{"paper_id":"3c8d8568-b0c8-4b8e-a0eb-8023815daddc","body_text":"How Interaction with Parents Affects Children's Mathematical Informal Learning and Verbal Skills | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article How Interaction with Parents Affects Children's Mathematical Informal Learning and Verbal Skills Na-Eun Lee, Biswajit Sarkar This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5885890/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Informal learning, characterized by unstructured, incidental, and everyday learning activities, shapes children's academic performance. This study leverages the Early Childhood Longitudinal Study, ECLS-K dataset, a nationally representative sample of U.S. kindergartners, to examine the relationship between parental engagement in informal learning activities and children's academic accomplishment. This study utilized data visualization to investigate tendencies of children’s scores based on each demographic and interaction factor. Further, the study attempts to apply XB-Boost which can analyze large-scale data and uncover complex, non-linear relationships. This study explores the potential of using AI methods to identify contributions of parental interaction with their children and demographic factors. STEM education Child Education Informal learning Various Activities Data visualization Random Forest Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 1. Introduction Informal learning refers to a type of learning that occurs outside of formal educational settings, like schools or training programs. Traditionally, we had considered learning to happen in school, kindergarten, and at home with conscious exercise and training(Folkestad (2006)). And this was expected to yield formal component of results such as grades, test scores and certificates. In contrast, informal learning refers to when the learning process takes place without the person being aware of it. This concept has gained growing attention these years. Such informal components consist of cultural backgrounds, extracurricular pursuits (Beard et al (2024))and communication within a belonging community. There are major characteristics in regard to process, location and setting, purpose, and content. According to Gong et al (2021), informal learning occurs incidentally during daily activities, often without any deliberate intent. People may participate in a learning process unknowingly. Additionally, informal learning is flexible, as it happens without strict limitations on time or form. It can also be an unintended byproduct of other actions. Rather than offering structured courses or formal training, it provides practical knowledge and insights. 1.1 Research Gap Data analysis has been necessary for quantitative research, providing objective understanding based on statistics. And there have been many attempts to understand children’s academic achievement, based on regression analysis and ANOVA(Analysis of Variance). Because Regression analysis provides the interpretability in correlation between variables, it can explain which factors and to what extent factors change academic performance. It is noted that there is not much research applying AI methods for the identification of informal learning environments. Remarkably, there is limited previous research investigating the contributions of informal learning environments combined with parental involvement. **Research gap 1; There is limited research analyzing the informal learning of activities between parents and children, that uses big dataset for empirical research, **Research gap 2; There are limited attempts to utilize advanced AI tools for analyzing the informal learning of activities between parents and children. 1.2 Literature review Informal learning has been acknowledged and there have been many attempts to identify informal learning processes in student’s daily lives and to implement informal learning into children’s learning processes. As Kelly et al (2023) mentioned, visiting and experiencing aquariums, museums, and zoos provides helpful educational resources to children and families. These activities in science-related places are considered to be one of the STEM informal education learning processes. In particular, many pedagogical instructions were suggested in the setting of museums to deliver engineering content. Also, daily grounded activities such as grocery shopping with caretakers are considered to promote interactions between adults and children. Moreover,(Caspe's,2009) book-sharing activity is regarded to contribute to building communication processes, social relationships, and interaction. Much research has examined informal learning within various activities and points out that informal learning practice affects children’s learning performance. LeFevre et al (2009) identified home numerical experiences are informal learning practices. Direct activities in numeracy development are explicitly designed to enhance quantitative skills. Contrarily, indirect activities such as playing cards or board games with numerical elements, cooking, or shopping also contribute significantly. It was found that indirect practices of daily life are positively correlated with child math performance. Informal learning practices have been widely studied across various activities, revealing their significant impact on children’s academic performance. For instance, LeFevre et al. (2009) explored the concept of home numerical experiences as a form of informal learning. They classified such activities into direct and indirect approaches. Direct practices, like targeted numeracy exercises, aim explicitly at developing quantitative skills. On the other hand, indirect activities—such as playing numerically-themed card or board games, cooking, or shopping—integrate learning naturally into daily life. These everyday practices, even when unstructured, have been positively linked to children’s mathematical abilities. Informal learning is highly influenced by race/ ethnicity, family income level, and parent’s education level. The value of the community or educator is embedded in daily lives, determining the frequency and attitudes of educational engagement for parents (Leyva et al, 2017). Also given a similar degree of maternal directiveness, the children’s outcome is different by ethnicity, suggesting the different acceptive attitudes that children have among diverse ethnicities (Leyva et al, 2017). Bermudez et al (2023) interviewed the Latin community and discovered that their cultural values shape STEM learning environments. These values are embedded in common everyday activities, such as cultural games, meal routines, and outdoor practices tied to their cultural heritage. Furthermore, the role of parents in fostering intellectual development is crucial. Research has shown that parenting styles, the types of questions parents pose, their engagement levels, and attitudes toward their children are key contributors to academic outcomes. For instance, Haden et al. (2014) emphasized the effectiveness of open-ended Wh-questions—such as why, how, and when—in enhancing children’s comprehension, memory, and ability to recall experiences. Moreover, Bingham et al. (2017) explored the relationships among parenting approaches, home literacy environments, and children's language development. They found the positive influence of authoritative parenting in contrast to the authoritarian style. Similarly, Caspe (2009) compared three types of maternal book-sharing styles of low-income Latino mothers and found that maternal educational style affects children's literacy longitudinally. As there is increasing attention on informal learning environments combined with family circumstances and structure and recent studies try to implement them into the daily lives of students, it is necessary to examine factors that consist of informal learning and influence children’s cognitive outcomes. Because of the significance of parental style and involvement, as they pass cultural values and help children to shape their learning agency, it is needed to understand effective activities between parents and children in the learning process in daily settings. Some previous research had implemented experiment models, setting one experiment group with a certain treatment and the other control group(s) without the treatment. This allowed them to observe the differences in results of children’s academic development between the experiment group and control group, with statistical tools like t-test or ANOVA(Analysis of Variance). Cohrssen and Niklas (2019) compared the intervention group with NT Preschool Math’s Games and the control group, by ANCOVA( Analysis of Covariance) and found that play-based games for mathematical concepts may enhance the child's mathematical skills. Haden et al (2014) used ANOVA to test differences among groups that received different instructions in museum educational settings. Aram et al (2013) asked the intervention parents group to read books to a child with guidance, while they let the control group read without guidance. The academic results of each group child were measured to check if they statistically differed from MANOVA, ANOVA, and t-test. Domitrovich et al (2012) conducted one-way ANOVA to check demographic similarities and compare the mathematic skills of the two preschool groups, one of which had a two-year program and the other with one year of the program. Gao and Want (2024) classified three groups with different parental and grandparental sensitivity and conducted ANOVA and t-test to compare each child’s emotional status in the learning process. It was noted that there is a positive correlation between caregiver’s sensitivity and children’s emotions. In AI technology fields, they made great strides in improving the accuracy of prediction. Various AI technologies were modeled and refined to make accurate predictions based on variables. As it had been a long-term aim to achieve high accuracy metrics, including F1, precision, and recall, many in the fields have been creative attempts to combine several AI tools for the best prediction. Combined with the educational topics, data mining, so-called educational data mining has been attempting to facilitate knowledge discovery. Different resampling techniques including ANN, KNN, SVM, and XG-Boost were compared to gain a higher accuracy indicator in predicting students’ performance(Coleman et al,2019). Hu and Song(2023) used the XG boost algorithm to classify students' scores and evaluate students' performance. Additionally, Cheng et al (2024) further employed the XG-Boost Classifier-EAEO hybrid model for classifying academic performance. The combined model proved to be the most efficient method with excellent precision and fast computational speed. Even though they succeeded in making high accuracy in the prediction of student performance, it is notable that there are limited approaches bridging the strong predictive model and the interpretation of the important educational contributors. 1.3 Contribution of this study The research focuses on home environments reflected in demographic factors and the way that parents interact with children during everyday activities, as the major components of informal learning conditions. The research addresses significant gaps in the existing literature by leveraging big data and advanced AI methodologies to analyze the role of informal learning activities between parents and children in children's academic achievement. Previous studies primarily relied on small or moderate-sized datasets. However, this research utilizes a large-scale dataset to conduct empirical analysis. This allows for more robust, generalizable insights into how informal learning activities between parents and children impact academic outcomes. Additionally, while traditional regression models have been widely used in prior studies for their interpretability, this research integrates cutting-edge machine learning tools that consider uncovering complex, non-linear relationships within big data. This approach seeks to improve the depth of understanding in identifying the influence of socio-economic factors and the most effective parental engagement strategies. These are the research questions that we are addressing in this paper: Can this study identify the effectiveness of parents’ engagement in informal learning for children’s mathematical academic achievement and of socio-cultural factors? Can this study effectively utilize AI methodology for the prediction of children’s mathematical academic achievement based on demographic factors and parent-children’s activities? 2. Data & Methodology Big data was sourced from public datasets, containing children’s information and parent’s interview-based data, and applied to data analysis. The AI method utilized was XG-Boost, for its high performance in predictive modeling and ability to handle complex, non-linear relationships in the data. 2.1 Data Sample The base-year sampling for the ECLS-K(Early Childhood Longitudinal Study - Kindergarten cohort) was designed to ensure a representative, precise, and detailed dataset capturing the early education experiences of U.S. children. 