Comparing the predictive power of involvement load hypothesis and technique feature analysis

This study aimed to compare the predictive power of Involvement Load Hypothesis (ILH) and Technique Feature Analysis (TFA) using five different vocabulary learning tasks that had similar and different rankings based on the two models. The experimental tasks included reading and true/false statements, reading and choosing definitions, reading and fill-in-the-blanks, and reading and rewording the sentences. The control task was a reading and comprehension questions task. Participants were 114 young adult EFL learners from five intact classes, assigned to one of the five tasks. During-task performance was assessed. Initial learning was measured by administering an L2/L1 word translation immediate posttest, and medium-term retention was measured by administering the same test with rearranged items one week later. ANOVA and hierarchical multiple-regression results showed that TFA was a more powerful predictor of both initial and medium-term learning. Pedagogical and theoretical implications of the findings will be discussed.