The relationship between sociocultural factors and autobiographical memoriesfrom childhood: the role of formal schooling

1 Memory. Special Issue: New Perspectives on Childhood Memory The relationship between sociocultural factors and autobiographical memories from childhood: the role of formal schooling Manuel L. de la Mata 1, Andrés Santamaría 1, Eva Mª Trigo 1, Mercedes Cubero 1, Samuel Arias 1, Radka Antalíková 2, Tia G.B. Hansen & Marcia L. Ruiz 3. 1. University of Seville, Seville, Spain; 2. Aalborg University, Aalborg, Denmark; 3. Autonomous University of Tamaulipas, Mexico. The research reported in this manuscript was supported by the Autonomous University of Tamaulipas (UAT), Mexico to Marcia L. Ruiz. The writing of the manuscript was supported by the Programa Estatal de Fomento de la Investigación, Científica y Técnica de Excelencia, Subprograma Estatal de Generación de Conocimiento, en el marco del Plan Estatal de Investigación Científica y Técnica y de Innovación 2013- 2016 [National Programme for the Promotion of Scientific and Technical Research of Excellence, within the National Plan for the Promotion of Scientific and Technical Research and Innovation 2013-2016] to Manuel de la Mata, Andrés Santamaría, Mercedes Cubero and Samuel Arias. The authors express their deep and abiding gratitude to the participants in the study, data from which are reported in this manuscript. Correspondence should be addressed to Manuel L. de la Mata Benítez, Department of Experimental Psychology, University of Seville. Camilo José Cela s/n, , 41018 Seville Spain mluis @us.es. “This is an Accepted Manuscript version of the following article, accepted for publication in Memory, de la Mata, M. L., Santamaría, A., Trigo, E. M., Cubero, M., Arias-Sánchez, S., Antalíková, R., Hansen, T. G. B., & Ruiz, M. L. (2019). The relationship between sociocultural factors and autobiographical memories from childhood: the role of formal schooling. Memory, 27(1), 103–114. https://doi.org/10.1080/09658211.2018.1515316. It is deposited under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way.” 2 Abstract Cross-cultural differences in autobiographical memory (AM) are associated with cultural variations. in models of the self and parental reminiscing style, but not many studies have analysed the relationship between AM and specific cultural practices such as formal schooling. Theoreticians like Greenfield (2009), Kağitçibaşi (2005) and Keller (2007) have considered formal schooling as an engine towards the model of independence; however, the empirical evidence in this regard is inconclusive: while some studies found evidence of a relation between formal schooling and characteristics of AM, others did not. To solve this inconsistency, the present study compared orally narrated childhood memories of Mexican adults with three different levels of education (from rudimentary literacy to university). Results support a relationship between formal schooling and AM in the predicted direction: More educated participants reported longer, more specific and more self-oriented memories than those with less schooling experience did. Some gender differences were also observed, with males generally reporting more individually and less socially oriented memories than females, except for university level participants. We conclude that these results support Greenfield’s theory about formal schooling as a sociocultural factor that promotes the cultural pathway to independence, as well as complexity and context-boundedness of gender differences in AM. Key-words. autobiographical memory, childhood memories, self, culture, education, gender Contact. mlu[email protected] 3 Cross-cultural research has provided extensive evidence of cultural variations in autobiographical memory (AM).Several studies have found that memories of individuals from Western cultures (European Americans, Europeans, etc.) are typically earlier, longer, more elaborated and include more references to emotions and internal states than memories reported by individuals from Asia and other non-Western cultures (Fivush, 2011; Han, Leichtman, & Wang, 1998; MacDonald, Uesiliana, & Hayne, 2000; Mullen, 1994; Nelson & Fivush, 2004; Wang, 2001, 2004, 2006; Wang, Conway, & Hou, 2004; Wang, Hutt, Kulkofsky, McDermott, & Wei, 2006). Studies also found that memories of Western individuals include fewer characters and are more focused on the self than memories from non-Westerners (Fivush, 2011; Wang, 2006). To account for the reported differences in AM, various cultural influences have been considered, notably self-construal and narrative environment (Leichtman, Wang, & Pillemer, 2003). The notion of self-construal refers to the culturally influenced ways in which “individuals define and make meaning of the self” (Cross, Harding, & Gercek-Swing, 2011, p. 143). The seminal study by Markus and Kitayama (1991) identified two distinct views of the self: the independent and the interdependent self-construal. While the independent self-construal represents the typical Western view of the self, the interdependent self-construal represents views of the self as always related to other selves. 