Results
Figure 1 provides the proportion of mind-wandering probes eliciting off-task thinking for the collaborative versus individual group during Recall 1 (left set of bars) and Recall 2 (right set of bars). The results indicate that participants in the collaborative versus individual group were less likely to report being off task during both Recall 1 [t(94) = 2.02, p = .023, d = .41] and Recall 2 [t(94) = 2.90, p = .002, d = .59]. These outcomes provide novel evidence that collaborative retrieval practice reduces the frequency of mind-wandering during learning. For exploratory purposes, we conducted a 2 (group: collaborative vs. individual) × 8 (mind-wandering probe: 1 vs. 2 vs. 3 vs. 4 vs. 5 vs. 6 vs. 7 vs. 8) repeated-measures ANOVA2 to provide a finer-grained analysis of mind-wandering rates across all eight probes for both groups (see Figure 2).
Figure. Mind-wandering rates during Recall 1 and Recall 2. Proportion of mind-wandering probes eliciting off-task thinking responses. Error bars report standard error of the mean. Cohen’s d effect size is reported in the box. (See the original article for the figure.)
Figure. Mind-wandering rates across all eight probes. Proportion of mind-wandering probes eliciting off-task responses. MW refers to the mind-wandering probe number. Error bars report standard error of the mean. (See the original article for the figure.)
Outcomes revealed a significant main effect of group [F(5.8, 543.7) = 10.47, p < .001, ηp2 = .100] and mind-wandering probes [F(1, 94) = 8.93, p = .004, ηp2 = .087]; the interaction was not significant (p = .157). These outcomes indicate mind-wandering increased across the experiment (for similar findings, see Seli et al., 2018) and the likelihood of mind-wandering was reduced for participants in the collaborative versus individual group.
Concerning recall, Figure 3 provides the number of words recalled during retrieval practice in Phase 2 (Recall 1 and Recall 2; see left two sets of bars) and on the final test in Phase 3 (see right set of bars) for the collaborative versus individual group. The results revealed that recall during practice was lower for participants in the collaborative versus individual group during Recall 1 [t(46) = 2.01, p = .025, d = .58] and Recall 2 [t(46) = 1.38, p = .088, d = .40]. These outcomes provide evidence of collaborative inhibition during practice, which replicates prior research (Blumen & Rajaram, 2008; Weldon & Bellinger, 1997). The results also showed that recall on the final test was significantly greater for the participants who previously engaged in collaborative (M = 25, SE = 1) versus individual (M = 20, SE = 1) retrieval practice, t(93) = 2.75, p = .004, d = .79. This outcome provides evidence of postcollaborative benefits, also replicating prior research (Blumen & Rajaram, 2008; Wissman & Rawson, 2015).
Figure. Words recalled during Recall 1, Recall 2, and final test. The number of exemplars recalled during Phase 2 and Phase 3. Error bars report standard error of the mean. Cohen’s d effect size is reported in the box. (See the original article for the figure.)
Finally, we examined responses to the three PEQs (see Table 1). Participants in the collaborative versus individual group reported significantly less difficulty in staying on task, t(94) = 2.31, p = .012, d = .47. Responses to PEQs regarding confidence and motivation did not differ across the two groups (p = .065 and p = .358, respectively).
Table. Participant responses to postexperiment questions (PEQ) (See the original article for the full table.)
CC BY 4.0. "Collaborative Retrieval Practice Reduces Mind-Wandering During Learning" by Knopps AG, Wissman KT, Experimental psychology (2023). https://doi.org/10.1027/1618-3169/a000589. Licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Peer-reviewed and open access; full text from PubMed Central (PMC10929687). Headings and layout adapted for reading; figures and tables are in the original.