Teacher discourse strategies used in kindergarten inquiry-based science learning

Various instructional strategies used in the generation of explanations by children have been examined by science educators. Numerous studies have shown that when teachers provide students with the appropriate scaffolding during learning, the children are likely to engage in the generation of explanation (Gagnon & Bell, 2008; King, 1994; McNeill, Lizotte, Krajcik & Max, 2006; Tabak & Reiser, 1999). However, most of these studies focus on middle school children and older rather than kindergarten children (King, 1994, McNeil et al., 2008; Tabak & Reiser, 1999). Studies that investigate the role of instructions in preschool children focus on linguistic development of explanatory discourse rather than on the development of categories or kinds of teacher discourse strategies. They do not specify forms of scientific explanations from non-scientific explanations (Beals, 1993; Peterson & French, 2008).

The purpose of this study is to examine teacher discourse strategies used in inquiry–based science learning. Being able to provide explanation is a key form of knowledge construction and is also regarded as a key goal of inquiry in order to understand natural phenomena (Sandoval & Reiser, 2004). During discourse, prompting can also be used as a part of instructional materials, for example prompts that are generated from computer assisted instruction as well those that are embedded within written task instructions. It must also be noted that these are not necessarily limited to prompts that are provided by teachers during discourse.

Theoretical Framework

This current study focuses mainly on teacher discourse strategies used in inquiry–based science learning. Several researchers have suggested that scaffolding is needed from teachers for students to generate scientific explanation successfully. For example; the teachers can provide students with prompts and a chance for the students to explain their thoughts (Bell, Semetana, & Binns, 2005; Gagnon & Abell, 2008). Ogbon, Kress, Martins, and McGillicuddy (1996) discussed various ways in which teachers can engage students in explanation during science learning, for instance, by showing counterintuitive results, using examples, and asking for clarification.

Walsh (2013) noted that theories are important and there are several types of theories which include grand theories, case theories, and mid-range theories. Lyons (2018) defines theory as a set of ideas that are organized which tries to provide explanation for a phenomenon. Smith and Hamon (2017) define a theory as a tool that is used for describing and understanding the world. They also stated that a theory is regarded as a general framework with ideas that explain how they are connected to each other. In addition, Smith and Hamon (2017) highlighted that theories could be used for asking and answering questions regarding specific phenomena. In another study, White et al., (2015) explained several types of theories and state that a scientific theory is as “a set of systematically related propositions that are empirically testable” (p.6). We do acknowledge that there are several theories that exit and are important to research. However for this particularly study we focus on constructivist theory and grounded theory. Constructivist theory is important to inquiry learning (Tillinger, 2013) and children successfully construct higher levels of knowledge when they are actively trying to master their world (Lightfoot et al., 2013). Several researchers (Glaser, 2017; Glaser, & Strauss, 1967; Denzin & Lincoln , 2003) have written extensively on grounded theory. Denzin and Lincoln (2003) explained that grounded theory is important in qualitative revolution and qualitative research provides a systematic social scientific inquiry. Denzin and Lincoln (2003) also mentioned that grounded theory methods include systematic inductive procedures that are used for the collection and analysis of data. They agreed that “grounded theory is durable because it accounts for variations; it is flexible because researchers can modify their emerging or established analysis as conditions change or future data are gathered” (p.252).

Teacher discourse strategies are important and should not be overlooked. However, there are little to no studies that examine the relationship between instructional strategies and students’ generation of scientific explanation during science learning in kindergarten. In an attempt to fill the gap in the literature, this study explores different kinds of teacher discourse strategies and examines how teacher discourse during inquiry-based kindergarten science instruction facilitates the development of scientific explanations among kindergarten students.

