View in EDS HTML Full Text PDF Full Text

Children Boost Their Cognitive Performance with a Novel Offloading Technique

Saved in:
Bibliographic Details
Title: Children Boost Their Cognitive Performance with a Novel Offloading Technique
Language: English
Authors: Armitage, Kristy L. (ORCID 0000-0002-1898-8622), Redshaw, Jonathan (ORCID 0000-0002-7729-1577)
Source: Child Development. Jan-Feb 2022 93(1):25-38.
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 14
Publication Date: 2022
Document Type: Journal Articles
Reports - Research
Descriptors: Children, Cognitive Style, Cognitive Processes, Problem Solving, Learning Strategies, Maps, Spatial Ability, Age Differences, Individual Differences
DOI: 10.1111/cdev.13664
ISSN: 0009-3920
Abstract: Ninety-seven children aged 4-11 (49 males, 48 females, mostly White) were given the opportunity to improve their problem-solving performance by devising and implementing a novel cognitive offloading strategy. Across two phases, they searched for hidden rewards using maps that were either aligned or misaligned with the search space. In the second phase, maps were presented on rotatable turntables, thus allowing children to manually align all maps and alleviate mental rotation demand. From age six onwards, children showed strong evidence of both mentally rotating misaligned maps in phase 1 and manually aligning them in phase 2. Older children used this form of cognitive offloading more frequently, which substantially improved performance and eliminated the individual differences observed in phase 1.
Abstractor: As Provided
Notes: https://osf.io/485ka
Entry Date: 2022
Accession Number: EJ1327043
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwEFy1jqTpBMQTkQmo2yjS6uAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDEzlgg2_PXyGEG4_qwIBEICBm4eiP_unwiuDLq5atKxlwIOYIS0zDmmRwaYai7yPbaYgEC6yslPX6XuvFn8i__K3nxNFCmuZ21idZ8S1QISWIqhcztKngIjuzWYowAFKQXIz540SudQ-pItXmhDvhLSFDwl0cm2qU_itxbF74nMwgxyCSDsgai7jWyFYP4PSaOxmNmd3AQxDzwp2Kqs5QRZJv5pyjUn9J_FX_WZm
Text:
  Availability: 1
  Value: <anid>AN0154716022;cdv01jan.22;2022Jan19.03:22;v2.2.500</anid> <title id="AN0154716022-1">Children boost their cognitive performance with a novel offloading technique </title> <p>Ninety‐seven children aged 4–11 (49 males, 48 females, mostly White) were given the opportunity to improve their problem‐solving performance by devising and implementing a novel cognitive offloading strategy. Across two phases, they searched for hidden rewards using maps that were either aligned or misaligned with the search space. In the second phase, maps were presented on rotatable turntables, thus allowing children to manually align all maps and alleviate mental rotation demand. From age six onwards, children showed strong evidence of both mentally rotating misaligned maps in phase 1 and manually aligning them in phase 2. Older children used this form of cognitive offloading more frequently, which substantially improved performance and eliminated the individual differences observed in phase 1.</p> <p></p> <ulist> <item> Abbreviations</item> <p></p> <item> GEEs generalized estimating equations</item> <p></p> <item> GLMMs generalized linear mixed models</item> </ulist> <p>A defining feature of human intelligence is the ability to devise and implement external means of bypassing our natural cognitive limits (Bocanegra et al., 2019; Clark, 2008; Clark & Chalmers, 1998). When we need to remember to perform a task in the future, for instance, we prevent ourselves from forgetting by writing reminders on our hands, storing information in calendars, or creating alarms on our phones. We use finger counting to avoid losing count, and calculators when confronted with more difficult mathematical problems. We use gestures and drawings to explain concepts, turn to Google for answers to our questions, and rely on navigational devices to find our way home. Evidently, the capacity to create and use external solutions for cognitive problems, or <emph>cognitive offloading</emph> (Risko & Gilbert, 2016), radically expands what we are able to achieve when compared to relying solely on our unaided internal capacities (Clark & Chalmers, 1998).</p> <p>Despite being an integral facet of human intelligence, the role of cognitive offloading in problem solving has only recently gained attention within cognitive and developmental science. Consider, for example, the history of mental rotation research. Since Shepard and Metzler's (1971) classic finding that mental rotation time varies linearly with the angle of rotation, an abundance of highly controlled studies have examined how adults and children perform internal cognitive transformations, and the importance of this capacity for the execution of behavioral tasks (e.g., Estes, 1998; Moè, 2009). What is far less researched, however, is our tendency to avoid effortful mental rotation in our everyday lives. For example, we tend to <emph>manually</emph> rotate an upside‐down piece of paper to its upright orientation rather than trying to mentally rotate the inverted text, or tilt our heads toward the angle of text alignment when manual rotation is not an option (Dunn & Risko, 2016; Risko et al., 2014). Via this process of <emph>external normalization</emph> (Risko et al., 2014), we offload such internal cognitive transformations into the world.</p> <p>Developmental psychology studies have established a clear relationship between mental and manual rotation. For instance, 5‐year‐olds who spontaneously use manual gestures to explain mental rotation tend to perform better on mental rotation tasks than their peers who do not use such gestures (Ehrlich et al., 2006), and both preschoolers and school‐aged children perform better on mental rotation tasks after a training phase involving analogous manual rotations (Ping et al., 2011; Wiedenbauer & Jansen‐Osmann, 2008). Furthermore, one study found that 5‐ and 8‐year‐olds, but not 11‐year‐olds and adults, showed impaired performance in a mental rotation task when performing a rotational hand movement in the direction opposite to the direction of mental rotation (Frick et al., 2009). These studies demonstrate how the use of manual rotation can both supplement and interfere with analogous mental rotation performance, especially in young children. Nonetheless, they do not explore the cognitive offloading decisions that children make when given the option to use internal or external means of solving such spatial transformation tasks.</p> <p>To date, only one study has directly examined children's offloading decisions in the context of mental rotation. Armitage et al. (2020) introduced 4‐ to 11‐year‐old children to a stimulus matching task, in which they had to identify whether two figures presented at angular disparities of 0–180 degrees were identical or mirrored. Critically, one of the figures was always presented on a rotatable turntable, which children could manually rotate to reduce the angular disparity between the figures and thus offload mental rotation demand. Few 4‐ and 5‐year‐olds rotated the turntable as a function of the degree of angular disparity, but the prevalence of this selective offloading behavior substantially increased in the 6‐ and 7‐year‐olds. Another sample of children completed a version of the task where manually rotating inverted stimuli would ostensibly benefit performance on some trials, but not others. Between 4 and 11 years, children became increasingly more likely to manually rotate inverted stimuli only on the trials where it would ostensibly benefit performance. Similarly, Bulley et al. (2020) found that 4‐ to 11‐year‐old children were more likely to use external reminders to offload memory demand as the amount of information to be recalled increased. Overall, these studies suggest that even young children are more likely to offload in situations where doing so will provide a greater performance benefit.</p> <p>A major limitation of these studies, however, is that the children were not typically required to first complete the tasks without the option of cognitive offloading, preventing any direct examination of the benefit to performance afforded by this option. Does children's overall cognitive performance improve after the opportunity to offload becomes available? In one of their experiments, Bulley et al. (2020) found that 4‐ to 11‐year‐old children's memory performance did improve after being given the option to offload cognitive demand, but in this case, the children were directly shown the target offloading behavior of reminder‐setting without needing to infer it themselves. Redshaw et al. (2018) similarly found that 6‐ to 13‐year‐old children's memory performance improved after they were directly taught a reminder‐setting behavior by an experimenter. Conversely, Berry et al. (2019) found that 8‐ to 10‐year‐old children's working memory performance did <emph>not</emph> improve after they were provided with the opportunity to devise their own cognitive offloading strategy. There is currently no evidence that children can improve on their unaided cognitive performance when given the option to infer a novel offloading behavior.</p> <p>Notably, these studies examined children's offloading propensities only on basic mental rotation (Armitage et al., 2020), short‐term memory (Bulley et al., 2020; Redshaw et al., 2018), and working memory tasks (Berry et al., 2019), rather than on tasks reflecting real‐world contexts in which children might be expected to devise and benefit from offloading strategies. One such context, in which much is known about children's internal cognitive capabilities, but little is known about their offloading behavior, is map‐reading.</p> <hd id="AN0154716022-2">Children's understanding and use of maps</hd> <p>To use a map successfully, an individual must understand that the locations of and spatial relations between the symbols on the map represent the locations of and spatial relations between real objects in the referent environment (Presson, 1982). In a series of classic experiments, DeLoache (1987, 1989) showed children a miniature toy being hidden in a small three‐dimensional model of a room, and found that 3‐year‐olds, but not 2.5‐year‐olds, were able to retrieve the larger version of the toy hidden in the corresponding location of the room itself. Subsequent studies have shown that even 2.5‐year‐olds can use geometric maps to locate hidden objects (Winkler‐Rhoades et al., 2013), and that 3‐year‐olds' understanding of spatial correspondence in map tasks significantly improves following a targeted intervention (Yuan et al., 2017). By age four, many children understand spatial correspondence well enough to distinguish between multiple identical objects represented on a map even in the absence of an intervention (Blades & Cooke, 1994), and can use both concrete and abstract maps to help solve maze tasks (Jirout & Newcombe, 2014).</p> <p>While the evidence on young children's map‐reading abilities is nuanced and varies across contexts, there is widespread consensus that children perform better when a map is aligned with the referent area than when it is misaligned (Blades & Cooke, 1994; Blades & Spencer, 1987a, 1987b, 1990; Bluestein & Acredolo, 1979; Levine et al., 1984; Liben & Downs, 1993; Vosmik & Presson, 2004). Studies have also reliably found that children's capacity to use misaligned maps to locate hidden objects improves with age (Blades & Cooke, 1994; Blades & Spencer, 1990; Bluestein & Acredolo, 1979), paralleling the developmental changes in performance on traditional timed mental rotation tasks (Estes, 1998; Frick et al., 2013). As in research with adults, however, developmental studies have rarely accounted for the fact that, when humans encounter handheld maps in our everyday lives, we often manually rotate them to match our orientation in space.