4 The distinction between independent vs. interdependent self-construal may account for the above-mentioned differences in AM between individuals from Western and Eastern cultures. While the independent self-construal would facilitate emergence and development of an organized and elaborate memory system for events that happen to the self, conceived as the main protagonist of the memories, the interdependent selfconstrual’s focus on communion (rather than agency) would de-emphasise individuality and promote social integration and interdependence (Wang, Conway, & Hou, 2004). Despite its great influence, Markus and Kitayama’s distinction has received numerous criticisms. Recent research has challenged some of its basic assumptions, notably that culture determines self-construal and that individuals from different cultures have reliable differences in self-construal. From this perspective, independent and interdependent self-construals are not mutually exclusive. Rather, they can coexist at the individual level and, hence, could be conceptualized more as a trend than as culturally predetermined categories. In support of this argument, authors such as Oyserman and colleagues (Oyserman, Coon, & Kemmelmeier, 2002; Oyserman & Lee, 2007), Kam, Xiaolin, Zhang and Ho (2012) and Vignoles et al. (2016) have reformulated the classical distinction between individualism and collectivism (Hofstede, 2001; Triandis, 1995) at the cultural level, and independent and interdependent self (Markus & Kitayama, 1991) at the individual level. Instead of two poles of a single dimension, these conceptual pairs are considered as orthogonal and domain-specific constructs that can be elicited by different social and situational cues. Rather than assuming the existence of a homogeneous notion of the self in a given country, these authors see a need to establish how different dimensions of the self are combined in specific social, historical, and cultural contexts. From this perspective, individuals may be characterized by specific 5 combinations of independent and interdependent trends modulated by social and cultural contexts and practices (Green, Deschamps, & Paez, 2005). Therefore, (cultural) self-construals would reflect the prevalent patterns in a specific cultural context, but not all members of a specific cultural context fit into this pattern due to individual differences. In this view, self-construals are dynamic and change across the diverse cultural practices and contexts that people are engaged in (Markus and Kitayama, 2010). A focus on the independent or interdependent type of self-construal would not deny the co-existence of its counterpart. So, the focuses on independence and interdependence in self-construal may vary across individuals and situations, influencing the content, style, and accessibility of autobiographical memories. Another mechanism of cultural influence on AM is narrative environment (Leichtman, Wang, & Pillemer, 2003). As a relatively extensive volume of research has demonstrated, the use of highly elaborative-evaluative parental reminiscing styles is associated with ways of remembering the personal past that are characterised by autonomy and self-centration (Fivush, 2011, 2014; Fivush, Haden, & Reese, 2006). A highly elaborative parental reminiscing style seems to be characteristic in middle-class mothers in Western societies (Fivush, 2007, 2011, but see also Leyva, Reese, Grolnick, & Price,., 2009, for evidence of this relationship in low-income families in the US), and cross-cultural research has evidenced its links to cultural models that emphasise autonomy (Schröder et al., 2013). In contrast, a less elaborative parental reminiscing style is characteristic in cultural groups that emphasise relatedness and tend to promote ways of remembering that are consistent with the model of self as interdependent. Research has also found gender differences in parental reminiscing style. In general, mothers (and fathers, too) make more references to emotions when conversing with their daughters than they do when conversing with their sons (Fivush, 2007). These 6 differences in parental reminiscing style are consistent with gender differences in the memories of children and adults regarding aspects such as the age in the earliest memory (see Davis, 1999; Hayne & MacDonald, 2003, but see also Wang, 2001, Wang et al., 2004) and the form and content of the memories, with females’ memories being longer, more detailed and more emotionally charged than males’ memories (Fivush, 2007, 2014; Niedzwienska, 2003). Moreover, females’ memories are characterised by a predominance of themes related to affiliation, in comparison with males whose memories, in turn, are more oriented towards autonomy and agency (Niedzwienska, 2003). However, it should be noted that gender differences may vary across cultures (Wang & Fivush, 2005), evidencing the complexity of the relationship between gender, culture and AM (Cala & de la Mata, 2010). The studies reviewed above indicate some of the processes by which culture influences AM. However, we think that there is an important gap in cross-cultural research, with too sparse attention paid to the relationship between specific cultural practices and AM. Among these practices, we are going to focus on formal schooling (Rogoff, 1981, 1990; Scribner & Cole, 1981). The need to study this relationship has both theoretical and empirical foundations. At a theoretical level, authors such as Kağitçibaşi and Greenfield consider formal schooling to be a factor that promotes the model of independence. For Kağitçibaşi (2005, 2007), an