Shared Book Reading in Early Childhood Education: Effect of Two Approaches on Children’s Language Competence, Story Comprehension, and Causal Reasoning

ABSTRACT

This study investigated the effect of traditional interactive reading and interactive reading using mindmaps on children’s language competence and their story comprehension and causal reasoning. Furthermore, we explored teachers’ dialogic scaffolding in the two approaches and investigated the relation between teachers’ dialogic scaffolding and children’s causal reasoning. In total, 7 early childhood teachers and their 176 pupils participated. Classrooms were randomly assigned to the traditional interactive reading group or the interactive reading using mindmaps group. Before and after a six-week intervention, children’s language competence was measured using tests for thematic vocabulary, narrative competence, and critical listening. In addition, story comprehension was measured with a short questionnaire and children’s causal reasoning and teachers’ dialogic scaffolding were assessed by coding videos of shared book reading activities. Outcomes indicated that children significantly improved their language competence over the course of the study, independent of the group to which they were assigned. In addition, findings showed that children’s causal reasoning was partly related to teachers’ dialogic scaffolding.

KEYWORDS:

• Causal reasoning
• early childhood education
• interactive book reading
• language competence
• mindmaps
• story comprehension

Shared book reading is an activity during which an adult reads a book with children and engages them in conversation about the story (e.g., Gámez et al., 2017; Milburn et al., 2014). In most early childhood classrooms, shared book reading is part of the daily routine. The popularity of shared book reading can be explained by its positive effect on children’s language competence (e.g., Piasta et al., 2012; Richman & Colombo, 2007; Zucker et al., 2013). There are, however, different approaches in shared book reading. The most popular approach is referred to as interactive book reading (Cabell et al., 2019). This approach is characterized by a large amount of interaction before, during, and after reading (Ard & Beverly, 2004; Blewitt et al., 2009; Wasik et al., 2016). During interactive book reading, teachers typically comment on the story, provide explanations, and ask children a variety of questions to encourage them to share their thoughts and ideas and to make inferences (Ard & Beverly, 2004; Blewitt et al., 2009; Mol & Bus, 2011; Mol et al., 2009).

Previous research has provided evidence for the positive effects of interactive book reading. It has been shown, for example, that extra textual talk (i.e., talk about literal, inferential, print, or phonological topics) during shared book reading in preschool predicts children’s vocabulary and literacy skills in kindergarten and 1st grade (Zucker et al., 2013). In addition, meta-analyses into the effect of interactive book reading have shown positive effects on children’s oral language skills, alphabetic knowledge, and reading skills (Mol & Bus, 2011; Mol et al., 2009). The effect of interactive book reading on children’s language competence depends on the specific approach. Specifically, multiple studies have shown that asking questions about target words facilitates children’s vocabulary learning (Ard & Beverly, 2004; Blewitt et al., 2009; Lenhart et al., 2019; Lenhart, Lenhard et al., 2020). Outcomes regarding the use of scaffolding in asking questions (i.e., first asking low demand questions and later asking high demand questions) are mixed (Lenhart et al., 2019; Lenhart, Suggate et al., 2020). In addition, compared to asking questions, the study of Ard and Beverly (2004) has demonstrated that commenting on a story might be even more effective in promoting children’s word knowledge.

A relatively new approach in shared book reading involves interactive reading using mindmaps. A mindmap is a graphic organizer in which information is visually represented and structured by distinguishing between main and subtopics (Buzan, 2018). The use of graphic organizers in general and mindmaps specifically can support learners in identifying connections between topics (Birbili, 2006; Budd, 2004; Buzan & Buzan, 2000; Wang et al., 2010). Based on the work of Buzan, Dutch preschool teacher Rianne Hofma developed an approach to use mindmaps during shared book reading in early childhood classrooms (Hofma, 2013; Van der Wilt et al., 2020). This approach consists of three reading sessions that focus on four questions: Session (1) Who is the story about?; Session (2) Where does it take place?; Session (3) What is the main problem and What is the main solution? The focus question is asked before reading the book and challenges children to identify connections and cause-effect relations between different elements of the story. Directly after reading the story, the teacher and children discuss the focus question and try to reach consensus on the answer. During this conversation, the teacher uses the input of the children to construct one of the branches (e.g., the “who-branch”) of the mindmap. Children’s input is represented in the mindmap by drawings and pictures.

Outcomes of studies comparing the effect of traditional interactive reading and interactive reading using mindmaps indicated that children’s level of language competence significantly improved over time, but no differences between conditions were found (Boerma et al., 2021; Van der Wilt et al., 2019). Apart from the aforementioned studies, research into the effect of interactive reading using mindmaps remains scarce. Besides, studies from Van der Wilt et al. (2019) and Boerma et al. (2021) did not report on classroom observations of interactive reading activities in which mindmaps were used. In the present study, we therefore re-investigated the effect of both traditional interactive reading and interactive reading using mindmaps on children’s language competence. Language competence, however, is a complex and multifaceted concept that consists of interrelated sub-skills (Conti-Ramsden & Durkin, 2012). In the current study, the following aspects of language competence were taken into account: (1) thematic vocabulary, (2) narrative competence, and (3) critical listening.

Thematic vocabulary refers to the knowledge of words that are focused on a specific theme (e.g., “competition”; Adan-Dirks, 2012). Narrative competence is the ability to tell a story that is structured and coherent (Pinto et al., 2017; Zanchi & Zampini, 2021). To be able to do so, children need to integrate linguistic, cognitive, and social skills (Pinto et al., 2017; Zanchi & Zampini, 2021). Critical listening refers to listening in such a way that one is able to both comprehend and evaluate the message (Bourdeaud’hui et al., 2021). Critical listening can be regarded as the highest level of listening, because it goes beyond discriminative listening (distinguishing and identifying) and comprehensive listening (retaining and remembering; Bourdeaud’hui et al., 2021). Multiple studies have demonstrated that traditional interactive reading in early childhood classrooms contributes to children’s vocabulary learning (e.g., Barnes et al., 2017), ability to produce well-structured narratives (e.g., Lever & Sénéchal, 2011), and listening comprehension skills (e.g., Sénéchal & LeFevre, 2002). Together, these studies indicate that traditional interactive reading is an effective means to promote children’s language competence.

