Influence of parents’ views about science on parent–child science talk at home

ABSTRACT This study explores parents’ views of science as a family sociocultural background that influences how parents support children’s science talk as they engage in a science activity together at home. We focus on Indonesian families as they have distinct sociocultural characteristics that may yield different parent–child interactions than their Western counterparts in informal science learning settings. A microethnography approach and thematic discourse analysis are employed to capture the parent–child science talk and interactions. Findings show that Indonesian parent–child interactions are directive and collectivist in nature. Parents tend to lead learning at home, and the other family members voluntarily participate. Through their science talks, Indonesian parents engaged children in the science activity using explanatory talk, corrective feedback, and real-world connections. Throughout the interactions, the parents emphasized particular science knowledge and practices based on their views of science. We present three cases where the parents viewed science as a hypothesis-testing practice, as knowledge related to everyday phenomena, and as an inference-making process. Their talks and support for children’s learning varied due to these different views of science. The study adds to the limited literature on parent–child interactions in informal science learning in non-Western contexts.


Introduction
Children develop scientific understanding from a myriad of experiences in various settings, including informal learning environments, such as homes, science museums, and out-of-school programs (Bell et al., 2009;Falk & Dierking, 2010). In such settings, scholars have regarded family as the most significant social learning group in which children learn science from everyday experience (Ellenbogen et al., 2004;Rogoff, 1990). While much research has studied children's learning with their families at science museums and afterschool programs, only limited research has explored family science learning in home settings. In fact, home is the first environment in which children develop scientific concepts from everyday activities with family members, as early as infancy and toddler ages (Sikder & Fleer, 2015).
At home, parents play a critical role in encouraging and supporting children's science learning experiences that later lead to their children's interest in science and ability to learn science (NSTA, 2009). Arguably, parents are well-suited to help children connect everyday experiences with relevant science concepts or practices because they have extensive information about their children's knowledge and shared experiences. Research has demonstrated that parents are capable of helping children learn science concepts from everyday experiences at home through playful and imaginative activities (Gomes & Fleer, 2019;Hao & Fleer, 2017) and spontaneous activities (Vedder-Weiss, 2017). Sikder and Fleer (2015) has suggested that non-structured, simple science activities at home, such as cooking, mirror play, discussing the day and night, or noticing the counterintuitive science observation of sunrise and sunset can lay the foundation for learning scientific concepts when parents help young children to notice such 'small science' concepts. Besides facilitating the development of science concepts, parents have also been shown to scaffold and guide children's participation in science activities at home by modeling and engaging children in scientific practices (Andrews & Wang, 2019;Vedder-Weiss, 2017).
Over the past two decades, researchers have specifically focused on how parents use language to support children's learning at home and in other informal settings (Ash, 2003;Callanan et al., 2017;Haden, 2010;Tenenbaum & Callanan, 2008). Language has been seen to play a pivotal role in parent-child interaction, not only as a tool to communicate and carry out activities but also as a means for making meaning and constructing knowledge together (Ash, 2003;Bakhtin, 1981). Parents have been found to use various strategies through their talks to support children's science learning, including encouraging predictions, providing explanations, asking productive questions, pretending play, and persuading inquiry behaviors (Siegel et al., 2007;Strickler-Eppard et al., 2019;Tenenbaum et al., 2005;Tenenbaum & Callanan, 2008). Numerous studies have demonstrated that the so-called 'science talk' between parents and children during informal science-related activities helped children develop scientific literacy (Tenenbaum et al., 2005), science competencies (Andrews & Wang, 2019), and science interest (Pattison & Dierking, 2019). Other studies showed that parents engaged children in scientific reasoning (Vandermaas-Peeler et al., 2019) and science inquiry (Strickler-Eppard et al., 2019) through such talk.
In the literature, the extent and types of parent-child science talk have been linked to the parents' sociocultural backgrounds. Parents' education or schooling level, ethnicities, and socioeconomic status are among the most frequently mentioned sociocultural backgrounds in the body of research on family learning. Parents with higher schooling (12-16 years of education) were found more likely to use science-principle-based explanations, to encourage children's predictions, and to provide positive encouragement (Dabney et al., 2016;Tenenbaum & Callanan, 2008). They were also more directive with their children compared to parents with less schooling (3-11 years of education) (Siegel et al., 2007). European-American parents asked more what, why, who, where, and how questions and talked more about STEM to their children than their African-American and Hispanic-American counterparts (Haden et al., 2014). Parents who could afford frequent visits to informal science places, such as science museums, were able to use their expertise (Zimmerman et al., 2008), prior knowledge (Zimmerman & McClain, 2014), metacognitive knowledge (Thomas & Anderson, 2013), and various learning strategies (Zimmerman et al., 2010) to make productive sensemaking talk.
Besides the demographic factors mentioned above, researchers have suggested family habitus and science capital as other potential factors that shape children's engagement in science (Archer et al., 2012(Archer et al., , 2015. Parents' possession of science capital, such as qualifications or professions in science, science knowledge, scientific literacy, and access to science-related resources, may contribute to how parents view science and science learning. Parents' views of science include their understandings of, attitudes towards, and beliefs about science and science learning. Prior studies indicated that individual differences in these implicit views of science might affect the extent to which parents engage their children in science discussions through cognitively demanding speech and explanations Szechter & Carey, 2009;Tenenbaum & Leaper, 2003). While these studies have investigated parent-child interactions in designed environments, such as in science centers and museums, little is known about how this specific parents' sociocultural background plays a role in ways that parents use science talk to support children's learning at home where parents exert great control over their children's activities and behaviors (Chak, 2010). This study seeks to fill this gap in the literature.
The present study also seeks to address the recent critiques in human research that often focuses on Western populations (Henrich et al., 2010;Nielsen et al., 2017). The majority of the research on children's science learning at home above has been conducted in Western contexts. Non-Western families, by contrast, have distinct sociocultural characteristics that may yield different ways of parent-child interactions at home. For example, non-Western families, such as in Indonesia, have a high degree of collectivism and a large power distancethe extent to which the less powerful members of the family accept and expect that power is distributed unequallybetween parents and children while their Western counterparts are deemed individualists and show a low parent-child power distance (Hofstede, 2011). Few existing studies on Indonesian families focused primarily on the broader and more general areas, such as parenting education (Tomlinson & Andina, 2015) and parental involvement in children's education (Yulianti et al., 2018(Yulianti et al., , 2019. Studies of Indonesian parent-child interactions and Indonesian parents' support for children's science learning at home are almost non-existent. Therefore, the presented study is designed to specifically answer the following questions: (1) How do Indonesian parents and children interact when engaged in informal science learning at home? (2) How do Indonesian parents support their children's science learning through their talk in these interactions? and (3) How do parents' views of science shape the science talk and interactions between the parents and children in informal science learning at home?
We begin this paper with a theoretical discussion from the sociocultural lens to highlight the critical role of parents in helping children develop science learning from everyday experiences at home. As the study focuses on the important role of language in the scaffolding process by parents, we then give an overview of science talk to discuss the literature on parents' use of science talk in supporting their children's learning. Next, we discuss two sociocultural contexts associated with the ways that the parents scaffold their children's learning at home through science talk. These contexts include parents' views of science shaped by their science capital and the unique cultural backgrounds in Indonesia. In the following sections, we report the study design and methodology, findings, and implications for informal science learning at home.

