Obtaining accurate measurements of L2 learners’ productive knowledge of collocations has proven difficult. The goal of the work reported in this thesis was to develop and test a means of eliciting from learners a reliable and representative sample of their productive knowledge of collocations. The two main methods typically used for this purpose are demonstrated to suffer from a number of drawbacks, yet one instrument is identified as having potential. This instrument, LexCombi, originally devised by Barfield (2009a), presents noun cues to learners and asks for three collocates in response to each cue, which are then evaluated as either canonical or not. In this thesis, LexCombi is taken forward and, through an iterative series of empirical studies, developed further. Specifically, after trialling LexCombi and exploring how learners interact with it, the format is adapted to more clearly guide respondents towards producing collocations; the scoring of learners’ responses is reviewed to gain a more complete picture of learners’ knowledge; and a new set of cue words is trialled and selected to resolve a number of issues identified with the original cues. After this development process, an empirical evaluation of the final form of the instrument, LexCombi 2, is conducted and its capacity to provide useful data on learners’ productive knowledge of collocations is evaluated. Following this, the empirical data is used to consider what can be learned about collocation knowledge using LexCombi 2. Explorations include the relationship between collocation knowledge and general L2 proficiency, the types of words that are used as responses to LexCombi 2, and how LexCombi 2 scores are affected by different conceptions of collocation. Finally, the thesis considers the overall significance of this work for our understanding of collocation knowledge more generally.
Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:761347 |
Date | January 2018 |
Creators | Brown, Dale |
Publisher | Cardiff University |
Source Sets | Ethos UK |
Detected Language | English |
Type | Electronic Thesis or Dissertation |
Source | http://orca.cf.ac.uk/117010/ |
Page generated in 0.0022 seconds