Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Secondary Education

Major Professor

John I. Liontas, Ph.D.

Committee Member

Dean Ann Cranston-Gingras, Ph.D.

Committee Member

Philip C. Smith, Ph.D.

Committee Member

Sanghoon Park, Ph.D.


Authentic Contexts, English for Academic Purposes, L2 idioms


Vivid Phrasal Idioms (VP Idioms) are non-compositional, figurative phrases that have limited flexibility, such as “once in a blue moon.” Research suggests that facility with idioms is an integral part of communicative competence in an L2 (Liontas, 2017; Sinclair, 1992). Despite the fact that they are ubiquitous throughout informal, spoken language (Karlsson, 2013; Liu, 2008), and occur less frequently in other registers of usage, newscasters and politicians employ VP Idioms for communication in what Myers (1996) has called the presentational mode. They also appeari in academic discourse (Liontas, 2008). Nevertheless, research suggests that the English language learners in formal educational contexts study idioms through textbooks that feature idioms in non-native speaker-like contexts or neglect them entirely (Khan & Daskin, 2014; Liontas, 2008, 2015). This dissertation investigates how the instructional interventions affect both comprehension and productions of VP Idioms. The participants, who were all English for Academic Purposes (EAP) students enrolled a single course, were guided through the comparison condition and the two instructional interventions. A website designed for the research was used to guide students throughout the intervention and included definitions and images for each of the nine target VP Idioms, instructions on how to complete tasks, and supplementary materials. For the first experimental condition, learners created their own digital depictions of the literal meaning of the idioms, as recommended by Liontas (2008, 2015, 2017) and Vasiljevic (2012). For the second condition, learners were exposed to VP Idioms in authentic audiovisual contexts as recommended by Cakir (2011) and Liontas (2008, 2015, 2017). The students wrote a short dialogue containing one target VP Idiom after each of the three conditions. Data analysis was mixed method. The quantitative data based upon the posttest scores was used to ascertain the relative effectiveness of each of the conditions. Based on the participants’ written dialogues, the researcher analyzed the qualitative data to ascertain the extent to which the students could use the target VP Idioms appropriately. The mixed method approach brought to light more complex interpretation of data than would have been possible through analysis using either quantitative or qualitative alone. Quantitative results reveal a statistical difference in the posttest scores between the Control and the Learner-Generated Images conditions only. Qualitative results are more nuanced, showing considerable individual differences in ability to produce idioms, with possible confounding variables, but students’ competent ability to produce two of the VP Idioms taught under the Learner-Generated Images condition and one presented under the Authentic Contexts condition. The discussion concludes with pedagogical implications based upon both the results of the current study and previous research by Liontas (2008, 2015, 2017b) and Vasiljevic (2015b).