Research into the effect of mindmaps on young children’s language development is scarce, but previous studies have shown positive effects of the use of mindmaps on other outcomes (Abi-El-Mona & Adb-El-Khalick, 2008; Budd, 2004; Farrand et al., 2002; Merchie & Van Keer, 2013, 2016a, 2016b). For example, the use of mindmaps has been found to be an effective study technique for medical students (Farrand et al., 2002) and students in economics (Budd, 2004). In addition, middle school students participating in a mindmap intervention obtained higher gains in science class than students in the control group (Abi-El-Mona & Adb-El-Khalick, 2008). Finally, several studies have indicated that children in 5th and 6th grade improved their graphical summarization skills as a result of participating in a mindmap intervention (Merchie & Van Keer, 2013, 2016a, 2016b).

The positive effect of mindmaps might be explained by the cognitive load theory (Sweller, 1994; also see, Paas et al., 2003), which indicates that learners’ working memory becomes overloaded if they receive a lot of new information (e.g., hearing a story for the first time). Learners whose working memory system is overloaded cannot concentrate on teacher instructions, experience difficulties in remembering information, and find it harder to stay on track during structured learning activities (Gathercole & Alloway, 2008), which might ultimately result in poor academic attainment (Holmes et al., 2010). The structured manner in which new information is visualized in a mindmap might reduce learners’ cognitive load and, thereby, support their learning. In addition, in line with Paivio’s dual coding theory, the use of a combination of words and pictures in mindmaps might result in deeper processing. According to the dual coding theory, words and pictures are stored and processed in two separated but connected cognitive subsystems: a verbal system and a non-verbal (imagery) system (Paivio, 2006). A mindmap might support learners in connecting the verbal and visual information and thereby facilitate their learning process (Merchie & Van Keer, 2013).

Although few studies have examined the effect of interactive reading using mindmaps in early childhood classrooms and these studies have found no additional effect of mindmaps (Boerma et al., 2021; Van der Wilt et al., 2019), we still have reasons to assume that the use of mindmaps can be particularly effective in promoting children’s language skills (compared to traditional interactive reading). First, previous research has shown that involving children in interactive discussions especially after reading the book (rather than before or during) was effective in enhancing children’s vocabulary (Gonzalez et al., 2014). The fact that the story is discussed afterward in the mindmap approach could therefore indicate that this approach is an effective means to promote children’s vocabulary. Second, it has been demonstrated that children’s narrative competence can be supported by engaging them in producing narratives (Silva et al., 2014). Since the mindmap approach encourages children to reproduce stories from the book (using the focus questions), this approach could support their development of narrative competence. Finally, it has been suggested that setting a specific goal for listening (e.g., by asking one or two focus questions) helps children channel their attention (Funk & Funk, 1989; Jalongo, 1996), although empirical evidence for the effect of listening instruction on listening skills is ambiguous (Beall et al., 2008). As children receive listening instructions during interactive reading using mindmaps, this approach might help children to focus better.

In addition to language competence, the current study examined the effect of traditional interactive reading and interactive reading using mindmaps on children’s story comprehension and causal reasoning. Story comprehension concerns the extent to which one understands the different elements in a story and the relations between these elements (Filiatrault-Veilleux et al., 2016), whereas causal reasoning refers to the ability to provide explanations and to make predictions and inferences about causal relations (Gosen, 2012; Reed et al., 2015; Zucker et al., 2010). Children’s levels of story comprehension and causal reasoning depend on the extent to which children are able to recognize the connections between different parts of a story and to identify cause-effect relations (Filiatrault-Veilleux et al., 2016). As such relations are explicitly addressed during interactive reading using mindmaps, it might be expected that this approach positively affects children’s story comprehension and causal reasoning. However, to our knowledge, no previous research exists in which the effect of interactive reading using mindmaps on children’s story comprehension and causal reasoning has been investigated. It is therefore unknown whether this approach supports children’s story comprehension and causal reasoning and whether traditional interactive reading and interactive reading using mindmaps might differ in their effect. Hence, the present study investigated the effect of traditional interactive reading and interactive reading using mindmaps not only on children’s language competence, but also on their story comprehension and causal reasoning.

Finally, this study explored teachers’ dialogic scaffolding within the traditional interactive reading approach and the interactive reading using mindmaps approach. Dialogic scaffolding refers to the implementation of conversational techniques that aim to encourage children to deepen their reasoning (Michaels & O’Connor, 2015; O’Connor & LaRusso, 2014; O’Connor & Michaels, 2007). It is, for example, well known that open questions have a different effect on the course of a conversation and on the contribution of children compared to closed questions (e.g., Nystrand et al., 2003). As previously mentioned, during traditional interactive reading, a variety of questions are asked (e.g., Mol et al., 2009). In contrast, the approach of interactive reading using mindmaps particularly focuses on four questions that direct children’s attention to different aspects of the story (i.e., characters, locations, and problems/solutions; Hofma, 2013). Hence, the conversational techniques teachers implement during shared book reading seem to depend on the approach they adopt. There are, however, no previous studies in which teachers’ use of dialogic scaffolding during different shared book reading approaches has been examined. In order to fill this gap, the current study explored teachers’ dialogic scaffolding during traditional interactive reading and interactive reading using mindmaps. In addition, as dialogic scaffolding is aimed to deepen children’s reasoning, the relation between teachers’ dialogic scaffolding and children’s causal reasoning was investigated as well.