21,260 kindergartners were selected in the base year (1998–99), following the three main sampling stages. Because it is crucial to ensure a nationally representative sample of U.S kindergarteners. In the first stage, primary sampling units (PSUs)—comprising counties or groups of counties—were selected based on the probability proportional to the number of 5-year-olds in the area. It is noted that they adjusted oversample Asian and Pacific Islander (API) populations. The PSUs included 24 self-representing (SR) areas selected with certainty and 76 non-SR areas stratified by census region, metropolitan status, minority population percentage, and economic indicators. In the second stage, public and private schools offering kindergarten programs were systematically sampled. To account for newly opened schools, a refreshed sampling frame was created. Schools were selected with a probability proportional to the number of enrolled kindergartners, yielding 1,413 participating schools (953 public and 460 private). In the third stage, kindergarten children were sampled within schools, with API children oversampled to ensure adequate subgroup representation. Targeting approximately 24 children per school, systematic sampling was used within two strata (API and all others). Parental consent was then obtained for each selected child, completing a rigorous and inclusive sampling process designed to capture early educational experiences comprehensively. The racial and ethnic composition of the selected children consists of mostly white(51.7%), followed by black (14.1%), Hispanic (16.5%), Asian (6%) and Pacific Islander children(1.7%). The sample was distributed geographically across the four main U.S. regions, from the Northeast region there are 18.8% of the total, Midwest 24.8%, South 32.9%, and West 23.5%. Additionally, parental education levels were recorded as follows: 2,027 parents (8.9%) had less than a high school education, 5,251 parents (23.2%) were high school graduates, 5,351 parents (23.6%) had completed some college, 4,004 parents (17.7%) were college graduates, and 1,429 parents (6.3%) held a master’s degree or higher. Trained evaluators assessed children in their schools, and they interviewed parents to collect information. The parents’ interviews were primarily conducted via computer-assisted telephone interviews (CAI). Interviews were also available in other languages, predominantly Spanish, to include non-English-speaking households in the study. The parent interview provided get children’s living environment and situations, Parental Involvement and School Interaction, Home Environment and Cognitive Stimulation, Child's Health and Well-being, Parent Characteristics, Family Processes, and Parental Expectations. 2.2 Measures The ECLS-K study employed Item Response Theory(IRT) to derive scale scores for assessing children’s abilities in reading, mathematics, and science. IRT scores are representative of children’s academic achievement and ability. IRT uses a statistical model to estimate a child’s proficiency by accounting for the difficulty and discrimination of test items. This offers a more precise measure than raw scores. Specifically, there are two-stage processes in the assessment: Firstly, all children took a routing test to gauge their general ability, followed by a skill-specific test tailored to their proficiency level. Then, IRT scale scores are placed on a continuous scale, allowing comparisons across grades and assessments while reflecting developmental progress over time. These scores quantify student ability, estimate the likelihood of mastering specific skills, and provide measures of precision, such as confidence intervals. This methodology enables the tracking of academic growth and the identification of learning patterns with high reliability, making it invaluable for longitudinal studies. For kindergarteners, the IRT scores indicate mastery of essential early learning skills such as number recognition, counting, and basic arithmetic in mathematics, as well as letter recognition, phonemic awareness, and simple word reading in literacy. In mathematics, IRT scores for kindergarten represent a child’s ability ranging from recognizing numbers and shapes to solving simple word problems. Higher IRT scores suggest greater proficiency and predict for advanced complex skills. In reading, the IRT scale captures pre-reading and emergent literacy skills, such as associating sounds with letters and reading simple sentences. These scores also predict future success in more advanced reading comprehension and critical thinking. The IRT scores for kindergarten are important as they serve as a baseline for longitudinal studies, enabling researchers to track children’s developmental trajectories since the first observed period. The longitude study allows us to analyze factors influencing growth in mathematics and reading cognitive capability over time. 2.3. Methodologies The study aims to pinpoint the relationship between parent-child interaction, demographic factors, and academic achievement and explore the possibility of AI methods to forecast students’ performance regarding information and their informal learning conditions. For this purpose, it is essential to preprocess the dataset, specially making balanced data. The research process began with an in-depth analysis of potential relationships between input and output variables using data visualization techniques. A scatter plot is a graphical representation used to visualize the relationship between two numerical variables, that reveals patterns and trends in the dataset and an initial understanding of correlations and interactions among variables. Additionally, Points that are far from the general pattern can be easily spotted as outliers. This allows informed subsequent analyses. The present study employed an XG-Boost model to further predict the academic outcomes regarding the demographic and activity variables. As a robust ensemble learning method, XG-Boost is well-suited for handling complex datasets and interactions. The performance of this algorithm is proven to be better than other machine learning algorithms such as ANN, SBM, and logistic algorithms for the prediction of students’ grades by previous studies (Cheng et al& Hu and Song). The study aimed to balance interpretability from visual data analysis and the predictive power of AI, offering a comprehensive understanding of the factors influencing children’s academic success. 3. Results 3.1 Scatter Plot Secondly, the present research drew the scatter plot of IRT Mathematical Academic Score and Reading activities with parents, to examine the associations between two variables. The IRT scores ranged from 10.96 to 93.23, and the average was 27.478 with a standard deviation of 9.423. The research encoded 0 for the parent's answer that they don’t do the certain activity with their child at all, and 1 for the answer when parents do the activity at least once a week. Since there is a large disproportion between group 0 (answered ‘not at all’) and group 1(answered ‘more than once a week) for the activities between parents and children, the code created a balanced dataset. As the visualization Fig. 1 suggests, when children had time with parents reading books (P1READBO), more people showed better performance in the Math grade. It is suggested that there is a possible positive relationship between parental activity involvement and children’s performance. Children who participate in the activity tend to achieve higher scores on average, as evident from the clustering of higher scores around group 1.0(who have parental activity with their parents at least once a week). The average score of Group 1 was 26.611235, compared to Group 2 of 19.904831. There is also a wider range of scores for group 1 as evident with higher standard variation value. Group 1 attained a higher standard deviation of 9.4204, compared to Group 2 standard deviation of 6.406. A notable number of children with high scores (close to or above 50) are associated with active parental activities, indicating that such activities could contribute to improved performance. However, the activities are predictors or determinants of academic performance. Since there are still many children who have such parental activities, but still score below the average score of 27.478. Children with no participation in such activities failed to achieve high scores, emphasizing that parental activity could be a necessary condition for better performance but not sufficient by itself to guarantee mathematical success. And the tendency is also the case for other activities including telling stories to the child, singing songs to the child, helping the child with arts and crafts, involving the child in household chores, and playing games or puzzles with the child. More children from group 1 who spend quality time with their parents doing such activities attained higher IRT scores, pinpointing the positive influence of these quality times with parents. However, it is notable that the difference in average score was largest when it comes to reading book activities. Additionally, the difference in standard deviation was small for the other activities. This can have two aspects. Firstly, reading books to children is such a common parental behavior that most parents caring for their children would engage in daily life interactions. It can be assumed that if parents don’t read books to their children at all, the parents spend way less quality time with their children, diminishing the chances for children to be exposed to educational informal resources. Secondly, reading books plays a critical role in providing opportunities for the enhancement of literacy for kindergarteners. The influence of activities such as talking about nature or science projects(P1NATURE), building or playing with construction toys, and doing sports(P1SPORT) appears to show a weaker positive tendency compared to other activities above. The differences in averages and standard deviation values are rather diminished, compared to the activities above. To be specific, group 0 and group 1 for nature experiences have an average of 26.10 and 27.75, respectively. Also, the standard deviation was not large enough to convey big amounts of difference. Examining scatter plots in depth, the research spotted unlike other activities a noticeable number of children who succeeded in achieving high scores above 50, despite the absence of parental engagement. A large proportion of children scored above average in both children groups, possibly indicating that these types of activities have a less direct or weaker influence on general academic performance. Still, the activities play as a predictor for the highest academic performance, as only the children who talked about nature scored above 80, whilst the other children failed to achieve higher than 80. Children without the activities can successfully achieve higher than average but are limited to gaining the highest score ranges. These activities are rather topic-specific, as the topic-specific talks about nature projects or building construction toys are related to explicit scientific topics. The research investigates parental care and attention by examining how often parents read picture books to their children. The number indicates the frequency with which parents read picture books to their children: 1 means \"Not at all,\" 2 means \"Once or twice a week,\" 3 means \"3 to 6 times a week,\" and 4 means \"Every day.\" More children who have parents more spending time, reading picture books to their child can possibly have great scores. Specifically, among the groups who scored above 60, there is a strong positive relationship between the frequency of parental reading and academic outcomes. Again, the frequency alone cannot be the sole predictor for the children’s academic outcomes, as a significant proportion of children who experienced frequent reading interactions with their parents still scored below average. This suggests that while frequent parental interaction is a necessary condition for achieving higher scores, it is not sufficient on its own. Other factors likely contribute to children’s academic performance as well. Also, the informal learning environment is examined, whether children have someone, other than their biological or adoptive mother, who is like a mother to them. 1 indicates they have whilst 2 indicates they don’t have. The spot plot shows there are no significant differences between children who have some like a mother or not, indicating that the closer primary caretaker is not a necessary condition for the excellent scores. Furthermore, the influence of financial factors including household income(W8INCCAT) and poverty level(WKPOV_R) is investigated. Firstly children below the poverty threshold (encoded as 1.0) failed to exceed the score level of 55. this is contrasted by the children who are not in poverty, as shown in the plot, where half of them achieved higher than average scores. When broken down into detailed brackets of income level, a positive relationship is observed between income levels and scores. Higher-income households are associated with more children achieving scores above 100, particularly in the $ 75,001- $ 100,000(encoded as 11) and $ 200,001 or more brackets(encoded as 13). Incomes categorized as 10 through 13, which is mid-to-upper income level display a broader spread of high scores, indicating better academic performance is achieved by children from wealthier households. For lower-income brackets (categories 1 to 5), scores are concentrated between 20 and 80, with very few children achieving scores above 100. This suggests that limited financial resources may hinder academic performance. Notably, a moderate-income level is necessary for achieving higher grades, but not sufficient condition. While most children who scored above 80 came from mid-to-upper income families, a significant proportion of children within these income brackets also recorded low scores. While high-income households provide a better informal environment and support, it does not guarantee academic success. The variability in middle- and lower-income groups suggests the influence of other social, educational, and environmental factors in shaping outcomes. This emphasizes that family income level alone is not the sole determinant of academic performance, and additional support systems may be critical for fostering success across income levels. Also, the education level(W8PARED) of parents seems to be functioning for helping children gain good IRT scores. As shown in the scatter plot, there is a clear distinction in the score ranges between children whose parents have less than a high school diploma and those whose parents have at least a high school diploma. Children whose parents did not complete high school failed to achieve scores higher than 60. In contrast, children whose parents attained an above or equivalent high school level of education exhibited a broader distribution of scores, with some achieving notably high scores. The group of parents with the education levels can be further divided into those below the college level (encoded as 3 and 4) and those with a college degree or higher (encoded as 5 or above). Despite slight differences, there is a pattern that children whose parents with college-level education or higher achieve higher scores more frequently than the other groups. Within these higher education groups, however, academic outcomes show little variation. It is inferred that once parents achieve a certain educational level, such as a high school diploma or college degree, it functions as a threshold for determining the highest score that children may achieve. Additionally, other factors may become more significant in shaping their children's academic success within each educational threshold group. Lastly, this research attempted to capture the effects of children’s race on the children’s academic success as race and nationality plays a large role in shaping the learning and cognitive agency. White, non-Hispanic children, who make up the largest proportion (25.6%), exhibit a wide distribution of scores. Due to the highest population, many of them also exceed excellent above 60. Black or African American children, a smaller group (3.88%), tend to have scores concentrated below the total average, with fewer high-performing individuals. Hispanic children show a clustering of scores near the total average, with relatively few high-scoring outliers compared to White children. Asian children, though representing only 2.13% of the sample, show strong academic performance, with the highest average of 31.08 among racial groups. In contrast, Native Hawaiian or Other Pacific Islander children (0.42% of the sample) have scores predominantly low average of 23.95, with limited variability. Similarly, American Indian or Alaska Native children (0.86%) exhibit scores average of 22.76, with few high-scoring individuals. 