A sociocultural perspective on family learning
We adapted Vygotskian sociocultural theory as the framework for our investigation of Indonesian parents' support of children's learning through informal science learning activities at home. First, we take into account Vygotsky's theory of everyday and scientific concept formation, as discussed in The collected works of L.S. Vygotsky: problems of general psychology (Vygotsky, 1987). As Vygotsky argued, everyday concepts resulting from day-to-day practices are dialectically related to scientific or academic concepts and support each other in the process of developing a more robust scientific understanding. Fleer et al. (2014) illustrated this dialectical relation as a coil spring, in which the everyday concepts start from the bottom and move upwards towards abstraction and generalization while the scientific concepts move downwards from abstract definition towards concrete representation. According to Vygotsky (1987), children's concrete experience of pushing a scooter, for example, facilitates learning of the scientific concept of force. Likewise, the concept of force helps explain the concrete experience of pushing things forward. In other words, everyday and scientific concepts support each other through this dialectical relation in learning science.
According to Vygotsky's theory, children gradually develop and recontextualise these everyday concepts into scientific concepts through collaboration with knowledgeable others (Vygotsky, 1978). From a sociocultural perspective, meaningful learning occurs in the social context of learners, and learning takes place with others in the community rather than as an individual activity. We considered Vygotsky's notion of the Zone of Proximal Development (ZPD) as the second important tenet of theory to inform this study. According to ZPD, what a child can acquire cognitively by themselves could be leveraged to his/her potential development level through support from knowledgeable others (Vygotsky, 1978). In the context of family, parents can serve as the more knowledgeable other to facilitate the development process of scientific concept formation and acquiring scientific competencies through scaffolding. Wood et al. (1976, p. 90) defined scaffolding as a process 'that enables a child to achieve a goal that would be beyond his unassisted efforts.' The scaffolding is apparent when parents, for example, help the child focus on the content, answer questions, create analogies, connect to prior knowledge, give explanations, and provide encouragement during the interactions. With regards to acquiring new skills such as scientific competencies in informal settings, Rogoff's (1990) notion of 'guided participation' explained that parents guide children's participation in day-to-day social activities, help them adapt their knowledge to new situations, and encourage them to try out new skills (Rogoff, 1990). Both parents' scaffolding and guided participation can help children notice everyday concepts and connect them to scientific concepts.
How do parents help with such a complex process of concept formation? This is explained by another tenet of sociocultural theory, which is mediation (Vygotsky, 1978(Vygotsky, , 1986. Mediation is a term that describes the psychological and cultural tools and processes that people use to make sense of experiences and events in their lives (Kozulin, 2002;Wertsch, 2007). Vygotsky suggested that tools such as signs, language, and other cultural symbols mediate all human activity. Language is particularly highlighted as the most important cultural tool that mediates children's co-construction of knowledge and participation in learning activities (Ash, 2003;Bakhtin, 1981;Wells, 1999). Through language, one constructs and shares an understanding of the world with others (Bakhtin, 1981). Language is also a central feature in informal settings because it is a means to carry out activities by learners in social groups (Ash, 2003). This theoretical framework informed our investigation of parents' scaffolding and guiding of their children to contextualize everyday experiences into science learning through language as a mediator. Such use of language in science learning is termed as science talk, which is discussed further in the following section.

