An Insight into developmental changes in reasoning skills among Indian Preschoolers: A cross-sectional study using a story-based approach

Introduction

Reasoning is a cognitive process of making inferences, drawing conclusions, or evaluating a proposed conclusion (Andrews, 2020). Reasoning skills enable individuals to understand and learn about the physical and social environment on a daily basis (Khemlani, 2018). The ability to appropriately reason across different life situations is essential for a successful professional and everyday life (Bronkhorst et al., 2020). It was long believed that reasoning skills do not develop until adolescence (Piaget, 1952). Although preschoolers are known as ‘little scientists’ with abundant curiosity and an innate drive to know the world (Alvarez & Booth, 2014; Piaget, 1952), reasoning skills are considered to be a higher-level cognitive skill that is beyond the reach of preschoolers (Whittaker & McMullen, 2014). Thinking abilities among preschoolers have often been described as “egocentric, pre-logical, affective, undifferentiated, pre-causal, personal, vague and unanalysed” (Piaget, 1952). However, some recent researchers indicate that the development of reasoning skills begins during the preschool years and continues through adolescence (Amsterlaw, 2006; Gopnik et al., 2004; Legare et al., 2010; Säre, Luik & Fisher, 2016a; Whittaker & McMullen, 2014).

It is essential to explore the development of reasoning skills during the preschool age, as it significantly influences the development of other critical skills like language and literacy (Bauer & Booth, 2019; Nussipzhanova et al., 2018). Understanding reasoning skills during preschool years is vital for several critical decisions, such as the timely preparation of children to perceive and understand real-life scenarios, the right time for school entry, and laying the foundation for academic achievement and success (Niklas et al., 2018; Pasnak et al., 2015).

The existing literature suggests that reasoning abilities among preschoolers have been explored in domains like reasoning associated with improbable and less logical events like growing money on a tree, going back in time (Shtulman & Carey, 2007), counterfactual events (Muller et al., 2007), judgments on human abilities (Heyman et al., 2003), temporal ordering of events (McCormack & Hoerl, 2005), teleological functions (Kelemen et al., 2003), consistent and inconsistent events (Legare et al., 2010), pre-energy reasoning (Koliopoulos et al., 2009), analogies (Simms et al., 2018), and causal reasoning concerning scientific literacy associated with space, structures, tools, etc. (Bauer & Booth, 2019). However, the findings from these studies remain inconclusive in understanding the general reasoning development in preschoolers that could be fostered through education during the preschool period (Säre, Luik, & Fisher, 2016a). These studies seem to focus less on how children verbally reason in a general living context and how education could be the mediator to foster reasoning among preschoolers in such contexts. We could find one such recent initiative by Säre, Luik, and Fisher (2016a), wherein a valid and reliable tool ‘Younger Children Verbal Reasoning Test’ was developed to aid educators in assessing the general verbal reasoning of older preschoolers between five and six years of age. The tool utilized scenarios related to a theme on ‘bravery’ with illustrations and prompted children to judge how brave the characters are in the given scenario based on their actions. However, the authors of the work highlighted the psychometric properties of the developed tool and provided limited description regarding application of their tool in understanding general reasoning skills development in preschoolers.

