Learning Strategy Differentially Impacts Memory Connections in Children And Adults(2)

3.4 Results of follow-up analyses hint at the emergence of integration among children told to integrate

The findings in the integrate condition are on their surface consistent with our original predictions: That even despite being instructed to integrate, children would not form A-C connections. However, we believe it would be premature to conclude an absence of integration in this group based on the present data. (Relatedly, it is again worth underscoring that we did not find significant evidence for the hypothesized interaction in the integrate condition or any other instruction condition, and as such we cannot conclude that integration is stronger in adults than children.)

Cistanche supplement near me-Improving Memory

Click here to view  Cistanche Improve Memory products

【Ask for more】 Email:cindy.xue@wecistanche.com /  Whats App:  0086 18599088692 /  Wechat:  18599088692

We performed two follow-up analyses to further interrogate evidence for integration among children told to integrate. The results of these analyses provide some indication that integration may exist, albeit weakly, in this group; and this may in part be because it is emerging over the sampled developmental period. First, given the lack of correspondence between implicit and explicit memory measures in children for the novel A-C connections (described below), we reasoned that it may be most appropriate to interrogate evidence for such priming in children across all trials—that is, irrespective of not only subsequent explicit AC inference performance but also irrespective of initial AB and BC direct pair learning success. This also serves to increase the number of trials contributing to the analysis among children, thereby affording more statistical power. Under this more lenient restriction, children now showed a statistical trend toward priming in the integrate condition (z = 1.819, p = 0.069; the effect also remained significant in adults, z = 2.014, p = 0.044). Therefore, it may be that restricting according to explicit memory exhibited in the 3AFC tests led us to underestimate these connections in children. Second, we asked whether priming varied across age (continuously) among children. The age × sequence type interaction was not significant (χ2(1) = 2.335, p = 0.126); however, we did find trend-level evidence for priming at the older (z = 1.887, p = 0.059) end of our age range that was not present at the younger (z = −0.813, p = 0.416). Together, these results are broadly consistent with the notion that some children in the integrated group—perhaps particularly those approaching age 10—may form A-C connections when told to engage in an integration strategy. However, given these findings provide only trend-level evidence for these ideas, future studies will be needed to replicate and extend upon the present data.

Effects of Cistanche-Improving Memory

3.5 Developmental differences in implicit-explicit correspondence

We next set out to assess the behavioral significance of the implicit connections exhibited during the priming task. Specifically, we reasoned that greater facilitative priming for an association might translate to more accurate or faster decisions on the corresponding explicit inference or memory test trial. Of note, we anticipated substantial person-to-person and even memory-to-memory variability in the tendency to connect interrelated memories (Pajkert et al., 2017; Schlichting & Preston, 2014, 2016; Schlichting et al., 2014, 2015; Zeithamova, Dominick, et al., 2012), instruction manipulation aside. Moreover, we reasoned that instruction manipulation would impact what is eligible for associative binding, not whether such binding occurs (Davis et al., 2021). Therefore, here we consider participants from all instruction conditions, reasoning that any strong item-item link exhibited during the priming task—perhaps regardless of when or how it was formed—could support performance on a subsequent explicit 3AFC test. Importantly given prior work (Bauer, Cronin-Golomb, Porter, Jaganjac & Miller, 2020), we also anticipated that the nature of the implicit-explicit relationship might be stronger in adults than children.

Benefits of cistanche tubulosa-Improving Memory

To assess the trial-by-trial link between the degree of priming and subsequent test performance, we needed experience-specific values representing the degree to which a response is particularly speeded or slowed for a given target object. However, we found that RTs significantly changed across the task (slopes in intact and rearranged, for children and adults; all z > 7.21, p < 0.001), such that participants became on average slower as the task went along. Therefore, rather than considering raw RTs, we instead developed an analytical approach to remove these overall trends (Figure 6A). Briefly, we regressed RTs to target objects on trial numbers for each participant. We then computed, for each target object, the degree to which the observed RT differed from the participants’ predicted average RT at that same trial number (i.e., residuals; predicted—observed). This approach allowed us to remove the effect of trial number, such that values for any particular object were not contaminated with when (i.e., at what trial number) it had been presented during the preference task. Positive values indicated a faster-than-average response, whereas negative values indicated a slower-than-average response. We then centered and scaled these values within-participant, thereby removing differences across individuals in the overall level and variability of priming to isolate the within-person relationship between implicit and explicit measures of associative memory.

