Evidence shows that the font size of study items significantly influences judgments of learning (JOLs) and that people’s JOLs are generally higher for larger words than for smaller words. Previous studies have suggested that font size influences JOLs in a belief-based way. However, few studies have directly examined how much people’s beliefs contribute to the font-size effect in JOLs. This study investigated the degree to which font size influenced JOLs in a belief-based way. In Experiment 1, one group of participants (learners) studied words with different font sizes and made JOLs, whereas another group of participants (observers) viewed the learners' study phase and made JOLs for the learners. In Experiment 2, participants made both JOLs and belief-based recall predictions for large and small words. Our results suggest that metamemory beliefs play an important role in the font-size effect in JOLs.

Numerous studies have provided experience-based or theory-based frameworks for the basis of judgment of learning (JOL). However, few studies have directly measured processing experience and beliefs related to the same cue in one experiment and examined their joint contribution to JOLs. The present study focused on font-size effects and aimed to examine the simultaneous contribution of processing fluency and beliefs to the effect of font size on JOLs. We directly measured processing fluency via self-paced study time. We also directly measured participants’ beliefs via two approaches: pre-study global differentiated predictions (GPREDs) as an indicator of preexisting beliefs about font size and memory and ease of learning judgments (EORs) as online generated item-specific beliefs about fluency. In Experiment 1, EORs partially mediated the font-size effect, whereas self-paced study time did not. In Experiments 2a and 2b, EORs mediated the font-size effect; at the same time, beliefs about font size and memory moderated the font-size effect. In summary, the present study demonstrates a major role of beliefs underlying the font-size effect.

The dual-basis theory of metamemory suggests that people evaluate their memory performance based on both processing experience during the memory process and their prior beliefs about overall memory ability. However, few studies have proposed a formal computational model to quantitatively characterize how processing experience and prior beliefs are integrated during metamemory monitoring. Here, we introduce a Bayesian inference model for metamemory (BIM) which provides a theoretical and computational framework for the metamemory monitoring process. BIM assumes that when people evaluate their memory performance, they integrate processing experience and prior beliefs via Bayesian inference. We show that BIM can be fitted to recall or recognition tasks with confidence ratings on either a continuous or discrete scale. Results from data simulation indicate that BIM can successfully recover a majority of generative parameter values, and demonstrate a systematic relationship between parameters in BIM and previous computational models of metacognition such as the stochastic detection and retrieval model (SDRM) and the meta-d′ model. We also show examples of fitting BIM to empirical data sets from several experiments, which suggest that the predictions of BIM are consistent with previous studies on metamemory. In addition, when compared with SDRM, BIM could more parsimoniously account for the data of judgments of learning (JOLs) and memory performance from recall tasks. Finally, we discuss an extension of BIM which accounts for belief updating, and conclude with a discussion of how BIM may benefit metamemory research.

Wanlin Zhao and Baike Li contributed equally to the present study.The data contained in this project and the pre-registration of Experiment 2 are publicly available at the Open Science Framework (OSF) at https://osf. io/7gcwp/.

No abstract