To summarize, the current study investigated the effect of traditional interactive reading and interactive reading using mindmaps on children’s language competence (i.e., thematic vocabulary, narrative competence, critical listening), and their story comprehension and causal reasoning. Furthermore, we explored teachers’ dialogic scaffolding during traditional interactive reading and interactive reading using mindmaps, and investigated the relation between teachers’ dialogic scaffolding and children’s causal reasoning. Although prior studies have found no additional effect of mindmaps, we still had reasons to assume that the use of mindmaps is particularly effective. Hence, it was expected that interactive reading using mindmaps would have a larger effect on children’s language competence than traditional interactive reading. In addition, the expectation was that the mindmap approach would have a larger effect on children’s causal reasoning compared to the traditional approach. Due to a lack of prior research, there was no specific expectation regarding the use of teachers’ dialogic scaffolding during the interactive reading approaches. Finally, it was expected that teachers’ dialogic scaffolding would be related to children’s causal reasoning.

Method

Ethical considerations

The present study was part of a larger research project on the effect of using mindmaps in early childhood classrooms. For the larger research project, ethical approval was obtained from the ethical committee of the authors’ university (details removed for review). Prior to the study, teachers were informed about its purpose and procedures during a meeting. In addition, pupils’ parents were asked for active consent. Children without consent did not participate and parents were given the opportunity to withdraw their child from the study at any time. The collected data were anonymously processed and saved, and only used for the purpose of this study. Only the main researchers had access to the data.

Participants

In total, N = 7 teachers and their N = 176 pupils participated in the present study. Classes were randomly assigned to one of two groups: the traditional interactive reading group (n = 3 teachers; n = 71 children) or the interactive reading using mindmaps group (n = 4 teachers; n = 105 children). Among the participating teachers, there were n = 6 female teachers and n = 1 male teacher. During the course of the study, teachers were employed at Dutch primary schools (n = 4) and taught in early childhood classrooms. In the Netherlands, early childhood classrooms are integrated into primary schools (children age 4–12 years), and are mostly mixed age groups (i.e., 4- to 6-year-olds combined in one group). Class sizes ranged from 21 to 28 children (M = 25.14, SD = 2.10).

The participating children were between 49 and 84 months old (M = 63.94, SD = 7.70) and the distribution of girls (n = 89, 50.6%) and boys (n = 87, 49.4%) was almost equal. For 72.7% of the children, information regarding their background was available. Most children were born in the Netherlands (69.9%). Other countries of birth were Aruba, Kenya, Norway, South Africa, and Thailand. For 66.5% of the children, Dutch was the main language spoken at home. Other home languages were a combination of Dutch and English (2.3%), Dutch and Aramaic (2.3%), Dutch and Arabic (1.1%), Dutch and Assyrian (0.6%), Dutch and Turkish (0.6%), and Aramaic (0.6%). Parents’ education levels were low (4.0%), average (30.7%), or high (36.4%).

Study design

The present study adopted a pretest-intervention-posttest design (see, Table 1 for details). Children were tested on several language tests before and after the intervention period. Besides, in each class a pre- and post-observation of a shared book reading activity was videotaped. Pre-observations were conducted at the start of the study. Teachers were asked to conduct a shared book reading activity (based on a book they selected themselves) that would represent their daily practice. Teachers scored the representativeness of the activity on a five-point-scale (0 = non-representative, 5 = representative). Scores ranged from 4 to 5 with an average of 4.30 (SD = 0.52).

Table 1. Overview of the study design.

	Week 1	Week 2	Week 3	Week 4	Week 5	Week 6	Week 7	Week 8	Week 9
Traditional interactive reading group	pretests and pre-observation	workshop traditional interactive reading	activity 1 (BR) interactive reading with focus on prediction	activity 2 (BR) interactive reading with focus on comprehension	activity 3 (BR) interactive reading with focus on feelings	activity 4 (TH) interactive reading with focus on prediction	activity 5 (TH) interactive reading with focus on comprehension	activity 6 (TH) interactive reading with focus on feelings (used as post-observation)	posttests
Interactive reading using mindmaps group	pretests and pre-observation	workshop interactive reading using mindmaps	activity 1 (BR) reading with focus question “Who is the story about?” and constructing “who-branch”	activity 2 (BR) reading with focus question “Where does the story take place?” and constructing “where-branch”	activity 3 (BR) reading with focus questions “What are the problems and solutions?” and constructing “problem-and solution-branch”	activity 4 (TH) reading with focus question “Who is the story about?” and constructing “who-branch”	activity 5 (TH) reading with focus question “Where does the story take place?” and constructing “where-branch”	activity 6 (TH) reading with focus questions “What are the problems and solutions?” and constructing “problem-and solution-branch” (used as post-observation)	posttests

Note. BR = The Black Rabbit; TH = The Competition of the Tortoise and the Hare

After conducting the pretests, all participating teachers engaged their pupils in six shared book reading activities. The first three activities were focused on the picture book The Black Rabbit (Leathers, 2013) and the last three on The Competition of the Tortoise and the Hare (Visser & Meirink, 2017; see Appendix A for a short synopsis of both books). The activities were spread over six weeks (one activity per week). Prior to the activities, teachers participated in a workshop during which the book reading activities were discussed and teachers practiced either with the interactive reading approach or the mindmap approach. In addition, teachers received a teacher manual in which each activity was described in detail and were asked to follow this manual as closely as possible. The final shared book reading activities (focused on the book The Competition of the Tortoise and the Hare) were used as post-observations.

As previously mentioned, classes were randomly assigned to the traditional group or the mindmap group. The teachers in the traditional group engaged their pupils in traditional interactive reading. Teachers were instructed to ask children questions about the story before, during, and after reading the picture book (e.g., What do you think will happen next?). Each reading activity focused on different types of questions (see Table 1).