Desert ginseng-Improving Memory

Both children and adults showed some evidence for a correspondence between implicit and explicit measures, yet it occurred for different types of associations: Adults showed an implicit-explicit relationship that was specific to the indirect A-C links, while children showed such correspondence only for direct AB pairs. Specifically in adults, the degree of A-C priming predicted the speed of the corresponding subsequent correct inference (t(76.41) = −2.40, p = 0.019; Figure 6B; no relationship for AC accuracy: z = 1.13, p = 0.259). Adults also did not show any significant trial relationships between these implicit and explicit assessments of memory for either AB or BC direct pairs in either accuracy or RT (all |z| < 0.72, p > 0.476). In contrast, children did not show any significant associations between A-C priming and inference in terms of either accuracy or speed (both p > 0.115). However, they did show significant correspondence between A-B priming and accuracy on the final explicit AB test (z = 2.07, p = 0.038; Figure 6C; no relationship with RT: t(93.23) = −0.61, p = 0.543). There were no significant relationships observed in children for BC associations (both I> 0.319).

What does cistanche do-Improving Memory

These results highlight that in adults—perhaps because their memories were disrupted throughout the priming task (see Supplementary Results)—only A-C connections (measured early in the priming task) mirrored later explicit inference decisions. Children also showed a correspondence between implicit and explicit memory behaviors. Yet, in contrast to adults, this relationship was only present for direct AB pairs. This finding is consistent with the idea that only the most established AB associations may give rise to implicitly detectable connections that are also accessible during the explicit test in children. By contrast, A-C links may be more tenuous, remaining unavailable to children during an explicit inference decision.

3.6 Integration offers protection from forgetting in children

Children exhibited substantial forgetting between initial and final direct pair tests (see Supplementary Results). Inspired by past work in adults showing that integration can offer protection from forgetting (Anderson & McCulloch, 1999; Radvansky, 2005), here we ask whether the same is true in children by comparing performance decrements across instruction conditions (Bauer, Esposito, et al., 2020; Varga & Bauer, 2013). We performed this analysis for children only, as the ceiling levels of performance in adults, we reasoned, would lead us to necessarily mis- (under-) estimate the size of their performance decrement.

FIGURE 6 Trial-by-trial correspondence between implicit (priming) and explicit (3AFC test) measures. (a) Analysis approach depicted for one sample participant (child). We fit a linear model to each participant’s RTs and computed the residuals (i.e., predicted—observed, such that observations below the solid line yield positive ‘RT deviance’ scores; lines) to yield an estimate of the degree to which a particular response was speeded (bottom of graph below-fitted line; blue zone on the color bar) or slowed (top of the graph; red zone on the colour bar) removing overall change across the task block. We then related this trial-to-trial variability in RT to explicit test performance. (b) The negative association between a and c priming and RT on the explicit AC inference test (dashed line) in adults (darker color). Priming was associated with faster subsequent inference. (c) The positive association between A-B priming and AB accuracy (solid line) in children (lighter color). In all panels, lines and 95% confidence bands are derived from mixed-effects models. * p < 0.05

FIGURE 7 Forgetting in children. Left, here we quantified the drop in memory performance from the initial (top AB, from the final repetition during learning in yellow; bottom BC, from the BC exposure test in blue to the final direct pair test in red). Difference scores were divided by initial memory (proportion correct) to correct for differences in memory at baseline. Right, direct pair (AB, BC combined) forgetting in children as a function of instruction condition (color). All conditions showed significant forgetting (here, values significantly below zero; color-coded asterisks), but this decline was attenuated in the integrated group (black asterisks denote pairwise comparisons). * p < 0.05