autonomous-related self is often found in traditional societies that used to emphasise relatedness between parents and children but are now beginning to value children’s autonomy, too. She has also suggested that this shift is caused by vast and rapid urbanisation and socio-economic development in these countries. The extension of formal education in the population is an important factor in this development. In a similar vein, Greenfield (Greenfield, 2009; Greenfield, Keller, Fuligni, & Maynard, 7 2003) has proposed two opposing cultural pathways of development, one emphasising individuation and independence, the other promoting group membership and interdependence. For Greenfield (2009), particular socio-demographic changes in society, such as the movement from rural to urban residence, from informal education in the family to formal education, and from subsistence economy and low-technology environments to highly technological environments, have resulted in a general shift towards independence. Research carried out by Keller and colleagues (Keller, 2007) has supported this claim, showing that formal education and urbanisation promote qualitative changes that mitigate the model of interdependence and emphasise the model of independence in different settings, which has important implications for individuals’ cognitive processes, including autobiographical memory (Demuth, Chaudary, & Keller, 2011; Keller, 2007). With regard to empirical evidence, the picture is varied and complex. At least three sources of evidence, coming from different research fields and traditions, must be considered. One important set of contributions are from cross-cultural studies about the relationship between formal schooling and memory strategies (Cole 1990, 1996; Rogoff, 1981). These studies have documented the existence of differences between individuals associated with their level of schooling. It was found that children and adults with more school experience tended to employ memory strategies that enhanced their recall, while people who were not exposed to Western schooling did not use such strategies (Cole, 1996; Rogoff, 1981, 1990). Moreover, such differences only appeared after the schooled participants had received several years of formal schooling (Mistry, 1997). Although these studies did not focus on autobiographical memory, they have provided a first piece of evidence of a relationship between formal schooling and memory. Moreover, some of the proposed mechanisms of the relationship (e.g., the 8 tendency of schooling activities to promote reflection) can be also involved in the relationship between schooling and AM. We shall return to this issue later. More recently, studies by Lai (2004) with Taiwanese and by Küntay and Atham (2004) with Turkish mothers and their 2-3 years old children have related formal schooling and parental reminiscing style. In both cases, the authors found significant correlations between parental reminiscing style and the educational experience of the mothers. More specifically, the mothers with more years of education used a highly elaborative reminiscing style when conversing with their children about shared past events more frequently than mothers with less educational background did. These mothers, in contrast, adopted a low elaborative style in these conversations in most cases. Given the importance of parental reminiscing style in explaining differences in the ways children and adults remember their personal past (Fivush, 2011, 2014; Fivush et al, 2006), we think that analysis of the links between parental reminiscing style and educational background of the mothers (and, eventually, of the fathers) is particularly relevant and necessary. The last set of empirical evidence comes from a few studies that have considered the relationship between formal schooling and AM. However, the findings of these studies are not easy to interpret. While some of them have found differences in AM characteristics associated with years of formal education (Heidenreich, Junghanns- Royack, & Stangier, 2007; Kingo, Berntsen, & Krøjgaard, 2013; Reid & Startup, 2010; Wessel, Meeren, Peeters, Arntz, & Merckelbach, 2001), others have not (Berna, Schönknecht, Seidl, Toro, & Schröder, 2012; Conway, Skitka, Hemmerich, & Kershaw, 2009; Janssen, Chessa, & Murre, 2005). In support of the relationship, Kingo et al., (2013) reported a significant effect of educational level (that varied from basic school to university education) on the age at 9 the earliest memory. Participants with 9-10 years of basic compulsory schooling reported later memories than those with higher levels of education. It is interesting to note that the age of the participants (ranging from 20 to 70 years old) had no significant effect on this variable. In a related vein, a few studies have found significant differences in AM specificity associated with years of schooling (Heidenreich et al, 2007; Reid & Startup, 2010; Wessel et al., 2001). Although the focus of these studies was not assessment of the relationship between formal schooling and AM specificity, in all three cases the participants’ schooling experience (measured in years of education) was considered a moderating or control factor and had a statistically significant effect on specificity. Finally, other studies have found no evidence of a relationship between years of education and certain AM