Laufer and Hulstijn (2001) proposed the involvement load hypothesis (ILH). The ILH
assumes that incidental vocabulary learning activities are more effective when the
combined involvement load index—need, search, and evaluation—is higher. According
to the ILH, activities with identical involvement load indices should yield similar results
in learning. The current study investigated this prediction using three activities:
composition-level gap filling (CGF), sentence-level gap filling (SGF), and multiplechoice
gloss (MCG). As these activities had equivalent involvement load indices, their
effects were predicted to be the same. The null hypothesis significance testing merely
confirms the absence of differences among the effects of the activities, rather than their
equivalence. To address this limitation, this study employed Bayesian estimation. Fiftynine
university students were randomly assigned to one of the CGF, SGF, and MCG
tasks to learn five low-frequency nouns. Results revealed neither practical differences
nor equivalence between the three tasks, showing that other variables, not included in
the involvement load index, is influencing incidental vocabulary acquisition.
抄録全体を表示