3.2 XG-Boost The study considered various activities (reading books, telling stories, singing songs, helping with arts and crafts, involving children in household chores, playing games or puzzles, discussing nature or science projects, building or constructing toys, and engaging in sports or exercise) as well as income level, parents’ highest education level, children’s race. The informal learning conditions were taken to predict the children’s IRT score as the output variable. As indicated in the scatter plots that show the distributions of scores according to the demographic, or parental activities, there is no linear regression between factors and IRT scores. Therefore, linear regression and logistic regression, which is grounded on linear relationship between factors are not suited for the given datasets, implying complex and non-linear associations between factors. XG-Boost, thus, was implemented for the classification of the children’s IRT scores whether they achieved higher than average scores or not. In classifying values, the focal point was to examine how effectively the model could predict whether a child would succeed academically, based on whether they achieved a higher-than-average score. The model’s accuracy was 65.87%, meaning that the model forecast two-thirds of children’s score accurately. The is showing a reasonable performance in predicting academic success. The other performance measure also shed light on the fair predictive power of the model, with recall (63.61%) and F1 score (60.08%). The relatively low precision (56.93%) suggests the model is prone to generating false positives. In conclusion, these metrics collectively suggest that the model performs moderately well, but some refinement would be needed for higher predictive accuracy and reliability. Additionally, XG-Boost provided metrics for evaluating the importance of input features, allowing us to identify the factors that had the greatest influence on predictions. By analyzing feature importance, the underlying mechanisms of XG-Boost are partially revealed. The most important features were parents’ educational level, family poverty, and income levels, with feature importance scores of 0.1459, 0.1288, and 0.1027, respectively. Race, particularly whether individuals were Hispanic or Black, also contributed moderately to the predictions, with feature importance values of 0.0997 and 0.0633. Parental activities, on the other hand, showed relatively low importance. For instance, reading books had a feature importance of 0.0139, and helping parents with chores scored 0.0206. Everyday activities displayed minimal influence on predictions, which may be attributed to high multicollinearity with other key features. For example, reading books exhibited an extremely high VIF of 58.44, indicating severe multicollinearity with poverty level, parent’s educational level, and other related activities. 4 Discussion In conclusion, this research explored the role of parental involvement in daily activities for informal learning and its impact on children's academic achievement. The activities studied ranged from verbal interactions, such as storytelling and singing songs, to home-related tasks like involving children in chores, and outdoor activities such as playing sports. Data visualization, and scatter plots, provided insights into the association between these activities and children’s academic performance. While these activities were not definitive predictors of academic success, they played a role in supporting higher academic scores. The children who gained exceptionally high scores were those who spent time with their parents engaging in activities like storytelling, reading books, singing songs, doing household chores, and playing games. In contrast, children who did not engage in these activities with their parents exhibited a dense cluster of scores around the average, failing to achieve higher academic outcomes. For other activities, such as talking about nature, engaging in science projects, or playing with constructive toys, they seemed to be less significant factors. Although some children who participated in these activities achieved higher scores, others who did not engage in them also managed to perform well. The results are in line with previous research as they The research also examined the importance of the frequency of parental involvement and academic performance. A scatter plot analyzing the frequency of reading books with parents and IRT scores revealed that frequent interactions were a necessary condition for higher scores but not a sufficient one. The findings highlight the importance of informal educational practices which involve parents in their children’s everyday activities. The time spent together empowers children to acquire the knowledge of various literacy, numeracy accountings, and mathematical thinking. These practices help children to achieve higher scores, but they are not definitive predictors of academic success, as a large proportion of children did not achieve high scores despite frequent parental involvement. If other factors, such as parental attitudes, school environments, study motivation, time spent on study, etc. are applied, it is expected to give a better understanding of academic outcomes. Demographic factors were found to play a similar role in influencing academic achievement. The study investigated the effects of family income, poverty level, parental education, and children’s race. For example, children from families living in poverty tended to have scores clustered around the average, whereas children from non-poverty families exhibited a wider range of scores, some of them achieving the highest. When examining in detail- family income level, children scoring above 40 came from mid-to-upper income families, meaning that a moderate level of income is critical for a high level of mathematical performance. However, a significant proportion of children within these income brackets also recorded low scores. Additionally, children scoring above 60 generally had parents with education levels beyond college, indicating whether parents have a moderate level of education plays a critical role. Differences in score distributions were also observed across racial groups. White, non-Hispanic children showed a wide range of scores, many of which exceeded the average. In contrast, Black or African American children’s scores were more concentrated below the average, while Hispanic children’s scores tended to cluster around the average. Societal, economic, and cultural factors clearly influence academic outcomes. However, those factors are not sufficient conditions to conclude the academic achievement of children. Therefore, it is necessary to find out what differentiates children’s outcomes within the same societal or racial groups. The study explored the application of advanced machine learning techniques, such as XG Boost, to predict academic outcomes. While these models are moderately accurate predictions, unlike the expectation, parental engagement played a nonsignificant role in determining the academic achievement of children. It can be inferred that the parents- children’s interactive activities are already highly multi-correlated with other demographic factors. To achieve higher accuracy in prediction, it is still needed to add other explanations to capture the complexities of academic achievement for the performance of the model. Future research should consider additional learning conditions, that would determine children's achievement among the same group of income level, race, or poverty level. Those factors are expected to enhance the predictive and explanatory power of these models. 5 Limitation This study emphasizes the influence of demographic factors on children’s academic performance and investigates the role of the way parents and children interact. While these factors provide valuable insights, improving the predictive capabilities of AI models requires other additional variables. Factors such as school environments, and access to resources, time spent on study, driven motivations, parental attitude, interactions with teachers could significantly enhance the explanatory power of these models. Furthermore, as classification methods proved more effective than regression in this study, future study might explore other decision tree-originated algorithms to classify outcomes into multiple categories rather than limiting analyses to binary classifications. This approach could offer a more nuanced understanding of how various factors contribute to different levels of academic achievement, ultimately supporting more targeted interventions. Declarations Author Contribution Naeun wrote the main manuscript text and Prof. Biswajit Sarkar drew main methodologies. All authours reviewed the manuscript References Adams, M., Rodríguez, L., & Smith, J. (2021). Designing frameworks for authentic equity in science teaching and learning: Informal learning environments and teacher education for STEM. Journal of STEM Education, 22 (3), 27–38. Alsabbagh, M., & Ibrahim, A. (2024). A hybrid model for the prediction of electrical energy consumption using hybrid LSTM and ML regressors. Energy Reports, 10 , 289–301. Aram, D., & Levin, I. (2001). Mother–child joint writing in low SES: Sociocultural factors, maternal mediation, and emergent literacy. Reading and Writing: An Interdisciplinary Journal, 14 (3), 213–233. Bachman, H., &Heather J. Bachman& Leanne Elliott &Shirley Duong & Laura Betancur& Monica G. Navarro& Elizabeth Votruba-Drzal& Melissa Libertus. (2020). Triangulating multi-method assessments of parental support for early math skills. Early Childhood Education Journal. Baroody, A. E., & Diamond, K. E. (2012). Measures of preschool children’s interest and engagement in literacy activities: Examining gender differences and construct dimensions. Barton, A. C., Tan, E., & Rivet, A. (2021). Co-designing for rightful presence in informal science learning environments. Journal of Research in Science Teaching, 58 (2), 145–168. Beard, C., Richardson, R., & Davison, P. (2024). Synergy between formal and informal education. Journal of Educational Research, 36 (4), 415–429. Beck, I. L., & McKeown, M. G. (2007). Increasing young low-income children's oral vocabulary repertoires through rich and focused instruction. The Elementary School Journal, 108 (4), 271–287. Bermudez, A., &Salazar, J &Garcia, L& Ochoa, K.D.& Ahn, J.& Bustamante, A.S. &Pesch, A. &Hirsh-Pasek, & Roldan, W. & Soto-Lara, S.& Gomez, W.& Rodriguez, R.( 2023). Designing culturally situated playful environments for early STEM learning with a Latine community. Journal of Early Childhood Education. Bingham, G. E., Pendergast, M. L., & Ross, R. (2017). Parenting styles and home literacy opportunities: Associations with children’s oral language skills. Early Education and Development, 28 (7), 894–912. Bird, Y. J., Murphy, K. R., & Martinez, D. S. (2023). Talking, praising, and teaching: How parent interaction during a learning task relates to children's early learning. Early Childhood Research Quarterly, 58 , 109–120. Carmiol, A., Rodríguez, P., & Vélez, L. (2021). Book sharing and reminiscing: Caregivers' conversational style and children's language and literacy development in low-income Costa Rican families. Journal of Early Childhood Literacy, 21 (4), 534–550. Caspe, M. (2009). Low-income Latino mothers’ booksharing styles and children's emergent literacy development. Early Childhood Education Journal, 37 (2), 97–106. Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , 785–794. Chu, S., Martin, Rachel &Quek, Francis& Berman, Alexander& Suarez, Mario &Zarei, Niloofar &Nam, Beth &Banigan, Colin (2017). Becoming makers: Examining ‘making’ literacy in the elementary school science classroom. Journal of Research in Science Teaching. Cheng, L., Wang, Z., & Zhang, Y. (2024). Evaluation of students’ performance during the academic period using the XG-Boost Classifier-Enhanced AEO hybrid model. Journal of Educational Data Mining, 12 (3), 121–135. Clements, D. H., & Sarama, J. (2024). Systematic review of learning trajectories in early mathematics. Mathematics Education Research Journal. Cohrssen, C., & Church, A. (2013). Purposeful pauses: Teacher talk during early childhood mathematics activities. Early Childhood Research Quarterly. Cohrssen, C., & Niklas, F. (2019). Using mathematics games in preschool settings to support the development of children’s numeracy skills. Early Education and Development. Combs, S., & Higgins, S. (2023). The relationship between shared picture book reading and language development in young children. Journal of Research in Childhood Education, 37 (2), 192–205. Cotto, J. A,, & Ribner, Andrew & Smith, Leann (2024). Understanding working memory and mathematics development in ethnically/racially minoritized children through an integrative theory lens. Behavioral Sciences (2076-328X). May2024, Vol. 14 Issue 5, p390. 20p. Deitcher, D., DeBaryshe, B. D., & Fujita, K. (2019). Alphabet books: Relations between aspects of parent-child shared reading, children’s motivation, and early literacy skills. Journal of Early Childhood Literacy. Dhiya'ulhaq, M. A., Sucipto, A., & Purnomo, M. A. (2024). Ocean wave prediction using Long Short-Term Memory (LSTM) and Extreme Gradient Boosting (XGBoost) in Tuban Regency for fisherman safety. Ocean Engineering, 242 , 110358. Ethridge, E. A., & King, J. R. (2005). Calendar math in preschool and primary classrooms: Questioning the curriculum. Early Childhood Education Journal, 32 (5), 291–296. Folkestad, G. (2006). Formal and informal learning situations or practices vs formal and informal ways of learning. European Educational Research Journal, 5 (1), 81–95. Fyfe, E. R,& Evans, Julia L. &Matz, Lauren Eisenband &Hunt, Kayla M. &Alibali, Martha W. (2017). Relations between patterning skill and differing aspects of early mathematics knowledge. In Cognitive Development October 2017 44:1-11 Ginsburg, H. P., & Lee, Joon Sun &Boyd, Judi Stevenson (2008). Mathematics education for young children: What it is and how to promote it. Society for Research in Child Development . 2008. Gong, J., Liu, H., & Zhang, Y. (2021). Informal learning in nature education promotes ecological conservation behaviors of nature reserve employees: A preliminary study in China. Global Ecology and Conservation, Vol 31, Iss , Pp e01814 Haden, C. A., Jant, Erin A. &Hoffman, Philip C. &Marcus, Maria &Geddes, Jacqueline R. &Gaskins, Suzanne (2014). Supporting family conversations and children’s STEM learning in a children’s museum. In Early Childhood Research Quarterly Q3 2014 29(3):333-344 Jodi K. Heidlage &Jennifer E. Cunningham & Ann P. Kaiser &Carol M. Trivette & Erin E. Barton &Jennifer R. Frey The effects of parent-implemented language interventions on child linguistic outcomes: A meta-analysis . Early Childhood Research Quarterly Volume 50, Part 1, 1st Quarter 2020, Pages 6-23 Megan Y. Roberts e (2020). The effects of parent-implemented language interventions on child linguistic outcomes: A meta-analysis. In Early Childhood Research Quarterly Q1 2020 50 Part 1:6-23 Hu, J., & Song, W. (2024). Research on XGBoost academic forecasting and analysis modelling. Journal of Physics: Conference Series, 14 October 2019, 1324(1)) Hussim, H., &Roslinda Rosli& Nurul.N & Siti.M & Muhammad.M& Zanaton.I& Azmin.R&Siti,M& Lilia.H &Kamisah.O & Ah .Lay (2024). A systematic literature review of informal STEM learning. European Journal of STEM Education. 2024 9(1) Kelly, L., Jones, S., & Brown, C. (2023). Smartphones and parent-child conversations during young children's informal science learning at an aquarium. In Computers in Human Behavior Reports May 2023 10 Khan, A., & Rodrigues, S. (2017). STEM for girls from low-income families: Making dreams come true. The Journal of Developing Areas, 2017 Apr 01. 51(2), 435-448. Leyva, D. (2019). How do low-income Chilean parents support their preschoolers’ writing and math skills in a grocery game? Early Education and Development. 2019 30(1):114-130. Leyva, D., Cates, C. B., & Yoshikawa, H. (2019). Maternal behaviors in a grocery game and first graders’ literacy and math skills in a low-income sample. Elementary School Journal. Jun 2019 119(4):629-650. Leyva, D., Tamis-LeMonda, C. S., Yoshikawa, H., & Cates, C. B. (2017). Grocery games: How ethnically diverse low-income mothers support children’s reading and mathematics. In Early Childhood Research Quarterly Q3 2017 40:63-76 Levin, I., & Aram, D. (2010). Mother-child joint writing as a learning activity. Reading & Writing; Jan2012, Vol. 25 Issue 1, p217-249, 33p Nancy, L., et al. (2020). Interested, disinterested, or neutral: Exploring STEM interest profiles and pathways in a low-income urban community. EURASIA Journal of Mathematics, Science and Technology Education. 2020 16(6). Nora, A., Callanan, M., & Jipson, J. (2010). Enhancing building, conversation, and learning through caregiver–child interactions in a children’s museum. Developmental Psychology. Mar 2010 46(2):502-515. Orosoo, M., Norouzi, M., & Behzad, M. (2025). Transforming English language learning: Advanced speech recognition with MLP-LSTM for personalized education. Alexandria Engineering Journal, Vol 111, Iss , Pp 21-32 (2025) Pace, A., Albright, L., & Mitchell, R. (2017). Identifying pathways between socioeconomic status and language development. Annual Review of Linguistics; 2017, 3 p285-p308, 24p. Pentimonti, J. M., Justice, L. M., & Piasta, S. B. (2012). A standardized tool for assessing the quality of classroom-based shared reading: Systematic Assessment of Book Reading (SABR). Early Childhood Research Quarterly. 3rd Qtr 2012 27(3):512-528. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-5885890\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":411490439,\"identity\":\"68a2e4f3-3d59-46dc-afc5-f02e007ab87b\",\"order_by\":0,\"name\":\"Na-Eun Lee\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7ElEQVRIie3QMQrCMBSA4ZQH6RLbtaXqGSIBR72KpeDUzUmEOgh18QIO3sHpzUrALh6g4GLt7NBB0M1UJ5e2boL5h0BCPsgLITrdL5YQIIyQLgXYn9WeWbVEvomwzTjgJaFNib9eHftOeVBLLGnm/BpGxjYd9ae3cNCmBLJLWkFcyYS/QQk8HY1PHQzUw6gQYQXhkoFs4Y4qcji5CIow6lUTM1ckYjz144mL8yaEiKCF4LgrSY0CZT0pZ+mpWbj6ZPAMTBiFmlmsZJk7V4zmMdhF8cDZ0DYXWV5FPgL2WpteLzPu39zW6XS6v+kJET1G/YarVTQAAAAASUVORK5CYII=\",\"orcid\":\"\",\"institution\":\"Yonsei University\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Na-Eun\",\"middleName\":\"\",\"lastName\":\"Lee\",\"suffix\":\"\"},{\"id\":411490440,\"identity\":\"d597fc85-0c95-4e99-9603-7ccd02651126\",\"order_by\":1,\"name\":\"Biswajit Sarkar\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Yonsei University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Biswajit\",\"middleName\":\"\",\"lastName\":\"Sarkar\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-01-23 07:08:29\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-5885890/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-5885890/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":75996862,\"identity\":\"2f7ffe99-dc0f-4d69-bd0f-57adc8c9c103\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:54:54\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":50582,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003e\\u0026lt;Reading books\\u0026gt;\\u003c/strong\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/de640c04ef8a6e1e273ba7d2.png\"},{\"id\":75995166,\"identity\":\"421a8f49-4359-4339-a251-462f6e86d60b\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:55\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":50305,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u0026lt;Singing songs\\u0026gt;\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/3f676b13188de10e0ec2ef10.png\"},{\"id\":75995129,\"identity\":\"f556f34e-1162-4ddc-b18a-9973b0b4522a\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:53\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":47104,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003e\\u0026lt;Playing games\\u0026gt;\\u003c/strong\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/6e465c92ee4c64775b540ba2.png\"},{\"id\":75995168,\"identity\":\"5a16ae70-f77d-4be2-a218-99abab6d176b\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:55\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":49290,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u0026lt;Helping chores\\u0026gt;\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/6e523678f7277e449e5689c7.png\"},{\"id\":75995169,\"identity\":\"ef2c8493-f730-4a2d-9cc9-e8b7d5ccaaf1\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:55\",\"extension\":\"png\",\"order_by\":5,\"title\":\"Figure 5\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":55014,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003e\\u0026lt;Nature projects\\u0026gt;\\u003c/strong\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/ec47eb8ccc88e176cd23d556.png\"},{\"id\":75995140,\"identity\":\"20f16559-58c3-41cf-a0b1-3cfec2e58646\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:54\",\"extension\":\"png\",\"order_by\":6,\"title\":\"Figure 6\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":53325,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u0026lt;Doing sport\\u0026gt;\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"6.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/ff301d67aad50ff68bf80bfb.png\"},{\"id\":75995137,\"identity\":\"1d6cffdb-9cb1-4087-9a42-a45a1159511d\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:54\",\"extension\":\"png\",\"order_by\":7,\"title\":\"Figure 7\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":109334,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u0026lt;frequency of reading picture books\\u0026gt;\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"7.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/48d127953ca40f4b50737116.png\"},{\"id\":75996863,\"identity\":\"220170b5-a8a7-4df4-abd6-4848683f05dd\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:54:54\",\"extension\":\"png\",\"order_by\":8,\"title\":\"Figure 8\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":54698,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003e\\u0026lt;Poverty level\\u0026gt;\\u003c/strong\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"8.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/5ee37cdd2b13b473e68515bb.png\"},{\"id\":75995132,\"identity\":\"d2b8105a-24a3-4ea2-ae44-e47af30f6691\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:54\",\"extension\":\"png\",\"order_by\":9,\"title\":\"Figure 9\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":103873,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u0026lt;Income level\\u0026gt;\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"9.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/3d988cc6f71c619b6956733f.png\"},{\"id\":75995133,\"identity\":\"7a7c7d85-674f-4750-8c36-2d45cbdc8e73\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:54\",\"extension\":\"png\",\"order_by\":10,\"title\":\"Figure 10\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":103689,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003e\\u0026lt;Highest education of parents\\u0026gt;\\u003c/strong\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"10.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/63451eb2d936ab9f13108379.png\"},{\"id\":75995145,\"identity\":\"f353e902-d5bc-40d9-a732-77019396c3fa\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:54\",\"extension\":\"png\",\"order_by\":11,\"title\":\"Figure 11\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":57432,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u0026lt;someone like mother\\u0026gt;\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"11.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/27c8a5dd5ecad31dd98902a7.png\"},{\"id\":75995148,\"identity\":\"1a8249f1-2d23-4090-b1bc-a62485f5b182\",\"added_by\":\"auto\",\"created_at\":\"2025-02-11 09:46:54\",\"extension\":\"png\",\"order_by\":12,\"title\":\"Figure 12\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":184281,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u0026lt;Children’s race\\u0026gt;\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"12.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/56698eeee949f6b69455fe70.png\"},{\"id\":82757554,\"identity\":\"3aef0a70-7e19-4197-9a7b-a090e8269178\",\"added_by\":\"auto\",\"created_at\":\"2025-05-15 01:53:30\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":2181928,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-5885890/v1/6dde1758-7b03-4c2e-b753-eced4912c83f.pdf\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"How Interaction with Parents Affects Children's Mathematical Informal Learning and Verbal Skills\",\"fulltext\":[{\"header\":\"1. Introduction\",\"content\":\"\\u003cp\\u003eInformal learning refers to a type of learning that occurs outside of formal educational settings, like schools or training programs. Traditionally, we had considered learning to happen in school, kindergarten, and at home with conscious exercise and training(Folkestad (2006)). And this was expected to yield formal component of results such as grades, test scores and certificates. In contrast, informal learning refers to when the learning process takes place without the person being aware of it. This concept has gained growing attention these years. Such informal components consist of cultural backgrounds, extracurricular pursuits (Beard et al (2024))and communication within a belonging community. There are major characteristics in regard to process, location and setting, purpose, and content. According to Gong et al (2021), informal learning occurs incidentally during daily activities, often without any deliberate intent. People may participate in a learning process unknowingly. Additionally, informal learning is flexible, as it happens without strict limitations on time or form. It can also be an unintended byproduct of other actions. Rather than offering structured courses or formal training, it provides practical knowledge and insights.