When analysing task performance, most authors use measures that somehow reflect their priorities; for instance, those primarily concerned with the negotiation work involved in the tasks use negotiation of meaning indices for evaluating task performance (Plough & Gass, 1993), while those interested in pragmatic development as a result of solving language tasks tend to use pragmatic appropriacy scales (Levkina, 2018). However, in order to enhance the comparability of task-based findings, Foster and Skehan (1996) argued for the use of more general measures of language performance. Drawing on a fundamental distinction between form and meaning, they called for the use of accuracy (a conservative orientation towards the focus on form), complexity (focus on form as using more elaborate language), and fluency (primacy of meaning while coping with real-time communication) measures and defined these variables operationally. Citing evidence from a factor analytical study (Skehan & Foster, 1997), they argued effectively for the independence of these three measures of language performance. Although CAF (standing for complexity, accuracy, fluency) measures have evolved over the years (Norris & Ortega, 2009; Skehan, 2009) and have also been subjected to criticism (Lambert & Kormos, 2014; Larsen-Freeman, 2009), they continue to be highly influential and widely used measures in task-based research (Housen, Kuiken, & Vedder, 2012; Michel, 2017). Accuracy, complexity, and fluency can be operationalised either as specific or as general measures. Foster and Skehan (1996) argued for using general measures as these are likely to be more sensitive and reflect weaker effects although the use of specific measures or a combination of general and specific measures can also be warranted in certain cases (Housen et al., 2012; Michel, 2017).