Directly comparing children’s performance between the initial exposure and final test revealed an overall decrease (z = 9.354, p < 0.0001) that nevertheless differed significantly by instruction condition (χ2(2) = 14.524, p = 0.0007). To compare forgetting across conditions, we divided the change in memory accuracy from the learning phase to the final test by the initial learning-phase memory ((final test—learning) / learning; Figure 7). Forgetting differed significantly across conditions (F(2,138) = 4.592, p = 0.011): Children instructed to integrate forgot proportionally fewer direct pairs (AB, BC combined) than did children instructed to either encode (t(138) = 2.926, p = 0.004) or retrieve (t(138) = 2.166, p = 0.032). Integration also offered more protection for AB memories relative to AB retrieval on its own (condition × task phase interaction on AB memory for retrieve vs. integrate: χ2(1) = 4.377, p = 0.036); and likewise, integrating A into the BC memory yielded more robust memories than encoding BC in isolation (condition × task phase interaction on BC direct pair memory for encode vs. integrate: χ2(1) = 8.210, p = 0.004). In sum, while children exhibited forgetting in all conditions, this decline was the smallest when they were encouraged to integrate during BC exposure. Therefore, in addition to offering the greatest flexibility by maximizing performance on the explicit inference test, attempting to combine the three memory elements into a single, coherent story during learning (as in the integrate condition) was also beneficial to children’s retention of the underlying direct associations.

4 DISCUSSION

We manipulated the learning strategy as children and young adults were presented with new (BC) information that could be related to prior (AB) knowledge. Children’s performance on explicit tests mirrored their focus during exposure, consistent with prior reports of the mnemonic rigidity in development (DeMaster et al., 2015; Ghetti & Fandakova, 2020).

Integration was overall the optimal strategy for children in terms of yielding the most robust (i.e., least likely to be forgotten) direct pair memories and best inference. Implicit signatures also revealed that children formed indirect A-C connections before the inference test in the encoding condition (and perhaps to a lesser extent in the integrate condition). As expected, adults also formed such links when told to integrate and yet showed signs that these links were eroded when the associative structure was disrupted in the priming task (Supplementary Results). We also found a relationship between implicit indirect A-C connections and explicit inference in adults but not children—despite the latter group showing a tight correspondence for well-learned direct AB pairs. Together, these results indicate that differential access to an advantageous learning strategy may indeed explain the bulk of developmental differences in explicit inference. However, developmental differences nevertheless remain in terms of the learning circumstances that best promote implicit memory-to-memory connections, as well as the accessibility of those connections for later explicit choice.

Both adults and children performed optimally on the explicit assessment that most closely matched their exposure instructions: BC memory was best for the encoding condition (though notably, integration also did not ‘cost’ children much in terms of BC learning; Bauer, Esposito, et al., 2020), and AC inference was best (in terms of accuracy for children, and RT for adults) for integrate. This difference in inference performance across conditions was especially pronounced in children—there was a whopping 24% accuracy difference in children (AC inference, encode versus integrate), yet a difference only in speed for adults. There are multiple mechanisms by which adopting different strategies during overlapping BC learning might result in different levels of AC inference. For instance, it might be the case that the similarity between the experience of BC exposure and the AC test (taking into account e.g., cognitive context) is especially low in the encoding condition, thereby disadvantaging children in particular as they rely more heavily on these sorts of cues (Ackerman, 1985; Levy-Gigi & Vakil, 2010). On the other hand, it might be that an integration strategy actively facilitates performance by directing participants’ attention to both similarities and differences between experiences (Wahlheim & Jacoby, 2013). We found that integration benefitted not only flexible inference but also protected direct pair memories from forgetting in children. This finding aligns with past work in adults showing that the engagement of such a strategy protects from varied sorts of memory failure (e.g., interference, retrieval-induced forgetting; Anderson & McCulloch, 1999; Radvansky, 2005); and work in children that integration does not blur underlying memories (Bauer, Esposito, et al., 2020), but instead might even promote their retention over longer delays (Varga & Bauer, 2013). Overall, our findings suggest that giving children access to an integration strategy during learning may largely close the performance gap between children and adults in explicit inference.