The shared book reading activities in the mindmap group were focused on the construction of a mindmap. Prior to each reading activity, teachers asked children a question in order to focus children’s attention on one aspect of the story. For example, during the second time the picture books were read, teachers asked children to focus on the different locations in the story (i.e., “Listen carefully to the story: where does the story take place?”; see Table 1). After reading the book, this focus question was discussed with the children, and teachers visualized this discussion by constructing a part of the mindmap. In order to ensure comparability between classes, teachers were provided with a pre-made mindmap that they used as an example and guideline. During each reading activity, a different part of the mindmap was constructed (see Figure 1 for an example of a mindmap on The Black Rabbit and Figure 2 for an example of a mindmap on The Competition of the Tortoise and the Hare).

Figure 1. Example of a mindmap on the picture book The Black Rabbit (Leathers, 2013).

Figure 2. Example of a mindmap on the picture book The Competition of the Tortoise and the Hare (Visser & Meirink, 2017).

Implementation fidelity

The design of this study was piloted in a previous study which had shown that the weekly reading sessions fit within the curriculum (Van der Wilt et al., 2019). To maximize fidelity of intervention delivery, teachers within the same group (i.e., traditional or mindmap) received the same workshop. Furthermore, during the intervention period, the authors telephoned every teacher twice to answer questions, collaboratively reflect on the reading sessions, and ensure teachers adhered to the manual. Besides, participating teachers were asked to fill out a short evaluation form shortly after each book reading session. In this form, teachers were asked to what extent they adhered to the manual and to report any abnormalities (though there were none). Finally, post-observations indicated that teachers closely followed the teacher manuals and were able to conduct the reading sessions accordingly.

Measures

Children’s language competence

Three indices were used to indicate children’s language competence: (1) thematic vocabulary, (2) narrative competence, and (3) critical listening.

Thematic vocabulary

Children’s thematic vocabulary was measured with the Thematic Vocabulary Assessment Test (Adan-Dirks, 2012; Van der Veen et al., 2016). In the current study, children were asked to explain the meaning of 20 words or expressions. These words were selected from the two picture books that were used: 10 words from The Black Rabbit (Leathers, 2013) and 10 words from The Competition of the Tortoise and the Hare (Visser, 2013; see Appendix B for the complete test). Tests were individually administered by a trained test assistant who asked children to shortly explain the meaning of the words. Audio-tapes of the test administrations enabled the test assistant to score the children’s answers afterward. That is, children’s answers were compared to descriptions that had been collaboratively constructed by the research team (consisting of both researchers and early childhood teachers). A correct answer (i.e., an answer that was in line with the description) was scored with one point whereas an incorrect answer was scored with zero points. To indicate children’s level of thematic vocabulary, scores were computed. Hence, total scores could range from 0 to 20. Depending on the test occasion, the internal consistency of the scale was questionable or acceptable (Cronbach’s alpha = 0.66 for the pretest and Cronbach’s alpha = 0.79 for the posttest). To assess the interrater reliability, the first author scored 10% of the tests. A Cohen’s kappa of 0.62 was found, indicating a substantial interrater reliability (McHugh, 2012).

Narrative competence

Narrative competence was assessed using the subscale Narrative Task of the standardized Cito Language Test for All Children (Verhoeven & Vermeer, 2006). Two versions were used: one for the pretest and one for the posttest. These versions are highly comparable in content, structure, and difficulty (Verhoeven & Vermeer, 2006). During the administration of this task, children were presented with a series of pictures. Together, these pictures formed a story. Children were asked to tell the story that would suit the series of pictures and to tell it in such a way that a person who could not see the pictures would still understand the story. These test administrations were also audio-taped. Children’s stories were evaluated by comparing the stories to 16 items which referred to the story’s events and to causal relations between events (see Appendix C). Items were dichotomously scored (1 point in case a child’s story was in line with the description in the item, 0 points if this was not the case) and total scores were obtained by computing children’s scores on the 16 items. Scores could therefore range from 0 to 16. The test has been found to be reliable with a Cronbach’s alpha of 0.90 (Verhoeven & Vermeer, 2006), although lower Cronbach’s alphas of 0.69 (pretest) and 0.64 (posttest) were found in the current study. The first author scored 10% of the tests, and a Cohen’s kappa of 0.64 indicated that the interrater reliability was substantial (McHugh, 2012).

Critical listening

The subscale Critical Listening of the Cito Language for Pre-schoolers was used to measure children’s ability to critically listen (Lansink, 2009). This test was administered by children’s own teacher. Children were provided with an answer sheet on which 14 items were presented. Each item consisted of four colored pictures. During the test administration, the teacher read short stories aloud and asked the children to indicate the matching picture by drawing a line under one of the four pictures. For example, one item goes as follows: “Karim and dad are outside. They play soccer. Karim stands in the goal and dad kicks the ball. Draw a line under that picture.” Children’s answers were dichotomously scored: 0 in case of an incorrect answer and 1 in case of a correct answer. Total scores were obtained by summing the number of correct answers. Scores could consequently range from 0 to 14. In the current study a Cronbach’s alpha of 0.56 was found for the pretest and a Cronbach’s alpha of 0.60 for the posttest.

Children’s story comprehension

A short task was developed in order to investigate whether children understood the two picture books that were used in the present study. As the book The Black Rabbit (Leathers, 2013) appeared to be richer and more complex than the book The Competition of the Tortoise and the Hare (Visser, 2013), three questions were based on the first book and two questions on the latter (see Appendix D). The purpose of the questions was to assess children’s understanding of what had happened in the story (action landscape) and of why it had happened (consciousness landscape; Bruner, 1986, also see, Curenton, 2011). For example, in the case of The Competition of the Tortoise and the Hare (Visser, 2013), children were asked who had won the competition (action landscape) and why this had happened (consciousness landscape). The story comprehension task was only administered during the posttests. Total scores were obtained by summing the number of correct answers (range = 0 to 5). With a Cronbach’s alpha of 0.39, the internal consistency of the five items was poor. The first author independently scored 10% of the tasks. A Cohen’s kappa of 0.69 was found, indicating that the interrater reliability was substantial.

