**3. Can preschoolers learn to generate science knowledge?**

50 Current Topics in Children's Learning and Cognition

salience of irrelevant pieces of information.

serving upper middle class families (Baker et al., 2011).

Grossen, & Miller, 1993).

while downplaying other relations (ones that are less relevant or misleading). Importantly, models represent predictive and explanatory rules, thus making visible the components of science phenomenon that are difficult to be perceived on the basis of phenomenological experience alone. As such, they make relevant science facts salient, in effect decreasing the

There are several studies that show the effectiveness of conceptual models in young children (e.g., Gobert & Buckley, 2000; Kenyon et al. 2008; Wiser & Smith, 2008; Baker, Haussmann, Kloos, & Fisher, 2011). An illustrative example uses the science domain of material density, a concept that is defined by the ratio of the two highly salient dimensions of mass and volume. Predictably, children often ignore density and use instead perceived heaviness of an object as the sole predictor of the object's buoyancy (e.g., Piaget & Inhelder, 1974; Kloos et al., 2010). To help children overcome this mistaken focus on an object's heaviness, a conceptual model of density was developed, also known as dot-per-box (e.g., Smith & Unger, 1997; Wiser & Smith, 2008). It involves a display in which the volume of an object is represented as a certain number of boxes, and mass is represented as number of dots inside the boxes. Thus, density is represented as the spacing between dots (the more crowded the dots, the more dense the material); and irrelevant variation of color, shape, and texture are omitted. Thus, density of the material is now similar in salience to that of mass or volume. Children indeed benefited from these abstract representations of density (for a discussion of these findings, see Wiser & Smith, 2008). Similar learning success was reported with 4- to 5 year-olds, whether children were recruited from Head Start preschools or from preschools

Introducing conceptual models early on might have a positive effect on learning as children get older. Support for this claim comes from research in the domain of evaporation and condensation, another domain that is a notoriously difficult area of instruction in science (Kenyon et al., 2008). Children between 6 and 8 years of age underwent a multi-week training on evaporation and condensation, which included observing the evaporation and condensation in a soda bottle, drawing diagrams to capture the system through various moments in time, testing their models through experiments, using tools to measure the amount of water in the air, and revising their models as needed. Findings show that the instructed students significantly outperformed the uninstructed students in their understanding of relevant concepts. Importantly, when students began the formal study of science in Grade 7, instructed students improve in their understanding of concepts much faster than uninstructed students. Clearly, the students who were helped to form basic science concepts in early grades had developed an understanding of the domain that continued to facilitate their meaningful learning, further developing their understandings and reducing their misconceptions (for related discussions, see Muthukrishna, Carnine,

In sum, research-based evidence points in a clear direction when it comes to promoting an understanding of science concepts. Unlike what a Piagetian stage model of abstract reasoning might imply, young children are indeed able to learn abstract concepts early on, even when the concepts run counter to what children already believe. Their learning The second aspect of science learning pertains to understanding the process by which science knowledge is generated. Rather than learning about established and accepted science facts and concepts, this aspect includes an understanding of how science facts are generated in the first place. This includes the ability to create settings that are sufficiently informative for science knowledge to be generated. And it includes the meta-cognitive understanding of how new information can change existing knowledge. It is important to note that the process of generating science concepts is in part affected by cultural norms. Norms pertain to constraints about what to count as an explanation of events (cf., Pearl, 2009), under what circumstances to abandon an existing theory (cf. Kuhn, 1996), or how to treat expected versus unexpected observations (cf., Popper, 1959). So far, these constraints have not been studied explicitly in the realm of early science learning. Instead, the emphasis is on understanding how children's everyday interactions with their environment can help children generate science knowledge (e.g., Zimmerman, 2000).