Parent-Child science talk
Science talk, or 'talking science', does not mean simply talking about science; it means doing science through the medium of language (Lemke, 1990). Parents can use science talk to engage children in science practices and help them make sense of scientific concepts from everyday experiences. Children's science talks with parents may support the children's early development of scientific learning (Crowley & Galco, 2001). Researchers have examined some emergent components of parent-child science talk in informal settings, including explanations, questioning, making connections, and engaging in scientific practices (Callanan et al., 2017;Haden et al., 2014;Strickler-Eppard et al., 2019;Tenenbaum & Callanan, 2008). Tenenbaum and Callanan (2008) studied science talk mostly as explanatory talk used by parents with their children. They focused on four types of talk: referencing prior knowledge, causal explanations, scientific principles explanations, and encouraging predictions. In another study, Callanan et al. (2017) found more strategies that parents used in science talk: (a) asking critical thinking questions; (b) giving explanations about scientific ideas or requesting for children to create such explanations; (c) using evidence to answer questions or requesting children to do so; (d) connecting personal events to the current activity for the child; and (e) comparing between the activity content and other information. Kaya and Lundeen (2010) observed that parents frequently used productive questioning in their science talk with their children during science activities at home. Likewise, Haden et al. (2014) demonstrated that parents asked open-ended questions (Who, What, Where, Why, and How) during interactions with their children to produce talks about science content. Children's responses to such questions have been linked to their later learning and retention of information (Hedrick et al., 2009). Science talk has also been used to engage children in scientific practices during everyday activities. Through science talk, parents encouraged their children to engage in inquiry behaviors and complex scientific reasoning, such as observing, predicting, comparing and contrasting, evaluating, drawing conclusions, and articulating explanations (Strickler-Eppard et al., 2019;Vandermaas-Peeler et al., 2019). In Vedder-Weiss (2017) study, where they described the process of guided participation in an informal science activity at home, the parents used science talk to model and engage in scientific practices together with the children. From synthesizing the studies around science talk, we used four elements of science talk (i.e. questioning, explanation, making connections, and science practice engagement) as our a priori codes in analysis, along with opening for any emergent element of the talk. In the following section, we discuss how parents' science talk is intertwined with various sociocultural aspects.

Sociocultural factors affecting parent-child science talk
In the introduction section, we have mentioned and reviewed that the types and extents of parentchild science talk in informal settings have been linked to family sociocultural factors, such as parent's education level, ethnicities, and socioeconomic status, according to research (Dabney et al., 2016;Haden et al., 2014;Siegel et al., 2007;Zimmerman et al., 2010). Here, we discuss other potential family sociocultural factors, such as parents' views of science and parents' cultural background. First, we consider the concept of family habitus and science capital. Family habitus, as defined by Archer et al. (2012), is 'the family environment within which young children are growing up and starting to develop their ideas about science and their relationships with science' (p. 885). Family habitus encompasses a broad spectrum of family resources, practices, values, cultural discourses, and science-related identifications.
Family habitus is shaped by the family's possession of particular economic, social, and cultural capital that is specifically related to scienceor 'science capital' (Archer et al., 2012). Science capital can be visualized as a large bag containing all the science-related knowledge, attitudes, experiences, and resources that we acquire through life and is grouped into four main 'pockets': (1) what science we know, (2) who we know (e.g. real-life scientists, parents who are interested in science), (3) how we think about science (our attitudes and dispositions), and (4) what we do (everyday engagement with science) (DeWitt et al., 2016). Accruing science capital might help families understand complex scientific texts, understand pathways into science careers or pursue science-based hobbies . Archer and colleagues also regarded science capital as an important factor influencing one's participation in science. However, the value of these science capitals varies depending on the ones who possess and use them under certain contexts (Archer et al., 2015). In this study, we are particularly interested to see how family habitus and science capital potentially influence how parents view science and, consequently, what they do and talk with their children as they help them learn in the informal science activity at home.
Another factor to consider is the cultural context of non-Westen families, particularly in the Indonesian context. Indonesia is a large country with a total population of over 276 million people across 17,508 islands. There is a wide and complex array of cultures, ethnicities, geography, and languages within its population. Such diversity makes it challenging to characterize Indonesian people. In general, however, Indonesians tend to value group harmony and conformity; and this attribute is present in parent-child and family relationships. As noted by Tomlinson and Andina (2015) in their report on Parenting Education in Indonesia, Indonesians have strong family networks that go beyond the nuclear family, which is an advantage for promoting children's success in education. Indonesians have high family loyalty and are considered a collectivist society, according to Hofstede (2011). Hofstede also suggested that the classic teacher-student dynamic and a large power distance applies to Indonesia. In the context of family as a social group, it means that the family members are dependent on hierarchy; that parents or grandparents are directive; and that there are unequal rights between parents and children. Parents are respected and expected to be in control. Nevertheless, it is important to note that such generalizations may not fit every family.
In the research literature, only a few studies have examined the areas of Indonesian family involvement in children's learning, and none has specifically discussed science learning at home. According to the limited existing research, Indonesian parents are strongly involved in their children's learning at home (Yulianti et al., 2018(Yulianti et al., , 2019. Parents with higher levels of education showed higher levels of involvement (Yulianti et al., 2018). Examples of these involvements include supervising their children when they watch TV or play on a computer, participating in learning activities with their children, and reading and discussing books together at home. Such involvement by Indonesian parents has been found to have small positive effects on their children's academic achievement (Yulianti et al., 2018). The present study, therefore, contributes to the meager literature on family science learning at home in the Indonesian context.