Children’s causal reasoning

The videos of the pre- and post-observations of the shared book reading activities were transcribed and analyzed. The average duration of the activities was M = 22.64 minutes (SD = 5.26, min = 16, max = 30). For each activity, children’s causal reasoning and teachers’ dialogic scaffolding was coded. The extent to which children reasoned during the shared book reading activities was investigated using the framework of Reed et al. (2015). This framework contains three non-causal (i.e., labeling, observation, and elaboration) and three causal categories (i.e., inference, predication, and explanation; see Table 2). Children’s turns were coded according to these categories. If a turn did not correspond with one of the categories, it was not coded. Interrater reliability of the coding framework was attained through independent scoring of 10% of the turns by a second rater. A strong reliability was found with a Cohen’s kappa of 0.81 (Landis & Koch, 1977; McHugh, 2012).

Table 2. Overview of the three non-causal and three causal categories.

Category	Definition	Examples
Labeling	Non-causal naming or identifying perceived features of the stimulus without further information	He climbs up the tree
Observation	Non-causal description or interpretation of perceptual features that goes beyond labeling	He looks a bit scared
Elaboration	Non-causal information about the stimulus that represents prior knowledge	He is afraid of heights
Inference	Causal statement about a connection between actions, events, characters, or psychological variables that is inferred from the stimulus	The hare jumped of the tree and then he broke his leg
Prediction	Causal expression of an expected outcome of a depicted or inferred action or event	They thought the hare would win
Explanation	Reference to a causal mechanism to explain a depicted or inferred action, event, or psychological state	He won, because he didn’t get distracted

Teachers’ dialogic scaffolding

In order to measure teachers’ dialogic scaffolding and to analyze possible differences in the role of the teacher in the traditional group and the mindmap group, teacher turns were coded using the Low Inference Discourse Observation tool (LIDO; O’Connor & LaRusso, 2014). The LIDO was developed to capture whole-class discussions and distinguishes three low-inference codes for dialogic scaffolds and three categories of general types of content questioning (see Table 3). Teacher turns that did not comply with one of these six LIDO categories were not coded. Interrater reliability of the LIDO was attained through independent scoring by a second rater of 10% of the coded teacher turns. A moderate reliability was found with a Cohen’s kappa of 0.72 (Landis & Koch, 1977; McHugh, 2012).

Table 3. Overview of dialogic teacher scaffolds of the LIDO.

Category	Definition	Examples
Responding	Teacher gets student(s) to respond to another student’s turn	What did Ann say? Have you heard, Tom?
Explaining	Teacher encourages students to explain/clarify themselves or to provide reasoning	I didn’t get that, could you say it again?
Continuing	Teacher encourages students to continue reasoning	But why is that so?
Open question	Teacher poses open question (answer contestable)	Who else feels strong sometimes? Nick, could you tell me a bit about that?
Semi-open question	Teacher poses semi-open question (circumscribed answer set)	What is the hare doing?
Closed question	Teacher poses closed question (answer is uncontestable)	Who won the competition? The tortoise or the hare?

Data analysis

Video transcripts were imported in Atlas.ti version 1.0.50. Each turn was first coded as a child turn or a teacher turn. Next, for each transcript, children’s turns were coded using the framework of Reed et al. (2015) and teachers’ turns were coded using the LIDO (O’Connor & LaRusso, 2014). To control for differences between the length of the shared book reading activities, categories of coded child turns and coded teacher turns were standardized by dividing them by the total minutes of the activity. Data obtained by coding the transcripts and scoring the tests were analyzed using the Statistical Package for Social Scientists (SPSS, version 26).

There were multiple missing values on pre- and posttests of the main variables. Percentages of missing data ranged from 10.91% to 13.84% on the pretests and from 8.04% to 8.52% on the posttests. Missing data was due to absence of children on the day of the test administrations (e.g., because of illness). Missing values were imputed using the commonly used Expectation-Maximization (EM) method in SPSS after finding no statistically reliable deviation from randomness (Little’s MCAR test X²(757) = 722.06, p = .815 on pretests and X²(621) = 634.64, p = .343 on posttests). The imputed dataset was used for analyses.

A visual inspection of the data indicated that the scores on the main variables were not normally distributed. Due to the relatively small sample of classrooms and teachers (i.e., N = 7), non-parametric techniques were used to analyze data that was extracted from the video transcripts. However, data on children’s language competence and story comprehension were analyzed using parametric techniques, because the sample size on child level was relatively large (i.e., N = 176) and it can be expected that the violation of the assumption of normality does not cause major problems with large sample sizes (e.g., Gravetter & Wallnau, 2004).

Results

Descriptive statistics

In Table 4, the means and standard deviations of pre- and posttests for each main variable are displayed. Unstandardized scores are reported in order to facilitate the interpretation of the descriptive statistics.

Table 4. Means and standard deviations of main variables.

