“…Besides the ways for researchers to operationalize learning gains (e.g., test format, test timing), numerous other variables are believed to contribute to or determine the role that frequency plays in vocabulary learning. The potential moderators include learner variables (e.g., age, gender, proficiency, motivation, working memory, or background knowledge), word characteristics (e.g., imageability, concreteness, cognateness, number of letters and syllables, or parts of speech), text characteristics (e.g., genre, text length, richness of contextual clues, or keyness of the word for comprehension), and methodological or treatment variations (e.g., spaced vs. massed treatment conditions, or use of nonwords vs. real words) (Elgort et al., ; Elgort & Warren, ; Malone, ; Paribakht & Wesche, ; Peters, Heynen, & Puimège, ; Peters & Webb, ; Webb, , ; Webb & Chang, ). Due to empirical motivations based on previous studies as well as to logistic reasons related to feasibility and reliability of variable identification and coding (e.g., general L2 proficiency), we limited our moderators to a total of 10 variables .…”