In adults, explicit performance differences between encode and integrate conditions were subtle, appearing in inference RT but not accuracy. These similar levels of accuracy, however, were supported by different underlying structures: Explicit inferences in the encoding but not integrated condition were made from separately stored initial AB and overlapping BC memories, therefore requiring additional test phase processing. Specifically, we found that inferences in the encode condition took relatively more time than did those in the integrate condition; moreover, saw no overall evidence of A-C linkage in the priming measure. These findings are generally consistent with past reports, which have also shown that encoding and integrating instructions yield minimal differences in accuracy despite robust effects in more sensitive metrics like brain response (Richter et al., 2015) and recall dependency (i.e., integration increases the likelihood of recalling A and C together given B; Burton et al., 2017). We also found that children between may become increasingly better equipped to accomplish the more time-consuming test-phase processing associated with the encode condition between ages 7–9

Comparing encode and integrate to retrieve also provided clear evidence that reactivation—despite being a necessary first step—is not sufficient for integrative encoding to occur. Even among children, reach motivation of AB coupled to ignore BC appeared to prevent memory linkage on average: Not only was explicit inference poor but neither children nor adults showed implicit evidence of connecting A and C items. Given that some (Burton et al., 2017; Howard et al., 2009; but not all, Richter et al., 2015; Zeithamova & Preston, 2017) past perspectives have suggested that co-activation of AB and BC traces might be all that is necessary for integration, this finding provides an interesting data point to the contrary. Our observation that even children—despite their relatively poorer attentional control (Plebanek & Sloutsky, 2019; Wendelken et al., 2011)—can follow instructions to ignore BC and prevent its inclusion in the memory trace is especially noteworthy. We anticipated that children, much in the way of ‘hyper binding’ in older adults (Campbell et al., 2010; Davis et al., 2021), might fail to ignore the present BC information and store the ABC relationship—including irrelevant C—in memory. However, there was no evidence for this phenomenon among either children or adult retrieve participants. It remains an open question as to whether slightly different instructions (e.g., to retrieve AB but make no mention of ignoring BC) would yield different effects.

Concerning implicit A-C connections, adults exhibited such priming only in integration—consistent with past findings that this strategy can enhance connections between indirectly related items (Burton et al., 2017). Children by contrast connected A and C in the encode condition (and perhaps marginally so in integration). While these results were not what we predicted, they might nevertheless be sensible in the context of other work on spreading activation showing connections can be formed between related memories under certain learning conditions as early as infancy (Barr et al., 2001, 2014; Hayne & Gross, 2017). Here, why might it be the case that children form such indirect connections when instructed to simply encode BC—particularly given that adults do not? One set of possibilities is related to the fact that the encode condition is the only one that does not have a strategy change between initial AB and overlapping BC exposure, and as such the cognitive contexts may be most similar in this condition (i.e., the task is identical between AB and BC learning phases, with only the materials themselves being different; Sahakyan & Delaney, 2003, 2005). Such similarity of mental state during AB and BC learning could yield source confusion (Cycowicz et al., 2001; Riggins, 2014) in children, such that A and C become linked because their associated memory traces are not easily distinguishable. Alternatively, it could be that this similarity in cognitive context provides additional cues that are enough to trigger spontaneous retrieval of AB—again, given children’s demonstrated sensitivity to the similarity between study and test (Ackerman, 1985; Levy-Gigi & Vakil, 2010)—despite no overt instruction to do so. It might be the case that then, having not been explicitly told to ignore either AB or BC in this condition, children engage in something akin to hyper binding such that A, B, and C connections are nevertheless formed (Davis et al., 2021); which is interestingly thought to be a purely implicit phenomenon in older adults (Campbell & Hasher, 2018). One paradigm-related reason why such (spontaneous) memory reactivation might be plausible here when it has not been reported in prior work using related paradigms (Bauer, Cronin-Golomb, et al., 2020; Schlichting et al., 2022) is that here, we purposely provided learners with ample AB experience to enable effortful retrieval (needed for both the retrieve and integrate conditions) during BC exposure. Therefore, spontaneous retrieval might have occurred given the well-learned nature of the AB pairs to support integration (Hayes-Roth, 1977; Schlichting et al., 2015); given past work (Krøjgaard et al., 2017), such spontaneous retrieval may be even more likely successful than strategic retrieval required in the integration condition. Future work will be needed to tease apart these two possibilities. Regardless of the reason for their emergence, it is worth underscoring that we observed links in the encode condition that were (1) triad-specific, rather than between A and C item types more generally; and (2) did not support explicit inference in children. Additional research will be required to develop a fuller picture of how an overt integration strategy might differentially impact implicit and explicit aspects of memory in childhood, given we saw only trend-level evidence for A-C connections in this condition.