The question of children's ability to generate science knowledge through observations is debated heavily, both in cognitive development and in educational research. In cognitive development, the ongoing debate centers on the question of whether young children are at all capable of engaging in appropriate knowledge-generating activities. Such activities, referred to as *scientific reasoning*, require the child to detect gaps in their existing knowledge base, ask questions in response to the identified gaps, carry out the experiments that could lead to an answer, and critically evaluate the evidence (cf., Klahr, 2005). Each one of these steps can be difficult for children (and even for adults), for several reasons: First, the mind is biased towards perceiving order, making it difficult to perceive disorder, missing information, or gaps (cf., e.g., Quinn, Eimas, & Rosenkrantz, 1993). Second, the mind is biased towards confirming already existing beliefs, rather than questioning their shortcomings, making it difficult to spontaneously challenge existing beliefs (e.g., Schauble, 1990). Under this view, scientific reasoning has to be trained explicitly.

In contrast to the cognitive-development debate, educational research already presupposes the child's ability to generate science knowledge, following the theoretical bent of constructivism (cf., Olson, 1996). For example, it is generally accepted that children can engage in *inquiry*, the processes of wondering, questioning, exploring, investigating, discussing, reflecting, and formulating ideas and theories (e.g., Kuhn, 2010). Indeed,

preschool education places strong emphasis on exploration, the idea being that exploration is at the center of what allows children to generate knowledge about the world (e.g., Luken, Carr, & Brown, 2011). The debate then centers on the question of how children's spontaneous inquiry can be supported by teacher interventions, explicit instruction, and/or feedback. In what follows, we review findings that speak to this question. In particular, we focus on the efficacy of three strategies that help children generate science knowledge. They include (1) engaging children in scientific discourse, (2) teaching them to keep track of their observations, and (3) helping them organize their knowledge.

Preschoolers Learning Science: Myth or Reality? 53

flowed around the circuit, both through the reading of factual books, and through one-onone interactions with the teachers as children explored the materials. In particular, after direct instruction, all children could talk about how the electricity continuously flowed around a simple electric circuit, an understanding they still held 2 months later. Without such direct instruction, and despite being able to connect the circuit, children's beliefs about electricity, measured through their entries in a 'notebook' did not change over the course of three months: All children with comparison data expressed exactly the same view of

As this previous study implies, a scientific discourse is aided when children can keep track of their explorations across time, for example in science journals. This is not only because children are exposed to richer information over time (e.g., to detect a contrast in an abstract dimension that is not available in snapshot events). Science journals also provide teachers with an opportunity to engage children in a targeted and individualized way (cf., Doris, 1991; Light & Simmons 1983). Recent findings show that even preschoolers can be taught to keep a science journal (e.g., Brenneman & Louro 2008). This is might involve over-simplified drawings and an explanation written by a teacher. Nevertheless, they provide an authentic medium for teachers to encourage scientific reasoning. For example, children's drawings make it possible for teachers to convey the difference between true observations and a child's imagination, the importance of dating the entries, and the focus on hidden features (e.g., the inside of a pumpkin). The outcome was a rich interaction between teachers and preschoolers, promoting children's ability to observe objectively, to record observations

So far, we have discussed the benefits of engaging children in science talk and encouraging them to make note of their observations. An even more targeted way of helping children generate science knowledge involves helping them organize their observations and beliefs in a systematic way. It involves the use of so-called concept maps (cf., Novak, 2010). Concept maps consist of nodes (to represent objects) and arrows between nodes (to represent events). The resulting flow charts (e.g., 'plants go in gardens') can be organized hierarchically to capture the relations between what children think and experience. For example, when presented with a targeted question about the cycle of growth, concepts maps make it possible for children to organize information about animals and plants in a way that yields both relevant science knowledge and an understanding of

The Young Florida Naturalist Program used such a concept-map strategy with 3- to 4-yearolds from urban early childhood center children (e.g., Hunter, Monroe-Ossi, & Fountain, 2008). The goal was to increase children's knowledge about the butterfly life cycle and plant growth over the course of an eight-week instructional period. A concept map was constructed first, using a set of pertinent pictures (e.g., tree roots, leaves, caterpillars, butterflies, cocoons), to capture children's initial understandings. The concept map was then posted in the classroom as a point of reference, and to allow for modifications as children learned new information. To stimulate children's thinking, a butterfly garden was planted on the center's grounds, and children engaged in various experiments with water, sunlight,

with some precision, and to become aware of patterns of change..

how such knowledge could be generated.