	Pretest			Posttest
	Total group	Traditional group	Mindmap group	Total group	Traditional group	Mindmap group
	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)
Children’s language competence
Thematic vocabulary	4.12 (2.47)	5.09 (2.47)	3.47 (2.26)	7.27 (3.67)	8.49 (3.42)	6.45 (3.62)
Narrative competence	8.18 (2.68)	8.58 (2.73)	7.91 (2.62)	9.11 (2.54)	9.30 (2.49)	8.98 (2.58)
Critical listening	11.62(2.28)	11.28 (2.56)	11.86 (2.06)	12.93 (1.83)	12.85 (1.51)	12.98 (2.02)
Children’s story comprehension	-	-	-	3.70 (1.06)	3.92 (1.06)	3.56 (1.03)
Children’s causal reasoning
Labeling	41.00 (31.78)	27.33 (5.86)	51.25 (40.86)	27.14 (13.26)	22.33 (19.63)	30.75 (7.37)
Observation	22.00 (12.95)	13.00 (3.61)	28.75 (13.60)	14.57 (7.14)	20.33 (6.11)	10.25 (4.35)
Elaboration	6.00 (4.83)	5.33 (5.51)	6.50 (5.07)	6.57 (4.28)	6.76 (3.06)	6.50 (5.51)
Inference	3.71 (3.86)	3.00 (5.20)	4.25 (3.30)	6.29 (2.98)	5.00 (1.00)	7.25 (3.78)
Prediction	9.29 (6.45)	12.33 (6.11)	7.00 (6.48)	2.43 (2.07)	2.67 (2.52)	2.25 (2.06)
Explanation	12.14 (8.75)	13.33 (7.77)	11.25 (10.50)	7.00 (3.06)	9.67 (1.53)	5.00 (2.16)
Teachers’ dialogic scaffolding
Responding	0.71 (1.50)	1.67 (2.08)	0.00 (0.00)	0.71 (1.11)	1.33 (1.53)	0.25 (0.50)
Explaining	2.43 (3.51)	3.67 (5.51)	1.50 (1.29)	2.57 (1.51)	3.33 (2.08)	2.00 (0.82)
Continuing	1.57 (1.51)	1.00 (1.00)	2.00 (1.83)	7.71 (5.85)	7.33 (5.86)	6.25 (6.70)
Open question	5.43 (4.20)	6.67 (4.73)	4.50 (4.20)	4.29 (4.15)	7.33 (4.93)	2.00 (1.41)
Semi-open question	13.00 (8.41)	14.67 (4.51)	11.75 (11.09)	19.00 (8.68)	26.67 (6.35)	13.25 (4.57)
Closed question	27.71 (12.26)	29.00 (12.53)	26.75 (13.89)	32.29 (18.14)	39.00 (28.00)	27.25 (7.50)

Hypothesis testing

To investigate the effect of traditional interactive reading and interactive reading using mindmaps on children’s language competence, mixed between-within subjects analyses of variance were conducted. A 2 × 2 design was defined with pre- and posttest as within-factor and the two reading approaches as between-factor. Besides, an independent t-test was conducted to examine whether there were group differences in the posttest of story comprehension.

For all language aspects, the interaction effect between the within and the between factor was not significant (see Table 5). This indicates that the effect of the two reading approaches on language competence did not differ. The between subject effect was statistically significant for thematic vocabulary, indicating that the scores on thematic vocabulary were significantly higher in the traditional group than in the mindmap group. Besides, the within subject effect was statistically significant for all language aspects: Children scored significantly higher on the posttests than on the pretests. Finally, outcomes of the independent t-test indicated that children in the traditional group and the mindmap group differed in their scores on story comprehension: t(174) = 2.23, p = .027, two-tailed). Children in the traditional group outperformed children in the mindmap group. To investigate whether this difference could be explained by the difference between the traditional group and the mindmap group in the pretest of thematic vocabulary, an analysis of variance was conducted as a posthoc analysis including story comprehension as the dependent variable, condition (traditional group versus mindmap group) as the fixed factor, and the pretest of thematic vocabulary as a covariate. Outcomes indicated no significant difference between groups in story comprehension, while controlling for the pretest of thematic vocabulary (see Table 5).

Table 5. Interaction effects, within-, and between-subjects effects for (1) thematic vocabulary, (2) narrative competence, and (3) critical listening, and between subjects effects for story comprehension.

		Wilks’ Lambda	F	df	p	Partial Eta Squared
Thematic vocabulary	Interaction	1.00	0.93	1, 174	.335	.005
	Within	.45	212.77	1, 174	<.001	.550
	Between		20.18	1, 174	<.001	.104
Narrative competence	Interaction	1.00	.494	1, 174	.483	.003
	Within	.93	13.14	1, 174	<.001	.070
	Between		2.52	1, 174	.114	.014
Critical listening	Interaction	1.00	2.05	1, 174	.154	.012
	Within	.70	74.82	1, 174	<.001	.301
	Between		1.66	1, 174	.199	.009
Story comprehension	Thematic vocabulary (pre)		14.95	1, 173	<.001	.080
	Between		4.98	1, 173	.342	.005

To investigate possible changes in children’s causal reasoning and teachers’ dialogic scaffolding, Wilcoxon Signed Rank Tests were performed to compare the pre- and posttests (see Table 6). Compared to the pretests, outcomes indicated that children made significantly less predictions on the posttests (large effect). There were no other significant differences between test occasions.

Table 6. Outcomes of Wilcoxon signed ranked tests investigating changes over time, independent of condition.

	Pretest	Posttest
	Md*	Md*	Z	p	Effect size r
Children’s causal reasoning
Labeling	1.75	1.06	−1.86	.063	0.50
Observation	0.98	0.55	−1.69	.091	0.45
Elaboration	0.26	0.32	−0.68	.499	0.18
Inference	0.20	0.28	−0.85	.398	0.23
Prediction	0.36	0.14	−2.20	.028	0.59
Explanation	0.56	0.25	−1.18	.237	0.32
Teachers’ dialogic scaffolding
Responding	0.00	0.00	−0.73	.465	0.20
Explaining	0.05	0.07	−0.51	.612	0.14
Continuing	0.06	0.27	−1.78	.075	0.48
Open question	0.18	0.11	−1.52	.128	0.41
Semi-open question	0.61	0.63	−1.01	.310	0.27
Closed Question	1.32	0.99	−0.85	.398	0.23

Note. * standardized

Finally, the relation between changes in teachers’ dialogic scaffolding and changes in children’s causal reasoning was explored by calculating spearman rank order correlations coefficients (rho). Results showed that changes in teachers’ attempts to ask children to continue and their use of open questions were related to changes in children’s use of inferences (see Table 7). That is, the increase in teachers’ attempts to ask children to continue and their decrease in the use of open questions was significantly correlated with the increase in children’s use of inferences (see Table 6). In addition, relations were found between changes in teachers’ use of open questions and semi-open questions and changes in children’s use of explanations. Specifically, teachers’ decrease in the use of open questions and increase in the use of semi-open questions was significantly correlated with the decrease in children’s use of explanations (see Table 6). There were no other significant correlations.

