Abstract:

The font-size effect refers to the phenomenon by which the judgments of learning (JOLs) are higher for words presented in a larger versus smaller font size. Recently, it has received a great deal of attention in the area of metacognition because the cognitive mechanism of this effect can provide a way to understand how individuals make judgments of learning, which has been a central question in metacognitive monitoring research. So far, there have been two hypotheses about the mechanism underlying this effect: (a) the fluency hypothesis, which claims that JOLs are higher for larger words because they are presumably easier to process and (b) the belief hypothesis, which argues that the font-size effect is caused by people’s beliefs (e.g., beliefs about how font size affects memory). Recently, Mueller and Dunlosky (2017) proposed a new account that the individual’s belief about processing fluency could produce the font-size effect and went on to provide indirect evidence of this. Building on the work of Mueller and Dunlosky, the present study aims to provide direct evidence supporting this idea that beliefs about processing fluency influence the font-size effect. Furthermore, the current study extended the work of Mueller and Dunlosky by splitting the beliefs about processing fluency into two components: the impact of font size on processing fluency and the impact of processing fluency on memory performance.
In this study, we conducted two experiments to investigate the influence of beliefs about processing fluency on the font-size effect via different instructions. Experiment 1 focused on the influence of beliefs about how font size impacts processing fluency on JOLs. Seventy-five participants were randomly assigned to three groups: group 1 received instructions about how the large font words were easier to process, group 2 received instructions about how small font words were easier to process, while the control group did not receive any such instructions. Then, all participants studied word pairs in large (48-point) or small (18-point) font sizes, made JOLs for each word pair and completed a cued-recall test. Experiment 2 focused on the influence of beliefs about how processing fluency impacts memory on JOLs. Eighty-nine participants were randomly assigned to three groups: group 1 received instructions stating that easier processing was positively associated with better memory performance, group 2 received instructions stating that processing fluency was unrelated to memory performance, and the control group did not receive any such instructions. All participants completed tasks similar to experiment 1.
There were two main results. First, JOLs were higher for large fonts when participants were led to believe that the large font was easier to process (group 1 in experiment 1) or the ease of processing was positively related to better memory performance (group 1 in experiment 2). Second, no difference in JOLs was observed when participants were instructed to believe that the small font was processed much more easily (group 2 in experiment 1) or processing fluency was irrelevant to memory performance (group 2 in experiment 2), i.e., there was no font-size effect.
In conclusion, our results provide direct evidence that beliefs about processing fluency can produce the font-size effect, and that they play a vital role in judgments of learning.

Key words: font-size effect, judgments of learning, beliefs about processing fluency