</p> <p>Only in several small‐scale studies have children been given the opportunity to manually rotate misaligned maps to match the referent environment, with the samples limited to 4‐ and 5‐year‐olds (Bremner & Andreasen, 1998; Vosmik & Presson, 2004) and 9‐ and 10‐year‐olds (Liben et al., 2013). Very few 4‐ and 5‐year‐olds chose to manually align maps (7 of the 40 participants in Bremner & Andreasen, 1998; 1 of the 12 participants in Vosmik & Presson, 2004), consistent with the relatively low levels of offloading observed in children of this age in basic mental rotation tasks (Armitage et al., 2020). Nonetheless, given that the maps were intentionally presented to children at a specific orientation without verbal permission to alter that orientation, some children may have been uncertain about whether they were allowed to manually rotate the maps during the tasks. In the 9‐ and 10‐year‐old sample, children were better able to navigate unfamiliar surroundings if they spontaneously used map‐space coordinating strategies, such as manually rotating a misaligned map or tracing a route with their fingers (Liben et al., 2013).</p> <p>The current study similarly investigated children's ability to interpret aligned and misaligned maps, and uniquely compared performance across phases where an offloading strategy was either unavailable or available. To capture the developmental trajectory of this ability, the sample included a broader age range than the previous map‐reading studies that allowed for cognitive offloading (Bremner & Andreasen, 1998; Liben et al., 2013; Vosmik & Presson, 2004), and accounted for permission issues by explicitly introducing children to the possibility of manual rotation. Nonetheless, as in real‐world cognitive offloading contexts, children were required to infer themselves that some maps were misaligned, and that they could manually rotate misaligned maps to match the search space.</p> <hd id="AN0154716022-3">The current study</hd> <p>Across two experimental phases, 4‐ to 11‐year‐old children were asked to find hidden stickers using maps that were either aligned (0‐degree angular disparity) or misaligned (90‐degree or 180‐degree angular disparity) with the referent environment. In phase 1, children were required to rely on their mental rotation abilities alone when using the misaligned maps to find the stickers. In phase 2, by contrast, all maps were presented on rotatable turntables, which children were given permission to rotate if desired. Critically, children were never told that some of the maps were misaligned, or how manual rotation could be used to align a misaligned map. Unlike in previous experiments comparing children's unaided and aided performance (Bulley et al., 2020; Redshaw et al., 2018), this design allowed us to directly examine the benefit to children's performance afforded by the option to infer a novel offloading strategy, and whether any such performance benefit varies with age.</p> <p>As even 4‐year‐old children appear competent with using aligned maps (Blades & Cooke, 1994; Bluestein & Acredolo, 1979), we expected that children of all ages would be faster and more accurate at locating targets when using aligned maps compared to misaligned maps in phase 1. In phase 2, however, we expected that children's speed and accuracy would be similar regardless of whether they were using an aligned or misaligned map, as long as they inferred and deployed the offloading strategy of manually rotating misaligned maps to match the search space. Given age‐related improvements in both mental rotation (Estes, 1998; Frick et al., 2013) and the understanding of spatial relations (Blades & Cooke, 1994), we also expected older children to perform better when using misaligned maps than younger children in both phases.</p> <p>As adults' cognitive offloading decisions are underpinned by metacognitive judgments about one's ability to perform the cognitive task unaided (Hu et al., 2019; Risko & Gilbert, 2016), we expected that children must first possess this metacognitive insight before being able to offload cognitive demand. We therefore also included a rough, categorical metacognitive measure, which examined whether children could differentiate between the difficulty of aligned and misaligned conditions both before and after the main search task. We expected children of all ages to perform above chance on this measure, as even 4‐year‐olds have shown evidence of basic metacognitive knowledge (Balcomb & Gerken, 2008; Neldner et al., 2015). Nonetheless, given the well‐established developmental lag between metacognitive knowledge and metacognitive control over behavior (Dufresne & Kobasigawa, 1989; Lockl & Schneider, 2004), and given previous findings on children's cognitive offloading in basic mental rotation tasks (Armitage et al., 2020), we also expected children to become more likely to offload mental rotation demand with increasing age in phase 2.</p> <hd id="AN0154716022-4">Exploratory questions</hd> <p>Our experimental design also permitted us to explore several other theoretically interesting questions about the propensities and functions of children's cognitive offloading, for which no specific hypotheses were formed prior to data collection. For instance, we explored whether children's use of cognitive offloading in phase 2 varied as a function of unaided performance failures in phase 1. Gilbert (2015) has shown that adults tailor their offloading decisions to their level of unaided competence, such that those who fail a cognitive task are more likely to use offloading strategies when subsequently given the opportunity. Previous developmental studies, however, have not examined whether children similarly offload as a function of unaided performance (Armitage et al., 2020; Berry et al., 2019; Bulley et al., 2020; Redshaw et al., 2018). We also explored whether children's search performance in phase 1 could predict their performance in phase 2, and whether this relationship was attenuated by the use of cognitive offloading. Finally, we explored children's spontaneous use of head tilting while looking at the maps, following previous observations of this behavior in mental rotation contexts in both adults (Dunn & Risko, 2016; Risko et al., 2014) and children (Armitage et al., 2020).</p> <hd id="AN0154716022-5">METHOD</hd> <p></p> <hd id="AN0154716022-6">Participants</hd> <p>Ninety‐seven children (49 males and 48 females) aged between 4.29 and 11.96 years (<emph>M </emph>= 7.98, SD = 2.29) were included in analyses. An additional 12 children were excluded due to experimenter error (<emph>n </emph>= 3), having a clinical diagnosis (<emph>n</emph> = 3), rotating the turntables before taking the cover off the maps (<emph>n </emph>= 4), being instructed to rotate the turntable by an observer (<emph>n </emph>= 1), and technical problems with video recordings (<emph>n</emph> = 1). The sample was mostly White and middle‐class, and from a medium‐sized Western and industrialized city. Age was consistently analyzed as a continuous variable, but, as specified in a preregistration of the study (see Note S1), all significant effects involving age were also followed up by splitting participants into four age groups. These groups were 4‐ and 5‐year‐olds (<emph>n </emph>= 25), 6‐ and 7‐year‐olds (<emph>n </emph>= 24), 8‐ and 9‐year‐olds (<emph>n </emph>= 27), and 10‐ and 11‐year‐olds (<emph>n</emph> = 21; see Table S1 for a breakdown of participants by age group and sex).</p> <p>Ethical approval was received prior to beginning recruitment, with all participants recruited through an existing university database. Verbal and written consent was obtained from children's guardians before testing, and children also provided verbal assent. Data were collected between March and May 2019.</p> <hd id="AN0154716022-7">Materials</hd> <p>In the main task, children used four maps to search for four stickers hidden under 36 identical white pots, presented in a doubly symmetric 6 × 6 grid in the center of the room (see Figure 1, left panel). A single red square mat was placed in the same corner of the room on every trial, acting as the only available landmark. Each map showed 36 circles in a 6 × 6 grid, representing the pots, and a red square in one corner, representing the landmark (see Figure 1, right panel). Maps were always placed on inverted buckets, such that they were at an appropriate height for children, and each map was initially covered by fabric. On each of the maps, one circle was bolded, indicating the location of one hidden sticker. Consequently, in order to locate all four stickers on each trial, children needed to use all four available maps. One map was located on each side of the search grid, and on each of the trials, only one of the four maps matched the child's orientation in space (0‐degrees). The other three maps were presented at rotated orientations—90‐degrees left, 90‐degrees right, and 180 degrees—which was only evident when comparing the location of the red square on the map to the red mat in the room. The hiding locations, alignment of each map, and order of presentation were counterbalanced (see Note S1 for full counterbalancing procedures).</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CDV/01jan22/cdev13664-fig-0001.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cdev13664-fig-0001.jpg" title="1 The left panel shows an example of the Phase 1 set‐up, with the search space comprised of the 36 white pots, the red mat acting as the landmark in the back left corner, and four maps placed on inverted buckets at different angular orientations. For example, the closest map is presented at a 180‐degree misalignment with the search space, which is evident when comparing the location of the red square on the map with the location of the landmark in the room. The yellow squares are standing mats, which children were instructed to remain on as they completed the task. The right panel shows an example of an aligned map, with the circles representing the 36 white pots, the red square representing the landmark, and the darkened circle representing the location of the hidden sticker." /> </p> <p></p> <hd id="AN0154716022-9">Measures and procedure</hd> <p></p> <hd id="AN0154716022-10">Initial training phase</hd> <p>After entering the testing room, children were instructed to stand on a yellow mat in front of a single map covered by fabric. They were then asked to remove the fabric and were told that they were looking at a map of the room. The map in this training phase always matched the child's orientation in space (0‐degrees). The experimenter explained that "the red square on the map shows the red square over in the corner" and "all these circles on the map are showing all these hiding places." During this explanation, the experimenter would point to each element on the map and the corresponding element in the room. Finally, the experimenter explained that "the dark circle is showing where the sticker is hidden." Children were then handed a pointing stick, which they could use to indicate where they wanted to search without needing to move away from the map, and were given as many attempts as necessary to find the single sticker.