Table 7. Correlations between teachers’ dialogic scaffolding and children’s causal reasoning (two-tailed).

	Children’s causal reasoning
	Labeling		Observation		Elaboration		Inference		Prediction		Explanation
Teachers’ dialogic scaffolding	rho	p	rho	p	rho	p	rho	p	rho	p	rho	p
Responding	−0.22	.632	−0.56	.195	0.33	.465	0.15	.751	0.04	.937	0.22	.632
Explaining	−0.18	.702	−0.18	.702	0.50	.253	0.50	.253	0.14	.760	0.61	.148
Continuing	0.11	.819	0.46	.294	0.18	.702	0.75	.052	0.07	.879	0.61	.148
Open question	−0.25	.589	0.43	.337	0.64	.119	0.89	.007	0.43	.337	0.96	<.001
Semi-open question	−0.21	.645	0.57	.180	0.61	.148	0.64	.119	−0.04	.939	0.82	.023
Closed Question	0.29	.535	0.57	.180	−0.36	.432	−0.01	.879	−0.25	.589	−0.14	.760

Note. Correlations between standardized difference scores on children’s causal reasoning and teachers’ dialogic scaffolding.

Discussion

The purpose of the present study was to investigate the effects of traditional interactive reading and interactive reading using mindmaps on children’s language competence, and their story comprehension and causal reasoning. Furthermore, we explored teachers’ use of dialogic scaffolding during traditional interactive reading and interactive reading using mindmaps, and investigated the relation between teachers’ dialogic scaffolding and children’s causal reasoning. Outcomes indicated that all children significantly improved their language competence over the course of the study, but they made less predictions (causal reasoning category) during the post-observation compared to the pre-observation. Regarding group differences, results revealed that children in the traditional group showed a significantly higher level of story comprehension on the posttest compared to children in the mindmap group, but this difference was not significant when controlling for differences in pretest scores on thematic vocabulary. Finally, outcomes of correlational analyses indicated that changes in teachers’ use of dialogic scaffolds (specifically, their use of (semi-)open questions and attempts to ask children to continue) were related to changes in children’s causal reasoning (i.e., their use of inferences and explanations).

How can our findings be explained? First, the finding that children’s thematic vocabulary, narrative competence, and critical listening were all significantly improved over the course of the study seems to coincide with studies showing the positive effect of shared book reading on children’s language development (e.g., Wasik et al., 2016). In addition, the outcome that there were no differences in the effect on children’s language competence between groups appears to be in line with previous research in which shared book reading had a significant effect on children’s language competence, independent of the specific approach in shared book reading (e.g., Boerma et al., 2021; Vaahtoranta et al., 2019; Van der Wilt et al., 2019). Together, the findings of previous studies and the present study could indicate that a short interactive reading intervention can positively affect children’s language competence, but that this effect is independent of the approach in shared book reading. In addition, findings of the present study might indicate that the possible effect of shared book reading also holds for aspects of language competence that are not related to the content of the shared book reading intervention. Specifically, although thematic vocabulary was measured by asking children to explain the meaning of words that were selected from the two picture books that were used during the intervention (i.e., near transfer), children’s narrative competence and critical listening were measured by assessments that were unrelated to the content of the intervention (i.e., far transfer). During the test administration of narrative competence, for example, children were asked to tell a story that was completely unrelated to the stories of the two picture books that were used during the intervention.

Second, the fact that the level of story comprehension of children in the traditional group was found to be higher compared to the level of children in the mindmap group was not in line with our expectations. In fact, we expected that it would be the other way around, because children are explicitly encouraged to identify connections and causal relations during interactive reading using mindmaps (Buzan & Buzan, 2000; Kim et al., 2004). Outcomes of a post-hoc analysis showed, however, that the difference in children’s story comprehension between the traditional group and the mindmap group could be explained by the group difference in children’s thematic vocabulary on the pretest. In addition, it is important to note that children’s story comprehension was only measured during the posttest, because children would not have been able to answer questions about the picture books that were used during the intervention on the pretests. As a consequence, it was impossible to assess whether the traditional group and the mindmap group differed in their effect on children’s story comprehension. Future research investigating the effect of traditional interactive reading and interactive reading using mindmaps on story comprehension should also take a pretest of story comprehension into account (cf., Lenhart, Lenhard et al., 2020).

Third, outcomes of the current study indicated that children made significantly less predictions on the posttest than on the pretest and that there were no other significant changes in children’s causal reasoning over the course of the study. As teachers’ interactive approach affects children’s reasoning (Cazden, 2001; Gosen, 2012; O’Connor & Michaels, 2007; Wells, 2007), the finding that teachers’ dialogic scaffolding did not change over time could explain the finding that children’s causal reasoning hardly changed over the course of the study. The stability of teachers’ dialogic scaffolding could be explained by the fact that the intervention was not specifically directed at teachers’ dialogic scaffolding. Another explanation might be the short intervention period. Although the duration of the intervention period might have been sufficient for improving children’s language competence, it might have been too short to observe change in teachers’ interactive approach. The suggestion that the findings regarding children’s causal reasoning could be explained by the findings regarding teachers’ dialogic scaffolding seems to be confirmed by the outcomes of the correlational analyses. That is, although the changes in children’s causal reasoning and in teachers’ dialogic scaffolding did not turn out to be significant (except for the extent to which children made predictions), there were significant relations between the changes in teachers’ dialogic scaffolding and children’s causal reasoning. Future research is required to investigate how and to what extent teachers’ dialogic scaffolding might support children’s causal reasoning.