'electricity' in both interviews.

Can young children engage in scientific discourse? What is also referred to as 'science talk' (Lemke, 1990) or exploratory language (Peterson & French, 2008), scientific discourse differs from a standard question-answer format for which the person asking the question already knows, expects to hear, and rewards the right answer. Scientific discourse instead promotes sense-making of events: accepted science terms, concepts, and methods are transmitted in the context of children asking scientifically valid questions (cf., Crowder, 1996). It is therefore a large part of inquiry and scientific reasoning. While scientific discourse is heavily studied in school-aged children (e.g., Kafai & Carter Ching, 2001), preschool teachers and parents of young children are likely to engage in such discourse naturally, as they guide children's explorations of topics relevant to their everyday life. The ideal context is likely to require an active and knowledgeable listeners and a shared science vocabulary that children are sufficiently comfortable with (cf., Crowley, Callanan, Jipson, et al., 2001; Fleer, 1996; Pramling & Pramling-Samuelson, 2001).

The effectiveness of science talk with preschoolers was demonstrated empirically with the domains of metamorphosis and plant growth (e.g., Witt & Kimple, 2008). Children were first given a pretest on these two science domains, asking questions such as *What is a cocoon?, What kind of food helps caterpillars and butterflies grow?, How does a plant soak up water?*, and *Do seeds grow faster in direct sunlight, darkness, or a mixture of both?.* Preschoolers then participated in two several-weeks-long activities, one pertaining to creating an environment for caterpillars and observing the metamorphosis, and one pertaining to planting seeds and observing the growth in different climates and environments (e.g., hot and sunny vs. cold and dark). Preschoolers were frequently engaged in conversations, allowing them to reflect on their observations and experience. Post-test performance showed remarkable improvement in science knowledge: Every child improved in their answers, and every question showed gains. For example, while none of the preschoolers knew the meaning of 'cocoon' or the conditions under which seeds grow best, all of them did so during the posttest.

Promising results for the effectiveness of science talk were also reported with the domain of electrical currents (Fleer, 1991; Fleer & Beasley, 1991). The task was to explore flashlights, find out what they are made out of and how they work, and construct their own flashlights using batteries, bulbs, and wires. With the use of guided interaction with the teachers, preschoolers and 1st-graders learned to formulate questions about the working of the flashlights, and they learned to report the findings of their own explorations. Even better results were obtained when children were given direct instruction on how the electricity flowed around the circuit, both through the reading of factual books, and through one-onone interactions with the teachers as children explored the materials. In particular, after direct instruction, all children could talk about how the electricity continuously flowed around a simple electric circuit, an understanding they still held 2 months later. Without such direct instruction, and despite being able to connect the circuit, children's beliefs about electricity, measured through their entries in a 'notebook' did not change over the course of three months: All children with comparison data expressed exactly the same view of 'electricity' in both interviews.

52 Current Topics in Children's Learning and Cognition

Pramling & Pramling-Samuelson, 2001).

posttest.

observations, and (3) helping them organize their knowledge.

preschool education places strong emphasis on exploration, the idea being that exploration is at the center of what allows children to generate knowledge about the world (e.g., Luken, Carr, & Brown, 2011). The debate then centers on the question of how children's spontaneous inquiry can be supported by teacher interventions, explicit instruction, and/or feedback. In what follows, we review findings that speak to this question. In particular, we focus on the efficacy of three strategies that help children generate science knowledge. They include (1) engaging children in scientific discourse, (2) teaching them to keep track of their