Methodology
Participants and context of the study Six Indonesian parents with children between the ages of 4 and 9 voluntarily participated in the study. We selected this age range based on the research that highlighted the age of 4 as the critical period where children exhibit signs of emerging interest in science (Alexander et al., 2015), and the interests decline as children move from elementary to middle school around the age of 9 (Sorge, 2007). Participants were a convenience sample recruited from the researcher's community in Jakarta and Lombok through a written broadcast message. The demographics of the participants are shown in Table 1.
In this study, parent-child dyads engaged in a hands-on science activity called 'The Strongest Pillar,' in which the participants built three vertical paper pillars with different shapes: triangular prism, rectangular prism, and cylinder. Participants investigated which of the three pillars was the strongest, supporting the most weight. Scientifically, the activity involved the concept of load or force distribution, influenced by the shape of an object. For prism pillars, the load is distributed and concentrated onto the vertices of the prismthe more vertices a shape has, the more strength the shape holds the load. For a cylinder with no vertices, the load is distributed equally throughout the whole pillar so it can hold more loads.
Participants received a digital handout containing the activity's procedures, optional worksheets, colloquial explanations of the scientific concepts behind the activity, connection to the real-world building structure, and some ideas for further challenges (see Supplementary Materials 1). The activity was designed to allow parents and children to work and learn together through manipulating materials and see how it translated into their everyday life phenomena. Research has shown that when meaningful and hands-on contexts were provided, parents became more supportive of their children's science learning (Shymansky et al., 2000;Solomon, 2003).

Data collection
To capture and analyze parent-child science talk and interactions during the activity at home, we employed a microethnography approach (Bloome et al., 2005). Microethnography capitalizes on sociocultural contexts to explain how family members use language and physical materials as cultural tools through embodied behaviors to support science learning at home. Data was collected from video recordings of parent-child interactions in a science activity at home, followed by online parent interviews one week later. For health and safety during the COVID-19 pandemic, participants videotaped activities using their cameras and then sent the videos to the researcher. The semi-structured parent interviews aimed to explore parent perspectives on the observed interactions, check the researcher's interpretation, and provide another data source for triangulation. The researcher also asked questions related to parents' views of science and science learning, such as 'What is science in your view?', 'What do you think scientists do?', 'How did you learn science?', and 'What do you hope for your child to learn about science?'. Validity was established by consulting other science education experts on the interview questions (see Supplementary Materials 2 for these questions).

Researcher's role in the study
The leading researcher is both an insider and an outsider to the study. As an insider, the leading researcher shares some parts of the identity and community with the participants. For example, he grew up and received education in the same Indonesian education system up to the undergraduate level and spoke Indonesian as his first language. Furthermore, the researcher has lived in Jakarta and Lombok, where the participants reside. Thus, he shares an understanding of the larger social and cultural contexts. As an outsider, the researcher let the participants freely record themselves without our presence to maximize their natural interactions. Young children, particularly, often act differently when someone outside of their social group is present.

Data analysis
Semantic content analysis or thematic discourse analysis (Lemke, 1990(Lemke, , 2003Yin, 2014) is employed to analyze the videos of parent-child interactions and parent interviews. All the data were first transcribed and translated from Indonesian to English by the lead researcher. The video data were analyzed through a hybrid thematic analysis utilizing both an inductive, data-driven approach by adding new themes that emerge from the data and a deductive approach using pre-determined codes from previous research on parent-child science talk. Deductive codes included questioning (Callanan et al., 2017), explanations (Tenenbaum & Callanan, 2008), making connections (Callanan et al., 2017), and scientific practices of the participants (Strickler-Eppard et al., 2019;Vandermaas-Peeler et al., 2019). An utterance-by-utterance analysis of the parentchild interaction was conducted (Bloome et al., 2005), focusing on how they acted and reacted to each other, constructed meaning, and engaged in scientific practices together (see Table 5 in the Findings section for an example segment of the talk analysis). An utterance constitutes one clause or sentence. Based on that, the talk was then categorized into bigger themes such as explanation, connection-making, corrective feedback, and science practices. We developed a coding scheme to categorize parents' support for children's science learning. A team of science education experts was engaged in iterative discussions and revisions about the coding scheme to ensure the reliability of the analysis (see Supplementary Materials 3 for the coding scheme). We also conducted member checks throughout the study. Participants were given the opportunity to discuss the researchers' interpretation of the video data. Furthermore, the participants also checked the written results and analysis, where they could indicate whether the interpretations were accurate and made sense.

Findings
The nature of Indonesian parent-child interactions in a science activity at home In all family participants, the parents dominated and led the interactions during the activity. Parents' utterances constitute 77% of the dyads' talk throughout the activity, on average (see Figure  1). Parents' talks were composed of statement-making, instructions, narrations of what was occurring, and questions to the child. The children talked to answer parents' questions, comment on what happened, ask questions a few times or talk about something else. On rare occasions when the child asked productive questions, such as 'why is the pillar not crumpled?', the questions were either answered shortly or not even responded to by the parent(s).
Across the family participants, although the informal science activity was done by the parentchild dyads, other family members voluntarily participated at some points. In Dit's family, the child excitedly ran to another room, calling his grandfather to join them. The father in Sep's family attended closely to the mother-child dyad's interactions and argued with the mother about the correct term for rectangular prism, which he understood differently. In Jun's father-child dyad, the mother assisted his husband using a kitchen scale to weigh the loads and helped the child connect the strength of pillars to the child's interest in superheroes. In Rom's family, the sister shared responsibility with the father in conducting the activity for her younger brother. The child in Far's family improvised the activity with his brother by changing the paper materials into cardboard so that the pillars held more loads. We will discuss in a later section how these findings may be related to the collectivist culture of Indonesia.

