Collaborative Retrieval Practice Reduces Mind-Wandering During Learning · Knopps AG, Wissman KT

Method

Undergraduates who participated for course credit (n = 106) were randomly assigned to one of two groups: collaborative or individual. Participants were excluded from final analysis due to noncompliance (i.e., responding to fewer than 75% of the mind-wandering prompts; n = 5). When there was noncompliance with one (collaborative or nominal) group member, data for the other group member were also excluded (n = 5). The final sample included n = 96 participants (55% female; average age = 19 years): n = 50 in the individual group and n = 46 in the collaborative group. The a priori sample size was n = 102, with power set to .80 and one-tailed α = .05 to detect a medium effect of d = .50 (Faul et al., 2009). This experiment’s pre-registration can be found in Open Science Framework (https://osf.io/u7xpe).

Materials

Materials included a 48-item word list, comprising six exemplars from eight categories (Van Overschelde et al., 2004). Materials also included mind-wandering probes (Peterson & Wissman, 2020; Unsworth & Robison, 2016), which asked participants “What were you just thinking about?” and provided the following six response options: (1) The current task. (2) My performance on the current task. (3) A memory from the past. (4) Something in the future. (5) Current state of being. (6) Other. Finally, three postexperiment questions (PEQs) asked about participants’ perception of task difficulty (from 1 = very easy to 5 = very difficult), confidence in staying on task (from 1 = extremely unconfident to 5 = extremely confident), and motivation to stay on task (from 1 = very unmotivated to 5 = very motivated).

Procedure

Prior to beginning the experiment, participants were given a brief overview, which included our interest in examining mind-wandering. Participants were told mind-wandering was a normal phenomenon and asked to respond honestly to the probes presented throughout the experiment (for a similar approach, see Peterson & Wissman, 2020). Participants were also informed they may be asked to work in a group at some point during the experiment (although it was not specified when collaboration would occur).

Phase 1 involved an initial study of the exemplars and was completed individually by all participants. Prior to being shown the exemplars to study, participants were informed that they would be tested on their memory for the words later in the experiment. Exemplars were presented in a random order two times, and participants were given 5 s to study each exemplar. Following Phase 1, participants were told to see the research assistant for further instructions and then informed whether they would work alone or in a group for the next phase. Phase 2 involved recall of the previously studied words and was completed either individually or collaboratively (as a dyad). During this phase, participants were asked to type words into a provided textbox and hit enter after each word. After hitting enter, the word was added to a textbox shown on the screen to remind participants which words they recalled. Participants working individually sat at their own computer to recall the words. Participants working collaboratively sat at one computer to recall the words; one group member was instructed to type in their responses, and they were told to work together however they saw fit. During Recall 1, participants were given 5 min and asked to recall as many words as they could remember. Participants then engaged in Recall 2, which occurred in the same way as Recall 1. All participants were asked to do their best to use the entire time to recall as many words as possible and told that the computer would automatically advance them when time was up. Critically, participants were presented with a total of eight mind-wandering probes across Recalls 1–2. All participants provided individual responses to mind-wandering probes via pencil-and-paper.1

Participants were given 10 s to respond to the probe. The mind-wandering probes occurred at four predetermined times during Recall 1 (37 s, 101 s, 156 s, and 268 s) and Recall 2 (79 s, 134 s, 203 s, and 240 s) to ensure that participants in both groups provided mind-wandering ratings at the same time. Following Phase 2, all participants completed an unrelated filler task for 3 min. Phase 3 was completed individually by all participants. During this phase, participants took the final test; they were given 7 min and asked to recall as many words as they could remember. Participants then answered the three PEQs about task experience.

Scoring

Consistent with mind-wandering literature, responses to mind-wandering probes were categorized as “on task” if a participant responded with “The current task” or “My performance on the current task”, with all other responses coded as “off task”; for interested readers, the breakdown of all six response categories can be accessed via OSF (https://osf.io/52v8w).

Consistent with collaborative memory literature, performance during Phase 2 compares collaborative recall (n = 25 groups) versus nominal recall (n = 23 groups), with nominal groups being formed by combing the protocols (i.e., pooling nonredundant items) of two participants in the individual group who completed the experiment on approximately the same day. For performance during Phase 3, outcomes are reported at the individual level.