Although the present study provided new insights into the effects of different approaches in shared book reading, it also suffered from several limitations. First, it is important to note that the present study did not include a control group in which children did not participate in an interactive reading intervention. As a result, we cannot be certain whether the improvement of children’s language competence is attributable to children’s participation in the interactive reading approaches (i.e., traditional interactive reading and interactive reading using mindmaps). In fact, the improvement of children’s language competence over time might be explained by other activities in which children participated over the course of the present study. Future studies into the effects of traditional interactive reading and interactive reading using mindmaps should therefore include a control group (e.g., book reading without interaction). Second, the intervention in the current study was not specifically directed at improving teachers’ dialogic scaffolding. Previous research, however, has indicated that the effect of shared book reading is largely dependent on the quality of the discussions on the book that is being read (Cabell et al., 2019). This quality is, in turn, highly influenced by teachers’ ability to familiarize themselves with and their use of dialogic scaffolds. Future research into the effect of different approaches in shared book reading should therefore not only focus the content of the intervention on shared book reading in general, but also should take the development of teachers’ dialogic scaffolding into account (Michaels & O’Connor, 2015; O’Connor & LaRusso, 2014; O’Connor & Michaels, 2007; Van der Veen et al., 2017).

Third, as previously mentioned, the intervention period might have been too short for teachers to change their behavior. Previous studies have indicated that teachers need sufficient time in order to be able to implement new conversational techniques (for example, O’Connor & Michaels, 2019). It is questionable whether the six-week intervention period was sufficient in order to have an effect on the use of teachers’ dialogic scaffolding and, consequently, on children’s causal reasoning. It seems, therefore, reasonable to suggest that future studies adopt a longer intervention period in order to provide teachers with enough time to practice with the new techniques, and for children to familiarize themselves with a new approach to shared book reading and classroom talk. In addition, in the current study, teachers only participated in a workshop, after which they were required to conduct the shared book reading activities on their own using a teacher manual. Previous research, however, has indicated that reflection sessions can be highly valuable for teachers’ professional development (Van der Veen et al., 2017; Van Veen et al., 2010). Future research should incorporate such reflection or video-coaching sessions in shared book reading interventions.

To conclude, the present study did not find added benefits of using mindmaps during interactive reading for supporting children’s language competence, but findings of the current study might confirm the importance of shared book reading for promoting children’s language competence. In addition, the present study seems to point to a significant relation between teachers’ use of dialogic scaffolding during interactive reading and children’s causal reasoning. Future research is required in order to further investigate this relation within the context of shared book reading in early childhood education.

Disclosure statement

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

Additional information

Funding

This work was supported by Nationaal Regieorgaan Onderwijsonderzoek (NRO) [filenumber 40.5.18500.023]

Appendix A. Short synopsis of the books that were read during the intervention

The Black Rabbit (Leathers, 2013)

The Black Rabbit is a children’s book about Rabbit, who is confronted with a large rabbit chasing him: Black Rabbit. Rabbit is afraid of Black Rabbit and tries to get rid of him, but Black Rabbit keeps following Rabbit. Who is this mysterious Black Rabbit? And why won’t he go away? From the pictures, it becomes clear that Black Rabbit is actually Rabbit’s own shadow. At the end of the story, when the Black Rabbit actually saves Rabbit form getting eaten by a Wolf, Rabbit and Black Rabbit become friends.

The Competition of Tortoise and Hare (Visser & Meirink, 2017)

The Competition of Tortoise and Hare is a retelling of one of Aesop’s fables. The story is about a Tortoise and a Hare who are doing a running competition. The Hare is confident of winning, so he is takes his time during the competition and does all sorts of physical exercises. Meanwhile, the Tortoise is walking slowly, but steadily. Near the end of the story, the Hare climbs a climbing frame. When he is all the way up, he is afraid of going back down. The Hare then closes his eyes, jumps down, and hurts his leg. While he is being cared for, against all expectations, the Tortoise wins the running competition.

Appendix B. Thematic vocabulary assessment test (assessing thematic vocabulary)

1. What is panic?

2. What is a finish?

3. If someone says: “Reveal yourself!,” what does (s)he mean?

4. What is bragging?

5. What is a shadow?

6. What is someone doing when (s)he scrambles up?

7. What is relief?

8. What is a medal?

9. What is someone doing when (s)he is on your heels?

10. What is curious?

11. What is a stopwatch?

12. If someone says: “Jill is nowhere to be found,” what does (s)he mean?

13. What is imperturbable?

14. What is a shore?

15. “The dog won’t budge”; what does that mean?

16. What is fear of heights*?

17. What is a referee?

18. If someone says: “Bugger off!”, what does (s)he mean?

19. What is protecting?

20. What is a starting line?

* In Dutch, fear of heights is called height fright (literally translated).

Appendix C. Part of the subscale narrative task (assessing narrative competence; Verhoeven & Vermeer, 2006)

1. Children show their dad an old pram

2. Dad constructs a cart/wagon

3. Children ride the cart/wagon

4. …

5. …

6. …

7. …

8. …

9. …

10. Mean 1 > goal 2

11. Cause 2 > effect 3

12. …

13. …

14. …

15. …

16. …

Appendix D. Story comprehension task (assessing story comprehension)

The Black Rabbit

1. How come Rabbit is being chased by Black Rabbit? (Why does Black Rabbit keep chasing Rabbit?)

2. Where was Black Rabbit when Rabbit was in the dark forest? (Does Rabbit see the Black Rabbit in the dark forest? Why (not)?)

3. Why does the Wolf not eat the Rabbit?

The Competition of the Tortoise and the Hare

1. When you think back: Who did you expect to win the competition? Why did you expect that?

2. How come the Tortoise won the competition?