The stimuli used to assess the reasoning abilities among preschoolers is a significant aspect to consider while discussing the development of general reasoning skills in this age group. In this context, Kendeou et al. (2019) put forth an integrated theoretical framework, inferential Language Comprehension (iLC), guiding and recommending the effective use of static or dynamic visual narratives (Stories) as a stimulus for evaluating reasoning skills among younger children. The study recommended that reasoning skills, such as inferencing opportunities, could be increased using static or dynamic visual narrative with questioning techniques in younger children who are non-readers. The story-based materials are reported to aid in better representation of general reasoning abilities in everyday communication contexts as they are more relatable to the natural learning context for the children (Killick & Boffey, 2012; O’Reilly et al., 2022). Stories also seem to provide an opportunity for children to provide explanations, form predictions, and make inferences that would help them to form logical, causally sequenced plots (Paris & Paris, 2003). Since a long, stories are known to constitute an inevitable part of childhood education, as children are intrinsically drawn to stories, and stories facilitate the development of thinking and learning in children (Paley, 2013). Recent research by Dawes et al. (2019) advocates the potential and sensitivity of a story-based approach to obtain insight into preschoolers’ developmental changes for skills such as inferential comprehension. The application of story-based stimuli to assess reasoning skills during preschool years has ranged from asking questions in an open-ended format after a short story sequence with picture cards (Muller et al., 2007; Shtulman & Carey, 2007) or after short fragments of story scenarios depicted through illustrations and objects (Bauer & Booth, 2019; Heyman et al., 2003; McCormack & Hoerl, 2005; Simms et al., 2018). A recent review by O’Reilly et al. (2022) highlighted the effectiveness of stories in fostering critical thinking, such as reasoning and problem-solving among preschoolers, and advocated the utility of story-based approaches in the field of critical thinking for future researchers.

Cognitive assessments, in general, have been recommended to be culturally and linguistically relevant to the age and community of the participants for yielding accurate findings (Tanveer et al., 2022). Likewise, though stories offer significant potential in assessing reasoning abilities among preschool age children, it is important to note that stories should be culturally and linguistically relevant for getting a true insight into the developmental trends of reasoning abilities among preschoolers (Rao et al., 2021; Sternberg, 2004).

Some other crucial aspects while exploring general reasoning skills among preschoolers, besides the stimuli, are the ways reasoning based responses are elicited with the children and the different tasks employed to assess their abilities. An interventional research on facilitating reasoning among preschoolers used undirected, shared picture book narration within a peer-group setting (Reed et al., 2015). The study qualitatively analyzed the narration discourse by preschoolers while they freely generated stories from picture books and observed the emergence of reasoning categories such as explanation, prediction, and inference. The findings of this research informed that spontaneously elicited reasoning categories in pre-schoolers without guidance could better represent the reasoning developmental changes in the preschool period. Regarding the tasks for assessing reasoning among preschoolers, different studies have employed tasks based on explanation, predictions, and inferencing. In the explanation-based reasoning tasks, participants are expected to explain the reason behind an outcome (Legare et al., 2010) whereas, in the prediction-based reasoning tasks, participants are required to predict the outcome of an event assuming a cause (Bonawitz et al., 2010). The inference-based reasoning task expects the participants to form a conclusion about a connection between an antecedent and an outcome with rational justification (Gopnik et al., 2004). Similar reasoning tasks based on explanation, prediction, and inference were utilized in preschoolers to explore reasoning skills related to the physical causality of objects by Blank et al. (1981). It seems most promising to incorporate explanation, prediction, and inference based reasoning tasks in a story-based approach while studying reasoning development during preschool years.

While realizing the need to explore the developmental trends of reasoning among preschoolers and the potential of culturally-linguistically relevant story-based approach for the same, the present research aimed at profiling the reasoning skills in typically developing Indian preschool children between 36 and 72 months using a story-based approach. The specific objectives of the research were to develop explanation, prediction, and inference based reasoning tasks around a story and assess reasoning skills in typically developing Indian preschool children using the same. The study did not expect the developmental trends of the reasoning abilities to be gender specific based on the evidence from previous literature (Ardila et al., 2011).

Methods

The study followed a cross-sectional research design following STROBE guidelines and commenced after obtaining Institutional Ethical Clearance from Kasturba Medical College, Mangalore, Manipal Academy of Higher Education (IEC KMC MLR 02-19/51). The data collection was conducted between June 2021 to January 2022 in Mangalore, India.

Results

The story-based tasks for assessing reasoning skills in preschoolers were developed. The details regarding of finalized story and the tasks are shown in Table 2. The psychometric properties of the tasks were determined using content validity and test-retest reliability measures. The content validity measures indicated appropriate validity as all the items of the reasoning tasks obtained an I-CVI of 0.83 or higher and S-CVI of 0.98 (Almanasreh et al., 2019; Polit & Beck, 2006). The test-retest reliability measures computed using ICC at 95% confidence interval with absolute agreement using two-way mixed model revealed an ICC of 0.76, indicating moderate test-retest reliability.