dimensions and characteristics. These dimensions include the age distribution of autobiographical memories (Janssen et al., 2005), the consistency of flashbulb memories about 911 (Conway, et al., 2009), and certain phenomenological characteristics of the memories of healthy vs. cognitively impaired adults (Berna et al., 2012). In concert, the studies summarised above have provided mixed and inconsistent evidence about the relationship between formal education experience and AM. How can we interpret these inconsistencies? It should be taken into account that these studies varied in many aspects, both conceptually (with diverse theoretical approaches to AM), and methodologically (design and techniques for data-collection and the specific characteristics of the autobiographical memories measured). However, a systematic analysis of these dimensions does not seem to account for the inconsistent results. Instead, we think that a promising aspect to be considered is the range of years of education of the participants. In this regard, some of the studies reporting positive effect 16 interpersonal relations, and sensitivity to the viewpoints of others. (“I love my daddy”; “She is my friend”). For this analysis, Kappa value obtained was K = .92. A self-focusing index was calculated for each participant by dividing the number of personal self-descriptions by the total number of self-descriptions (personal + social self-descriptions). This index was between 0 and 1, and higher scores indicated a greater degree of personal self-focusing. Results All the dependent variables involved in the hypotheses related to education level and gender (memory volume, specificity, content, individual/social orientation, autonomous orientation, and self-focusing) were analysed through a factorial ANOVA with educational level (literacy, basic education, and university) and gender (male, female) as between-group factors. First, the homoscedasticity assumption was tested using the Levene test. When the assumption was not met, we also calculated the heteroscedastic Welch F in a one-way design, with educational level as the unique factor, because the Welch F is not available for our factorial design. A Bonferroni adjustment was applied on 3 of the measures (p * 3), because Pearson correlations between them were above .30 (medium effect size). Post-hoc Tukey tests were employed for pairwise comparisons of educational levels when the general test found statistically significant differences with α = .05. Finally, we calculated generalized η2 as the effect size index (Bakeman, 2005, Olejnik & Algina, 2003, Trigo & Martínez Cervantes, 2016) for two non-experimental factors. This index was evaluated in accordance with the levels defined by Cohen (1988): small (.01), medium (.06) and high (.14). 17 Table 1 shows the statistics for all factorial ANOVAs. As can be seen in the table, the main effect of educational level was statistically significant for memory volume, with a medium effect size. Tukey tests revealed statistically significant differences between university level and the other two levels: literacy (p = .001) and basic level (p = .029). As hypothesized, the memories of university students had a higher volume than the memories from the other two levels did (see Table 2). Contrary to expectations, we found no gender difference in memory volume. Here table 1 The same pattern was obtained for memory specificity. We found a significant main effect of educational level, with a high effect size in this case (see Table 1), and statistically significant differences in the pairwise comparisons university-literacy level (p < .001), and university-basic level (p = .004), with higher specificity in the university level group (see Table 2), as predicted. Here table 2 With regard to the dependent variable content (social vs individual), although two main effects were statistically significant with the normal F, only gender had a significant effect with the heteroscedastic Welch F, with a medium effect size (see Table 1). In this case, women’s autobiographical memories presented higher levels of social themes than those of men (see Table 2). In the analysis of autonomous orientation, we found two statistically significant effects, a main effect of gender and an interaction between gender and educational level, 18 which modulated the former (see Table 1). As Figure 1 shows, while the index of autonomous orientation was higher for men than women at the literacy and basic level, in the university levels the pattern was the opposite, with women presenting more autonomous orientation than men. With regard to individual-social orientation (self/others ratio), the two main effects of educational level and gender were statistically significant (see Table 1). However, only the variable educational level reached a medium effect size. Tukey tests revealed that the difference between university and literacy level was statistically significant (p = .001), with a higher mean at university than literacy level (see Table 2). Differences between the other levels were not significant. The data were not conclusive regarding the hypothesis of women being more socially oriented than men, because of a small effect size. The last dependent variable analysed was the self-focusing index. Statistically significant differences were obtained for this variable, with a high effect size