</p> <hd id="AN0154716022-11">Prospective metacognitive awareness measure</hd> <p>After receiving the initial instructions, and immediately before completing phase 1, children were simultaneously presented with two maps—one at 0‐degrees orientation with the room and the other at 180‐degrees. Children were asked to select the map that they believed would be easier to use when searching for a sticker, serving as an indication of whether they were able to differentiate between easier and harder trials (see Figure S1). After children had provided an answer, they were asked to sit facing the wall and the experimenter engaged them in conversation while setting up the following study phase, checking regularly to ensure the child was not watching the hiding events. The same procedure was used between all trials during the testing period.</p> <hd id="AN0154716022-12">Phase 1</hd> <p>Before the first phase, four maps were placed on the top of inverted buckets and covered with fabric. At the beginning of the phase, children were explicitly told that they were not able to move or change the orientation of the maps. As only one map in each trial matched the child's orientation in space (0‐degrees), all other maps needed to be mentally rotated to allow the child to accurately locate the stickers. Importantly, children were only able to search for the sticker once per map, discouraging them from simply guessing the location repeatedly until they found a sticker.</p> <p>Both phase 1 and phase 2 were composed of two trials (each with four maps and four hidden stickers). At the beginning of every trial across both phases, the experimenter would point to and verbally remind children about the location of the red mat in the room. Children were able to uncover the maps in whichever order they desired, with the constraint that they had to search for the depicted sticker before uncovering another map. For each map, children's search accuracy (whether they correctly located the same hidden sticker depicted on the map) and speed (the amount of time between removing the map cover and selecting a search location) were measured from video recordings.</p> <p>Notably, a previous investigation of children's offloading in basic mental rotation tasks revealed that children sometimes chose to tilt their heads when studying rotated stimuli, even when doing so did not facilitate performance (Armitage et al., 2020). We preregistered our intent to code similar instances of children spontaneously tilting their heads if this behavior was observed during testing. Head tilting was coded using video recordings of testing sessions and was only scored when it occurred after the child removed the cover from the map (i.e., head tilting was not scored if the child simply had their head in a tilted position before uncovering the map).</p> <hd id="AN0154716022-13">Phase 2</hd> <p>Before the second phase, the experimenter placed four rotatable turntables on the four inverted buckets. Maps were then placed on these turntables and covered with fabric, as in phase 1. At the beginning of each of the two trials, the experimenter told the child that they were allowed to rotate the turntables whenever they wanted during the task. The experimenter also rotated one of the turntables, still covered with fabric, as a demonstration. Although they were given permission to do so, children were never explicitly directed to use manual rotation on any trial, nor were they told how manual rotation could be used to align the misaligned maps with the search space. These measures were taken to ensure that children had to infer themselves that aligning the red square symbol on the map with the landmark in the room eliminated the need for mental rotation.</p> <p>Any use of manual rotation was recorded, but it was only scored as cognitive offloading if it served to align a rotated map with the search space (cognitive offloading was therefore not possible on 0‐degree trials). The final orientation of the map was recorded based on the closest orientation at the moment the child selected a pot to search (e.g., if the final orientation was closer to 0 degrees than 90 degrees left, it was coded as 0 degrees). Search accuracy, speed, and head tilting were also recorded, as in phase 1.</p> <hd id="AN0154716022-14">Retrospective metacognitive awareness measure</hd> <p>After finishing phase 2, children were again simultaneously presented with two identical maps—one at 0‐degrees orientation with the room and the other at 180‐degrees, as for the prospective metacognitive awareness measure. This time, however, rather than indicating which map they <emph>believed</emph> would be easier to use, children were instead asked to select the map that they actually found easier to use when searching for stickers during the task. Finally, they were thanked for their time, compensated with a small prize, and the caregiver was provided with a written and verbal debrief.</p> <hd id="AN0154716022-15">RESULTS</hd> <p></p> <hd id="AN0154716022-16">Analysis strategy</hd> <p>Data from both phases were analyzed using generalized linear mixed models (GLMMs) with a random intercept for each participant, using the statistical program SAS 9.4. Bonferroni corrections were applied to <emph>p</emph> values for all follow‐up tests involving multiple comparisons. All models were run once treating map orientation as a categorical variable (0, 90‐left, 90‐right, 180) and again treating map orientation as a linear variable (0–180). Only the categorical models are reported below, as we were primarily interested in children's behavior on misaligned trials compared to aligned trials (i.e., on trials with and without mental rotation demand). Accordingly, all significant effects involving map orientation were followed‐up by comparing performance on the aligned (0‐degree) trials to the average performance on misaligned (all 90‐ and 180‐degree) trials. As our intent to analyze map orientation as a categorical variable was not made explicit in our preregistration, we have also reported all linear models in a separate supplementary file for transparency. Full model details for the reported categorical models can be found in Notes S2 and S3.</p> <hd id="AN0154716022-17">Preregistered and exploratory analyses</hd> <p>Although our preregistered analysis plan specified only cross‐phase analyses (i.e., analyses incorporating data from <emph>both</emph> phase 1 and phase 2), we also report analyses from each phase separately in order to provide a fuller picture of children's unaided and aided performance patterns. In particular, we analyzed all dependent variables (accuracy, speed, head tilting, and rotation) first within each phase, and then across phases as preregistered (excluding rotation, which was not possible in phase 1). Any models examining the effects outlined in the Exploratory Questions sub‐section from the introduction have been explicitly labeled as exploratory both in text and in the supporting information, as these were not included in the preregistration.</p> <hd id="AN0154716022-18">Assumption checking</hd> <p>Inspection of the main effects models indicated that random intercepts were approximately normally distributed for all dependent variables, with the exception of speed, which had positively skewed intercepts in both phases and thus violated an assumption of GLMM analyses. We therefore also analyzed speed using marginal models (generalized estimating equations; GEEs), which, unlike conditional models (GLMMs), include no random effects (Muff et al., 2016). Patterns of statistical significance were identical across both the GLMM and GEE analyses, consistent with the robustness of mixed‐effects modeling to violations of assumptions (Schielzeth et al., 2020). For consistency across models, we therefore report GLMM results for all our DVs. Our full dataset and syntax are available on the Open Science Framework (https://osf.io/485ka/) for independent reproduction.</p> <hd id="AN0154716022-19">Reliability coding</hd> <p>A second rater independently coded behavior during the main task for 100% of the sample. Inter‐rater reliability analyses indicated very high agreement for phase 1 search accuracy (<emph>κ </emph>= .97), phase 1 speed (<emph>r </emph>= .926, <emph>p </emph>= .001), phase 2 search accuracy, (<emph>κ </emph>= .98), phase 2 speed (<emph>r </emph>= .988, <emph>p </emph>< .001), any use of manual rotation (<emph>κ </emph>= .98), and the use of manual rotation to the correct orientation (i.e., cognitive offloading; <emph>κ </emph>= .99). There was substantial agreement for phase 1 head tilting (<emph>κ </emph>= .68) and moderate agreement for phase 2 head tilting (<emph>κ </emph>= .59), but the main findings were consistent across both phases for both coders (see Note S1), and we therefore report the results from the primary coder only.</p> <hd id="AN0154716022-20">Phase 1</hd> <p></p> <hd id="AN0154716022-21">Accuracy</hd> <p>Phase 1 accuracy was modeled as a function of age, map orientation, trial, and use of head tilting. Children were significantly more likely to find the hidden stickers with increasing age, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0001" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref1">674</reflink>, _I_N _i_= 97) = 50.92, <emph>p </emph>< .001, <emph>w</emph> = 0.72 (Cohen, 1992), and on the second trial compared to the first, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0002" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref2">674</reflink>, _I_N_i_ = 97) = 12.77, <emph>p</emph> < .001, <emph>w</emph> = 0.36, indicating that across trials, they were becoming more likely to infer and internally compensate for the orientation of the maps. A significant main effect of map orientation, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0003" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref3">674</reflink>, _I_N _i_= 97) = 138.84, <emph>p </emph>< .001, <emph>w</emph> = 1.20, was followed up to reveal that children were also more likely to find the hidden stickers when using aligned maps than misaligned maps, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0004" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref4">674</reflink>, _I_N_i_ = 97) = 138.59, <emph>p</emph> < .001, <emph>w</emph> = 1.20 (see Figure 2). This effect of map orientation on accuracy did not significantly interact with age, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0005" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib671" id="ref5">671</reflink>, _I_N _i_= 97) = 5.62, <emph>p</emph> = .132, <emph>w</emph> = 0.24, contrary to the prediction that older children would perform better than younger children when using misaligned maps but not aligned maps. However, the effect of map orientation did vary by sex, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0006" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib671" id="ref6">671</reflink>, _I_N_i_ = 97) = 15.93, <emph>p</emph> = .001, <emph>w</emph> = 0.41, with females being significantly more accurate on the 0‐degree trials than males, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0007" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib671" id="ref7">671</reflink>, _I_N_i_ = 97) = 7.72, <emph>p </emph>= .022, <emph>w </emph>= 0.28. No sex differences were detected for the misaligned trials (all <emph>p</emph>s > .211, all <emph>w</emph>s < 0.20), or across any other dependent variables in the dataset.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CDV/01jan22/cdev13664-fig-0002.