Parent supports for children's learning through science talk
How do parents support their children's science learning through their talk during informal science activities? We found that Indonesian parents capitalized on the use of explanations and connection of the science activity with prior knowledge and experiences. In their explanatory talk, the parents in each family tended to provide simple causal explanations 1-3 times during the activity, which lasted for 4-17 minutes. Only two out of six parents used scientific explanations frequently, accounting for 7-10 times throughout the activity. Besides providing explanations, we observed that parents used open-ended and close-ended questions to request and encourage their children to build explanations. When the children initiatively asked for explanations, however, the parents delayed their responses until the activity concluded. With regards to strategies, connecting the activity to real-world objects or events in their everyday lives was found to be the most common science talk during the at-home activity. Parents accomplished this strategy by initiating the connection for the children and encouraging them to make the connections through guided questions. Furthermore, parents built on their children's prior knowledge to introduce new concepts. For instance, the parents built on children's understanding of a triangle (a 2D shape) to introduce a triangular prism (a 3D shape) and explained that a triangular prism is a triangle with a space or a triangle but tall.
While explanations and making connections are also common strategies used by Western and Mexican parents (Callanan et al., 2017;Fender & Crowley, 2007;Siegel et al., 2007;Van Schijndel & Raijmakers, 2016;Willard et al., 2019), we observed frequent corrective feedback in Indonesian parents' science talk. Across all the families, we counted that the parents used such feedback 35 times, with five to six times on average per family. Examples and frequencies of corrective feedback, making connections, and explanatory talk used by Indonesian parents in this study are presented in Table 2.
These strategies to support children's understanding were clear in all participants' science talks. However, the parents used particular strategies to a different extents. The ways the parents engaged children in the science activity also varied across families (see Table 3 for a complete list of identified parent supports to engage children in the activity). One factor that might contribute to this variation is the parents' views of science and science learning. We noted parents' view of science as one of the important family sociocultural backgrounds that shaped how they carried out informal science activities with their children. In what follows, we present three family cases in which the parents have different views of science and science learning, and, therefore, we observed different interactions, parent supports, and science practices emphasized throughout the activity. The three families were selected to illustrate that despite having a considerable amount of science capital, the ways the parents viewed science are demonstrated in unique interactions and science talks during the activity.

Bud family
The parent in Bud's family has a bachelor's degree in chemistry education and is an experienced elementary school teacher and teacher trainer. Recently, he was awarded a Teacher Prize by a nationally recognized education company. He and his family had done many science activities together at home. In other words, he possesses a high science capital.
As the activity started, the dyad was sitting side-to-side on the floor behind a coffee table with three different paper pillars and books on the table. The parent began by stating a task they would engage in, 'we will do an experiment'. They then talked about the materials used in the activity, and the parent guided his child to put books on top of three different paper pillars. After the last standing pillar collapsed, they counted and compared the number of books held by each pillar before it collapsed. Finally, they concluded by identifying the strongest pillar based on the number of books the pillars could hold without collapsing. The last part of the activity was the parent's explanatory talk and action planning for future activities. During this five-anda-half-minute interaction, the parent's utterances constitute 78% of the total 128 exchanges between the dyad. The parent guided his child in science practices, such as doing observations, making predictions, testing the predictions, analyzing results using mathematics, and drawing conclusions (see Table 4 for examples).  So, which one is the strongest? Really? Which one is the strongest and can hold how many books?
Asking C to make a conclusion Requesting C to use evidence as a part of the conclusion Note. P = Parent. C = Child.
Throughout the interactions, we noted the parent reinforced the idea of proving predictions through science experiments. He used the word 'prove' and encouraged the child to make predictions to be proven later. From the beginning, the parent highlighted that the activity, which they called an 'experiment', would prove the child's prediction (see Table 5 for detailed analysis). Rhetorical questions were then used throughout the experiment to encourage the child to make predictions. For example, 'Now, how about two books?' and 'Let's try to put more. How many books can it hold?' At the end of the experiment, the parent suggested a follow-up activity to see the shape of pillars in buildings, in which he encouraged the child to make and prove his prediction in a real-life setting through an observation (see the exchanges below). P: Next time, we will try to see Do those buildings use pillars with a shape like this? C: I don't think so P: You don't think so P: We will see them next times Why is this idea of science practices as a way to prove predictions or hypotheses well-grounded in this dyad's science talk? We found that family habitus plays a role in what the parent did and talked about throughout the activity. In the interview, the parent noted, 'In our previous activities together, we used to do the O-I-C approach: Observe, Imagine, and Check.' The O-I-C approach can be translated into science practices of doing observations, making predictions or hypotheses, and testing the hypotheses. The way the parent views science and scientific works was related to the O-I-C approach the family employed in science activities. When asked what scientists do, the parent responded, ' … usually related to proving something. They can also combine something Highlight the experiment will prove a prediction P restates the activity's objective; Experiment will reveal C's prediction Note. P = Parent, C = Children.
to make something new.' This view of science was related to the experience he gained from obtaining his qualification in chemistry education. I used to study science, starting with 'why' questions. From these questions, I tried an experiment to prove the questions. After doing the experiment, I got the answer. Then there was proof. From the evidence (proof), an evaluation was carried out. (Parent interview) The parent's view of science translated further into his efforts to support the child in learning science during this pillar activity and their prior informal science activities together. Moreover, as the parent mentioned his hope for the child regarding learning science, I hope when he studies science, he can finally become skilled at proving things by himself. So, when he finds a problem, I hope he can try to find out why it happens. So he can think even more critically.
In the activity, Bud's parent emphasizes learning science as doing experiments and testing hypotheses. Using repeated questions asking for the child to make a guess of what would happen and followed by 'why' questions, he assisted the child in making sense of the everyday idea of guessing and proving into a more abstract scientific concept of hypothesis testing. At the same time, the parent used the word 'experiment' to label the activity they engaged in, or as Vygotsky's (1987) termed, 'descending scientific concepts to the concrete.' The specific parent's support and emphasis are unique to this family, where the parent acquired and held such a view of science from his accumulative training, experiences, and beliefs as a science educator.