for educational level, and medium effect size for gender (see Table 1). The Tukey test revealed statistically significant differences between university level and the other two levels: literacy (p = .002) and basic level (p < .001). As predicted, university level participants had a higher self-focusing index than participants from the other two levels did, and the self-focusing index was higher in men than in women (see Table 2). Here Figure 1 Discussion 19 The aim of this study was to provide evidence of a relationship between formal schooling and AM. An additional goal was to examine gender differences in AM and interaction between formal schooling and gender in this regard. For that purpose, we asked Mexican women and men to orally narrate a memory from their childhood. The level of education of the participants ranged from a very low level (the literacy level, with participants that were learning to read and write), to university background (participants with a university degree), and including an intermediate (basic) level (participants that had concluded the six years of compulsory primary education). In support of our hypotheses, we found statistically significant relations between level of education and most dimensions of AM (volume, specificity, content, individualsocial orientation and self-focusing index). In the case of autonomous orientation, we did not find the expected effect of education. Instead, we found a significant interaction between education and gender. Men at the literacy and basic level had higher levels of autonomous orientation than men did, whereas at the university level, women scored higher than men. A further examination of the data revealed that the differences associated to education background were always between the university level and the other two levels, with no significant differences between basic and literacy level. In this regard, the memories of the more educated participants (university level) were more voluminous (including more words), more specific, and more focused on the self (rather than on others) than the memories of literacy and basic level participants. Moreover, the effect sizes for all these variables were medium or high. These results confirm the findings of the study by de la Mata et al., (2016). In their study, the authors found significant differences between the same educational levels (from literacy to university level) in some characteristics of the earliest memory 20 (age, social vs. individual content and self-focusing index). The present study represents a replication and extension of the previous one. Besides the differences in the sample size, the type of memory studied (three memories from childhood in the present study) and the AM characteristics analysed, the findings of both studies are consistent. They are also consistent with the results reported by Kingo et al. (2013), as well as with the evidence from studies showing a positive correlation between years of schooling and AM specificity (Heidenreich et al.,2007; Reid & Startup, 2010; Wessel et al., 2001). In a broader perspective, these results are also convergent with the findings of classical cross-cultural research about formal schooling and memory (Cole, 1990; Mistry, 1997; Rogoff, 1981; Scribner & Cole, 1981). Among all the cultural factors studied in this area (cultural group, urban vs, rural residence, formal schooling and literacy), formal schooling experience seems to be the best predictor of differences in memory skills (Rogoff, 1981, Scribner & Cole, 1981). Although the memory skills assessed in those studies are quite different to AM, we think that our data may be considered as an extension of the findings of classical cross-cultural research about memory. Some other studies, however, have not found a significant relation between education and AM (Berna et al., 2012; Conway et al., 2009; Jansen et al., 2005). To account for these inconsistent findings about the possible role of formal schooling in AM, we have considered the range of years of education of the participants. As stated above, the differences may only appear when groups with short experience in formal schooling are studied (e.g. literacy and basic level). Other authors have advocated for alternative explanations, pointing to the possible influence of different factors that may co-vary with education, such as the number of siblings (see Bender & Chasiotis, 2011), or the socio-economic status (SES) of the families. In this regard, we must not forget that formal education experience cannot be 21 experimentally controlled. Indeed, that goes for all cultural factors such as educational background, SES and urban vs. rural residence. Moreover, and from a more theoretical perspective, they all comprise a large set of experiences that are difficult (and very often impossible) to disentangle. We therefore need to be cautious when interpreting differences and cannot establish causal relationships. Nevertheless, we think that formal schooling is a cultural activity that needs to be considered to account for at least part of the observed differences. There are several reasons to do so. With regard to other variables, such as the number of siblings, we have argued elsewhere (de la Mata et al., 2016) that unlike Bender and Chasiotis (2011), our study included participants with almost no schooling experience and this may increase