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cdev13664-fig-0002.jpg" title="2 The average accuracy and use of head tilting for each orientation in phase 1, split according to age group. On average, children were more accurate in their searches when using aligned compared to misaligned maps, and with increasing age. The older children were also more likely to tilt their heads while studying the maps than the younger children." /> </p> <p></p> <p>To explore whether children's errors could be explained by failures to infer that many of the maps were misaligned, a "treat as aligned" variable was calculated. For each trial, children were scored as either having treated the map as aligned or not, as determined by comparing their chosen search location to the location that they should have searched if they had treated the map as aligned with the search space. This variable was then modeled as a function of age, map orientation, and trial. The analysis revealed a significant effect of map orientation, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0008" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib675" id="ref8">675</reflink>, _I_N_i_ = 97) = 112.90, <emph>p </emph>< .001, <emph>w </emph>= 1.08, which interacted with age, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0009" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib672" id="ref9">672</reflink>, _I_N_i_ = 97) = 44.37, <emph>p </emph>< .001, <emph>w </emph>= 0.68. Follow‐up analyses showed that children aged 6 years and older appropriately treated the maps as aligned significantly more frequently on the 0‐degree compared to misaligned trials (all <emph>p</emph>s < .001, all <emph>w</emph>s > 0.96), whereas the 4‐ and 5‐year‐olds showed no evidence that they treated the misaligned maps any differently from the aligned maps, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0010" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib666" id="ref10">666</reflink>, _I_N_i_ = 25) = 3.18, <emph>p </emph>= .299, <emph>w</emph> = 0.36.</p> <hd id="AN0154716022-23">Speed</hd> <p>Phase 1 speed was modelled as a function of age, map orientation, trial, and use of head tilting. Older children spent significantly more time looking at the maps before selecting a hiding place than younger children, <emph>F</emph>(<reflink idref="bib1" id="ref11">1</reflink>, 674) = 5.46, <emph>p </emph>= .020, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.01. A significant main effect of map orientation, <emph>F</emph>(<reflink idref="bib3" id="ref12">3</reflink>, 674) = 20.13, <emph>p </emph>< .001, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.08, was followed up to reveal that children spent longer on the misaligned maps than aligned maps, <emph>F</emph>(<reflink idref="bib1" id="ref13">1</reflink>, 674) = 53.64, <emph>p </emph>< .001, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.07. A further, exploratory analysis showed that children's search times also significantly increased from 90‐degree trials (collapsing across both 90‐degree right and 90‐degree left trials) to 180‐degree trials, <emph>F</emph>(<reflink idref="bib1" id="ref14">1</reflink>, 674) = 6.22, <emph>p </emph>= .013, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.01, suggesting that children did indeed tend to use mental rotation when forced to solve the task internally (cf. Shepard & Metzler, 1971).</p> <p>The effect of map orientation on speed significantly varied with age, <emph>F</emph> (<reflink idref="bib3" id="ref15">3</reflink>, 671) = 4.60, <emph>p</emph> = .003, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.02, with follow‐up analyses revealing that 6‐ to 11‐year‐olds spent significantly longer on misaligned than aligned maps (all <emph>p</emph>s < .001, all <emph>η<subs>p</subs></emph><sups>2</sups>s > 0.02), but 4‐ and 5‐year‐olds did not, <emph>F</emph>(<reflink idref="bib1" id="ref16">1</reflink>, 665) = 0.08, <emph>p </emph>> .999, <emph>η<subs>p</subs></emph><sups>2</sups> < 0.01. Overall, children's search speed was relatively consistent across the two trials, <emph>F</emph>(<reflink idref="bib1" id="ref17">1</reflink>, 674) = 1.26, <emph>p </emph>= .262, <emph>η<subs>p</subs></emph><sups>2</sups> < 0.01.</p> <hd id="AN0154716022-24">Head tilting</hd> <p>Although they were unable to manually rotate the maps in phase 1, children often spontaneously tilted their heads. Phase 1 head tilting was modeled as a function of age, map orientation, and trial. Children tilted their heads more frequently with increasing age, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0011" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mrow><mspace width="0.166667em" /><mi>χ</mi></mrow><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib675" id="ref18">675</reflink>, _I_N_i_ = 97) = 26.53, <emph>p </emph>< .001, <emph>w </emph>= 0.52 (see Figure 2). A main effect of map orientation, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0012" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib675" id="ref19">675</reflink>, _I_N _i_= 97) = 24.25, <emph>p</emph> < .001, <emph>w </emph>= 0.50, was followed up to reveal that children also tilted their heads more frequently when looking at misaligned than aligned maps, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0013" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib675" id="ref20">675</reflink>, _I_N _i_= 97) = 23.26, <emph>p </emph>< .001, <emph>w </emph>= 0.49. Descriptive data showed that the majority of children who tilted their heads either to the left or right on the 90‐degree trials did so to the correct side (to the left on the 90‐left trials, and to the right on the 90‐right trials), although there was substantial variability (62.64% correct, 38.36% incorrect; see Note S2).</p> <p>There was no significant difference in the frequency of head tilting between trials, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0014" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib675" id="ref21">675</reflink>, _I_N_i_ = 97) = 0.10, <emph>p </emph>= .751, <emph>w</emph> = 0.03. Head tilting was associated with children spending more time on the relevant map, <emph>F</emph>(<reflink idref="bib1" id="ref22">1</reflink>, 674) = 96.63, <emph>p </emph>< .001, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.13 (see <emph>Speed</emph> sub‐section above for model details), but was not associated with search accuracy, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0015" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref23">674</reflink>, _I_N_i_ = 97) = 0.34, <emph>p </emph>= .562, <emph>w</emph> = 0.06 (see <emph>Accuracy</emph> sub‐section above for model details).</p> <hd id="AN0154716022-25">Phase 2</hd> <p></p> <hd id="AN0154716022-26">Rotation and accuracy</hd> <p>Manual rotation of the turntables was modeled as a function of age, map orientation, trial, and use of head tilting. A significant main effect of map orientation, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0016" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref24">674</reflink>, _I_N _i_= 97) = 139.38, <emph>p</emph> < .001, <emph>w </emph>= 1.20, was followed up to reveal that children used manual rotation more frequently on searches with misaligned maps than aligned maps, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0017" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref25">674</reflink>, _I_N _i_= 97) = 139.24, <emph>p </emph>< .001, <emph>w </emph>= 1.20. This effect interacted with age, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0018" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib671" id="ref26">671</reflink>, _I_N _i_= 97) = 38.51, <emph>p </emph>< .001, <emph>w</emph> = 0.63, and was significant for children aged 6 years and older (all <emph>p</emph>s < .001, all <emph>w</emph>s > 1.17). However, the effect did not cross the significance threshold for the 4‐ and 5‐year‐olds, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0019" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib665" id="ref27">665</reflink>, _I_N _i_= 25) = 6.02, <emph>p </emph>= .057, <emph>w </emph>= 0.49. In other words, only children aged 6 years and older showed significant evidence of cognitive offloading, such that they rotated the misaligned maps more frequently than the aligned maps (see Figure 3). Manual rotation on the misaligned trials was uncommon among the 4‐ and 5‐year‐old children (16% of misaligned trials), with many more children demonstrating this behavior in the 6‐ and 7‐year‐old group (on 51% of misaligned trials), the 8‐ and 9‐year‐old group (on 81% of misaligned trials), and the 10‐ and 11‐year‐old group (on 75% of misaligned trials).</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CDV/01jan22/cdev13664-fig-0003.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cdev13664-fig-0003.jpg" title="3 The average accuracy, use of any rotation (including 0‐degree trials and rotation to an incorrect orientation) and head tilting for each orientation in phase 2, split according to age group. The 4‐ and 5‐year‐old children demonstrated lower average levels of accuracy and rotation, but with increasing age, children's averages on both variables increased, with the oldest two age groups demonstrating similarly high rates of accuracy and rotation. The 10‐ and 11‐year‐old children also continued to engage in head tilting far more frequently than younger age groups." /> </p> <p></p> <p>Across all searches, the frequency of manual rotation decreased on the second trial in the phase compared to the first, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0020" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref28">674</reflink>, _I_N _i_= 97) = 7.26, <emph>p </emph>= .007, <emph>w</emph> = 0.27. However, in an exploratory analysis that only included instances of manual rotation to the correct orientation (i.e., excluding manual rotation on 0‐degree trials or to an incorrect orientation), the effect became non‐significant, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0021" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib481" id="ref29">481</reflink>, _I_N _i_= 97) = 0.06, <emph>p</emph> = .800, <emph>w</emph> = 0.02. In other words, the frequency of manual rotation on 0‐degree trials or to an incorrect orientation decreased across trials, whereas the frequency of manual rotation to the correct orientation remained consistent across trials.</p> <p>To investigate the effect of cognitive offloading on accuracy, phase 2 accuracy was modeled as a function of age, map orientation, head tilting, and manual rotation to the correct orientation. Aligned‐map trials were excluded from this analysis, given that offloading was not possible on these trials. When children used manual rotation to align a misaligned map with the search space, their accuracy was significantly greater, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0022" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib480" id="ref30">480</reflink>, _I_N _i_= 97) = 46.00, <emph>p </emph>< .001, <emph>w</emph> = 0.69, suggesting that such action did indeed offload internal cognitive demand. Whether the map was rotated at 90‐ or 180‐degrees had no effect on accuracy, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0023" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib480" id="ref31">480</reflink>, _I_N _i_= 97) = 0.89, <emph>p</emph> = .641, <emph>w </emph>= 0.10, as children were able to simply alter the orientation via manual rotation. Children were significantly more accurate in their searches with increasing age, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0024" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib480" id="ref32">480</reflink>, _I_N _i_= 97) = 22.99, <emph>p </emph>< .001, <emph>w </emph>= 0.49 (see Figure 3), and on the second trial of the phase compared to the first, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0025" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib480" id="ref33">480</reflink>, _I_N _i_= 97) = 6.48, <emph>p </emph>= .011, <emph>w</emph> = 0.26.