Teacher Scaffolding and Modeling of Scientific Explanations.
Research has documented the importance of investigating the construction of scientific explanation (Faye, 2014) with the aid of teacher’s scaffolding. For example, Tabak and Reiser (1999) conducted a study which examined how teacher instructional strategies related to high school student’s construction of biological explanations. They found that establishing opportunities and expectations for the construction of explanation are very important during classroom science discourse. Further, Tabak and Reiser (1999) argued that teachers need to assist students in generating explanations. They noted that the teacher in their study used various strategies to support students’ construction of explanation during learning. These included using a set of norms to guide the production of high quality scientific explanations. Specific instructional strategies that were used included guided prompts, general elaboration prompts, specific elaboration prompts, restating driving questions, critiquing and questioning students’ initial statements, encouraging causal explanations, synthesizing and re-voicing the remarks of students. Their results showed that students can be trained to generate scientific explanations although they depend heavily on the teacher to assist with the explanation construction process.

In a similar study, Renkel (2002) investigated the benefits of learning when high school students are trained to produce scientific explanations of high quality when they are provided with examples such as instructional explanations. The participation in his study included 48 students who were working on how to solve problems dealing with probability. There were 12 males and 36 females in the study who completed both pre- and post- test problems on probability. The experimental group was given self-explanation activity supplemented with instructional explanations (SEASITE) principles, which is an explanation training program that uses a group of instructional principles that are used in example based learning as well as the developmental of instructional explanations (Renkl, 2002). Similar to the experimental group, the control group was also involved in self-explanation however; this group did not receive the SEASITE principles. The results of Rankle’s study showed no differences exist between the two groups of students. He concluded that students in the self-explanation group did just as well as those who were given supplemental instruction that includes good examples.

Several studies have investigated how teacher prompts can facilitate the generation of explanation by students during science learning (e.g., Chi, DeLeeuw, Chuiu & LaVancher, 1994; Chi, 2000; Chi, Siler, Jeong, Yamauchi & Hausmann, 2001; McNamara, 2004; Nokes et al., 2011; Palincsar & Brown, 1984; Van Meter, 2001). Chi and colleagues’ (1994) study showed that compared to students who were not prompted to provide self-explanation, students who were provided with prompts to self-explain learned better and showed a deeper understanding of the course content. It must be noted that prompts can be incorporated into instructional materials for example computer assisted instructions as well as written task instructions and are not restricted to prompts provide by teachers during discourse.

Chi (2000) investigated college physics and college biology courses in order to determine whether there were differences among the two courses as it relates to instructional prompts and students’ self-explanation. Chi found that the students were prompted in generating self-explanations more in biology in comparison to physics course. Chi (2000) noted that in both physics and biology classes, students who produced self-explanation showed better performance in the course than those students that did not produce self-explanation.

Aleven, Koedinger and Cross (1999) used the PACT geometry tutor to determine how students produced explanation (Aleven et al., 1999). The PACT geometry tutor is a curriculum which is made up of a variety of geometry topics which included angles, circles as well as Pythagorean Theorem. The PACT geometry tutor was designed to help students in reasoning and explanation of answers. Aleven et al. (1999) stated that students’ comprehension of geometry will improve if they are trained to give explanations for answers. Both of the experiments assessed the effectiveness of teaching students how to explain their answers during student learning. In the first experiment 41 high school students enrolled in a high school geometry who were placed either in an experimental group (reason condition) or control group (answer only condition). Aleven et al. (1999) concluded that there were benefits to teaching students how to provide explanations for their answers. Students who were in the experimental group (i.e., students who were asked to give reasons for their answers) performed and learned better during problem solving in comparison to the control group (i.e., students who were not asked to give reasons for their answers) group. Due to time constraints in the first experiment the control group completed the problem in a shorter period because they were not required to generate explanations. Alvin and colleagues’ (1999) second experiment consisted of 53 high school students who were in two geometry courses that were controlled by time factor which involve a seven hour time period for students in both experimental and control groups. The results revealed that those students from the experimental group had gained significantly from the pre-to posttest. They also performed much better on problems that required reasoning than the control group.

Bielaczyc, Pirolli and Brown (1995) investigated self-explanation strategies as well as self-regulation that students used while using the Lisp Tutor. The Lisp Tutor is a program which is made up of variety of programming instruction as well as exercises which allow the students to write Lisp code. There were 24 students who recently graduated from a university that participated in the study. There students were divided into instructional and control groups. The instructional group consisted of 11 participants while the control group consists of 13 participants. During the exercise the participants were provided with prompts in order to read aloud as well as state their ideas verbally.

