Spaced repetition and list-strength in recognition memory

Intuition and classic mathematical models suggest that "strong" list items should see a competitive recognition-memory advantage when mixed with "weak" list items within a single list. This holds for manipulations of reading aloud but not other manipulations. Study time per item (e.g., 500 vs. 2000 ms), can even invert the effect, with a larger strength effect between pure-strength lists than within mixed lists. This inversion remained when the longer study time comprised three massed repetitions. An attentional subsetting theory account explained the inverted list-strength effect by assuming weak items had primarily shallow features encoded, and participants disregarded those on pure-strong lists, when the better, deeper features were available. Here we asked whether spaced repetition is an exception, rendering disregarding ineffective, or like massed repetition, produces an inverted list-strength effect if presentation time is short. In three experiments, spaced repetition produced a null or upright list-strength effect. Under equated conditions, massed repetition and long duration produced inverted or null effects. Thus, despite being within an experimental regime that has robustly produced inverted list-strength effects, spaced repetition appears resilient. These boundary conditions constrain models, including attentional subsetting theory and Retrieving Effectively from Memory, and we discuss how various theories may need to adjust to explain the full range of list-strength data.