the relative importance of this factor in explaining the differences. At the same time, there is extensive evidence that educational background of individuals and families are prevailing factors in explanations of both cognitive achievement of children (Dubow, Boxer, & Huesmann, 2009) and cognitive functioning in the elderly (Jefferson et al., 2011). Finally, as we have discussed above, cross-cultural research about formal schooling and memory has demonstrated that among the cultural factors and experiences studied in this field, formal schooling seems to be the one most clearly associated with memory (Cole, 1990; Rogoff, 1981, Scribner & Cole, 1981). All in all, we think that this evidence supports the consideration of the relationship between formal schooling and AM. The analysis of our data showed that when a statistically significant effect of education was found, the difference was always between the university level and the other two levels (literacy and basic level). How can we interpret this finding? To answer that question some theoretical and methodological issues must be taken into account. First, our methodology was based on the analysis of narratives reported by the 22 participants. These narratives were reported orally (the participants from the literacy level could hardly read and write) and it seems very likely that university graduates would possess better narrative skills than participants from the other two levels. Therefore, whether the differences should be attributed to the autobiographical narratives or to the autobiographical memories of the participants remains an open question. Further studies might examine this issue by investigating differences in nonverbal or non-narrative aspects of autobiographical memories as a function of educational background. Second, our analysis of the autobiographical narratives focused on dimensions related to autonomy and self-orientation versus relatedness and other-orientation. According to Greenfield (2009) and Kağitçibaşi (2007) formal schooling can be considered a cultural factor, which promotes a model of the self that emphasises autonomy and self-determination. From this theoretical perspective, it is not surprising that autobiographical memories of people with high levels of formal education are individually oriented and focused on the self. A second independent variable in our study was gender. As discussed in the introduction, previous evidence about gender differences in AM is rather complex and sometimes unclear. While some studies have reported gender differences in AM characteristics such as the age at the earliest memory, themes and social orientation, other studies have not found such differences (see Fivush, 2011; Grysman & Hudson, 2013; Niedzwieska, 2003, for a review of these studies). In still other cases, the differences were not as expected, including culture by gender interactions (Wang, 2004). In our study, we found gender differences in some AM dimensions (themes, autonomous orientation, individual-social orientation and self-focusing index). In all cases, the differences observed supported our hypotheses. The memories reported by 23 female participants were more socially oriented than those reported by males, while the memories reported by the males were more oriented towards autonomy and more focused on the self than were those reported by the females. However, as in most previous studies, the effect sizes were small. In the specific case of autonomous orientation, the main effect of gender was modulated by the educational background of the participants. While men from the literacy and basic level scored higher than women in this variable, the pattern observed in the university level participants was the opposite, with women presenting higher values of autonomous orientation than men. It is interesting to note that in this case, the size effect was medium. From our perspective, these results show the complexity and contextual nature of gender differences in AM. According to Grysman, & Hudson (2013), factors such as the specific instructions given to the participants, various aspects of the context (including the researcher’s gender), gender salience, and the measures of AM used can affect and in many cases mitigate or even eliminate gender differences. Besides these factors, it is also necessary to consider the relationship between gender identity and AM as well as developmental changes in these relations. In this vein, Grysman, and Fivush (2016) argue that the relatively inconsistent findings in the literature about AM and gender may be due to the focus of these studies on categorical gender, rather than on gender identity (the fit with gender stereotypes). Moreover, most studies have examined gender differences in emerging adults and, especially, in university students, Since issues of professional identity may become central in this period, it is expectable to find a lesser degree of endorsement of gender-typical dimensions (such as relatedness for women). The results obtained by Grysman and Fivush (2016) and by Grysman, Merrill and Fivush (2017) support this view. In these studies, the authors found that gender identity (not least the fit with feminine gender stereotypes), rather than categorical gender, 24 predicted AM phenomenology and the use of emotion expressions in autobiographical memories. Moreover, in both studies