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec2\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003e1.1 Research Gap\\u003c/h2\\u003e \\u003cp\\u003eData analysis has been necessary for quantitative research, providing objective understanding based on statistics. And there have been many attempts to understand children\\u0026rsquo;s academic achievement, based on regression analysis and ANOVA(Analysis of Variance). Because Regression analysis provides the interpretability in correlation between variables, it can explain which factors and to what extent factors change academic performance. It is noted that there is not much research applying AI methods for the identification of informal learning environments. Remarkably, there is limited previous research investigating the contributions of informal learning environments combined with parental involvement.\\u003c/p\\u003e \\u003cp\\u003e**Research gap 1; There is limited research analyzing the informal learning of activities between parents and children, that uses big dataset for empirical research,\\u003c/p\\u003e \\u003cp\\u003e**Research gap 2; There are limited attempts to utilize advanced AI tools for analyzing the informal learning of activities between parents and children.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec3\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003e1.2 Literature review\\u003c/h2\\u003e \\u003cp\\u003eInformal learning has been acknowledged and there have been many attempts to identify informal learning processes in student\\u0026rsquo;s daily lives and to implement informal learning into children\\u0026rsquo;s learning processes. As Kelly et al (2023) mentioned, visiting and experiencing aquariums, museums, and zoos provides helpful educational resources to children and families. These activities in science-related places are considered to be one of the STEM informal education learning processes. In particular, many pedagogical instructions were suggested in the setting of museums to deliver engineering content. Also, daily grounded activities such as grocery shopping with caretakers are considered to promote interactions between adults and children. Moreover,(Caspe's,2009) book-sharing activity is regarded to contribute to building communication processes, social relationships, and interaction. Much research has examined informal learning within various activities and points out that informal learning practice affects children\\u0026rsquo;s learning performance. LeFevre et al (2009) identified home numerical experiences are informal learning practices. Direct activities in numeracy development are explicitly designed to enhance quantitative skills. Contrarily, indirect activities such as playing cards or board games with numerical elements, cooking, or shopping also contribute significantly. It was found that indirect practices of daily life are positively correlated with child math performance.\\u003c/p\\u003e \\u003cp\\u003eInformal learning practices have been widely studied across various activities, revealing their significant impact on children\\u0026rsquo;s academic performance. For instance, LeFevre et al. (2009) explored the concept of home numerical experiences as a form of informal learning. They classified such activities into direct and indirect approaches. Direct practices, like targeted numeracy exercises, aim explicitly at developing quantitative skills. On the other hand, indirect activities\\u0026mdash;such as playing numerically-themed card or board games, cooking, or shopping\\u0026mdash;integrate learning naturally into daily life. These everyday practices, even when unstructured, have been positively linked to children\\u0026rsquo;s mathematical abilities.\\u003c/p\\u003e \\u003cp\\u003eInformal learning is highly influenced by race/ ethnicity, family income level, and parent\\u0026rsquo;s education level. The value of the community or educator is embedded in daily lives, determining the frequency and attitudes of educational engagement for parents (Leyva et al, 2017). Also given a similar degree of maternal directiveness, the children\\u0026rsquo;s outcome is different by ethnicity, suggesting the different acceptive attitudes that children have among diverse ethnicities (Leyva et al, 2017). Bermudez et al (2023) interviewed the Latin community and discovered that their cultural values shape STEM learning environments. These values are embedded in common everyday activities, such as cultural games, meal routines, and outdoor practices tied to their cultural heritage. Furthermore, the role of parents in fostering intellectual development is crucial. Research has shown that parenting styles, the types of questions parents pose, their engagement levels, and attitudes toward their children are key contributors to academic outcomes. For instance, Haden et al. (2014) emphasized the effectiveness of open-ended Wh-questions\\u0026mdash;such as why, how, and when\\u0026mdash;in enhancing children\\u0026rsquo;s comprehension, memory, and ability to recall experiences. Moreover, Bingham et al. (2017) explored the relationships among parenting approaches, home literacy environments, and children's language development. They found the positive influence of authoritative parenting in contrast to the authoritarian style. Similarly, Caspe (2009) compared three types of maternal book-sharing styles of low-income Latino mothers and found that maternal educational style affects children's literacy longitudinally.\\u003c/p\\u003e \\u003cp\\u003eAs there is increasing attention on informal learning environments combined with family circumstances and structure and recent studies try to implement them into the daily lives of students, it is necessary to examine factors that consist of informal learning and influence children\\u0026rsquo;s cognitive outcomes. Because of the significance of parental style and involvement, as they pass cultural values and help children to shape their learning agency, it is needed to understand effective activities between parents and children in the learning process in daily settings.\\u003c/p\\u003e \\u003cp\\u003eSome previous research had implemented experiment models, setting one experiment group with a certain treatment and the other control group(s) without the treatment. This allowed them to observe the differences in results of children\\u0026rsquo;s academic development between the experiment group and control group, with statistical tools like t-test or ANOVA(Analysis of Variance). Cohrssen and Niklas (2019) compared the intervention group with NT Preschool Math\\u0026rsquo;s Games and the control group, by ANCOVA( Analysis of Covariance) and found that play-based games for mathematical concepts may enhance the child's mathematical skills. Haden et al (2014) used ANOVA to test differences among groups that received different instructions in museum educational settings. Aram et al (2013) asked the intervention parents group to read books to a child with guidance, while they let the control group read without guidance. The academic results of each group child were measured to check if they statistically differed from MANOVA, ANOVA, and t-test. Domitrovich et al (2012) conducted one-way ANOVA to check demographic similarities and compare the mathematic skills of the two preschool groups, one of which had a two-year program and the other with one year of the program. Gao and Want (2024) classified three groups with different parental and grandparental sensitivity and conducted ANOVA and t-test to compare each child\\u0026rsquo;s emotional status in the learning process. It was noted that there is a positive correlation between caregiver\\u0026rsquo;s sensitivity and children\\u0026rsquo;s emotions.\\u003c/p\\u003e \\u003cp\\u003eIn AI technology fields, they made great strides in improving the accuracy of prediction. Various AI technologies were modeled and refined to make accurate predictions based on variables. As it had been a long-term aim to achieve high accuracy metrics, including F1, precision, and recall, many in the fields have been creative attempts to combine several AI tools for the best prediction. Combined with the educational topics, data mining, so-called educational data mining has been attempting to facilitate knowledge discovery. Different resampling techniques including ANN, KNN, SVM, and XG-Boost were compared to gain a higher accuracy indicator in predicting students\\u0026rsquo; performance(Coleman et al,2019). Hu and Song(2023) used the XG boost algorithm to classify students' scores and evaluate students' performance. Additionally, Cheng et al (2024) further employed the XG-Boost Classifier-EAEO hybrid model for classifying academic performance. The combined model proved to be the most efficient method with excellent precision and fast computational speed. Even though they succeeded in making high accuracy in the prediction of student performance, it is notable that there are limited approaches bridging the strong predictive model and the interpretation of the important educational contributors.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec4\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003e1.3 Contribution of this study\\u003c/h2\\u003e \\u003cp\\u003eThe research focuses on home environments reflected in demographic factors and the way that parents interact with children during everyday activities, as the major components of informal learning conditions. The research addresses significant gaps in the existing literature by leveraging big data and advanced AI methodologies to analyze the role of informal learning activities between parents and children in children's academic achievement. Previous studies primarily relied on small or moderate-sized datasets. However, this research utilizes a large-scale dataset to conduct empirical analysis. This allows for more robust, generalizable insights into how informal learning activities between parents and children impact academic outcomes. Additionally, while traditional regression models have been widely used in prior studies for their interpretability, this research integrates cutting-edge machine learning tools that consider uncovering complex, non-linear relationships within big data. This approach seeks to improve the depth of understanding in identifying the influence of socio-economic factors and the most effective parental engagement strategies.\\u003c/p\\u003e \\u003cp\\u003eThese are the research questions that we are addressing in this paper:\\u003c/p\\u003e \\u003cp\\u003e \\u003cul\\u003e \\u003cli\\u003e \\u003cp\\u003eCan this study identify the effectiveness of parents\\u0026rsquo; engagement in informal learning for children\\u0026rsquo;s mathematical academic achievement and of socio-cultural factors?\\u003c/p\\u003e \\u003c/li\\u003e \\u003cli\\u003e \\u003cp\\u003eCan this study effectively utilize AI methodology for the prediction of children\\u0026rsquo;s mathematical academic achievement based on demographic factors and parent-children\\u0026rsquo;s activities?\\u003c/p\\u003e \\u003c/li\\u003e \\u003c/ul\\u003e \\u003c/p\\u003e \\u003c/div\\u003e\"},{\"header\":\"2. Data \\u0026 Methodology\",\"content\":\"\\u003cp\\u003eBig data was sourced from public datasets, containing children\\u0026rsquo;s information and parent\\u0026rsquo;s interview-based data, and applied to data analysis. The AI method utilized was XG-Boost, for its high performance in predictive modeling and ability to handle complex, non-linear relationships in the data.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec6\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003e2.1 Data Sample\\u003c/h2\\u003e \\u003cp\\u003eThe base-year sampling for the ECLS-K(Early Childhood Longitudinal Study - Kindergarten cohort) was designed to ensure a representative, precise, and detailed dataset capturing the early education experiences of U.S. children. 21,260 kindergartners were selected in the base year (1998\\u0026ndash;99), following the three main sampling stages. Because it is crucial to ensure a nationally representative sample of U.S kindergarteners.\\u003c/p\\u003e \\u003cp\\u003eIn the first stage, primary sampling units (PSUs)\\u0026mdash;comprising counties or groups of counties\\u0026mdash;were selected based on the probability proportional to the number of 5-year-olds in the area. It is noted that they adjusted oversample Asian and Pacific Islander (API) populations. The PSUs included 24 self-representing (SR) areas selected with certainty and 76 non-SR areas stratified by census region, metropolitan status, minority population percentage, and economic indicators. In the second stage, public and private schools offering kindergarten programs were systematically sampled. To account for newly opened schools, a refreshed sampling frame was created. Schools were selected with a probability proportional to the number of enrolled kindergartners, yielding 1,413 participating schools (953 public and 460 private). In the third stage, kindergarten children were sampled within schools, with API children oversampled to ensure adequate subgroup representation. Targeting approximately 24 children per school, systematic sampling was used within two strata (API and all others). Parental consent was then obtained for each selected child, completing a rigorous and inclusive sampling process designed to capture early educational experiences comprehensively.\\u003c/p\\u003e \\u003cp\\u003eThe racial and ethnic composition of the selected children consists of mostly white(51.7%), followed by black (14.1%), Hispanic (16.5%), Asian (6%) and Pacific Islander children(1.7%). The sample was distributed geographically across the four main U.S. regions, from the Northeast region there are 18.8% of the total, Midwest 24.8%, South 32.9%, and West 23.5%. Additionally, parental education levels were recorded as follows: 2,027 parents (8.9%) had less than a high school education, 5,251 parents (23.2%) were high school graduates, 5,351 parents (23.6%) had completed some college, 4,004 parents (17.7%) were college graduates, and 1,429 parents (6.3%) held a master\\u0026rsquo;s degree or higher.