Relating implicit to explicit assessments of A-C connections revealed an intuitive correspondence in adults: Across all learning conditions, inference problems associated with more priming in the preceding preference task were solved faster than those exhibiting less priming. However, this same pattern was not found in children. These findings are highly consistent with prior work: Specifically, one past study (Bauer, Cronin-Golomb, et al., 2020) also showed that children’s implicit connections across related experiences (stem facts) did not relate to their explicit performance—in that study, children’s ability to combine related facts on an open-ended test. Consistent with the idea that more tentative links may exist, however, the researchers did find evidence of correspondence when explicit links were assessed using an alternative-forced-choice test (Bauer, Cronin-Golomb, et al., 2020). Here, we extend this work by showing that children’s apparent difficulty accessing such implicit connections might be specific to those that are weaker and/or span memories: We found that children did show an implicit-explicit relationship in the expected direction for the well-established and directly experienced AB pairs. Therefore, future work will be needed to better understand children’s difficulty leveraging such existing implicit links during an explicit test. Particularly for emergent or weak links (such as indirect A-C connections), it may be the case that providing further support at tests or using a different sort of memory assessment could increase children’s ability to access their memories on demand There are several limitations to the present study. One is that participants—and perhaps children especially—may have experienced fatigue since the task was relatively long and completed online. This issue might have been exacerbated given our data collection was completed during the COVID-19 pandemic, and as such most participants were also using their devices for work or school. This fatigue might have yielded suboptimal performance in, for example, the fast-paced priming task. Such factors may have reduced our ability to detect implicit A-C connections that truly exist (e.g., perhaps in the integrated condition among children). However, given that we would not expect such effects to differ between instruction conditions within each age group, nor to create artifactual evidence for connections, we suggest that this concern would not have greatly influenced our conclusions. A second limitation is that due to the online nature of our study, participation was limited to individuals who had access to a computer and internet connection in their homes. Our participants were therefore likely of relatively high socioeconomic status given they had these resources readily available. Finally, our task was quite challenging, which led to the exclusion of 51 participants for failure to meet our performance-based criteria. Integration appeared especially challenging for children, with exclusion rates due to low catch trial performance being higher for children than adults only in this instruction condition. While we can be confident that those participants we retained did understand and were able to perform the task as intended, it is nevertheless important to keep in mind that our results reflect this high-performing sample. Given this study required much of participants in terms of both resource access and performance, it is possible that the results we observed would not generalize to the broader population—a limitation that might be particularly salient in the integration condition.

Together, our results suggest that children’s lack of knowledge of an integration strategy during learning may explain a large share of the developmental differences reported on explicit tests of memory-based inference. This finding has strong significance: Contrary to our predictions, children aged 7–9 years are not fundamentally incapable of integration, and it may be that the demands of explicit memory and inference tests exaggerate behavioral differences between children and adults. Children do form memory-to-memory connections— though they might be more tenuous, less accessible, and therefore less influential for explicit behaviors. This finding presents a notable challenge as many key behaviors require the ability to reference one’s memories in a self-guided, controlled way—even for children, such as when writing a test in school. Future work is needed to determine whether interventions during explicit tests might be more effective than the learning-phase manipulation we deploy here, and perhaps support children in their ability to access and use their stored knowledge to its fullest extent.