</p> <hd id="AN0154716022-28">Cross‐phase performance comparisons</hd> <p>Accuracy across both phases was modeled as a function of age, map orientation, and phase. Children demonstrated greater accuracy during phase 2 compared to phase 1, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0026" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (1451, _I_N _i_= 97) = 168.81, <emph>p</emph> < .001, <emph>w </emph>= 1.32. However, a significant interaction between phase and age, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0027" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (1450, _I_N _i_= 97) = 37.87, <emph>p </emph>< .001, <emph>w </emph>= 0.62, revealed that this was only the case for the 6‐ to 11‐year‐olds (all <emph>p</emph>s < .001, all <emph>w</emph>s > 1.54). By contrast, the 4‐ and 5‐year‐old children performed comparably across phases, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0028" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (1448, _I_N _i_= 25) = 1.88, <emph>p </emph>= .681, <emph>w </emph>= 0.27, indicating that these children did not significantly benefit from having the option to offload (likely because of their limited use of manual rotation compared to older age groups; see Figure 3). An exploratory analysis revealed that children's performance on misaligned trials substantially increased across phases when they manually rotated the map in phase 2, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0029" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (1062, _I_N _i_= 97) = 155.78, <emph>p </emph>< .001, <emph>w </emph>= 1.27, but not when they did not manually rotate the map in phase 2, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0030" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (1062, _I_N _i_= 97) = 3.28, <emph>p </emph>= .140, <emph>w </emph>= 0.18. In other words, children did not simply improve with experience, but instead as a result of cognitive offloading.</p> <p>To explore whether children's unaided performance on misaligned trials in phase 1 predicted their use of cognitive offloading in phase 2, a "phase 1 accuracy on misaligned trials" variable was calculated for each child by summing their accuracy scores for each misaligned trial type (90‐left, 90‐right and 180) across the two trials in phase 1. We then modeled manual rotation (to the correct orientation) as a function of age, map orientation, and phase 1 accuracy on misaligned trials. This exploratory analysis indicated that children's accuracy on misaligned trials in phase 1 did not predict their likelihood of manually rotating the maps to the correct orientation in phase 2, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0031" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib482" id="ref34">482</reflink>, _I_N _i_= 97) = 0.04, <emph>p </emph>= .844, <emph>w </emph>= 0.02, but this effect did significantly interact with age, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0032" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib481" id="ref35">481</reflink>, _I_N _i_= 97) = 7.95, <emph>p </emph>= .005, <emph>w </emph>= 0.29. Follow‐up analyses revealed that 4‐ to 9‐year‐olds showed a positive association between phase 1 accuracy and cognitive offloading in phase 2 (<emph>w</emph> values ranged from 0.14 to 0.24), whereas 10‐ and 11‐year‐olds showed a negative association between phase 1 accuracy and cognitive offloading in phase 2 (<emph>w </emph>= −0.38). The interaction, therefore, indicates that, with increasing age, there was a significant increase in children's likelihood of manually rotating the turntable as a function of failing the misaligned trials in phase 1.</p> <p>To explore whether phase 1 accuracy predicted phase 2 accuracy on misaligned trials, the latter was modeled as a function of age, map orientation, manual rotation (to the correct orientation), and phase 1 accuracy on misaligned trials. This exploratory analysis showed that children's search accuracy in phase 1 did not significantly predict their search accuracy in phase 2, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0033" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib481" id="ref36">481</reflink>, _I_N _i_= 97) = 3.71, <emph>p </emph>= .054, <emph>w</emph> = 0.20, but critically, the nature of this effect significantly varied according to whether children manually rotated the map to the correct orientation, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0034" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib480" id="ref37">480</reflink>, _I_N_i_ = 97) = 4.40, <emph>p </emph>= .036, <emph>w </emph>= 0.21. When children <emph>did not</emph> rotate the map to the correct orientation, their phase 1 performance was highly predictive of phase 2 performance on identical trials, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0035" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib480" id="ref38">480</reflink>, _I_N _i_= 97) = 7.02, <emph>p </emph>= .016, <emph>w </emph>= 0.27, likely because they were using the same strategy across phases (whether that was performing the same mental rotation operation, treating all maps as aligned, or simply guessing the answer). When children <emph>did</emph> rotate the turntable to the correct orientation, by contrast, their phase 1 performance was not a reliable indicator of their phase 2 performance on identical trials, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0036" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib480" id="ref39">480</reflink>, _I_N _i_= 97) = 0.25, <emph>p </emph>> .999, <emph>w</emph> = 0.05. In other words, children's decisions to offload mental rotation demand eliminated the effect of individual differences in mental rotation ability on search performance.</p> <hd id="AN0154716022-29">Speed</hd> <p>Phase 2 speed on misaligned trials was modeled as a function of age, map orientation, trial, use of head tilting, and manual rotation to the correct orientation, again excluding 0‐degree trials where cognitive offloading was not possible. No significant age effects were detected, but there was a significant main effect of map orientation, <emph>F</emph>(<reflink idref="bib2" id="ref40">2</reflink>, 480) = 3.66, <emph>p </emph>= .026, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.02. Follow‐up analyses revealed that children's search times in phase 2 again significantly increased from 90‐degree trials to 180‐degree trials, <emph>F</emph>(<reflink idref="bib1" id="ref41">1</reflink>, 480) = 5.02, <emph>p </emph>= .026, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.01. Contrary to hypotheses, manually rotating the turntable to the correct orientation did not result in faster performance, <emph>F</emph>(<reflink idref="bib1" id="ref42">1</reflink>, 480) = 0.29, <emph>p </emph>= .592, <emph>η<subs>p</subs></emph><sups>2</sups> < 0.01.</p> <hd id="AN0154716022-30">Cross‐phase performance comparisons</hd> <p>Speed across both phases was modelled as a function of age, map orientation, and phase. Children completed phase 2 in a significantly shorter amount of time than phase 1, <emph>F</emph>(<reflink idref="bib1" id="ref43">1</reflink>, 1451) = 55.78, <emph>p </emph>< .001, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.04 (see Figure 4). However, this effect varied by age, <emph>F</emph>(<reflink idref="bib1" id="ref44">1</reflink>, 1450) = 51.01, <emph>p </emph>< .001, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.03, revealing that, unlike the older children, the speed at which 4‐ and 5‐year‐olds, <emph>F</emph>(<reflink idref="bib1" id="ref45">1</reflink>, 1488) = 1.49, <emph>p </emph>= .887, <emph>η<subs>p</subs></emph><sups>2</sups> < 0.01, and 6‐ and 7‐year‐olds, <emph>F</emph>(<reflink idref="bib1" id="ref46">1</reflink>, 1448) = 2.98, <emph>p </emph>= .339, <emph>η<subs>p</subs></emph><sups>2</sups> < 0.01, completed the task was unaffected by the availability of manual rotation. An exploratory analysis revealed that, on misaligned phase 2 trials, children were significantly faster (compared to phase 1) when they rotated the map to the correct orientation, <emph>F</emph>(1062, _I_N _i_= 97) = 142.00, <emph>p </emph>< .001, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.12, but they were not significantly faster (compared to phase 1) when they did not rotate the map to the correct orientation, <emph>F</emph>(1062, _I_N _i_= 97) = 0.20, <emph>p </emph>> .999, <emph>η<subs>p</subs></emph><sups>2</sups> < 0.01. Thus, when operationalized in this manner, cognitive offloading did indeed decrease search time.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CDV/01jan22/cdev13664-fig-0004.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cdev13664-fig-0004.jpg" title="4 The average time taken by child per trial (in seconds) in each phase. In phase 1, children took longer with increasing age, and when the maps were presented at higher degrees of angular disparity, in line with the classic finding that mental rotation time increases as the degree of rotation increases. The age effect was no longer evident in phase 2. Aside from the 4‐ and 5‐year‐old children, all other age groups were faster in phase 2 than in phase 1." /> </p> <p></p> <hd id="AN0154716022-32">Head tilting</hd> <p>Even when given the opportunity to manually rotate the maps in phase 2, some children still adopted the head tilting strategy. Phase 2 head tilting was modeled as a function of age, map orientation, trial, and manual rotation. As in phase 1, older children continued to tilt their heads more frequently than younger children, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0037" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref47">674</reflink>, _I_N _i_= 97) = 12.87, <emph>p </emph>< .001, <emph>w </emph>= 0.36 (see Figure 3), even though head tilting again had no effect on accuracy, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0038" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib480" id="ref48">480</reflink>, _I_N_i_ = 97) = 0.01, <emph>p </emph>= .941, <emph>w </emph>= 0.01 (see <emph>Accuracy</emph> sub‐section above for model details), and was associated with a significantly longer amount of time at the relevant map, <emph>F</emph>(<reflink idref="bib1" id="ref49">1</reflink>, 480) = 33.12, <emph>p </emph>< .001, <emph>η<subs>p</subs></emph><sups>2</sups> = 0.06 (see <emph>Speed</emph> sub‐section above for model details).</p> <p>Children's use of head tilting was not significantly predicted by their use of manual rotation, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0039" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref50">674</reflink>, _I_N_i_ = 97) = 0.23, <emph>p </emph>= .628, <emph>w </emph>= 0.05. Head tilting also no longer varied across aligned and misaligned trials as it did in phase 1, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0040" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref51">674</reflink>, _I_N _i_= 97) = 3.93, <emph>p </emph>= .269, <emph>w</emph> = 0.20, likely because children had the opportunity to manually alter the orientation of the misaligned maps in phase 2. Children tilted their heads less frequently on the second trial of the phase than the first, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0041" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (<reflink idref="bib674" id="ref52">674</reflink>, _I_N _i_= 97) = 5.80, <emph>p </emph>= .016, <emph>w </emph>= 0.24, and cross‐phase comparisons revealed that head tilting was less common in phase 2 than in phase 1, <ephtml> <math altimg="urn:x-wiley:00093920:media:cdev13664:cdev13664-math-0042" xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>χ</mi><mn>2</mn></msup></math> </ephtml> (1451, _I_N_i_ = 97) = 45.79, <emph>p </emph>< .001, <emph>w </emph>= 0.69.