Sep family
The mother in Sep's family also has a bachelor's degree in chemistry education. She was a former high school chemistry teacher for four years before she quit the job when her first son was born. In her words, 'I chose to stay home to accompany my children to learn.' Currently, she is still teaching science and math for K-12 children through private tutoring occasionally. Her husband earned a bachelor's degree in chemistry and currently works at a high school science laboratory. The child has routine science learning activities with his mother at home and is often brought by his father to the laboratory and many science events. Parents' significant possession of science capital shapes a family habitus that surrounds the child with science.
During the pillar activity, the interactions between the mother-child dyad lasted for about seventeen minutes with 232 verbal utterances, among which 71% were parent utterances. Unlike other families in which the parents had already built the paper pillars before the activity, this dyad started the activity by making three pillars from scratch. The parent outlined the papers for the child to fold into three different shapes. During this pre-activity, the parent questioned the child about 2D shapes and built on the child's answers to explain 3D shapes. Once the pillars were ready, they engaged in similar science practices as Bud's family above: making predictions, testing predictions, and drawing a conclusion. However, they are the only dyad who talked about the detailed reasons for the cylinder pillar being the strongest. The mother used the last part of the activity to explain the scientific reasons behind this phenomenon, as shown below. In the parent interview, we checked with the mother whether such an explanation was unique to this activity. As we found out, their previous science activities also involved such explanations. The mother frequently explained a science phenomenon with a 'simple explanation'. According to her, the child commonly asked 'how' and 'why' questions, demanding explanations from the parents. As she stated in the interview, It [explanation] is common for him. Previously, we did a Rainbow Sugar experiment. He asked, how can the sugar become a rainbow? So, I explained it. I didn't explain the density concept. Just an explanation that was enough for him to understand: because the amounts of sugars were different. A simple explanation. Or when we did the Volcano experiment, he asked, how can the foam go out? I was like, when the baking soda is mixed with the vinegar, they will react and form those foams. Simple explanation.
The way the parents in Sep's family supported their child to make sense of science phenomena through explanations was also related to how they viewed science and science learning. According to the parents, they see science as 'knowledge related to phenomena around our everyday lives'. As the parent believed, scientists produce science knowledge from experiments to see how something happens. When explicitly asked about her intention and hope for her child in learning science, the mother said, I'm more focused on teaching science for him to understand natural phenomena that happen and why they happen. For example, eggs can float and sink. That's because of the density. So, we make science activities for him to satisfy his curiosity.
In this parent's view, science explains interesting phenomena in daily life. This view creates a necessity for the parent to explain why the cylinder is the strongest structure. It also serves as a motive to connect science phenomena to real-world situations through her talks and follow-up activities until the child satisfies his curiosity. We found 25 out of 51 making connection snippets presented in Table 2 produced from this family. Such talk was evident when the parent tried to explain the academic concept of three-dimensional prisms by building on the child's existing knowledge of twodimensional shapes and the real-world objects, such as tunnels, to illustrate the concept of space. Furthermore, without mentioning an abstract concept of 'force or load distribution,' the parent explained and showed the causal relationship between the number of vertices in a shape and its strength to 'hold the load of your books.'