The validated story-based tasks were then used to assess the reasoning abilities of 63 typically developing Indian preschool children. The responses obtained from the participants across the explanation, prediction, and inference based reasoning tasks were quantitatively and qualitatively analyzed. The mean and standard deviation of the total reasoning scores and the percentage of categories of responses for the explanation, prediction, and inference based reasoning tasks across the age groups are illustrated in Figure 1(a,b), Figure 2(a,b), and Figure 3(a-c) respectively.

Figure 1. (a) Mean and standard deviation of the reasoning scores and (b) the percentage of categories of reasoning responses for the explanation task across the age groups.

Figure 2. (a) Mean and standard deviation of the reasoning scores and (b) the percentage of categories of reasoning responses for the prediction task across the age groups.

Figure 3. (a) Mean and standard deviation of the reasoning scores under inference, (b) justification of inference, and (c) the percentage of categories of reasoning responses for the justification of inference task across the age groups.

A comparison of the mean scores of the explanation based reasoning tasks revealed that the explanation based reasoning scores increased with age, F (2,60)=17.91, p<0.001, irrespective of the gender (t[61]=0.401,p=0.690). The post hoc pairwise comparison showed that the older groups (Groups II and III) performed significantly better (p<0.01) than the younger group (Group I); however, there was no statistically significant difference between the older groups (p=0.08). Receiver Operating Characteristic (ROC) analysis revealed a cut-off score of 2.5 discriminating between the younger (Group I) and older groups (Groups II and III) for explanation-based reasoning performance with 81% sensitivity, 62% specificity, and 0.8 Area under the curve (AUC) that is statistically significant (p<0.05).

The qualitative analysis using the percentage of categories of reasoning responses showed that with the increase in age, most children reasoned appropriately, integrating story content with prior knowledge. The codes identified under appropriate reasoning categories and their frequency across the age groups under the explanation task are presented in Table 3.

The comparison of the mean scores for the prediction based reasoning task also showed a significant increase with age F (2,60)=26.58, p<0.001, irrespective of the gender t (61)=0.564, p=0.575. Post hoc pairwise comparison revealed a significant difference in prediction scores between all three age groups (p<0.01). The ROC analysis revealed a cut-off prediction score of 4.5 discriminating between Group I and Group II with 52% sensitivity, 90% specificity, and 0.8 statistically significant AUC (p<0.001), and a cut-off prediction score of 5.5 discriminating Group II from Group III with 52% sensitivity, 86% specificity, and 0.7 AUC, that is statistically significant (p<0.05).

The qualitative analysis showed that similar to the explanation task, most children reasoned integrating story content with prior knowledge as their age increased. The codes identified under appropriate reasoning categories and their frequency across the age groups under the prediction task are presented in Table 4.

For the inference-based reasoning tasks, comparison of the mean score revealed that the inference scores remained almost similar across the age groups, F (2,60)=0.206, p=0.82, irrespective of the gender (t [61]=0.167, p=0.868). However, the scores pertaining to the justification for inference were found to increase with age, F (2,60)=6.04, p<0.05, without gender as a factor (t [61]=1.767, p=0.082). Post hoc pairwise comparison revealed significant differences in the inference justification performance between older groups (Groups II and III) and younger group (Group I) (p<0.05). There was no statistically significant difference between the older groups (Group II and III) (p=0.99). ROC analysis determined the cut-off score as 0.5, discriminating between the younger (Group I) and older groups (Groups II and III) for inference justification scores with 62% sensitivity, 86% specificity, and 0.7 statistically significant AUC (p<0.05).

Qualitative analysis showed that most of the responses belong to the ‘No reason’ category across the age groups and a smaller percentage of responses under ‘appropriate reasoning’ categories. Among the ‘appropriate categories,’ majority of the preschoolers reasoned under the category of ‘appropriate reasoning with story content.’ The codes identified under the smaller percentage of appropriate reasoning categories and their frequency across the age groups under the justification of inference task are presented in Table 5.

