments, with or without text, where basic subtasks may be observed, mainly adding new text, changing existing text, and searching for information. Accumulated experience and planning allow translators to lump strategically similar activities together, in order to spare efforts and task switching costs while maximizing efficiency. Hence, task segments may contain activities of just one such subtask or many. Translation fluency is a key notion of the TSF, operationalized through many indicators such as typing speed, prior pause length, TS (task segment) length in events, text length as full words, number of typos and respites (=mid inter-keystroke intervals), subtask(s), and the like. The approach seems particularly sensitive to translation expertise levels and may be applied with variations to other multilectal mediated communication tasks. This article lays down the conceptual basis of the TSF and summarizes its basic notions and constructs.

Traducción literal vs. traducción por defecto. Un desafío para los dos constructos mediante la traducción del egipcio clásico como caso extremo

“…Schaeffer et al define word translation entropy as follows: the degree of uncertainty regarding which lexical TT item(s) are chosen given the sample of alternative translations for a single ST word: if the probabilities are distributed equally over a large number of items, the word translation entropy is high and there is a large degree of uncertainty regarding the outcome of the translation process. If, however, the probability dis- The Task Segment Framework (TSF, Muñoz & Cardona Guerra, 2019; Muñoz & Apfelthaler, 2022) provides an alternative to the methodological problems regarding the analysis of translation processes as keylogged that have been discussed in many publications, such as using fixed, arbitrary pause thresholds (Muñoz & Martín de León, 2018;Muñoz & Cardona Guerra, 2019;Muñoz & Apfelthaler, 2022). The TSF identifies task segments in the keylogged translation process and distinguishes between the following subtasks: (1) adding new text to the text written so far, i.e., ADD; (2) changing the text written so far, i.e., CHANGE; (3) searching for information, i.e., SEARCH, and ( 4) interacting with the computer for purposes other than those immediately related to the task at hand (such as lowering the music volume, changing the font, etc.…”