\\u003c/p\\u003e \\u003cp\\u003eTrained evaluators assessed children in their schools, and they interviewed parents to collect information. The parents\\u0026rsquo; interviews were primarily conducted via computer-assisted telephone interviews (CAI). Interviews were also available in other languages, predominantly Spanish, to include non-English-speaking households in the study. The parent interview provided get children\\u0026rsquo;s living environment and situations, Parental Involvement and School Interaction, Home Environment and Cognitive Stimulation, Child's Health and Well-being, Parent Characteristics, Family Processes, and Parental Expectations.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec7\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003e2.2 Measures\\u003c/h2\\u003e \\u003cp\\u003eThe ECLS-K study employed Item Response Theory(IRT) to derive scale scores for assessing children\\u0026rsquo;s abilities in reading, mathematics, and science. IRT scores are representative of children\\u0026rsquo;s academic achievement and ability. IRT uses a statistical model to estimate a child\\u0026rsquo;s proficiency by accounting for the difficulty and discrimination of test items. This offers a more precise measure than raw scores. Specifically, there are two-stage processes in the assessment: Firstly, all children took a routing test to gauge their general ability, followed by a skill-specific test tailored to their proficiency level. Then, IRT scale scores are placed on a continuous scale, allowing comparisons across grades and assessments while reflecting developmental progress over time. These scores quantify student ability, estimate the likelihood of mastering specific skills, and provide measures of precision, such as confidence intervals. This methodology enables the tracking of academic growth and the identification of learning patterns with high reliability, making it invaluable for longitudinal studies. For kindergarteners, the IRT scores indicate mastery of essential early learning skills such as number recognition, counting, and basic arithmetic in mathematics, as well as letter recognition, phonemic awareness, and simple word reading in literacy.\\u003c/p\\u003e \\u003cp\\u003eIn mathematics, IRT scores for kindergarten represent a child\\u0026rsquo;s ability ranging from recognizing numbers and shapes to solving simple word problems. Higher IRT scores suggest greater proficiency and predict for advanced complex skills. In reading, the IRT scale captures pre-reading and emergent literacy skills, such as associating sounds with letters and reading simple sentences. These scores also predict future success in more advanced reading comprehension and critical thinking. The IRT scores for kindergarten are important as they serve as a baseline for longitudinal studies, enabling researchers to track children\\u0026rsquo;s developmental trajectories since the first observed period. The longitude study allows us to analyze factors influencing growth in mathematics and reading cognitive capability over time.\\u003c/p\\u003e \\u003c/div\\u003e\\n\\u003ch3\\u003e2.3. Methodologies\\u003c/h3\\u003e\\n\\u003cp\\u003eThe study aims to pinpoint the relationship between parent-child interaction, demographic factors, and academic achievement and explore the possibility of AI methods to forecast students\\u0026rsquo; performance regarding information and their informal learning conditions. For this purpose, it is essential to preprocess the dataset, specially making balanced data.\\u003c/p\\u003e \\u003cp\\u003eThe research process began with an in-depth analysis of potential relationships between input and output variables using data visualization techniques. A scatter plot is a graphical representation used to visualize the relationship between two numerical variables, that reveals patterns and trends in the dataset and an initial understanding of correlations and interactions among variables. Additionally, Points that are far from the general pattern can be easily spotted as outliers. This allows informed subsequent analyses.\\u003c/p\\u003e \\u003cp\\u003eThe present study employed an XG-Boost model to further predict the academic outcomes regarding the demographic and activity variables. As a robust ensemble learning method, XG-Boost is well-suited for handling complex datasets and interactions. The performance of this algorithm is proven to be better than other machine learning algorithms such as ANN, SBM, and logistic algorithms for the prediction of students\\u0026rsquo; grades by previous studies (Cheng et al\\u0026amp; Hu and Song). The study aimed to balance interpretability from visual data analysis and the predictive power of AI, offering a comprehensive understanding of the factors influencing children\\u0026rsquo;s academic success.\\u003c/p\\u003e\"},{\"header\":\"3. Results\",\"content\":\"\\u003cdiv id=\\\"Sec10\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003e3.1 Scatter Plot\\u003c/h2\\u003e \\u003cp\\u003eSecondly, the present research drew the scatter plot of IRT Mathematical Academic Score and Reading activities with parents, to examine the associations between two variables. The IRT scores ranged from 10.96 to 93.23, and the average was 27.478 with a standard deviation of 9.423. The research encoded 0 for the parent's answer that they don\\u0026rsquo;t do the certain activity with their child at all, and 1 for the answer when parents do the activity at least once a week. Since there is a large disproportion between group 0 (answered \\u0026lsquo;not at all\\u0026rsquo;) and group 1(answered \\u0026lsquo;more than once a week) for the activities between parents and children, the code created a balanced dataset.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003eAs the visualization Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e suggests, when children had time with parents reading books (P1READBO), more people showed better performance in the Math grade. It is suggested that there is a possible positive relationship between parental activity involvement and children\\u0026rsquo;s performance. Children who participate in the activity tend to achieve higher scores on average, as evident from the clustering of higher scores around group 1.0(who have parental activity with their parents at least once a week). The average score of Group 1 was 26.611235, compared to Group 2 of 19.904831. There is also a wider range of scores for group 1 as evident with higher standard variation value. Group 1 attained a higher standard deviation of 9.4204, compared to Group 2 standard deviation of 6.406. A notable number of children with high scores (close to or above 50) are associated with active parental activities, indicating that such activities could contribute to improved performance. However, the activities are predictors or determinants of academic performance. Since there are still many children who have such parental activities, but still score below the average score of 27.478. Children with no participation in such activities failed to achieve high scores, emphasizing that parental activity could be a necessary condition for better performance but not sufficient by itself to guarantee mathematical success. And the tendency is also the case for other activities including telling stories to the child, singing songs to the child, helping the child with arts and crafts, involving the child in household chores, and playing games or puzzles with the child. More children from group 1 who spend quality time with their parents doing such activities attained higher IRT scores, pinpointing the positive influence of these quality times with parents. However, it is notable that the difference in average score was largest when it comes to reading book activities. Additionally, the difference in standard deviation was small for the other activities. This can have two aspects. Firstly, reading books to children is such a common parental behavior that most parents caring for their children would engage in daily life interactions. It can be assumed that if parents don\\u0026rsquo;t read books to their children at all, the parents spend way less quality time with their children, diminishing the chances for children to be exposed to educational informal resources. Secondly, reading books plays a critical role in providing opportunities for the enhancement of literacy for kindergarteners.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003eThe influence of activities such as talking about nature or science projects(P1NATURE), building or playing with construction toys, and doing sports(P1SPORT) appears to show a weaker positive tendency compared to other activities above. The differences in averages and standard deviation values are rather diminished, compared to the activities above. To be specific, group 0 and group 1 for nature experiences have an average of 26.10 and 27.75, respectively. Also, the standard deviation was not large enough to convey big amounts of difference. Examining scatter plots in depth, the research spotted unlike other activities a noticeable number of children who succeeded in achieving high scores above 50, despite the absence of parental engagement. A large proportion of children scored above average in both children groups, possibly indicating that these types of activities have a less direct or weaker influence on general academic performance. Still, the activities play as a predictor for the highest academic performance, as only the children who talked about nature scored above 80, whilst the other children failed to achieve higher than 80. Children without the activities can successfully achieve higher than average but are limited to gaining the highest score ranges. These activities are rather topic-specific, as the topic-specific talks about nature projects or building construction toys are related to explicit scientific topics.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e The research investigates parental care and attention by examining how often parents read picture books to their children. The number indicates the frequency with which parents read picture books to their children: 1 means \\\"Not at all,\\\" 2 means \\\"Once or twice a week,\\\" 3 means \\\"3 to 6 times a week,\\\" and 4 means \\\"Every day.\\\" More children who have parents more spending time, reading picture books to their child can possibly have great scores. Specifically, among the groups who scored above 60, there is a strong positive relationship between the frequency of parental reading and academic outcomes. Again, the frequency alone cannot be the sole predictor for the children\\u0026rsquo;s academic outcomes, as a significant proportion of children who experienced frequent reading interactions with their parents still scored below average. This suggests that while frequent parental interaction is a necessary condition for achieving higher scores, it is not sufficient on its own. Other factors likely contribute to children\\u0026rsquo;s academic performance as well. Also, the informal learning environment is examined, whether children have someone, other than their biological or adoptive mother, who is like a mother to them. 1 indicates they have whilst 2 indicates they don\\u0026rsquo;t have. The spot plot shows there are no significant differences between children who have some like a mother or not, indicating that the closer primary caretaker is not a necessary condition for the excellent scores.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003eFurthermore, the influence of financial factors including household income(W8INCCAT) and poverty level(WKPOV_R) is investigated. Firstly children below the poverty threshold (encoded as 1.0) failed to exceed the score level of 55. this is contrasted by the children who are not in poverty, as shown in the plot, where half of them achieved higher than average scores. When broken down into detailed brackets of income level, a positive relationship is observed between income levels and scores. Higher-income households are associated with more children achieving scores above 100, particularly in the \\u003cspan\\u003e$\\u003c/span\\u003e75,001-\\u003cspan\\u003e$\\u003c/span\\u003e100,000(encoded as 11) and \\u003cspan\\u003e$\\u003c/span\\u003e200,001 or more brackets(encoded as 13). Incomes categorized as 10 through 13, which is mid-to-upper income level display a broader spread of high scores, indicating better academic performance is achieved by children from wealthier households. For lower-income brackets (categories 1 to 5), scores are concentrated between 20 and 80, with very few children achieving scores above 100. This suggests that limited financial resources may hinder academic performance. Notably, a moderate-income level is necessary for achieving higher grades, but not sufficient condition. While most children who scored above 80 came from mid-to-upper income families, a significant proportion of children within these income brackets also recorded low scores. While high-income households provide a better informal environment and support, it does not guarantee academic success. The variability in middle- and lower-income groups suggests the influence of other social, educational, and environmental factors in shaping outcomes. This emphasizes that family income level alone is not the sole determinant of academic performance, and additional support systems may be critical for fostering success across income levels.