Can young children engage in scientific discourse? What is also referred to as 'science talk' (Lemke, 1990) or exploratory language (Peterson & French, 2008), scientific discourse differs from a standard question-answer format for which the person asking the question already knows, expects to hear, and rewards the right answer. Scientific discourse instead promotes sense-making of events: accepted science terms, concepts, and methods are transmitted in the context of children asking scientifically valid questions (cf., Crowder, 1996). It is therefore a large part of inquiry and scientific reasoning. While scientific discourse is heavily studied in school-aged children (e.g., Kafai & Carter Ching, 2001), preschool teachers and parents of young children are likely to engage in such discourse naturally, as they guide children's explorations of topics relevant to their everyday life. The ideal context is likely to require an active and knowledgeable listeners and a shared science vocabulary that children are sufficiently comfortable with (cf., Crowley, Callanan, Jipson, et al., 2001; Fleer, 1996;

The effectiveness of science talk with preschoolers was demonstrated empirically with the domains of metamorphosis and plant growth (e.g., Witt & Kimple, 2008). Children were first given a pretest on these two science domains, asking questions such as *What is a cocoon?, What kind of food helps caterpillars and butterflies grow?, How does a plant soak up water?*, and *Do seeds grow faster in direct sunlight, darkness, or a mixture of both?.* Preschoolers then participated in two several-weeks-long activities, one pertaining to creating an environment for caterpillars and observing the metamorphosis, and one pertaining to planting seeds and observing the growth in different climates and environments (e.g., hot and sunny vs. cold and dark). Preschoolers were frequently engaged in conversations, allowing them to reflect on their observations and experience. Post-test performance showed remarkable improvement in science knowledge: Every child improved in their answers, and every question showed gains. For example, while none of the preschoolers knew the meaning of 'cocoon' or the conditions under which seeds grow best, all of them did so during the

Promising results for the effectiveness of science talk were also reported with the domain of electrical currents (Fleer, 1991; Fleer & Beasley, 1991). The task was to explore flashlights, find out what they are made out of and how they work, and construct their own flashlights using batteries, bulbs, and wires. With the use of guided interaction with the teachers, preschoolers and 1st-graders learned to formulate questions about the working of the flashlights, and they learned to report the findings of their own explorations. Even better results were obtained when children were given direct instruction on how the electricity As this previous study implies, a scientific discourse is aided when children can keep track of their explorations across time, for example in science journals. This is not only because children are exposed to richer information over time (e.g., to detect a contrast in an abstract dimension that is not available in snapshot events). Science journals also provide teachers with an opportunity to engage children in a targeted and individualized way (cf., Doris, 1991; Light & Simmons 1983). Recent findings show that even preschoolers can be taught to keep a science journal (e.g., Brenneman & Louro 2008). This is might involve over-simplified drawings and an explanation written by a teacher. Nevertheless, they provide an authentic medium for teachers to encourage scientific reasoning. For example, children's drawings make it possible for teachers to convey the difference between true observations and a child's imagination, the importance of dating the entries, and the focus on hidden features (e.g., the inside of a pumpkin). The outcome was a rich interaction between teachers and preschoolers, promoting children's ability to observe objectively, to record observations with some precision, and to become aware of patterns of change..

So far, we have discussed the benefits of engaging children in science talk and encouraging them to make note of their observations. An even more targeted way of helping children generate science knowledge involves helping them organize their observations and beliefs in a systematic way. It involves the use of so-called concept maps (cf., Novak, 2010). Concept maps consist of nodes (to represent objects) and arrows between nodes (to represent events). The resulting flow charts (e.g., 'plants go in gardens') can be organized hierarchically to capture the relations between what children think and experience. For example, when presented with a targeted question about the cycle of growth, concepts maps make it possible for children to organize information about animals and plants in a way that yields both relevant science knowledge and an understanding of how such knowledge could be generated.

