<p>Language learner use of concordances attracts considerable research interest. Concordances are now being conceptualized as presenting language learners with multiple affordances (Leńko-Szymańska & Boulton, 2015), affordances being the latent value of an object determined by the use made of it and not necessarily by its design. However, empirical research has typically operationalized concordances in simple, monolithic terms, and thus, from an experimental design perspective, they are implicitly seen as providing a single set of language learning affordances. That is, the majority of research conducted on concordancing treatments is discussed in terms of concordancing in fairly simple undifferentiated terms (see overviews of research in Boulton, 2010a; Cobb & Boulton, 2015; C. Yoon, 2011). Thus, previous research has contributed to an understanding of whether concordances can facilitate language learning, but it has rarely addressed the issue of how the operationalisation of concordances in pedagogical contexts interacts with the language learning affordances provided. Insufficient engagement with this issue is important because, alongside studies that have shown strong facilitative effects, there are many studies of learner use of concordances that show little to no facilitative effect of the concordancing condition. To address this issue, this thesis addresses the construct of learner use of concordances from an experiment design perspective, helping to define the construct of concordancing and examining potential variables in learner use of concordances. The thesis reports a series of studies that examine the construct of concordancing from both the perspective of concordance users’ concordancing preferences and the perspective of the lexical qualities of concordances as texts. First, a quasi-experimental, quantitative survey of concordance users’ concordancing preferences showed the construct of concordancing to comprise four distinguishable operational parameters: citation format, type of corpus concordanced, citation order and reading style. It then found correlations between these parameters and three of four user groupings: grouping by frequency of concordance use, by linguistic relationship to concordance language, and by field of concordance use. It revealed that different types of concordance user can be characterized by preferences for different types of concordance use, and vice versa. In two further studies, quantitative analysis of concordances showed that manipulation of factors in concordance generation resulted in concordances with significant differences in measures of word frequency and type token ratio. These analyses showed the extent to which the affordances of concordances vary in relation to two key factors in concordance generation: which corpus is concordanced and level of corpus generality. The findings of these four studies are discussed in relation to the definition of concordancing presented in Chapter 2 of the thesis and the affordances of concordances that have been discussed in the literature. Together the findings indicate that effective learner use of concordances is likely to be dependent on matching the operationalization of concordancing in pedagogical contexts to learner profiles. For this reason, research on learner use of concordances needs to adopt experimental designs that can account for variation in concordancing as a treatment condition. The findings are also discussed in relation to their practical implications for effectively operationalising concordancing in pedagogical practice and the development of pedagogical concordancers and concordance-based language learning materials.</p>
MALL language learning outcomes: A comprehensive meta-analysis 1994–2019