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003eAlso, the education level(W8PARED) of parents seems to be functioning for helping children gain good IRT scores. As shown in the scatter plot, there is a clear distinction in the score ranges between children whose parents have less than a high school diploma and those whose parents have at least a high school diploma. Children whose parents did not complete high school failed to achieve scores higher than 60. In contrast, children whose parents attained an above or equivalent high school level of education exhibited a broader distribution of scores, with some achieving notably high scores. The group of parents with the education levels can be further divided into those below the college level (encoded as 3 and 4) and those with a college degree or higher (encoded as 5 or above). Despite slight differences, there is a pattern that children whose parents with college-level education or higher achieve higher scores more frequently than the other groups. Within these higher education groups, however, academic outcomes show little variation. It is inferred that once parents achieve a certain educational level, such as a high school diploma or college degree, it functions as a threshold for determining the highest score that children may achieve. Additionally, other factors may become more significant in shaping their children's academic success within each educational threshold group.\\u003c/p\\u003e \\u003cp\\u003e Lastly, this research attempted to capture the effects of children\\u0026rsquo;s race on the children\\u0026rsquo;s academic success as race and nationality plays a large role in shaping the learning and cognitive agency. White, non-Hispanic children, who make up the largest proportion (25.6%), exhibit a wide distribution of scores. Due to the highest population, many of them also exceed excellent above 60. Black or African American children, a smaller group (3.88%), tend to have scores concentrated below the total average, with fewer high-performing individuals. Hispanic children show a clustering of scores near the total average, with relatively few high-scoring outliers compared to White children. Asian children, though representing only 2.13% of the sample, show strong academic performance, with the highest average of 31.08 among racial groups. In contrast, Native Hawaiian or Other Pacific Islander children (0.42% of the sample) have scores predominantly low average of 23.95, with limited variability. Similarly, American Indian or Alaska Native children (0.86%) exhibit scores average of 22.76, with few high-scoring individuals.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec11\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003e3.2 XG-Boost\\u003c/h2\\u003e \\u003cp\\u003eThe study considered various activities (reading books, telling stories, singing songs, helping with arts and crafts, involving children in household chores, playing games or puzzles, discussing nature or science projects, building or constructing toys, and engaging in sports or exercise) as well as income level, parents\\u0026rsquo; highest education level, children\\u0026rsquo;s race. The informal learning conditions were taken to predict the children\\u0026rsquo;s IRT score as the output variable. As indicated in the scatter plots that show the distributions of scores according to the demographic, or parental activities, there is no linear regression between factors and IRT scores. Therefore, linear regression and logistic regression, which is grounded on linear relationship between factors are not suited for the given datasets, implying complex and non-linear associations between factors. XG-Boost, thus, was implemented for the classification of the children\\u0026rsquo;s IRT scores whether they achieved higher than average scores or not. In classifying values, the focal point was to examine how effectively the model could predict whether a child would succeed academically, based on whether they achieved a higher-than-average score. The model\\u0026rsquo;s accuracy was 65.87%, meaning that the model forecast two-thirds of children\\u0026rsquo;s score accurately. The is showing a reasonable performance in predicting academic success. The other performance measure also shed light on the fair predictive power of the model, with recall (63.61%) and F1 score (60.08%). The relatively low precision (56.93%) suggests the model is prone to generating false positives. In conclusion, these metrics collectively suggest that the model performs moderately well, but some refinement would be needed for higher predictive accuracy and reliability.\\u003c/p\\u003e \\u003cp\\u003eAdditionally, XG-Boost provided metrics for evaluating the importance of input features, allowing us to identify the factors that had the greatest influence on predictions. By analyzing feature importance, the underlying mechanisms of XG-Boost are partially revealed. The most important features were parents\\u0026rsquo; educational level, family poverty, and income levels, with feature importance scores of 0.1459, 0.1288, and 0.1027, respectively. Race, particularly whether individuals were Hispanic or Black, also contributed moderately to the predictions, with feature importance values of 0.0997 and 0.0633. Parental activities, on the other hand, showed relatively low importance. For instance, reading books had a feature importance of 0.0139, and helping parents with chores scored 0.0206. Everyday activities displayed minimal influence on predictions, which may be attributed to high multicollinearity with other key features. For example, reading books exhibited an extremely high VIF of 58.44, indicating severe multicollinearity with poverty level, parent\\u0026rsquo;s educational level, and other related activities.\\u003c/p\\u003e \\u003c/div\\u003e\"},{\"header\":\"4 Discussion\",\"content\":\"\\u003cp\\u003eIn conclusion, this research explored the role of parental involvement in daily activities for informal learning and its impact on children's academic achievement. The activities studied ranged from verbal interactions, such as storytelling and singing songs, to home-related tasks like involving children in chores, and outdoor activities such as playing sports. Data visualization, and scatter plots, provided insights into the association between these activities and children\\u0026rsquo;s academic performance. While these activities were not definitive predictors of academic success, they played a role in supporting higher academic scores. The children who gained exceptionally high scores were those who spent time with their parents engaging in activities like storytelling, reading books, singing songs, doing household chores, and playing games. In contrast, children who did not engage in these activities with their parents exhibited a dense cluster of scores around the average, failing to achieve higher academic outcomes. For other activities, such as talking about nature, engaging in science projects, or playing with constructive toys, they seemed to be less significant factors. Although some children who participated in these activities achieved higher scores, others who did not engage in them also managed to perform well. The results are in line with previous research as they\\u003c/p\\u003e \\u003cp\\u003eThe research also examined the importance of the frequency of parental involvement and academic performance. A scatter plot analyzing the frequency of reading books with parents and IRT scores revealed that frequent interactions were a necessary condition for higher scores but not a sufficient one. The findings highlight the importance of informal educational practices which involve parents in their children\\u0026rsquo;s everyday activities. The time spent together empowers children to acquire the knowledge of various literacy, numeracy accountings, and mathematical thinking. These practices help children to achieve higher scores, but they are not definitive predictors of academic success, as a large proportion of children did not achieve high scores despite frequent parental involvement. If other factors, such as parental attitudes, school environments, study motivation, time spent on study, etc. are applied, it is expected to give a better understanding of academic outcomes.\\u003c/p\\u003e \\u003cp\\u003eDemographic factors were found to play a similar role in influencing academic achievement. The study investigated the effects of family income, poverty level, parental education, and children\\u0026rsquo;s race. For example, children from families living in poverty tended to have scores clustered around the average, whereas children from non-poverty families exhibited a wider range of scores, some of them achieving the highest. When examining in detail- family income level, children scoring above 40 came from mid-to-upper income families, meaning that a moderate level of income is critical for a high level of mathematical performance. However, a significant proportion of children within these income brackets also recorded low scores. Additionally, children scoring above 60 generally had parents with education levels beyond college, indicating whether parents have a moderate level of education plays a critical role. Differences in score distributions were also observed across racial groups. White, non-Hispanic children showed a wide range of scores, many of which exceeded the average. In contrast, Black or African American children\\u0026rsquo;s scores were more concentrated below the average, while Hispanic children\\u0026rsquo;s scores tended to cluster around the average. Societal, economic, and cultural factors clearly influence academic outcomes. However, those factors are not sufficient conditions to conclude the academic achievement of children. Therefore, it is necessary to find out what differentiates children\\u0026rsquo;s outcomes within the same societal or racial groups.\\u003c/p\\u003e \\u003cp\\u003eThe study explored the application of advanced machine learning techniques, such as XG Boost, to predict academic outcomes. While these models are moderately accurate predictions, unlike the expectation, parental engagement played a nonsignificant role in determining the academic achievement of children. It can be inferred that the parents- children\\u0026rsquo;s interactive activities are already highly multi-correlated with other demographic factors. To achieve higher accuracy in prediction, it is still needed to add other explanations to capture the complexities of academic achievement for the performance of the model. Future research should consider additional learning conditions, that would determine children's achievement among the same group of income level, race, or poverty level. Those factors are expected to enhance the predictive and explanatory power of these models.\\u003c/p\\u003e\"},{\"header\":\"5 Limitation\",\"content\":\"\\u003cp\\u003eThis study emphasizes the influence of demographic factors on children\\u0026rsquo;s academic performance and investigates the role of the way parents and children interact. While these factors provide valuable insights, improving the predictive capabilities of AI models requires other additional variables. Factors such as school environments, and access to resources, time spent on study, driven motivations, parental attitude, interactions with teachers could significantly enhance the explanatory power of these models. Furthermore, as classification methods proved more effective than regression in this study, future study might explore other decision tree-originated algorithms to classify outcomes into multiple categories rather than limiting analyses to binary classifications. This approach could offer a more nuanced understanding of how various factors contribute to different levels of academic achievement, ultimately supporting more targeted interventions.\\u003c/p\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003ch2\\u003eAuthor Contribution\\u003c/h2\\u003e\\u003cp\\u003eNaeun wrote the main manuscript text and Prof. Biswajit Sarkar drew main methodologies. All authours reviewed the manuscript\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eAdams, M., Rodr\\u0026iacute;guez, L., \\u0026amp; Smith, J.\\u003c/strong\\u003e (2021). Designing frameworks for authentic equity in science teaching and learning: Informal learning environments and teacher education for STEM. \\u003cem\\u003eJournal of STEM Education, 22\\u003c/em\\u003e(3), 27\\u0026ndash;38.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eAlsabbagh, M., \\u0026amp; Ibrahim, A.\\u003c/strong\\u003e (2024). A hybrid model for the prediction of electrical energy consumption using hybrid LSTM and ML regressors. \\u003cem\\u003eEnergy Reports, 10\\u003c/em\\u003e, 289\\u0026ndash;301.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eAram, D., \\u0026amp; Levin, I.\\u003c/strong\\u003e (2001). Mother\\u0026ndash;child joint writing in low SES: Sociocultural factors, maternal mediation, and emergent literacy. \\u003cem\\u003eReading and Writing: An Interdisciplinary Journal, 14\\u003c/em\\u003e(3), 213\\u0026ndash;233.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eBachman, H., \\u0026amp;Heather J. Bachman\\u0026amp; Leanne Elliott \\u0026amp;Shirley Duong \\u0026amp; Laura Betancur\\u0026amp; Monica G. Navarro\\u0026amp; Elizabeth Votruba-Drzal\\u0026amp; Melissa Libertus.\\u003c/strong\\u003e (2020). Triangulating multi-method assessments of parental support for early math skills. \\u003cem\\u003eEarly Childhood Education Journal.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eBaroody, A. E., \\u0026amp; Diamond, K. E.\\u003c/strong\\u003e (2012). Measures of preschool children\\u0026rsquo;s interest and engagement in literacy activities: Examining gender differences and construct dimensions.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eBarton, A. C., Tan, E., \\u0026amp; Rivet, A.\\u003c/strong\\u003e (2021). Co-designing for rightful presence in informal science learning environments. \\u003cem\\u003eJournal of Research in Science Teaching, 58\\u003c/em\\u003e(2), 145\\u0026ndash;168.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eBeard, C., Richardson, R., \\u0026amp; Davison, P.\\u003c/strong\\u003e (2024). Synergy between formal and informal education. \\u003cem\\u003eJournal of Educational Research, 36\\u003c/em\\u003e(4), 415\\u0026ndash;429.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eBeck, I. L., \\u0026amp; McKeown, M. G.