Why does material that feels easy during study often fail to stick?
During study, learners often judge their learning by how smoothly the material goes down—how easy it is to read, how familiar it feels, how quickly an answer comes to mind. These feelings of fluency are vivid and immediate, but they are not reliable signals of whether the information will be available after a delay. A practical system for spaced repetition notes that people's judgments of learning can be distorted by what feels easy during study, even when that ease does not translate into durable retention [1].
The problem is that fluency can come from sources unrelated to long-term memory. In a study with 202 children around age 10, text printed in hard-to-read fonts was recalled worse than fluent text, yet a mixed-font condition led to more accurate performance judgments [5]. In other words, the feeling of ease or difficulty while reading did not track actual memory, but varying the font may have pushed learners to rely less on a simple fluency cue. This is consistent with the broader idea that study-time judgments are built from whatever cues are salient, and those cues can be misleading [1][3].
Why does when you judge your learning change how accurate the judgment is?
The timing of a judgment of learning matters because immediate and delayed judgments draw on different information. According to the Monitoring Dual Memories account described in a 2024 study, immediate judgments are primarily based on short-term memory information—such as font size or momentary familiarity—while judgments made after a delay rely more on long-term memory information [3]. That shift helps explain why delayed judgments tend to be more accurate: they are less anchored in the fleeting feeling of the study moment.
A 2012 experiment tested this directly by having participants study paired associates and then either restudy the items or make a judgment of learning for each item, followed by a final cued recall test after either 5 minutes or a longer interval [2]. The design separates the act of judging from the act of restudying, allowing the researchers to ask whether making a judgment itself changes later retention. The finding that delayed judgments can promote long-term retention suggests that the timing of self-assessment is not a neutral measurement—it can shape how well information is later recalled [2].
What makes learners' judgments line up better with actual recall?
One route to better alignment is to delay the judgment. When learners judge after a delay, they have to retrieve the information rather than simply recognize that it feels familiar, and that retrieval attempt provides a more memory-relevant basis for the judgment [3]. The 2024 study aims to replicate the effect that judgments of learning are more accurate when made after a delay, and it also examines cue utilization—specifically the font size effect—under immediate versus delayed conditions [3]. The underlying idea is that delayed judgments are less likely to be fooled by surface cues like font size and more likely to reflect whether the information is actually accessible in long-term memory [3].
Another route is to change the study conditions so that fluency is less misleading. In the children's font study, the mixed-font condition produced more accurate judgments of performance even though disfluent fonts did not improve—and in fact impaired—recall [5]. That dissociation is important: making judgments more accurate is not the same as making memory better. The study concludes that applying disfluency principles in education is premature without further research into mechanisms and moderators in young learners [5].
Does making a judgment of learning always improve later performance?
Not necessarily. A 2023 study on delayed metacomprehension judgments found that delayed judgments did not directly improve learning from texts after a one-week delay [4]. The abstract reports differences between groups—for example, a delayed judgment group compared with a no-retrieval group—but the overall conclusion is that the judgment itself is not a reliable learning booster [4]. This qualifies the picture: delaying a judgment may make it more accurate as a monitor of memory, but accuracy in monitoring does not automatically translate into stronger retention.
The practical implication is that learners should treat their study-time feelings of ease as weak evidence. Judging after a delay, or varying conditions so that fluency is not the only cue, can make self-assessment more informative [3][5]. But the 2023 text study cautions that improved judgment accuracy is not the same as improved learning, and the 2012 paired-associates study suggests that any retention benefit depends on how the judgment is embedded in study [2][4]. The divergence between judgment and delayed recall is therefore best understood as a cue problem—one that can be reduced, but not assumed to disappear, by changing when or how learners judge their own learning [1][3].
Sources used in this answer
The Spaced Study Engine: A Practical System for Spaced Repetition, Active Recall, Interleaving, and Long-Term Retention
People's judgments of learning can be distorted by what feels easy during study, and a spaced-repetition system should lengthen intervals when delayed recall is strong and shorten or repair practice when it is not.
Study for now, but judge for later: Delayed judgments of learning promote long‐term retention
In a paired-associates experiment, making a judgment of learning instead of restudying was followed by a final cued recall test after either 5 minutes or a longer interval, and delayed judgments of learning were found to promote long-term retention.
Cue-Utilization When Making Delayed Judgments of Learning (JOL)
Delayed judgments of learning are expected to be more accurate than immediate ones because immediate judgments rely on short-term memory cues such as font size, while delayed judgments draw on long-term memory information.
Delayed metacomprehension judgments do not directly improve learning from texts
Delayed metacomprehension judgments did not directly improve learning from texts after a one-week delay, with group differences observed but no overall learning benefit from the judgment itself.
Switching between fluent and disfluent fonts improves young learners’ estimation of memory effectiveness but not memory itself
In 202 children around age 10, disfluent fonts did not improve recall and recall was poorer for disfluent text, but a mixed-font condition led to more accurate judgments of performance.