</p> <hd id="AN0154716022-33">Metacognitive beliefs</hd> <p>We used binomial tests to investigate whether the proportion of children in each age group answering the prospective and retrospective metacognitive questions correctly (i.e., identifying that finding hidden stickers would be/was easier using the aligned map than the misaligned map) differed from chance levels. These tests revealed that 4‐ and 5‐year‐old children performed above chance on the prospective metacognitive task, <emph>p </emph>= .043, but not on the retrospective task, <emph>p </emph>= .230. Children aged 6 and older, however, consistently performed above chance level on both tasks, (all <emph>p</emph>s < .001; see Table S2). The proportion of children answering correctly appeared to increase with age for both the prospective task (4‐ and 5‐year‐olds = 72.00%, 6‐ and 7‐year‐olds = 91.67%, 8‐ and 9‐year‐olds = 92.86%, 10‐ and 11‐year‐olds = 95.24%) and the retrospective task (4‐ and 5‐year‐olds = 64.00%, 6‐ and 7‐year‐olds = 87.50%, 8‐ and 9‐year‐olds = 88.88%, 10‐ and 11‐year‐olds = 95.24%). Point‐biserial correlations between age (continuous) and responses to the metacognitive questions confirmed that this age effect was significant for both the prospective task, <emph>r</emph>(<reflink idref="bib95" id="ref53">95</reflink>) = .26, <emph>p </emph>= .010, and the retrospective task, <emph>r</emph>(<reflink idref="bib95" id="ref54">95</reflink>) = .28, <emph>p</emph> = .005.</p> <p>For each child, we calculated a difference score between their use of manual rotation that (i) did classify as cognitive offloading (scored from 0 to 6 across the phase 2 trials, excluding the 0‐degree trials) and (ii) did not classify as cognitive offloading (i.e., rotation to the incorrect orientation or on 0‐degree trials, scored from 0 to 8 across the phase 2 trials). Age‐partialled correlations were then used to explore the association between children's metacognitive beliefs and their use of cognitive offloading. These analyses revealed non‐significant correlations for both prospective metacognitive beliefs, <emph>r</emph>(<reflink idref="bib94" id="ref55">94</reflink>) = −.03, <emph>p </emph>= .755, and retrospective metacognitive beliefs, <emph>r</emph>(<reflink idref="bib94" id="ref56">94</reflink>) = .17, <emph>p </emph>= .095.</p> <hd id="AN0154716022-34">DISCUSSION</hd> <p>Previous studies have failed to demonstrate that children can improve on their unaided cognitive performance when given the opportunity to infer and deploy a novel cognitive offloading strategy. Here, we have shown that with increasing age, children became more likely to devise and implement an offloading strategy to alleviate mental rotation demand while reading maps, and that the availability of this strategy led to improvements on unaided performance in children aged 6 years and older. By contrast, the 4‐ and 5‐year‐olds showed little evidence of offloading, in line with earlier findings (Armitage et al., 2020; Bremner & Andreasen, 1998; Vosmik & Presson, 2004), and they performed comparably across phases. Our results also show, for the first time, that offloading can eliminate pre‐existing individual differences in children's cognitive performance, and that older children are more likely than younger children to offload as a function of unaided performance failures.</p> <p>Across ages, children performed considerably better on aligned trials than misaligned trials in phase 1. When the opportunity for cognitive offloading became available in phase 2, therefore, many of the older children chose to avoid the internal demand of mental rotation by manually rotating misaligned maps to the match the search space. This behavior led to substantial improvements on unaided performance, unlike in Berry et al.'s (2019) study where children showed no working memory performance gains after being given the opportunity to devise an offloading strategy. The discrepancy is likely explained by the fact that children's unaided performance in Berry et al.'s (2019) study was very similar across both easy and hard conditions, and therefore cognitive offloading offered little benefit in the hard condition. By contrast, our misaligned trials were clearly harder than the aligned trials, such that children had much to gain from transforming a misaligned map into an aligned map.</p> <p>Our results largely accord with those of Bulley et al. (2020), who found that children's memory performance benefitted following explicit demonstration of an offloading behavior by an experimenter, with one notable difference. Whereas Bulley et al. (2020) found that even 4‐ and 5‐year‐olds' performance greatly improved from the unaided to the aided phase, here we found significant performance benefits only for children aged 6 years and older. In phase 1 of our task, 4‐ and 5‐year‐olds demonstrated lower levels of search accuracy than older children, especially in the misaligned trials, such that they had even more to gain from devising an offloading strategy than the older children. When the opportunity for manual rotation became available in phase 2, however, these 4‐ and 5‐year‐olds did not rotate the misaligned maps significantly more often than aligned maps, and they failed to perform significantly better than in phase 1.</p> <p>One likely reason that the youngest children in our sample rarely devised their own offloading strategies is that they were not yet inclined to perform the analogous internal cognitive operations. In phase 1, the 4‐ and 5‐year‐olds were seemingly inclined to treat all maps as if they were aligned with the search space, regardless of the orientation of each specific map. Thus, despite receiving numerous reminders about the location of the landmark, it appears that children of this age typically failed to infer that the maps were often misaligned, and so it is unlikely that they recognized the utility of either mental or manual rotation in this context. By contrast, 6‐ and 7‐year‐old children showed strong evidence of both mental and manual rotation of misaligned maps. This developmental pattern may suggest that children become inclined to offload cognitive operations around the same time as or not long after they become inclined to perform these operations internally. Indeed, among the 4‐ to 9‐year‐olds, children who were better able to perform mental rotation operations in phase 1 were also more likely to offload these operations in phase 2.</p> <p>Although the 4‐ and 5‐year‐olds in our sample showed little evidence of cognitive offloading, a significant proportion of them were able to appropriately differentiate between easier (aligned) and harder (misaligned) conditions before the task. One possible explanation is that these children simply used a "same" or "different" judgment and could recognize only that the aligned map correctly matched the layout of the room, without understanding why the misaligned map did not, or that one could compensate for such misalignment via mental or manual rotation. Such a developmental delay between acquiring basic metacognitive awareness about task difficulty and becoming able to transform such awareness into metacognitive control over behavior has been widely documented (Dufresne & Kobasigawa, 1989; Dunlosky & Metcalfe, 2008; Flavell, 2000; Nelson & Narens, 1990; Redshaw et al., 2018; Schraw & Moshman, 1995).</p> <p>Indirect evidence for the continued development of metacognitive control throughout middle childhood comes from our novel finding that, with increasing age, children were increasingly more likely to offload in phase 2 after making errors in phase 1. Indeed, only the 10‐ and 11‐year‐olds showed a negative association between phase 1 performance and phase 2 offloading, as adults do (Gilbert, 2015), rather than the positive association shown by the 4‐ to 9‐year‐olds. This finding could be explained by the fact that older children were more likely to tailor their offloading decisions to their awareness about their unaided cognitive abilities, in line with a metacognitive account of cognitive offloading (Gilbert, 2015; Risko & Gilbert, 2016). To test for this possibility directly, future studies could examine associations between children's <emph>predictions</emph> about their unaided performance and their use of cognitive offloading (Gilbert, 2015).</p> <p>Whatever the underpinnings of children's offloading decisions, our results clearly demonstrate how such decisions can fundamentally change the cognitive skill set required to solve a task. In particular, we found that cognitive offloading not only substantially improved search performance from phase 1 to phase 2, but also eliminated the association between phase 1 accuracy and phase 2 accuracy, such that it was not possible to accurately predict how children who rotated the turntable would perform based on their mental rotation ability alone. Our results therefore demonstrate the power of cognitive offloading both in allowing children to achieve cognitive feats that may have otherwise been out of their reach, and in leveling out natural individual differences in children's cognitive aptitude.</p> <p>One interesting secondary finding was that, across both phases, children frequently tilted their heads while looking at the maps. They did so despite the fact that head‐tilting had no significant effect on accuracy in either phase, and was also associated with a decrease in overall performance efficiency (i.e., an increase in search time). One possibility is that children's head tilting was an embodied simulation of the mental rotation process (Hostetter & Alibali, 2008, 2018). Spontaneous rotational gestures have been documented in other mental rotation tasks, with findings typically showing that adults gesture more frequently under high than low cognitive load (Chu & Kita, 2011), even when doing so does not benefit performance (Chandrasekharan et al., 2010). However, given that children appeared to find the task easier with increasing age, older children should have tilted their heads less frequently than younger children if they were doing so as a function of cognitive load (setting aside the 4‐ and 5‐year‐olds, who may not have inferred that some maps were misaligned). Instead, older children consistently tilted their heads more frequently, with this age effect persisting across both phases.</p> <p>A plausible alternative, therefore, is that children were attempting to offload mental rotation demand through head tilting, just as they were attempting to (and succeeding at) offloading mental rotation demand using the turntables. Accordingly, children typically tilted their heads in the same direction as the 90‐degree maps were rotated, rather than the opposite direction. As children tilted their heads, however, the subjective angular difference between the map and search space remained the same, such that they may not have offloaded mental rotation demand at all. Children may have believed this behavior would be useful based on past experience in situations where head tilting does indeed serve to offload mental rotation demand, such as in contexts where a comparison between stimuli is not required (e.g., while reading rotated text; Dunn & Risko, 2016; Risko et al., 2014). A similar phenomenon has been observed in adults, whereby they will engage in external manipulations of task parameters not only when doing so benefits task performance, but also in instances where they incorrectly <emph>believe</emph> that doing so will benefit task performance (Dunn & Risko, 2016; Gilbert, 2015; Risko & Gilbert, 2016; Weis & Wiese, 2018). This may suggest that the use of external manipulations is not particularly well‐calibrated with genuine offloading of cognitive demand in either children or adults. Instead, it may be that as they age, children increasingly acquire both accurate and erroneous metacognitive beliefs about the usefulness of external manipulations aimed at offloading cognitive demand, with these beliefs persisting into adulthood.