Jun family
The father in Jun's family has a master's degree in mathematics education and is a lecturer at the local university. He occasionally did science-related activities with his child at home. However, the activities were spontaneous rather than pre-planned or structured. He saw some science activities for kids on the internet and tried them with his child only when he recalled them later and had the materials in hand. For example, when they had a jar of candy at home, he showed his child how candies dissolve differently in various kinds of solutions (e.g. cold water, hot water, oil). Another example is when he was about to take a vitamin C tablet, he called up his child and showed him how the tablet slowly dissolves in water.
Showing the child unique science phenomena seems typical for this dyad's interactions in their informal science activities. Similarly, during the pillar activity, the parent demonstrated the activity to the child. Throughout the ten minutes of interactions, the parent did all the important steps: arranging the pillars, putting books on top of them, weighing the books on a kitchen scale, writing down the results, and stating the conclusion. Parent utterance constitutes 83% of the total 92 utterances between the dyad. The vignette below shows a typical interaction between this dyad during the pillar activity. Such interaction, in which the parents did the activity's procedure, accounted for 87.5% of the total 40 counts of the dyad's engagement with the materials during the activity. As shown, the parent wanted the child to observe the science activity he demonstrated. The child was an observer and sometimes talked aloud about what he observed (see lines 17, 19, and 21 in the vignette). Line 27 further illustrated that the parent expected the child to observe what happened and delayed the child's participation at that moment as he assumed the child could halt the activity before it showed the intended phenomenon. However, when the moment was appropriate, the parent allowed the child to participate. For example, in the below excerpt, the child read the readings on the kitchen scale when the parent measured the weight of the cylinder pillar's loads (see line 68). This type of interaction in informal learning activities can be seen as what Paradise and Rogoff (2009) called 'learning by observing and pitching in.' In this approach to learning, the child is included in the activity where he is actively watching and perceiving, and then learning by contributing to the ongoing activity. Paradise and Rogoff (2009) noted that this way of learning is particularly common in indigenous communities. In the parent interview, our interpretation was confirmed when the parent expressed his views about science. As shown in the quote below, the parent viewed observations as a critical part of science practices. From his perspective, science is a series of observations, analyses, and conclusions.
Science is observations. Scientists observe something and then make a conclusion. For example, he sees an object falling, like Newton; why this object falls to the ground? From that observation, he then analyzes. If needed, he runs an experiment. Then make a conclusion. Science is observing, analyzing, and making a conclusion.
According to the parent, a child engaged in science activities so that he could successfully draw a conclusion from the events that he observed. He stressed a causal relationship behind a phenomenon as an appropriate conclusion from a science activity. In the pillar activity, he expected the child to infer a causal relationship between the number of books and the rate at which the pillars collapse. In his words, I wanted to teach him to observe a cause-and-effect relationship from an event. For example, when objects don't have a load, they stand. When they're given a load, one falls, but the other ones stand. When we add more loads, it (eventually) falls. So, he can make a conclusion that with heavier loads, the buildings or structures will fall. Increasing loads makes them fall.
The parent hoped that the child could think and make such a conclusion in his mind from close observations of the falling pillar phenomena the parent demonstrated to him in this activity. His view that science knowledge is generated from careful observations contributes to his science talk and non-verbal behaviors during the activity. He focused heavily on making sure the child paid attention to what happened, which he thought was a foundation for scientific observations. He also thought the observation was necessary to further understand the abstract concept of a cause-and-effect relationship between loads and strengths. However, such abstraction was not communicated to the child yet in this activity. Considering that scientific concepts are developed over time across various settings (Vygotsky, 1987), we assume that further engagement in science activities with the parents would eventually bring the abstract scientific concept closer to bear with the everyday phenomena that the child observed in this activity.

Discussion
Indonesian parents' directive support in children's learning As the findings showed, Indonesian parents are capable of facilitating and contributing to the process of their children's conceptual development from everyday activities at home through their science talk. During the activity, we found that the parents dominated the interactions and provided a high level of directive support to their children through their talks. There are several possible explanations for this directiveness. These directive interactions may be related to Indonesian culture, which is deemed a collectivist society and has a large power distance (Hofstede, 2011). A large power distance means the less powerful members of institutions accept and expect the leaders' power. In the context of family as a social institution, the child accepts, obeys, and expects their parent's lead. This may be one possible explanation for the unequal frequencies of utterances from the parents and children and the nature of the children's utterances merely as a response to their parents. In a study with Brazilian children, who also have a large power distance according to Hofstede's country scores, Scott et al. (2006) observed that the parent-child exchanges were also dominated by the parents, in which the parent demanded his child to take his point of view toward the science lesson he presented. The fact that other family members were voluntarily involved in the parent-child dyads' science activity in our study shows the collectivist nature of Indonesian society. These family members supported the children's informal science learning at home spontaneously.
In collectivist cultures, people are integrated into strong, cohesive groups that extend beyond nuclear families (e.g. aunts, uncles, and grandparents) and continue to share loyalty to each other (Hofstede, 2011).
The directive nature of parent-child science talk, however, may also be related to other factors such as parent's education level and children's age. Parents with higher education tend to be more directive than those with basic education because they may be socialized in 'school-like' behaviors that may, in turn, structure their interactions with children (Siegel et al., 2007). Parents were also found to be more directive with younger children, but they were more collaborative with older ones (Siegel et al., 2007). In our study, while the parents with younger children read the activity handout by themselves beforehand, the parent of the older child (the nine years old) shared the handout with the child to read and follow the instructions. Parents may be tempted to scaffold younger children more than older ones in the learning interactions at home. As one parent mentioned in the study, 'I think it [the instruction] was not for his age; therefore, I helped him on the parts that were probably too hard for him.' Such perceptions were also communicated by other parents in the study in the interviews.
Parents' perceptions of their children's abilities may also be a key factor determining the high level of directive supports they provide to their children. In parent-child dyads, parents have extensive information about their children's knowledge, abilities, and shared experiences (Guberman, 2003). However, using this information to optimize the level of scaffolding was found to be challenging for parents (Salonen et al., 2007). With optimal scaffolding, parents would 'flexibly regulate their level of directiveness according to moment-to-moment progressions or regressions of the child's independent functioning' (Salonen et al., 2007, p. 80). Conversely, parents may otherwise form rigid patterns of interactions based on their initial assessment of children's abilities. In such cases, termed as 'regulatory mismatch', researchers have pointed out unfavorable socio-cognitive developmental consequences from either parental over-directiveness (too direct cueing) or under-directiveness (too vague directions) (Mattanah et al., 2005). Thus, it is important for parents to find a balance in providing assistance to their children. While scaffolding can help children going through their ZPD (Vygotsky, 1987), too much directive may undermine children's cognitive selfregulation (Neitzel & Stright, 2003).
Another apparent finding from this study is that Indonesian parents used corrective feedback frequently in the science talk while also employing other strategies commonly found in Western families, such as explanation and making connections. On the one hand, providing immediate feedback can sometimes be an effective way in science teaching to correct a child's misconceptions or incorrect understanding (Taconis et al., 2001). On the other hand, providing corrective information is classified as directive feedback by Black and Wiliam (1998), and such short and quick feedback usually terminates a productive dialogue with conclusive information. Perhaps, the long-regarded concept of wait-time in formal instructions can be borrowed by parents in informal science education settings, allowing them to step back and let their child say their thoughts and do spontaneous work (Rowe, 1974).

