Generation and L2 learning: an investigation into generation as a desirable difficulty in L2 learning contexts among university learners in Japan

Duplice, John ORCID: 0000-0002-0237-1273 (2024). Generation and L2 learning: an investigation into generation as a desirable difficulty in L2 learning contexts among university learners in Japan. University of Birmingham. Ph.D.

Preview

Duplice2024PhD.pdf
Text
Available under License Creative Commons Attribution.
Download (3MB) | Preview

Abstract

This thesis investigated the efficacy of Generation used as a desirable difficulty (DD) in a second language (L2) context. This was a modular thesis that consisted of three modules, the present module being the third and final. Building upon an initial study conducted in Module Two, this final module describes three studies conducted among university English language learners (ELLs) in Japan. This thesis examined the impact of Generation on L2 learning, focusing on vocabulary, reading comprehension, and self-feedback in academic writing.

The first study looked at the efficacy of Generation on L2 vocabulary meaning recall over three and 15-week periods. This study investigated the impact of Generation tasks on long-term learning among 38 university ELLs in Japan. The study compared the Generation of novel sentences using target L2 vocabulary to a baseline of recall tasks. This was done to investigate whether the benefits found in L1 contexts would carry over to L2 vocabulary learning. While the study did show some interactions that show a trend to possible benefits of Generation, the study did not find a significant clear benefit to using Generation in the L2 context among the participants studied.

Study 2 investigated the application of Generation tasks in L2 reading comprehension, examining both L1 and L2 Generation usage through prompts embedded in the reading text. The study consisted of 44 university ELLs in Japan and explored the impact of varying cognitive load demands through two distinct Generation tasks and a control condition. This novel approach aimed to understand the influence of Generation on L2 reading comprehension and the comparative effects of employing Generation tasks in L1 versus L2. The results did not indicate a benefit from Generation in either L1 or L2. Moreover, L2 Generation incurred a cost compared to the control condition of rereading the text.

Study 3 explored the role of self-reflection in L2 academic writing among 39 ELL university students, utilizing Generation checklists. The study analyzed three aspects of writing - Formatting, Grammar, Spelling, Punctuation, and Capitalization (GSPC), and Content - across four essays. The results indicated that Generation led to improvements in all three writing aspects compared to the control condition, suggesting that Generation tasks through self-reflection can enhance L2 academic writing. The study also examined the carryover effect of Generation tasks from one essay to the next. Content scores significantly improved in the final two essays, indicating that the deeper thinking invoked by Generation tasks in earlier essays may have enhanced subsequent learning and ability.

The findings from these present studies revealed that Generation’s efficacy in L2 contexts differs from previous results found in L1 contexts and can vary within L2 contexts and tasks. While Generation was found ineffective as an initial DD in vocabulary learning and for L2 reading comprehension, it showed benefits when used as self-feedback in L2 writing.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):