\\u003c/strong\\u003e (2007). Increasing young low-income children\\u0026apos;s oral vocabulary repertoires through rich and focused instruction. \\u003cem\\u003eThe Elementary School Journal, 108\\u003c/em\\u003e(4), 271\\u0026ndash;287.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eBermudez, A., \\u0026amp;Salazar, J \\u0026amp;Garcia, L\\u0026amp; Ochoa, K.D.\\u0026amp; Ahn, J.\\u0026amp; Bustamante, A.S. \\u0026amp;Pesch, A. \\u0026amp;Hirsh-Pasek, \\u0026amp; Roldan, W. \\u0026amp; Soto-Lara, S.\\u0026amp; Gomez, W.\\u0026amp; Rodriguez, R.(\\u003c/strong\\u003e2023). Designing culturally situated playful environments for early STEM learning with a Latine community. \\u003cem\\u003eJournal of Early Childhood Education.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eBingham, G. E., Pendergast, M. L., \\u0026amp; Ross, R.\\u003c/strong\\u003e (2017). Parenting styles and home literacy opportunities: Associations with children\\u0026rsquo;s oral language skills. \\u003cem\\u003eEarly Education and Development, 28\\u003c/em\\u003e(7), 894\\u0026ndash;912.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eBird, Y. J., Murphy, K. R., \\u0026amp; Martinez, D. S.\\u003c/strong\\u003e (2023). Talking, praising, and teaching: How parent interaction during a learning task relates to children\\u0026apos;s early learning. \\u003cem\\u003eEarly Childhood Research Quarterly, 58\\u003c/em\\u003e, 109\\u0026ndash;120.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eCarmiol, A., Rodr\\u0026iacute;guez, P., \\u0026amp; V\\u0026eacute;lez, L.\\u003c/strong\\u003e (2021). Book sharing and reminiscing: Caregivers\\u0026apos; conversational style and children\\u0026apos;s language and literacy development in low-income Costa Rican families. \\u003cem\\u003eJournal of Early Childhood Literacy, 21\\u003c/em\\u003e(4), 534\\u0026ndash;550.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eCaspe, M.\\u003c/strong\\u003e (2009). Low-income Latino mothers\\u0026rsquo; booksharing styles and children\\u0026apos;s emergent literacy development. \\u003cem\\u003eEarly Childhood Education Journal, 37\\u003c/em\\u003e(2), 97\\u0026ndash;106.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eChen, T., \\u0026amp; Guestrin, C.\\u003c/strong\\u003e (2016). XGBoost: A scalable tree boosting system. \\u003cem\\u003eProceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining\\u003c/em\\u003e, 785\\u0026ndash;794.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eChu, S., Martin, Rachel \\u0026amp;Quek, Francis\\u0026amp; Berman, Alexander\\u0026amp; Suarez, Mario \\u0026amp;Zarei, Niloofar \\u0026amp;Nam, Beth \\u0026amp;Banigan, Colin\\u003c/strong\\u003e(2017). Becoming makers: Examining \\u0026lsquo;making\\u0026rsquo; literacy in the elementary school science classroom. \\u003cem\\u003eJournal of Research in Science Teaching.\\u0026nbsp;\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eCheng, L., Wang, Z., \\u0026amp; Zhang, Y.\\u003c/strong\\u003e (2024). Evaluation of students\\u0026rsquo; performance during the academic period using the XG-Boost Classifier-Enhanced AEO hybrid model. \\u003cem\\u003eJournal of Educational Data Mining, 12\\u003c/em\\u003e(3), 121\\u0026ndash;135.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eClements, D. H., \\u0026amp; Sarama, J.\\u003c/strong\\u003e (2024). Systematic review of learning trajectories in early mathematics. \\u003cem\\u003eMathematics Education Research Journal.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eCohrssen, C., \\u0026amp; Church, A.\\u003c/strong\\u003e (2013). Purposeful pauses: Teacher talk during early childhood mathematics activities. \\u003cem\\u003eEarly Childhood Research Quarterly.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eCohrssen, C., \\u0026amp; Niklas, F.\\u003c/strong\\u003e (2019). Using mathematics games in preschool settings to support the development of children\\u0026rsquo;s numeracy skills. \\u003cem\\u003eEarly Education and Development.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eCombs, S., \\u0026amp; Higgins, S.\\u003c/strong\\u003e (2023). The relationship between shared picture book reading and language development in young children. \\u003cem\\u003eJournal of Research in Childhood Education, 37\\u003c/em\\u003e(2), 192\\u0026ndash;205.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eCotto, J. A,, \\u0026amp; Ribner, Andrew \\u0026amp; Smith, Leann\\u003c/strong\\u003e (2024). Understanding working memory and mathematics development in ethnically/racially minoritized children through an integrative theory lens. \\u003cem\\u003eBehavioral Sciences (2076-328X). May2024, Vol. 14 Issue 5, p390. 20p.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eDeitcher, D., DeBaryshe, B. D., \\u0026amp; Fujita, K.\\u003c/strong\\u003e (2019). Alphabet books: Relations between aspects of parent-child shared reading, children\\u0026rsquo;s motivation, and early literacy skills. \\u003cem\\u003eJournal of Early Childhood Literacy.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eDhiya\\u0026apos;ulhaq, M. A., Sucipto, A., \\u0026amp; Purnomo, M. A.\\u003c/strong\\u003e (2024). Ocean wave prediction using Long Short-Term Memory (LSTM) and Extreme Gradient Boosting (XGBoost) in Tuban Regency for fisherman safety. \\u003cem\\u003eOcean Engineering, 242\\u003c/em\\u003e, 110358.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eEthridge, E. A., \\u0026amp; King, J. R.\\u003c/strong\\u003e (2005). Calendar math in preschool and primary classrooms: Questioning the curriculum. \\u003cem\\u003eEarly Childhood Education Journal, 32\\u003c/em\\u003e(5), 291\\u0026ndash;296.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eFolkestad, G.\\u003c/strong\\u003e (2006). Formal and informal learning situations or practices vs formal and informal ways of learning. \\u003cem\\u003eEuropean Educational Research Journal, 5\\u003c/em\\u003e(1), 81\\u0026ndash;95.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eFyfe, E. R,\\u0026amp; Evans, Julia L. \\u0026amp;Matz, Lauren Eisenband \\u0026amp;Hunt, Kayla M. \\u0026amp;Alibali, Martha W.\\u003c/strong\\u003e (2017). Relations between patterning skill and differing aspects of early mathematics knowledge. \\u003cem\\u003eIn Cognitive Development October 2017 44:1-11\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eGinsburg, H. P., \\u0026amp; Lee, Joon Sun \\u0026amp;Boyd, Judi Stevenson\\u003c/strong\\u003e (2008). Mathematics education for young children: What it is and how to promote it. \\u003cem\\u003eSociety for Research in Child Development\\u003c/em\\u003e. 2008.\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eGong, J., Liu, H., \\u0026amp; Zhang, Y.\\u003c/strong\\u003e (2021). Informal learning in nature education promotes ecological conservation behaviors of nature reserve employees: A preliminary study in China. \\u003cem\\u003eGlobal Ecology and Conservation, Vol 31, Iss , Pp e01814\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eHaden, C. A., Jant, Erin A. \\u0026amp;Hoffman, Philip C. \\u0026amp;Marcus, Maria \\u0026amp;Geddes, Jacqueline R. \\u0026amp;Gaskins, Suzanne\\u003c/strong\\u003e(2014). Supporting family conversations and children\\u0026rsquo;s STEM learning in a children\\u0026rsquo;s museum. \\u003cem\\u003eIn Early Childhood Research Quarterly Q3 2014 29(3):333-344\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eJodi K. Heidlage \\u0026amp;Jennifer E. Cunningham \\u0026amp; Ann P. Kaiser \\u0026amp;Carol M. Trivette \\u0026amp; Erin E. Barton \\u0026amp;Jennifer R. Frey\\u003cbr\\u003e\\u003c/strong\\u003eThe effects of parent-implemented language interventions on child linguistic outcomes: A meta-analysis\\u003cem\\u003e. Early Childhood Research Quarterly Volume 50, Part 1, 1st Quarter 2020, Pages 6-23\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eMegan Y. Roberts e\\u003c/strong\\u003e(2020). The effects of parent-implemented language interventions on child linguistic outcomes: A meta-analysis. \\u003cem\\u003eIn Early Childhood Research Quarterly Q1 2020 50 Part 1:6-23\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eHu, J., \\u0026amp; Song, W.\\u003c/strong\\u003e (2024). Research on XGBoost academic forecasting and analysis modelling. \\u003cem\\u003eJournal of Physics: Conference Series, 14 October 2019, 1324(1))\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eHussim, H., \\u0026amp;Roslinda Rosli\\u0026amp; Nurul.N \\u0026amp; Siti.M \\u0026amp; Muhammad.M\\u0026amp; Zanaton.I\\u0026amp; Azmin.R\\u0026amp;Siti,M\\u0026amp; Lilia.H \\u0026amp;Kamisah.O \\u0026amp; Ah .Lay\\u003c/strong\\u003e(2024). A systematic literature review of informal STEM learning. \\u003cem\\u003eEuropean Journal of STEM Education. 2024 9(1)\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eKelly, L., Jones, S., \\u0026amp; Brown, C.\\u003c/strong\\u003e (2023). Smartphones and parent-child conversations during young children\\u0026apos;s informal science learning at an aquarium. \\u003cem\\u003eIn Computers in Human Behavior Reports May 2023 10\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eKhan, A., \\u0026amp; Rodrigues, S.\\u003c/strong\\u003e (2017). STEM for girls from low-income families: Making dreams come true. \\u003cem\\u003eThe Journal of Developing Areas, 2017 Apr 01. 51(2), 435-448.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eLeyva, D.\\u003c/strong\\u003e (2019). How do low-income Chilean parents support their preschoolers\\u0026rsquo; writing and math skills in a grocery game? \\u003cem\\u003eEarly Education and Development. 2019 30(1):114-130.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eLeyva, D., Cates, C. B., \\u0026amp; Yoshikawa, H.\\u003c/strong\\u003e (2019). Maternal behaviors in a grocery game and first graders\\u0026rsquo; literacy and math skills in a low-income sample. \\u003cem\\u003eElementary School Journal. Jun 2019 119(4):629-650.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eLeyva, D., Tamis-LeMonda, C. S., Yoshikawa, H., \\u0026amp; Cates, C. B.\\u003c/strong\\u003e (2017). Grocery games: How ethnically diverse low-income mothers support children\\u0026rsquo;s reading and mathematics. \\u003cem\\u003eIn Early Childhood Research Quarterly Q3 2017 40:63-76\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eLevin, I., \\u0026amp; Aram, D.\\u003c/strong\\u003e (2010). Mother-child joint writing as a learning activity. \\u003cem\\u003eReading \\u0026amp; Writing; Jan2012, Vol. 25 Issue 1, p217-249, 33p\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eNancy, L., et al.\\u003c/strong\\u003e (2020). Interested, disinterested, or neutral: Exploring STEM interest profiles and pathways in a low-income urban community. \\u003cem\\u003eEURASIA Journal of Mathematics, Science and Technology Education. 2020 16(6).\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eNora, A., Callanan, M., \\u0026amp; Jipson, J.\\u003c/strong\\u003e (2010). Enhancing building, conversation, and learning through caregiver\\u0026ndash;child interactions in a children\\u0026rsquo;s museum. \\u003cem\\u003eDevelopmental Psychology. Mar 2010 46(2):502-515.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003eOrosoo, M., Norouzi, M., \\u0026amp; Behzad, M.\\u003c/strong\\u003e (2025). Transforming English language learning: Advanced speech recognition with MLP-LSTM for personalized education. \\u003cem\\u003eAlexandria Engineering Journal, Vol 111, Iss , Pp 21-32 (2025)\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003ePace, A., Albright, L., \\u0026amp; Mitchell, R.\\u003c/strong\\u003e (2017). Identifying pathways between socioeconomic status and language development. \\u003cem\\u003eAnnual Review of Linguistics; 2017, 3 p285-p308, 24p.\\u003c/em\\u003e\\u003c/li\\u003e\\n \\u003cli\\u003e\\u003cstrong\\u003ePentimonti, J. M., Justice, L. M., \\u0026amp; Piasta, S. B.\\u003c/strong\\u003e (2012). A standardized tool for assessing the quality of classroom-based shared reading: Systematic Assessment of Book Reading (SABR). \\u003cem\\u003eEarly Childhood Research Quarterly. 3rd Qtr 2012 27(3):512-528.\\u003c/em\\u003e\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"STEM education, Child Education, Informal learning, Various Activities, Data visualization Random Forest\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-5885890/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-5885890/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eInformal learning, characterized by unstructured, incidental, and everyday learning activities, shapes children's academic performance. This study leverages the Early Childhood Longitudinal Study, ECLS-K dataset, a nationally representative sample of U.S. kindergartners, to examine the relationship between parental engagement in informal learning activities and children's academic accomplishment. This study utilized data visualization to investigate tendencies of children\\u0026rsquo;s scores based on each demographic and interaction factor. Further, the study attempts to apply XB-Boost which can analyze large-scale data and uncover complex, non-linear relationships. This study explores the potential of using AI methods to identify contributions of parental interaction with their children and demographic factors.\\u003c/p\\u003e\",\"manuscriptTitle\":\"How Interaction with Parents Affects Children's Mathematical Informal Learning and Verbal Skills\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-02-11 09:46:48\",\"doi\":\"10.21203/rs.3.rs-5885890/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"e5c87751-439a-4823-a015-9683b5f1bfe5\",\"owner\":[],\"postedDate\":\"February 11th, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2025-05-15T01:53:14+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2025-02-11 09:46:48\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-5885890\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-5885890\",\"identity\":\"rs-5885890\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}