Bielaczyc et al. (1995) also investigated the role explanation plays in learning. They investigated the effect of instructional strategies which facilitated students’ explanations during learning at a college level course in programming. The study consisted of several stages. At the introductory stage, students were allowed to practice and think out loud. The second stage, which is called the pre-intervention stage, involved the collection of data on students’ explanation and performance while participants’ study as well as explained the help sections in the instructional manual. The third stage which is called the instructional stage (received special training in self-regulation and self-explanation as well as Lisp tutoring) or involves the assignment of participants to an intervention or control group (received Lisp tutoring). The intervention group was given explicit training strategies which consisted of structured one-to-one interaction by the experimenter and each student. The learning strategies entail the elaboration and identification of the existence of relationships between major point and during the use of text. It also involves the generation of meaning and forms in order to code the Lisp as well as connect concepts with the aid of examples and text. A variety of questions were asked on self-explanation as well as self-regulation strategies (e.g., students were asked to provide explanations and descriptions of certain application features, provide explanation on the usefulness of strategies based on specific category, presentation of methods for the application of self-interrogative strategies, give explanation of when, and why as well as how to used the strategies and provide discussion on self-regulation strategies). After the completion of the instructional stage the post intervention stage was next, this involves the collection of data and final programming performance as well as explanation. The data were collected from both the intervention and the control groups. The students were allowed to ask as well as answer questions verbally. The pre-and post instructional stages consisted of an encoding stage which describes students’ explanation and study of instructional materials. For the last stage, students were provided with problem solving activity which includes novel programming.

Bielaczyc et al. (1995) concluded that the performance of the intervention group was better than the control group in several areas. For example, the intervention group provided better explanation on major points during the programming activities as well as explain Lisp codes. Based on the data that was collected from the instructional and post instructional stage the interventional group generated more explanation in comparison to the control group.

King (1994) investigated strategies that teachers use to teach 4th and 5th
graders on how to produce scientific explanations. The study consisted of three groups: guided questioning- explaining, lesson based questioning with explanation, and unguided questioning with explanation. Guided questioning- explaining involve the engagement of students in discussions where questions are used to make connections with the lesson. Lesson based questioning with explanation involve the engagement of students in discussion that are guided by questions and explanations which utilizes lesion and experience based questions. Unguided questioning with explanation involves the engagement of the control group in untrained questioning. The group consisted of 30 grade five students and 28 grade four students. The data was in the form of video and audio tape lessons. It also consisted of pre and post test of the three groups. The results revealed that the performance of the participants of the question-based (Group 1) group was better than the participants of the experience-based group (Group 2) and control group (Group 3).

Sandoval and Reiser (2004) investigated student’s inquiry and how explanation supports inquiry using a qualitative approach. The participants for their study consisted of 69 students. The students were enrolled in a biology course from three 9 grade classes who used the curriculum (Explanation Constructor) developed by the researchers on evaluation. The Explanation Constructor provides online assistance and prompts to students as they engaged in explanation constructions during inquiry. The researcher collected data over a four week period. They used audiotape, videotape, field notes and observation for their data collection. The analysis also included data from a focus group session that was conducted with four of the students. The results showed that, on average, students generated less than two explanations per problem during inquiry. The authors concluded that the Explanation Constructor assisted students in generating explanations and evaluate the progress of their explanations in relation to the students’ inquiry questions.

McNeill et al. (2006) examined the effects of various types of explanations scaffolding (continuous scaffolding or faded scaffolding) on how they impact students’ learning and scientific explanation. They also developed a curriculum which assisted students at the seventh grade level in their understanding of scientific explanations. The results of McNeill et al’s study showed an increase in scientific explanations and better explanations throughout classroom learning when students’ were provided with continuous scaffolding and written explanations, as well as teacher scaffolding and modeling of explanation. In addition, students who received the faded scaffolding condition (i.e., those students who were given similar instructional assistance for explanation from the initial stage of learning and then gradually reduced during learning) produced less scientific explanations.