</p> <p>One important caveat of recent cognitive offloading research, including the current study, is that the child and adult participants have all been sampled from WEIRD populations (see Henrich et al., 2010) with abundant access to external thinking tools like smart phones and other computers. Although cognitive offloading could easily be mistaken for an exclusively modern phenomenon, it in fact has an ancient history. Examples can be found across cultures, for instance in the mnemonic devices used by pre‐colonial Incans (Ascher & Ascher, 1981), Polynesians (Best, 1921), and Luba in the Congo (Roberts & Roberts, 1996), a 9000‐year‐old "town plan" map uncovered at Çatalhöyük in Turkey (Schmitt et al., 2014), and the compasses independently invented in ancient China and Mesoamerica to assist with spatial orienting (Carlson, 1975). We therefore implore the importance of cross‐cultural research, especially in populations where digital thinking tools are less ingrained in everyday life, to obtain a more complete understanding of the development of and factors affecting cognitive offloading across the lifespan (see Nielsen et al., 2017).</p> <p>Another fruitful avenue for future research will be to explore whether it is possible to speed up and enhance the calibration of children's cognitive offloading strategies. Our results not only showed that many children tilted their heads without improving their performance, but also that some children rotated the maps away from the angle of alignment. Ideally, children should be appropriately <emph>selective</emph> in the use of external manipulations and thinking tools, in that they should differentiate between circumstances in which an external manipulation simplifies a cognitive task, or extends upon what can be achieved unaided, and circumstances in which it does not (Armitage et al., 2020). There has been some success in training metacognitive abilities, at least in adults, with evidence that this improvement can translate across cognitive domains (Carpenter et al., 2019). Improving the accuracy of children's metacognitive judgments about behaviors aimed at offloading cognition may prove to be critical in reducing or preventing any possible deleterious consequences of relying too heavily on these behaviors (Carr, 2008; Hejtmánek et al., 2018; Lu et al., 2020; Ruginski et al., 2019).</p> <hd id="AN0154716022-35">ETHICAL APPROVAL STATEMENT</hd> <p>Ethical approval was obtained through the University of Queensland's Faculty of Health and Behavioural Sciences Ethics Committee (Clearance ID: 2019000267).</p> <hd id="AN0154716022-36">ACKNOWLEDGMENTS</hd> <p>This work was supported by an Australian Research Council (ARC) Discovery Early Career Researcher Award granted to Jonathan Redshaw (DE210100005).</p> <hd id="AN0154716022-37">CONFLICT OF INTEREST</hd> <p>The authors hereby declare no conflicts of interest.</p> <hd id="AN0154716022-38">DATA AVAILABILITY STATEMENT</hd> <p>All datasets are available on the Open Science Framework at https://osf.io/485ka/.</p> <p>GRAPH: Supplementary Material</p> <ref id="AN0154716022-39"> <title> REFERENCES </title> <blist> <bibl id="bib1" idref="ref11" type="bt">1</bibl> <bibtext> Armitage, K. L., Bulley, A., & Redshaw, J. (2020). Developmental origins of cognitive offloading. Proceedings of the Royal Society B: Biological Sciences, 287 (1928), 20192927. 10.1098/rspb.2019.2927</bibtext> </blist> <blist> <bibl id="bib2" idref="ref40" type="bt">2</bibl> <bibtext> Ascher, M., & Ascher, R. (1981). Code of the Quipu: A study in media, mathematics, and culture. University of Michigan Press.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref12" type="bt">3</bibl> <bibtext> Balcomb, F. K., & Gerken, L. A. (2008). Three‐year‐old children can access their own memory to guide responses on a visual matching task. Developmental Science, 11 (5), 750 – 760. 10.1111/j.1467‐7687.2008.00725.x</bibtext> </blist> <blist> <bibl id="bib4" type="bt">4</bibl> <bibtext> Berry, E. D. J., Allen, R. J., Mon‐Williams, M., & Waterman, A. H. (2019). Cognitive offloading: Structuring the environment to improve children's working memory task performance. Cognitive Science, 43 (8), e12770. 10.1111/cogs.12770</bibtext> </blist> <blist> <bibl id="bib5" type="bt">5</bibl> <bibtext> Best, E. (1921). Polynesian mnemonics: Notes on the use of quipus in Polynesia in former times; also some account of the introduction of the art of writing. New Zealand Journal of Science and Technology, 4, 67 – 74.</bibtext> </blist> <blist> <bibl id="bib6" type="bt">6</bibl> <bibtext> Blades, M., & Cooke, Z. (1994). Young children's ability to understand a model as a spatial representation. Journal of Genetic Psychology, 155 (2), 201 – 218. 10.1080/00221325.1994.9914772</bibtext> </blist> <blist> <bibl id="bib7" type="bt">7</bibl> <bibtext> Blades, M., & Spencer, C. (1987a). The use of maps by 4–6‐year‐old children in a large‐scale maze. British Journal of Developmental Psychology, 5 (1), 19 – 24. 10.1111/j.2044‐835X.1987.tb01037.x</bibtext> </blist> <blist> <bibl id="bib8" type="bt">8</bibl> <bibtext> Blades, M., & Spencer, C. (1987b). Young children's strategies when using maps with landmarks. Journal of Environmental Psychology, 7 (3), 201 – 217. 10.1016/S0272‐4944(87)80030‐0</bibtext> </blist> <blist> <bibl id="bib9" type="bt">9</bibl> <bibtext> Blades, M., & Spencer, C. (1990). The development of 3‐year‐olds to 6‐year‐olds map using ability—The relative importance of landmarks and map alignment. The Journal of Genetic Psychology, 151, 181 – 194. 10.1080/00221325.1990.9914653</bibtext> </blist> <blist> <bibtext> Bluestein, N., & Acredolo, L. (1979). Developmental changes in map‐reading skills. Child Development, 50 (3), 691 – 697. 10.2307/1128934</bibtext> </blist> <blist> <bibtext> Bocanegra, B. R., Poletiek, F. H., Ftitache, B., & Clark, A. (2019). Intelligent problem‐solvers externalize cognitive operations. Nature Human Behaviour, 3 (2), 126 – 142. 10.1038/s41562‐018‐0509‐y</bibtext> </blist> <blist> <bibtext> Bremner, J. G., & Andreasen, G. (1998). Young children's ability to use maps and models to find ways in novel spaces. British Journal of Developmental Psychology, 16 (2), 197 – 218. 10.1111/j.2044‐835X.1998.tb00919.x</bibtext> </blist> <blist> <bibtext> Bulley, A., McCarthy, T., Gilbert, S. J., Suddendorf, T., & Redshaw, J. (2020). Children devise and selectively use tools to offload cognition. Current Biology, 30 (17), 3457 – 3464. 10.1016/j.cub.2020.06.035</bibtext> </blist> <blist> <bibtext> Carlson, J. B. (1975). Lodestone compass: Chinese or Olmec primacy? Multidisciplinary analysis of an Olmec hematite artifact from San Lorenzo, Veracruz, Mexico. Science, 189 (4205), 753 – 760. 10.1126/science.189.4205.753</bibtext> </blist> <blist> <bibtext> Carpenter, J., Sherman, M. T., Kievit, R. A., Seth, A. K., Lau, H., & Fleming, S. M. (2019). Domain‐general enhancements of metacognitive ability through adaptive training. Journal of Experimental Psychology: General, 148 (1), 51 – 64. 10.1037/xge0000505</bibtext> </blist> <blist> <bibtext> Carr, N. (2008). Is google making us stupid? Yearbook of the National Society for the Study of Education, 107 (2), 89 – 94. 10.1111/j.1744‐7984.2008.00172.x</bibtext> </blist> <blist> <bibtext> Chandrasekharan, S., Athreya, D., & Srinivasan, N. (2010). Investigating prospective action control: When and how actions supporting cognition are generated. International Journal of Mind, Brain & Cognition, 1, 93 – 114.</bibtext> </blist> <blist> <bibtext> Chu, M., & Kita, S. (2011). The nature of gestures' beneficial role in spatial problem solving. Journal of Experimental Psychology: General, 140 (1), 102 – 116. 10.1037/a0021790</bibtext> </blist> <blist> <bibtext> Clark, A. (2008). Supersizing the mind. Oxford University Press.</bibtext> </blist> <blist> <bibtext> Clark, A., & Chalmers, D. (1998). The extended mind. Analysis, 58 (1), 7 – 19. 10.1111/1467‐8284.00096</bibtext> </blist> <blist> <bibtext> Cohen, J. (1992). A power primer. Psychological Bulletin, 112 (1), 155 – 159. 10.1037/0033‐2909.112.1.155</bibtext> </blist> <blist> <bibtext> DeLoache, J. S. (1987). Rapid change in the symbolic functioning of very young children. Science, 238 (4833), 1556 – 1557. 10.1126/science.2446392</bibtext> </blist> <blist> <bibtext> DeLoache, J. S. (1989). Young children's understanding of the correspondence between a scale model and a larger space. Cognitive Development, 4 (2), 121 – 139. 10.1016/0885‐2014(89)90012‐9</bibtext> </blist> <blist> <bibtext> Dufresne, A., & Kobasigawa, A. (1989). Children's spontaneous allocation of study time: Differential and sufficient aspects. Journal of Experimental Child Psychology, 47 (2), 274 – 296. 10.1016/0022‐0965(89)90033‐7</bibtext> </blist> <blist> <bibtext> Dunlosky, J., & Metcalfe, J. (2008). Metacognition. Sage Publications.</bibtext> </blist> <blist> <bibtext> Dunn, T. L., & Risko, E. F. (2016). Toward a metacognitive account of cognitive offloading. Cognitive Science, 40 (5), 1080 – 1127. 10.1111/cogs.12273</bibtext> </blist> <blist> <bibtext> Ehrlich, S. B., Levine, S. C., & Goldin‐Meadow, S. (2006). The importance of gesture in children's spatial reasoning. Developmental Psychology, 42 (6), 1259 – 1268. 10.1037/0012‐1649.42.6.1259</bibtext> </blist> <blist> <bibtext> Estes, D. (1998). Young children's awareness of their mental activity: The case of mental rotation. Child Development, 69 (5), 1345 – 1360. 10.2307/1132270</bibtext> </blist> <blist> <bibtext> Flavell, J. H. (2000). Development of children's knowledge about the mental world. International Journal of Behavioral Development, 24 (1), 15 – 23. 10.1080/016502500383421</bibtext> </blist> <blist> <bibtext> Frick, A., Daum, M. M., Walser, S., & Mast, F. W. (2009). Motor processes in children's mental rotation. Journal of Cognition and Development, 10 (1–2), 18 – 40. 10.1080/15248370902966719</bibtext> </blist> <blist> <bibtext> Frick, A., Hansen, M. A., & Newcombe, N. S. (2013). Development of mental rotation in 3‐ to 5‐year‐old children. Cognitive Development, 28 (4), 386 – 399. 10.1016/j.cogdev.2013.06.002</bibtext> </blist> <blist> <bibtext> Gilbert, S. J. (2015). Strategic use of reminders: Influence of both domain‐general and task‐specific metacognitive confidence, independent of objective memory ability. Consciousness and Cognition, 33, 245 – 260. 10.1016/j.concog.2015.01.006</bibtext> </blist> <blist> <bibtext> Hejtmánek, L., Oravcová, I., Motýl, J., Horáček, J., & Fajnerová, I. (2018). Spatial knowledge impairment after GPS guided navigation: Eye‐tracking study in a virtual town. International Journal of Human‐Computer Studies, 116, 15 – 24. 10.1016/j.ijhcs.2018.04.006</bibtext> </blist> <blist> <bibtext> Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioural Brain Sciences, 33 (2–3), 61 – 83. 10.1017/S0140525X0999152X</bibtext> </blist> <blist> <bibtext> Hostetter, A. B., & Alibali, M. W. (2008). Visible embodiment: Gestures as simulated action. Psychonomic Bulletin & Review, 15 (3), 495 – 514. 