Discussion

The evaluation of psychometric properties of the story-based reasoning tasks revealed adequate content validity and test-retest reliability. The developed stimuli consisting of a story and reasoning tasks in a question format align with the theoretical framework proposed for exploring reasoning skills among preschoolers (Kendeou et al., 2019). The use of story as the assessment material is also consistent with the other research that supports the efficacy of story-based approaches among preschoolers (Dawes et al., 2019; O’Reilly et al., 2022). Besides adding culturally and linguistically relevant data on reasoning skills to the existing literature in this domain among preschoolers (Muller et al., 2007; Shtulman & Carey, 2007), the current study has offered additional depth by exploring the development of reasoning skills across explanation, prediction and inference domains using a single story. Moreover, the present format of these tasks, wherein domains of the reasoning abilities could be assessed on spontaneously emerged output of typically developing preschoolers, is aligned with the earlier perspectives in this regard (Reed et al., 2015). Further, the components of the story, such as the characters, sequences, objects, dress, pictures, vocabulary, and task formats, were designed to reflect the real-life Indian context, which could offer appropriate familiarity and suitability to Indian preschoolers and aid in accurate representation of their abilities (Rao et al., 2021; Sternberg, 2004).

The developed story-based tasks assessed the reasoning abilities of typically developing Indian preschool children. The comparison of mean reasoning scores across the age groups in the explanation, prediction, and justification of inference task revealed that the preschooler’s reasoning skills improved significantly with age. In most tasks (explanation and justification of inference), a significant change in performance was observed from Group I (36–48 months) to Group II (49–60 months) and III (61–72 months) than between Group II (49–60 months) and III (61–72 months), indicating the rapid development of reasoning skills during the initial phase of the preschool period. The results agree with the previous literature that younger preschoolers show a significant improvement in reasoning skills with age than older preschoolers (Hong et al., 2005). The results also imitate the age bands of thinking differences (Symbolic function substage: 2–4 years, Intuitive thought substage: 4 to 7 years) in Piaget’s preoperational stage of cognitive development (Piaget, 1952). However, the results obtained in the prediction task showed evidence for age-wise developmental changes than in age bands, with significant improvement in performance with age in years. The improved performance in reasoning skills with age across the tasks was not surprising as the preschool period is known to be critical for most of the developmental domains, including executive function skills (Garon et al., 2008).

Contrary to the expectation, the age-related difference in reasoning performance was not found in the inference task. More specifically, the findings from the inference task suggested that Group I (36–48 months) children performed like the Group II (49–60 months) and III (61–72 months) children. These results indicate that children as young as 36–48 months of age (Group I) can make inferences. Previous literature supports the findings attesting that preschoolers make logical inferences easily and earlier in development (Moshman, 2004). However, we could not exclude the fact that the inference task’s yes/no question nature might also have influenced the extent of this lack of age-related differences. In yes/no questions, researchers opine that younger preschoolers might show a yes/no bias (all yes responses, i.e., yes bias and vice versa) due to social pressure or difficulty inhibiting yes responses (Okanda & Itakura, 2010). Therefore, there is a chance that the younger participants in this study might have also undergone such yes/no bias in the inference task influencing the lack of age-related differences. Hence, it indicates that, though yes/no questions have been recommended for reasoning skills assessment (Säre et al., 2017), we need to be cautious regarding the chance of yes/no bias in younger preschoolers. However, despite the probability of yes/no bias in this study, the preschooler’s early inferential ability observed in the previous literature adds substantial weight to the conclusion of inferential ability in younger preschoolers.