REFERENCES

Ackerman, B. P. (1985). Children’s use of context and category cues to retrieve episodic information from memory. Journal of Experimental Child Psychology, 40, 420–438. https://doi.org/10.1016/0022-0965(85) 90075-X

Addis, D. R., Cheng, T., Roberts, R. P., & Schacter, D. L. (2011). Hippocampal contributions to the episodic simulation of specific and general future events. Hippocampus, 21, 1045–1052 Available at: https://www.pubmedcentral.nih.gov/articlerender. fcgi?artid=3815461&tool=pmcentrez&rendertype=abstract [Accessed March 27, 2014] https://doi.org/10.1002/hipo.20870

Anderson, M. C., & McCulloch, K. C. (1999). Integration as a general boundary condition on retrieval-induced forgetting. Journal of Experimental Psychology. Learning, Memory, and Cognition, 25, 608–629. Available at: https://psycnet.apa.org/journals/xlm/25/3/608/ [Accessed July 23, 2014] https://doi.org/10.1037/0278-7393.25.3.608

Barr, R., Vieira, A., & Rovee-Collier, C. (2001). Mediated imitation in 6- month-olds: Remembering by association. Journal of Experimental Child Psychology, 79, 229–252. https://doi.org/10.1006/jecp.2000.2607

Barr, R., Walker, J., Gross, J., & Hayne, H. (2014). Age-related changes in spreading activation during infancy. Child Development, 85, 549–563. https://doi.org/10.1111/cdev.12163

Bates, D., Mächler, M., Bolker, B. M., & Walker, S. C. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67, 1–48. https://doi.org/10.18637/jss.v067.i01

Bauer, P. J., Cronin-Golomb, L. M., Porter, B. M., Jaganjac, A., & Miller, H. E. (2020a). Integration of memory content in adults and children: Developmental differences in task conditions and functional consequences. Journal of Experimental Psychology General, 150, 1259–1278.

Bauer, P. J., Esposito, A. G., & Daly, J. J. (2020b). Self-derivation through memory integration: A model for accumulation of semantic knowledge. Learning and Instruction, 66, 101271. Available at: https://doi.org/10.1016/j.learninstruc.2019.101271 https://doi.org/10.1016/j.learninstruc.2019.101271

Bauer, P. J., & San Souci, P. (2010). Going beyond the facts: Young children extend knowledge by integrating episodes. Journal of Experimental Child Psychology, 107, 452–465. Available at: https://www.pubmedcentral.nih.gov/articlerender.fcgi?artid= 2922488&tool=pmcentrez&rendertype=abstract [Accessed September 26, 2011] https://doi.org/10.1016/j.jecp.2010.05.012

Bauer, P. J., & Varga, N. L. (2017). Similarity and deviation in event segmentation and memory integration: commentary on Richmond, gold, & Zacks. Journal of Applied Research in Memory and Cognition, 6, 124–128. Available at: https://doi.org/10.1016/j.jarmac.2017.01.006

Bauer, P. J., Varga, N. L., King, J. E., Nolen, A. M., & White, E. A. (2015). Semantic elaboration through integration: Hints both facilitate and inform the process. Journal of Cognition and Development, 16, 351–369. Available at: https://www.tandfonline.com/doi/abs/10.1080/ 15248372.2013.849707 [Accessed March 3, 2014] https://doi.org/10. 1080/15248372.2013.849707

Burton, R. L., Lek, I., & Caplan, J. B. (2017). Associative independence revisited: competition between conflicting associations can be resolved or even reversed in one trial. Quarterly Journal of Experimental Psychology, 70, 832–857. https://doi.org/10.1080/17470218.2016.1171886

Campbell, K. L., & Hasher, L. (2018). Hyper-binding is only apparent under fully implicit test conditions. Psychology and Aging, 33, 176–181. https:// doi.org/10.1037/pag0000216

Campbell, K. L., Hasher, L., & Thomas, R. C. (2010). Hyper-binding: A unique age effect. Psychological Science, 21, 399–405. https://doi.org/10.1177/ 0956797609359910

Carpenter, A. C., Thakral, P. P., Preston, A. R., & Schacter, D. L. (2021). Reinstatement of item-specific contextual details during retrieval supports recombination-related false memories. Neuroimage, 236, 118033. Available at: https://doi.org/10.1016/j.neuroimage.2021.118033 https://doi. org/10.1016/j.neuroimage.2021.118033

Coughlin, C., Lyons, K. E., & Ghetti, S. (2014). Remembering the past to envision the future in middle childhood: Developmental linkages between prospection and episodic memory. Cognitive development, 30, 96–110. Available at: https://doi.org/10.1016/j.cogdev.2014.02.001