10.3758/pbr.15.3.495</bibtext> </blist> <blist> <bibtext> Hostetter, A. B., & Alibali, M. W. (2018). Gesture as simulated action: Revisiting the framework. Psychonomic Bulletin & Review, 26 (3), 721 – 752. 10.3758/s13423‐018‐1548‐0</bibtext> </blist> <blist> <bibtext> Hu, X., Luo, L., & Fleming, S. M. (2019). A role for metamemory in cognitive offloading. Cognition, 193, 104012. 10.1016/j.cognition.2019.104012</bibtext> </blist> <blist> <bibtext> Jirout, J. J., & Newcombe, N. S. (2014). Mazes and maps: Can young children find their way? Mind, Brain, and Education, 8 (2), 89 – 96. 10.1111/mbe.12048</bibtext> </blist> <blist> <bibtext> Levine, M., Marchon, I., & Hanley, G. L. (1984). The placement and misplacement of you‐are‐here maps. Environment and Behavior, 16 (2), 139 – 157. 10.1177/0013916584162001</bibtext> </blist> <blist> <bibtext> Liben, L. S., & Downs, R. M. (1993). Understanding person‐space‐map relations: Cartographic and developmental perspectives. Developmental Psychology, 29 (4), 739 – 752. 10.1037/0012‐1649.29.4.739</bibtext> </blist> <blist> <bibtext> Liben, L. S., Myers, L. J., Christensen, A. E., & Bower, C. A. (2013). Environmental‐scale map use in middle childhood: Links to spatial skills, strategies, and gender. Child Development, 84 (6), 2047 – 2063. 10.1111/cdev.12090</bibtext> </blist> <blist> <bibtext> Lockl, K., & Schneider, W. (2004). The effects of incentives and instructions on children's allocation of study time. European Journal of Developmental Psychology, 1 (2), 153 – 169. 10.1080/17405620444000085</bibtext> </blist> <blist> <bibtext> Lu, X., Kelly, M. O., & Risko, E. F. (2020). Offloading information to an external store increases false recall. Cognition, 205, 104428. 10.1016/j.cognition.2020.104428</bibtext> </blist> <blist> <bibtext> Moè, A. (2009). Are males always better than females in mental rotation? Exploring a gender belief explanation. Learning and Individual Differences, 19 (1), 21 – 27. 10.1016/j.lindif.2008.02.002</bibtext> </blist> <blist> <bibtext> Muff, S., Held, L., & Keller, L. F. (2016). Marginal or conditional regression models for correlated non‐normal data? Methods in Ecology and Evolution, 7 (12), 1514 – 1524. 10.1111/2041‐210X.12623</bibtext> </blist> <blist> <bibtext> Neldner, K., Collier‐Baker, E., & Nielsen, M. (2015). Chimpanzees (Pan troglodytes) and human children (Homo sapiens) know when they are ignorant about the location of food. Animal Cognition, 18 (3), 683 – 699. 10.1007/s10071‐015‐0836‐6</bibtext> </blist> <blist> <bibtext> Nelson, T. O., & Narens, L. (1990). Metamemory: A theoretical framework and new findings. Psychology of Learning and Motivation, 26, 125 – 173. 10.1016/S0079‐7421(08)60053‐5</bibtext> </blist> <blist> <bibtext> Nielsen, M., Haun, D., Kärtner, J., & Legare, C. H. (2017). The persistent sampling bias in developmental psychology: A call to action. Journal of Experimental Child Psychology, 162, 31 – 38. 10.1016/j.jecp.2017.04.017</bibtext> </blist> <blist> <bibtext> Ping, R., Ratliff, K., Hickey, E., & Levine, S. (2011). Using manual rotation and gesture to improve mental rotation in preschoolers. Proceedings of the Annual Meeting of the Cognitive Science Society (p. 33).</bibtext> </blist> <blist> <bibtext> Presson, C. C. (1982). The development of map‐reading skills. Child Development, 53 (1), 196 – 199. 10.2307/1129653</bibtext> </blist> <blist> <bibtext> Redshaw, J., Vandersee, J., Bulley, A., & Gilbert, S. J. (2018). Development of children's use of external reminders for hard‐to‐remember intentions. Child Development, 89 (6), 2099 – 2108. 10.1111/cdev.13040</bibtext> </blist> <blist> <bibtext> Risko, E. F., & Gilbert, S. J. (2016). Cognitive offloading. Trends in Cognitive Sciences, 20 (9), 676 – 688. 10.1016/j.tics.2016.07.002</bibtext> </blist> <blist> <bibtext> Risko, E. F., Medimorec, S., Chisholm, J., & Kingstone, A. (2014). Rotating with rotated text: A natural behavior approach to investigating cognitive offloading. Cognitive Science, 38 (3), 537 – 564. 10.1111/cogs.12087</bibtext> </blist> <blist> <bibtext> Roberts, M. N., & Roberts, A. F. (1996). Memory: Luba art and the making of history. African Arts, 29 (1), 23 – 103. 10.2307/3337444</bibtext> </blist> <blist> <bibtext> Ruginski, I. T., Creem‐Regehr, S. H., Stefanucci, J. K., & Cashdan, E. (2019). GPS use negatively affects environmental learning through spatial transformation abilities. Journal of Environmental Psychology, 64, 12 – 20. 10.1016/j.jenvp.2019.05.001</bibtext> </blist> <blist> <bibtext> Schielzeth, H., Dingemanse, N. J., Nakagawa, S., Westneat, D. F., Allegue, H., Teplitsky, C., Réale, D., Dochtermann, N. A., Garamszegi, L. Z., & Araya‐Ajoy, Y. G. (2020). Robustness of linear mixed‐effects models to violations of distributional assumptions. Methods in Ecology and Evolution, 11 (9), 1141 – 1152. 10.1111/2041‐210X.13434</bibtext> </blist> <blist> <bibtext> Schmitt, A. K., Danišík, M., Aydar, E., Şen, E., Ulusoy, İ., & Lovera, O. M. (2014). Identifying the volcanic eruption depicted in a neolithic painting at Çatalhöyük, Central Anatolia. Turkey. Plos One, 9 (1), e84711. 10.1371/journal.pone.0084711</bibtext> </blist> <blist> <bibtext> Schraw, G., & Moshman, D. (1995). Metacognitive theories. Educational Psychology Review, 7 (4), 351 – 371. 10.1007/BF02212307</bibtext> </blist> <blist> <bibtext> Shepard, R. N., & Metzler, J. (1971). Mental rotation of three‐dimensional objects. Science, 171 (3972), 701 – 703. 10.1126/science.171.3972.701</bibtext> </blist> <blist> <bibtext> Vosmik, J. R., & Presson, C. C. (2004). Children's response to natural map misalignment during wayfinding. Journal of Cognition and Development, 5 (3), 317 – 336. 10.1207/s15327647jcd0503_2</bibtext> </blist> <blist> <bibtext> Weis, P. P., & Wiese, E. (2018). Using tools to help us think: Actual but also believed reliability modulates cognitive offloading. Human Factors: The Journal of Human Factors and Ergonomics Society, 61 (2), 243 – 254. 10.1177/0018720818797553</bibtext> </blist> <blist> <bibtext> Wiedenbauer, G., & Jansen‐Osmann, P. (2008). Manual training of mental rotation in children. Learning and Instruction, 18 (1), 30 – 41. 10.1016/j.learninstruc.2006.09.009</bibtext> </blist> <blist> <bibtext> Winkler‐Rhoades, N., Carey, S. C., & Spelke, E. S. (2013). Two‐year‐old children interpret abstract, purely geometric maps. Developmental Science, 16 (3), 365 – 376. 10.1111/desc.12038</bibtext> </blist> <blist> <bibtext> Yuan, L., Uttal, D., & Gentner, D. (2017). Analogical processes in children's understanding of spatial representations. Developmental Psychology, 53 (6), 1098 – 1114. 10.1037/dev0000302</bibtext> </blist> </ref> <aug> <p>By Kristy L. Armitage and Jonathan Redshaw</p> <p>Reported by Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib674" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib671" firstref="ref5"></nolink> <nolink nlid="nl3" bibid="bib675" firstref="ref8"></nolink> <nolink nlid="nl4" bibid="bib672" firstref="ref9"></nolink> <nolink nlid="nl5" bibid="bib666" firstref="ref10"></nolink> <nolink nlid="nl6" bibid="bib665" firstref="ref27"></nolink> <nolink nlid="nl7" bibid="bib481" firstref="ref29"></nolink> <nolink nlid="nl8" bibid="bib480" firstref="ref30"></nolink> <nolink nlid="nl9" bibid="bib482" firstref="ref34"></nolink> <nolink nlid="nl10" bibid="bib95" firstref="ref53"></nolink> <nolink nlid="nl11" bibid="bib94" firstref="ref55"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1327043
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Children Boost Their Cognitive Performance with a Novel Offloading Technique
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Armitage%2C+Kristy+L%2E%22">Armitage, Kristy L.</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-1898-8622">0000-0002-1898-8622</externalLink>)<br /><searchLink fieldCode="AR" term="%22Redshaw%2C+Jonathan%22">Redshaw, Jonathan</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-7729-1577">0000-0002-7729-1577</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Child+Development%22"><i>Child Development</i></searchLink>. Jan-Feb 2022 93(1):25-38.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 14
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2022
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Children%22">Children</searchLink><br /><searchLink fieldCode="DE" term="%22Cognitive+Style%22">Cognitive Style</searchLink><br /><searchLink fieldCode="DE" term="%22Cognitive+Processes%22">Cognitive Processes</searchLink><br /><searchLink fieldCode="DE" term="%22Problem+Solving%22">Problem Solving</searchLink><br /><searchLink fieldCode="DE" term="%22Learning+Strategies%22">Learning Strategies</searchLink><br /><searchLink fieldCode="DE" term="%22Maps%22">Maps</searchLink><br /><searchLink fieldCode="DE" term="%22Spatial+Ability%22">Spatial Ability</searchLink><br /><searchLink fieldCode="DE" term="%22Age+Differences%22">Age Differences</searchLink><br /><searchLink fieldCode="DE" term="%22Individual+Differences%22">Individual Differences</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1111/cdev.13664
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0009-3920
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Ninety-seven children aged 4-11 (49 males, 48 females, mostly White) were given the opportunity to improve their problem-solving performance by devising and implementing a novel cognitive offloading strategy. Across two phases, they searched for hidden rewards using maps that were either aligned or misaligned with the search space. In the second phase, maps were presented on rotatable turntables, thus allowing children to manually align all maps and alleviate mental rotation demand. From age six onwards, children showed strong evidence of both mentally rotating misaligned maps in phase 1 and manually aligning them in phase 2. Older children used this form of cognitive offloading more frequently, which substantially improved performance and eliminated the individual differences observed in phase 1.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: Note
  Label: Notes
  Group: Note
  Data: https://osf.io/485ka
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2022
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1327043
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1327043
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1111/cdev.13664
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 14
        StartPage: 25
    Subjects:
      – SubjectFull: Children
        Type: general
      – SubjectFull: Cognitive Style
        Type: general
      – SubjectFull: Cognitive Processes
        Type: general
      – SubjectFull: Problem Solving
        Type: general
      – SubjectFull: Learning Strategies
        Type: general
      – SubjectFull: Maps
        Type: general
      – SubjectFull: Spatial Ability
        Type: general
      – SubjectFull: Age Differences
        Type: general
      – SubjectFull: Individual Differences
        Type: general
    Titles:
      – TitleFull: Children Boost Their Cognitive Performance with a Novel Offloading Technique
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Armitage, Kristy L.
      – PersonEntity:
          Name:
            NameFull: Redshaw, Jonathan
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2022
          Identifiers:
            – Type: issn-print
              Value: 0009-3920
          Numbering:
            – Type: volume
              Value: 93
            – Type: issue
              Value: 1
          Titles:
            – TitleFull: Child Development
              Type: main
ResultId 1