Another interesting finding comparing mean reasoning scores was that the preschoolers performed poorly on the justification of inference task than on the other tasks. Though preschoolers could make inferences, the findings contend that they find it difficult to justify their inferences. This difficulty could be because the ability to justify the inferences is a higher quality of thinking requiring metacognition (Moshman, 2004). Though metacognition was not explicitly studied in this research, previous literature findings indicate that children require metacognitive awareness to justify their inferences (Whittaker & McMullen, 2014), which is beyond the preschool age in development (Veenman et al., 2006), lend support for the poorer performance in the current study. However, it is also noteworthy that a smaller percentage of older preschoolers (Group II, 49–60 months and Group III, 61–72 months) could provide appropriate responses (20%) in the justification of inference tasks using story content. Though it could be due to individual differences, it might also indicate the traces of metalogical development for reasoning from 4 years of age using the linguistic content, suggesting the scope of improvement in justification if provided with support.

On qualitative analysis, results from the explanation and prediction task revealed that among the appropriate reasons made by the preschoolers, younger preschoolers (Group I, 36–48 months) predominantly reason under the category of ‘reasoning with story content,’ and the older preschoolers (Group II, 49–60 months and Group III, 61–72 months) reason in the category of ‘reasoning integrating story content with prior knowledge.’ The results indicate that while younger preschoolers reason with reference to the inference based on the story content premise, older preschoolers undergo a higher thinking process of coordinating the logical inferences from the story and prior knowledge to conclude and reason. The results support the previous literature findings that preschoolers integrate inferences from linguistic context (Florit et al., 2011) and access prior knowledge while reasoning (Gopnik et al., 2004). The findings also align with the comprehension development process, which activates and integrates linguistic knowledge with background knowledge to connect the story contents and form a mental representation to comprehend (Kendeou et al., 2019).

Another interesting finding in qualitative analysis was with respect to the justification of the inference task. The results from the justification of inference task revealed that preschoolers depend on the source of knowledge from the linguistic contents to justify their inference. We have observed that older preschoolers tried to justify inference with respect to the source of knowledge from the story contents rather than the prior knowledge premise. It was also noted that most preschoolers provided empty reasons such as ‘I know’ for justification questions, claiming they had always known the inference. The findings are supported by the previous literature (Kuhn et al., 2004; Taylor et al., 1994) that children tend to give such responses when unsure or unaware of the answer. This could be because the metacognitive awareness of reasoning is in the emerging stage during the preschool period (Veenman et al., 2006). Based on the pattern of development observed in other tasks of reasoning, it could be expected that with increase in age over the preschool period, with the development of metacognitive awareness of reasoning, children might justify their inference majorly by integrating the story content with their prior knowledge and own interpretation, thereby following the similar pattern of reasoning development.

All the developmental changes in reasoning obtained in this study could be attributed to the biological maturational change of the brain and reconfiguration of frontoparietal networks (Andrews, 2020) and facilitated by the everyday interactional context (Niklas et al., 2018; Säre, Luik, & Tulviste, 2016b), working memory (Andrews, 2020), language (Richland & Burchinal, 2013) and theory of mind development (Taggart et al., 2005; Whittaker & McMullen, 2014). Consistent with existing literature gender wise differences were not found in all the reasoning tasks (Ardila et al., 2011). The study’s findings need to be cautiously generalized, as reasoning responses are dynamic.

In conclusion, the study describes the general reasoning skill development in typically developing Indian preschool children between 36 to 72 months of age using story-based tasks. The study put forth a valid and reliable story-based task for the purpose, and the assessment concluded that reasoning skills increase with age during preschool in a pattern of moving from reasoning based on story content to reasoning integrating the story content with prior knowledge. The study findings provide clinical and educational implications with insights on general reasoning development in preschoolers and the utility of age-appropriate story-based explanation, prediction, and inference question tasks to assess and foster reasoning responses in preschoolers. The findings also guide educators to scaffold reasoning in preschoolers, and the developed stimuli aid in early identification of reasoning deficits and guide planning intervention for Indian preschoolers with reasoning deficits. Though the study findings provide insight into general reasoning development in preschoolers, the role of contributing factors such as working memory, theory of mind, and social interaction was not explored, and it remains a limitation of the present study. Therefore, future studies considering such factors could provide an in-depth understanding of reasoning development to assess and foster reasoning during preschool years.