Downloads & Free Reading Options - Results

Prior research has shown that retrieval can enhance learning of new (i.e., not-previously presented) information. This forward testing effect (FTE; for reviews, see Pastötter & Bäuml, 2014; Yang et al., 2018, 2021) has already been demonstrated with various study materials (e.g., single words, key-term definitions, texts and videos; for a meta-analysis, see Chan et al., 2018b). It has been suggested that metacognitive judgements, specifically Judgments of Learning (JOLs), are based on at least partial covert retrieval attempts when not all information is available (Dougherty et al., 2018), and thereby can produce retrieval-like effects. Thus, JOLs do not only index metacognitive monitoring (Rhodes, 2015), but can also enhance learning (Tauber & Witherby, 2019). Only a few prior studies examined the extent to which JOLs can enhance new learning of information (see Kubik et al., 2022; Lee & Ha, 2019). In the present study, we aim to investigate the relative advantage of JOLs (perhaps eliciting covert retrieval) and overt retrieval on new episodic and inductive learning compared to restudy. Lee and Ha (2019) examined forward effects of metacognitive judgments on inductive learning in two sections. In Section A, participants learned the painting styles of six artists by studying 36 painting–artist pairs (i.e., six paintings per artist). They were then asked to restudy these prior 36 pairs, take a cued-recall test on them (i.e., using the painting as a cue to retrieve the artist’s name), or evaluate the pairs’ learning status by performing a JOL task: Participants were either asked to perform item-based JOLs (i.e., assessing probability that they could identify the artist of a particular painting in a later test when the painting–artist pair is presented); category-based JOLs (i.e., assessing the probability that they could identify new paintings of an artist when the artist’s last name was shown); or global JOLs (in response to a single prompt that solicit an aggregate estimation of their future overall learning performance). In Section B, they learned six painting styles by studying 36 painting–artist pairs. Subsequently, they received a final cumulative transfer test, in which 48 new paintings (i.e., 4 of each artist) were presented in a single-choice test; participants’ task was to choose the correct artists’ last name. The authors demonstrated that category-based JOLs enhanced performance in a later transfer test identifying new paintings of the same artists, but item-based judgments did not. The latter finding in Experient 1 of Lee and Ha (2019) might be due to the fact that the task consisted of participants evaluating their learning of specific painting–artist pairs, which was inconsistent with the objective of learning the artists’ general painting styles and applying them in a transfer test. In addition, item-based judgments were solicited based on the complete cue–target pairs, that is, including both the cue (painting) and the target (last name of the artist). Lee and Ha (2019) argued that item-based JOLs cannot enhance new inductive learning of painting styles. We hypothesize that presenting the cue–target pairs rather than only the cue reduced the likelihood that participants engaged in covert retrieval when making JOLs. Based on the two-stage theory (Son & Metcalfe, 2005), making JOLs on complete information may only involve a first pre-retrieval stage of a quick familiarity assessment, while JOLs based on incomplete information likely covert retrieval attempts and thereby to enhance new inductive learning, similarly to overt retrieval practice. That is, participants are more likely to attempt complete retrieval attempts from memory when only the painting without the artist’s last name is presented than when both the cue (painting) and the target (last name of the artist) are presented. A first indication was provided by Kubik et al. (2022) using simple lists of words that JOLs based on incomplete information (i.e., cue-only JOLs) but not on complete information (i.e., cue–target JOLs) can enhance new learning. However, analogous research has not been conducted with more complex materials or inductive learning. To this end, we aim to replicate and extend the study by Lee and Ha (2019). We investigate and compare the benefits of JOLs and retrieval practice on future learning using pairs of paintings and artists. However, in addition to groups of cue–target JOLs, overt retrieval practice and restudy, we added an additional experimental group of cue-only JOLs.

Prior research has shown that retrieval can enhance learning of new (i.e., not-previously presented) information. This forward testing effect (FTE; for reviews, see Pastötter & Bäuml, 2014; Yang et al., 2018, 2021) has already been demonstrated with various study materials (e.g., single words, key-term definitions, texts and videos; for a meta-analysis, see Chan et al., 2018b). It has been suggested that metacognitive judgements, specifically Judgments of Learning (JOLs), are based on at least partial covert retrieval attempts when not all information is available (Dougherty et al., 2018), and thereby can produce retrieval-like effects. Thus, JOLs do not only index metacognitive monitoring (Rhodes, 2015), but can also enhance learning (Tauber & Witherby, 2019). Only a few prior studies examined the extent to which JOLs can enhance new learning of information (see Kubik et al., 2022; Lee & Ha, 2019). In the present study, we aim to investigate the relative advantage of JOLs (perhaps eliciting covert retrieval) and overt retrieval on new episodic and inductive learning compared to restudy. Lee and Ha (2019) examined forward effects of metacognitive judgments on inductive learning in two sections. In Section A, participants learned the painting styles of six artists by studying 36 painting–artist pairs (i.e., six paintings per artist). They were then asked to restudy these prior 36 pairs, take a cued-recall test on them (i.e., using the painting as a cue to retrieve the artist’s name), or evaluate the pairs’ learning status by performing a JOL task: Participants were either asked to perform item-based JOLs (i.e., assessing probability that they could identify the artist of a particular painting in a later test when the painting–artist pair is presented); category-based JOLs (i.e., assessing the probability that they could identify new paintings of an artist when the artist’s last name was shown); or global JOLs (in response to a single prompt that solicit an aggregate estimation of their future overall learning performance). In Section B, they learned six painting styles by studying 36 painting–artist pairs. Subsequently, they received a final cumulative transfer test, in which 48 new paintings (i.e., 4 of each artist) were presented in a single-choice test; participants’ task was to choose the correct artists’ last name. The authors demonstrated that category-based JOLs enhanced performance in a later transfer test identifying new paintings of the same artists, but item-based judgments did not. The latter finding in Experient 1 of Lee and Ha (2019) might be due to the fact that the task consisted of participants evaluating their learning of specific painting–artist pairs, which was inconsistent with the objective of learning the artists’ general painting styles and applying them in a transfer test. In addition, item-based judgments were solicited based on the complete cue–target pairs, that is, including both the cue (painting) and the target (last name of the artist). Lee and Ha (2019) argued that item-based JOLs cannot enhance new inductive learning of painting styles. We hypothesize that presenting the cue–target pairs rather than only the cue reduced the likelihood that participants engaged in covert retrieval when making JOLs. Based on the two-stage theory (Son & Metcalfe, 2005), making JOLs on complete information may only involve a first pre-retrieval stage of a quick familiarity assessment, while JOLs based on incomplete information likely covert retrieval attempts and thereby to enhance new inductive learning, similarly to overt retrieval practice. That is, participants are more likely to attempt complete retrieval attempts from memory when only the painting without the artist’s last name is presented than when both the cue (painting) and the target (last name of the artist) are presented. A first indication was provided by Kubik et al. (2022) using simple lists of words that JOLs based on incomplete information (i.e., cue-only JOLs) but not on complete information (i.e., cue–target JOLs) can enhance new learning. However, analogous research has not been conducted with more complex materials or inductive learning. To this end, we aim to replicate and extend the study by Lee and Ha (2019). We investigate and compare the benefits of JOLs and retrieval practice on future learning using pairs of paintings and artists. However, in addition to groups of cue–target JOLs, overt retrieval practice and restudy, we added an additional experimental group of cue-only JOLs.