the effect sizes were large. For the authors, these findings confirm the need to focus on gender identity and the complexity of the relationship between gender, identity and AM. The results of our study are consistent with a contextual and identity-focused conception of the relationship between gender and AM. More specifically, they call for the consideration of the role of education as a mediating factor in that relationship. The interaction between gender and educational level observed for autonomous orientation can be interpreted as an example of that complexit and may not be unique to AM. In this regard, we have found a similar pattern of interaction between gender and education in argumentation (Cala & de la Mata, 2004). In the same way as the present study, Cala and de la Mata found that gender differences were modulated by educational background, with fewer differences in the university level group than in the others. At a general theoretical level, our study provides additional support to the consideration of formal education as a sociocultural activity that promotes the model of independence (Greenfield, 2009; Greenfield et al., 2003; Kağitçibaşi, 2005, 2007; Keller (2007). The fact that the memories of participants with a higher level of education (university level) was characterised by more self-focusing and less orientation towards others, in comparison with those of the literacy and primary school-level participants, seems to support this interpretation. However, accepting the relationship between formal schooling and AM does not answer the question of how this sociocultural practice may promote forms of autobiographical remembering that are characterised by an emphasis on individuality and self-focusing. The study reported in this paper does not provide evidence in this regard. We can only speculate about the possible processes by which formal schooling 25 activities may influence AM. One possible avenue is likely to be parental reminiscing style. As we have argued above, there is extensive evidence of the links between highly elaborative styles in parent-child conversation about past events and children’s ways of remembering. Western middle-class mothers and fathers typically use a more elaborative reminiscing style than non-Western parents (Fivush, 2011; Schröder et al., 2013) and, consequently, their children tend to recall earlier and more elaborative autobiographical memories. Moreover, these narratives are usually characterised by a greater emphasis on autonomy, with the self as the central protagonist. In contrast, autobiographical memories of non-Western children are typically dated later, are less elaborated (including less references to child’s mental states), more socially oriented ((Fivush, 2007, 2011; Fivush et al., 2006; Reese, Haden & Fivush, 1993) and more didactic, placing the child’s individual experiences in the context of the group and moral behaviour (Fivush, 2011). These studies are based on comparisons of mother-child dyads from different cultures (Western and Easter cultures, typically). The studies developed by Heidi Keller and colleagues (see Schröder et al., 2013, for a review) that analyse reminiscing situations and memories of parent-child dyads, which represent the cultural models of autonomy, relatedness and autonomy-relatedness, mark a significant advancement in disentangling how culture modulates AM. In our view, it is necessary to also analyse how cultural practices like formal schooling may relate to AM via reminiscing situations. In this vein, results of the studies by Lai (2004) and Küntay and Atham (2004) demonstrating a positive relationship between mothers’ education and highly elaborative reminiscing style, provide one step further in this direction and support the idea of parental reminiscining style as a privileged link between formal education and AM. 32 Küntay, A. & Atham, B. (2004). Annelerin Cocuklariyla Geçmis Hakkindaki Konusmalarinin Anne Egitim Düzeyiyle Iliskisi. Türk Psikoloji Dergisi, [Effect of maternal education on Turkish mothers' styles of reminiscing with their children. Turkish Journal of Psychology], 19, 19-31. Lai, W. F. (2004). Co-constructing narratives with young children: A study of the relationships between Taiwanese mothers' discourse styles and mothers' education, family income, and children's age. Unpublished Doctoral Dissertation: Boston University. Leichtman, M. D., Wang, Q., & Pillemer, D. B. (2003). Cultural variations in interdependence and autobiographical memory: Lessons from Korea, China, India and the United States. In R. Fivush & C. A. Haden (Eds.), Autobiographical memory and the construction of a narrative self. Developmental and cultural perspectives (pp. 73-97). Mahwah, N.J.: Lawrence Erlbaum Associates. Leyva, D., Reese, E., Grolnick, W., & Price, C. (2009). Elaboration and Autonomy Support in Low-Income Mothers' Reminiscing: Links to Children's Autobiographical Narratives. Journal of Cognition and Development, 9(4), 363-389. http://doi.10.1080/15248370802678158. MacDonald, S., Uesiliana, K., & Hayne, H. (2000). Cross-cultural and gender differences in childhood amnesia. Memory, 8(6), 365-376. http://dx.doi.10.1080/09658210050156822. Markus, H. R. & Kitayama, S. (1991). Culture and the self: Implications for Cognition, Emotion, and Motivation. Psychological Review, 98(2), 224-253. http://dx.doi.10.1037/0033-295X.98.2.224. 