Parents' views of science and their supports to children's learning
We have presented three cases from the families to illustrate the relationship between parents' views of science and the scaffolding supports they provided to their children through science talk and interactions during at-home activities. The parent who viewed science as experimenting scaffolded the child to engage in the everyday practice of guessing and proving, which would facilitate the child's understanding of hypothesis-testing (Brod, 2021;Gredebäck et al., 2018). The Parent who viewed science as a body of knowledge related to everyday phenomena scaffolded the child by explaining and connecting formal science with informal ideas. The parent who viewed science as an inference-making process demonstrated the activity for the child to observe, hoping that his children could infer a cause-and-effect relationship from the observation. There may be other views of science beyond the three presented here. These individual differences in the parents' views of science are likely shaped by their different possessions of science capital, including their beliefs of what was important from science and how they learned science in the past. Parents' hopes for their children when learning science is related to what they believe as important aspects of science and, thus, what they emphasise when supporting their children's science learning. Parents' consistent emphasis on particular science learning aspects through informal learning activities may form specific epistemological framings towards science learning that children would prefer. Epistemological framings refer to the stances or approaches that individuals use in learning activities to meet their expectations towards the activities (Hutchison & Hammer, 2010). When learning science, children may prefer to look for explanations, make predictions and test them, or expect interesting science demonstrations because such a particular stance is what they acquire from everyday experiences with their families.
As a limitation, we acknowledge that the parents in this study have a relatively high level of formal schooling, as is typical in many informal science education studies. The three families also have notable science capital, and the parents are either current or former STEM educators. Our intention in this research report is to illustrate the variations of parent-child interactions even among families who possess relatively high science capital. Therefore, future studies on families with a low level of science capital for comparison will better illuminate how this capital may shape parents' view of science and their family habitus, which, in turn, may lead to certain parent-child interactions in informal settings. Longitudinal studies will enrich our understanding of the dialectical relation between everyday concepts and scientific concepts in the process of concept formation.
Science capital is not a static sociocultural characteristic (Archer et al., 2012). Individuals may acquire more or reduce their science capital throughout their lives. In other words, families can build their science capital over time, and parents' views of science may change as their science capital changes dynamically. Informal science education environments can play a critical role in building family science habitus and science capital by providing families with multiple opportunities to engage in free-choice science activities outside of school and work settings (Bell et al., 2009).

Conclusion
This study provides examples and evidence of how Indonesian parents support children's science learning through talk and interactions in an informal science activity at home. These findings expand our understanding of how non-Western families of collectivist cultures approach informal science learning. The parents in the study demonstrated directive support to their children by providing explanations, making connections, and giving corrective feedback. Vignettes from the families show that through these science talks, the parents helped children connect everyday concepts with the scientific concepts in the activity. However, as we have discussed, it is important that parents adjust their scaffolding by considering children's progression moment-by-moment. Overdirectiveness may undermine children's cognitive self-regulation, while minimal scaffolding would not move children into their Zone of Proximal Development (Vygotsky, 1987). With the advancement of information technology and the availability of online resources, parents can easily find science activities on the internet and try them out with their children at home. However, informal science educators and educational content creators may also need to educate parents with strategies for facilitating learning, such as using wait-time and asking questions that encourage children to build their own explanations or make connections.
The findings also highlight the importance of parents' views of science on children's engagement in informal science activities. We have seen some variations in parent-child interactions and parents' support for their children during the at-home science activity due to parents' unique views of science, even within high-science-capital families. Parents in the three presented families viewed science differently as hypothesis-testing practice, a body of knowledge surrounding everyday phenomena, and inference-making from observations. There may be other ways how parents view science beyond these three examples, and they may uniquely shape the interactions between parent and child in informal science learning. Parents were found to emphasize learning the aspects that they believe as important from science. Consistent emphasis on particular science learning at home may form specific epistemological framings towards effective science learning that children prefer. Then, educators must be aware of these variations in stances towards science learning to provide educational support and resources accordingly. In informal settings such as science museums, for example, the programs and exhibitions need to be designed to accommodate various science learning approaches that family visitors may take. In classroom settings, teachers' awareness of the diverse learning approaches that students bring from their everyday experiences may better prepare them to address potential tensions between the students and their own epistemological framing toward science learning. Consequently, it is important for science teacher education programs to equip the teachers with the knowledge, skills, and tools needed to identify and address students' diverse stances toward science learning.

Ethics statement
This research was approved by the Ohio State University Institutional Review Board (approval #2020B0205).

Disclosure statement
No potential conflict of interest was reported by the author(s).