In another study McNeil et al. (2008) investigated the effects of the adoption of an explanation framework on students’ science learning while they received science instruction. Their study consisted of thirteen teachers who taught at the 7th grade level, and 1,197 students who were from an urban and a suburban school. Data collection was carried out over an eight week period during an implementation of a chemistry unit in every classroom. The lessons were videotaped along with both pre-and posttest evaluations of students’ comprehension and chemistry explanations. The researcher collaborated with teachers to assist them in providing support during the evaluation and generation of students’ explanations of chemistry. Comparison of the pre- and post test showed that the students exhibit significant learning outcomes for scientific explanation. In addition, the data from the classroom lessons that were videotaped indicated that students’ generation of scientific explanation were increased during the instructional unit. The teacher instructional practices that supported students’ scientific explanation also varied significantly. The authors concluded that the variation was in relation to the systematic differences of the amount as well as the quality of scientific explanations generated by the students throughout learning outcomes based on the posttest.

Many researchers have investigated the function of inscriptional tools for example, science notebooks, three dimensional models, and diagrams in fostering scientific explanation development of students (Ainsworth & Loizou, 2003; Haefner, Zembal-Saul & Avraamida, 2002).

Ainsworth and Loizou (2003) agreed with the claim that students learn better when they are provided with diagrams that have information on the human circulatory system, than students who were provided with text. High school students were selected randomly in their study and they were given text and diagram materials in one group and text only in another group to study. All of the students received pre-and posttest to determine their knowledge about circulatory system. The results showed that students from the text and diagram group provided more causal self-explanations as well as scored higher during the posttest in comprising to the text only group.

Several research have been conducted at the preschool level (Meacham, Vukelich, Han, & Buell, 2014; Brenneman, 2009; Conezio, & French, 2002; Leslie, 2013; Maherally, 2014). However many of the researchers have referred to children as natural scientist (Gropik, 2012; Gopnik, Meltzoff, & Bryant, 1997; Graaf, Segers, & Verhaegen, 2015; Worth, 2010). Gropik (2012) explained that “children use data to formulate and test hypotheses and theories in much the same way that scientists do” (p.1625). She noted that when children watch others they can learn about casual relationships and based on the evidence that they received from teachers they can draw different conclusions rather form the evidence that they have gather themselves. In other studies it has shown that young children have often regarded as natural curiosity (Conezio, & French, 2002; Klahr, Zimmerman, & Jirout, (2011; Gropik, 2012; Tillinger, 2013; Worth, 2010). Conezio and French (2002) explained that preschooler’s shows curiosity and often wonder about the world. Gropik (2012) noted that the brilliance and natural curiosity that children poses when incorporated with science could be used to help them become better at science teaching and learning. Therefore it is vital to know that activities that encourage play, present anomalies, and ask for explanations can prompt scientific thanking way better that the use of direct instructions (Gropik 2012). Ernst-Slavit and Pratt (2017) noted the importance of asking questions in the science classroom and how the questions that the teachers asked can be used as models for the types of questions that they would like their students to ask. Brown and Minnesota (2017) explained that science is taught at the K-12 level where understanding and explaining the natural world are practices that are accepted and that asking questions about the world are considered universal. Nevertheless, it is important that everyone is aware that even though young children can engage in scientific thinking many of the children have left school and have not learned much about science (Klahr, Zimmerman, & Jirout, 2011).

Peterson and French (2008) investigated the co-construction of explanations throughout classroom discourse with preschools between the ages of three and four as well as their teachers while they worked and learned about color mixing unit as part of the ScienceStart curriculum over a five weeks period (Peterson and French 2008). Based on the approach of Beals (1991, 1993), and Callanan, Shrager and Moore (1995) studies, this work looked at the linguistic perceptive of children’s development of explanatory discourse, such as children development and used of causal connectives. Peterson and French (2008) noted that young children generated responses that were more topic-relevant, utilized terms that were more standard color and employed more causal connectives towards the ending of the unit.

The above discussion illustrates that instructional support is useful for students to generate scientific explanations of the highest quality. However the frequency and nature, of how teacher discourse strategies are used during science inquiry learning to influence students generating of explanations remains unknown. In the following section, I will describe details of the study.