33 Markus, H.R. & Kitayama, S. (2010). Cultures and selves: A cycle of mutual constitution. Perspectives on Psychological Science, 5(4), 420-430. . http://doi. 10.1177/1745691610375557. Mistry, J. (1997). The development of remembering in cultural contexts. In N. Cowan & Ch. Hulme (eds.), The development of memory in children, (pp. 343-367). Hove. Psychology Press. Mullen, M. K. (1994). Earliest recollections of childhood: A demographic analysis. Cognition, 52(1), 55-79. http://dx.doi.org/10.1016/0010-0277(94)90004-3. Nelson, K., & Fivush, R. (2004). The emergence of autobiographical memory: A social cultural developmental theory. Psychological Review, 111(2), 486-511. http://doi.10.1037/0033-295X.111.2.486. Niedzwienska, A. (2003). Gender Differences in Vivid Memories. Sex Roles, 49(7-8), 321-331. http://dx.doi:10.1023/A:1025156019547. Nourkova, V., & Dnestrovskaya, M. (2013). Dynamics of autonomy and relatedness across lifespan in autobiographical narratives: Semantic and susceptibility. Procedia Social and Behavioral Sciences, 46, 475-481. http://dx.doi.10.1016/j.sbspro.2013.08.600. Olejnik, S. & Algina, J. (2003). Generalized eta and omega squared statistics: measures of effect size for some common research designs. Psychological Methods, 8(4), 434– 447. http://doi.org/10.1037/1082-989X.8.4.434. Olson, D. R. (1994). The world on paper. New York : Cambridge University Press. Olson, D. R. (2016). The mind on paper: Reading, consciousness and rationality. New York : Cambridge University Press. 34 Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128(1), 3-72. http://doi.10.1037//0033-2909.128.1.3. Oyserman, D., & Lee, W. S. (2007). Priming “culture”: Culture as situated cognition. In S. Kitayama and D. Cohen (Eds.). Handbook of Cultural Psychology (pp. 255-275). New York: Guilford. Reese E., Haden C. A., & Fivush, R. (1993). Mother-child conversations about the past: Relationships of styles and memory over time. Cognitive Development, 8(4), 403– 430. Reid, T. & Startup, M. (2010). Autobiographical memory specificity in borderline personality disorder: Associations with co-morbid depression and intellectual ability. British Journal of Clinical Psychology, 49(3), 413-420. http://doi.10.1348/014466510X487059. Rogoff, B. (1981). Schooling and the development of cognitive skills. In H. C. Triandis & A. Heron (Eds.), Handbook Cross-Cultural Psychology, Vol. 4: (pp. 233-294). Boston: Allyn and Bacon. Rogoff, B. (1990). Apprenticeship in thinking. Cognitive development in social context. Oxford: Oxford University Press. Schröder, L., Keller, H., Kärtner, J., Kleis, A., Abels, M., Yovsi, R. D., Chaudary, N., Jensen, H., & Papaligoura, Z. (2013). Early reminiscing in cultural contexts: Cultural models, maternal reminiscing styles, and children’s memories. Journal of Cognition and Development, 14(1), 10-34. http://dx.oi.10.1080/15248372.2011.638690. Scribner, S., & Cole, M. (1981). The psychology of literacy. Cambridge, Mass.: Harvard University Press. Triandis, H. C. (1995). Individualism and collectivism. San Francisco, CA: Westview 35 Press. Trigo, M. E. & Martínez Cervantes, R. J. (2016). Generalized eta squared for multiple comparisons on between-groups designs. Psicothema, 28(3), 340–345. http://doi.org/10.7334/psicothema2015.124. Vignoles, V., Owe, E., Becker, M., Smith, P. B., Easterbrook, M. J., Brown, R., ... & Lay, S. (2016). Beyond the ‘East–West’ Dichotomy: Global Variation in Cultural Models of Selfhood. Journal of Experimental Psychology: General, 145(8), 966- 1000. http://dx.doi.org/10.1037/xge0000175. Wang, Q. (2001). Culture effects on adult’s earliest childhood recollection and selfdescription: Implications for the relation between memory and the self. Journal of Personality and Social Psychology, 81(2), 220-233. http://dx.doi.10.1037/0022- 3514.81.2.220. Wang, Q. (2004). The emergence of cultural self-construct: Autobiographical memory and self-description in American and Chinese children, Developmental Psychology, 40(1) 3–15. http://doi.10.1037/0012-1649.40.1.3 Wang, Q. (2006). Earliest recollections of self and others in European American and Taiwanese young adults. Psychological Science, 17(8), 708-714. http://dx.doi.org/10.1111/j.1467-9280.2006.01770.x. Wang, Q., Conway, M. A., & Hou, Y. (2004). Infantile amnesia: A cross-cultural investigation. Cognitive Science, 1, 123-135. Wang, Q., & Fivush, R. (2005). Mother-child conversations of emotionally salient events: exploring the functions of emotional reminiscing in European-American and Chinese families. Social Development, 14(3), 473-495. http://dx.doi.org/10.1111/j.1467-9507.2005.00312.x 36 Wang, Q., Hutt, R., Kulkofsky, S., McDermott, M., & Wei, R (2006). Emotion situation knowledge and autobiographical memory in Chinese, immigrant Chinese, and European American 3-year-olds, Journal of Cognition and Development, 7(1) 95– 118. http://doi.10.1207/s15327647jcd0701_5. Wang, Q. & Leichtman, M.D. (2000). Same beginnings, different stories: A comparison of American and Chinese children´s narratives. Child Developmental, 71(5), 1329- 1346. http://doi.org/10.1111/1467-8624.00231. Wessel, L, Meeren, M., Peeters, F, Arntz, A., & Merckelbach, H. (2001). Correlates of autobiographical memory specificity: The role of depression, anxiety and childhood trauma. Behaviour Research and Therapy, 39(4), 409-421. http://doi.10.10l6/S0005- 7967(00)00011-5.