Repetition learning is neither a continuous nor an implicit process

Abstract

Learning advances through repetition. A classic paradigm for studying this process is the Hebb repetition effect: Immediate serial recall performance improves for lists presented repeatedly as compared to nonrepeated lists. Learning in the Hebb paradigm has been described as a slow but continuous accumulation of long-term memory traces over repetitions [e.g., Page & Norris, Phil. Trans. R. Soc. B 364 , 3737–3753 (2009)]. Furthermore, it has been argued that Hebb repetition learning requires no awareness of the repetition, thereby being an instance of implicit learning [e.g., Guérard et al., Mem. Cogn. 39 , 1012–1022 (2011); McKelvie, J. Gen. Psychol. 114 , 75–88 (1987)]. While these assumptions match the data from a group-level perspective, another picture emerges when analyzing data on the individual level. We used a Bayesian hierarchical mixture modeling approach to describe individual learning curves. In two preregistered experiments, using a visual and a verbal Hebb repetition task, we demonstrate that 1) individual learning curves show an abrupt onset followed by rapid growth, with a variable time for the onset of learning across individuals, and that 2) learning onset was preceded by, or coincided with, participants becoming aware of the repetition. These results imply that repetition learning is not implicit and that the appearance of a slow and gradual accumulation of knowledge is an artifact of averaging over individual learning curves.

Testing expectations and retrieval practice modulate repetition learning of visuospatial arrays.

“…Recent simulation work has shown that this is a necessary assumption for computational models of episodic memory like the “retrieving effectively from memory” (REM) model (Shiffrin & Steyvers, 1997), to successfully account for several memory phenomena (Ensor et al, 2021). In Musfeld et al (2023), we have found strong empirical evidence for this hypothesis in the Hebb paradigm: Participants’ performance in a visuospatial and a verbal repetition learning task only improved once they recognized the repeating information.…”

“…Our findings also add another puzzle piece to our understanding of the mechanisms underlying long-term learning by repetition. As we have argued in Souza and Oberauer (2022) and Musfeld et al (2023), repetition learning is likely to depend on a two-stage process: an initial stage of recognizing the repetition and a second stage of forming a stable representation of the repeated information once the repetition was recognized. In the initial phase of a repetition learning experiment, every trial encoded during the working memory test leaves a new memory trace in episodic long-term memory (Hintzman, 1984; Jamieson et al, 2022; Logan, 2002; Shiffrin & Steyvers, 1997).…”

“…We also provide new insights into the mechanisms of how testing influences long-term learning by repetition. We propose that the way how information is encoded into memory in anticipation of the test mainly influences the chance of recognizing the repeated information, which we have found to be necessary to initiate the learning process (Musfeld et al, 2023). Providing more opportunities to practice retrieval of the whole array during a full recall test favors the strengthening of the memory trace as a more integrated representation and thereby further boosts learning.…”

“…This explanation assumes that learning through repetition requires participants to recognize previous encounters of the same information while performing the working memory task (Ensor et al, 2021; Musfeld et al, 2023). An alternative account is that long-term memory traces can still accumulate over repetitions (Page & Norris, 2009), but are not retrieved during the working memory task because the task is not demanding enough to make people draw on long-term memory (Oberauer et al, 2017).…”

Hebb repetition effects in complex and simple span tasks are based on the same learning mechanism.

“…Hebb repetition learning is considered a laboratory analog of vocabulary and language acquisition (Mosse & Jarrold, 2008;Page & Norris, 2008Page et al, 2013;St-Louis et al, 2019;Szmalec et al, 2009Szmalec et al, , 2012. Although it has been mostly studied using verbal materials, Hebb repetition learning has also been found with visual (Horton et al, 2008;Johnson & Miles, 2019;Johnson et al, 2017;Musfeld et al, 2023), spatial (Couture & Tremblay, 2006;Sukegawa et al, 2019), and visuospatial (Souza & Oberauer, 2022) materials, demonstrating the generalizability of the effect.…”