The Young Florida Naturalist Program used such a concept-map strategy with 3- to 4-yearolds from urban early childhood center children (e.g., Hunter, Monroe-Ossi, & Fountain, 2008). The goal was to increase children's knowledge about the butterfly life cycle and plant growth over the course of an eight-week instructional period. A concept map was constructed first, using a set of pertinent pictures (e.g., tree roots, leaves, caterpillars, butterflies, cocoons), to capture children's initial understandings. The concept map was then posted in the classroom as a point of reference, and to allow for modifications as children learned new information. To stimulate children's thinking, a butterfly garden was planted on the center's grounds, and children engaged in various experiments with water, sunlight,

soil, etc. to explore plant growth. At the end of the period, children's understanding was assessed in semi-structured interviews to infer a child's individual concept map. For example, children were asked to sort and organize a set of pictures, they were encouraged to talk about what they know about plants and butterflies, and they were asked about their understandings of the final class concept map (e.g., 'what do the pictures tell you about plants?"). Results document that a large majority of both 3- and 4-year-olds could make higher-order propositions, and they could recall terms and concepts relevant to parts of plants and aspects of butterfly transformation.

Preschoolers Learning Science: Myth or Reality? 55

children's already existing ideas about science (cf., Davis & Krajcik, 2005). Second, science learning is aided when intentional teaching is incorporated with play, such that teaching practices not only become purposeful and thoughtful, but also engage young children with topic-specific phenomena and inquiry (cf., Bodrova & Leong, 2007; Crowley & Jacobs, 2002; Copple & Bredekamp, 2009; NAEYC, 2009). The promise is to make accessible relevant science concepts to young children – even abstract concepts and those that run counter to already existing beliefs – forming the foundation upon which young learners will construct

Research on early science learning also highlights the gaps that still remain in our understanding of children's learning (cf., Davis, 2009). In fact, existing efforts to measure early science learning might be merely a first step. A more complete understanding calls for findings on how to best organize a child's science education throughout the curriculum, how to measure their progress across science domains, how to harness individual differences among children, and what kind of early exposure leads to long-term gains in science learning. Related, empirical questions still remain about how inquiry and explorations interface with direct instructions of science concepts, and how a child's attitude towards science learning both affects and is affected by learning of science. Without research-based findings to speak to these issues, our intuitions about early science learning, while fueling arguments among various viewpoints, might nevertheless jeopardize

Working on this paper was partly funded though awards from the National Science foundation to HK (DRL 0723638) and to VC, RB, and HK (DRL 1114674). Please address

Acher, A., Arca, M., Sanmarti, N., (2007) Modeling as a teaching learning process for understanding materials: a case study in primary education. *Wiley InterScience*. *9*, 1398-

their ideas later in life (cf., Lucas, 1993).

progression the area of early science learning.

Heidi Kloos, Heather Baker and David Pfeiffer

Eleanor Luken and Victoria Carr

*Department of Psychology, University of Cincinnati, Cincinnati OH, USA* 

*Early Childhood Education, University of Cincinnati, Cincinnati OH, USA* 

*Educational Studies, University of Cincinnati, Cincinnati OH, USA* 

correspondence to Heidi Kloos (heidi.kloos@uc.edu).

**Author details** 

Rhonda Brown

**5. References** 

418.

**Acknowledgement** 

Taken together, these results show that young children can benefit from quite sophisticated adult support in order to generate science-relevant knowledge. That is to say, children's natural curiosity can be harnessed to help them explore their surroundings in scientifically appropriate ways. Adult support can range from merely providing children with a context for explorations to engaging them in guided discussions about their explorations, to helping them document their findings and organize their thoughts. As part of this process, young children are likely to learn the cultural norms of what is considered science, what counts as a worthwhile phenomenon, how it should be explored and evaluated, and what kind of knowledge construal would be acceptable (i.e., what can be ignored and what must be included). Open questions pertain to the relative benefit of allowing children to develop their own representations of what they observe versus working with representations provided to them in a top-down fashion.
