What Can Young Children Really Do? Pre-Service Teachers’ Contradictory Beliefs and Implications for Professional Teacher Education

Abstract

Pre-service teachers’ (PTs) beliefs about young children’s cognitive abilities shape both their instructional practices and their developing understandings of teaching and learning. This study examined PTs’ beliefs about preschool children’s cognitive abilities, focusing on cognitive operations, conceptual change, and learning processes, in relation to emerging professional identity development. A cross-sectional comparative design was employed with a convenient sample of 241 students from Early Childhood Education Departments who completed the Childhood and Cognitive Abilities Questionnaire. The findings revealed statistically significant differences between participants with and without practicum experience, with the former reporting more sophisticated beliefs, aligned with constructivist learning approaches. However, many participants simultaneously endorsed child-centered perspectives and traditional transmission-based conceptions of teaching, indicating the coexistence of contradictory beliefs. Correlation and cluster analyses further suggested that participants’ beliefs formed broader but only partially coherent belief systems rather than consistent conceptual profiles. These findings may reflect tensions within PTs’ emerging professional identities and suggest that practicum-related experience may coincide with opportunities for reflection on and restructuring of prior beliefs, processes associated with a coherent professional identity. Overall, this study highlights the importance of teacher education programs systematically addressing misconceptions about children’s cognitive abilities, while fostering coherent, research-informed professional identities and evidence-based instructional practices in early childhood education.

1. Introduction

Teachers’ and pre-service teachers’ beliefs about knowledge, learning, and teaching, as well as their perceptions of students’ abilities, appear to shape how they view their students and also their instructional practices (Fives & Buehl, 2016; Sherwood, 2001). Such beliefs are formed early, shaped by teachers’ own experiences as students, and subsequently act as a filter for both the decisions they make during the instructional process and for their interpretations of classroom interactions (Buehl & Beck, 2015; Fives & Buehl, 2012; Pajares, 1992). Contemporary research indicates that these beliefs may be organized into complex and multidimensional belief systems, in which contradictory beliefs—such as “teacher-centered” and “learner-centered” educational beliefs—may coexist (Ferguson & Bråten, 2018; Kyriakopoulou & Skopeliti, 2024, 2025; Vosniadou, 2017). Notably, teachers are often unaware of these deeply rooted, implicit beliefs, and therefore may fail to recognize how such beliefs influence the instructional process—either supporting it or, more frequently, hindering it (Vosniadou et al., 2020, 2023).

For instance, educators, who view learning ability as neither innate nor fixed, tend to interact more meaningfully with their students and encourage them to participate actively in processes of knowledge construction (Bråten & Strømsø, 2005; Chan & Elliott, 2004; Ferguson, 2020; Yeager & Dweck, 2012). Such constructivist-oriented approaches conceptualize children as active participants in the learning process, emphasizing the role of prior knowledge, active engagement, social interaction, and the gradual construction and reorganization of understanding through interaction with the environment. Those who support the personal construction of knowledge for their students also believe that learning involves not only a deep processing of new information but also an emotional and personal engagement with it (Sinatra & Kardash, 2004). By contrast, educators who conceive learning primarily as a product of their own instruction rather than as a distinct process, are more likely to assume that students learn in specific, predetermined ways and to view the transmission of knowledge and information as their main task (Ferguson & Bråten, 2018; Vosniadou et al., 2020, 2021). Within this transmission-oriented perspective, learning is viewed mainly as the direct transfer of knowledge from teacher to learner, with the teacher regarded as the primary source of knowledge and the learner taking comparatively a more passive role.

Teachers’ expectations also play a critical role in shaping their educational practices, consequently, in shaping students’ academic outcomes. Low or biased expectations may lead students to very low achievement, reinforcing educational inequities (Alvidrez & Weinstein, 1999; Rimm-Kaufman et al., 2000; Rubie-Davies & Hattie, 2025). These expectations are often shaped by broader social and contextual factors, including gender, socioeconomic background, and ethnicity (Baker et al., 2015; Donovan & Cross, 2002; Doyle et al., 2023; Gentrup et al., 2024). For example, gender-related beliefs about academic domains (e.g., mathematics) have been shown to influence both teachers’ expectations and students’ self-perceptions of their own competence, often to the disadvantage of girls (Lindner et al., 2022). In addition to influencing instructional practices directly, such expectations may also be conveyed to students—either implicitly or explicitly—contributing to the internalization of stereotype-related beliefs. These internalized beliefs, in turn, affect students’ cognitive and motivational processes. Longitudinal evidence indicates that endorsement of academic stereotypes is associated with lower cognitive engagement, stronger fixed mindset beliefs, and reduced academic performance over time (Z. Wang & Pan, 2025; M. T. Wang et al., 2022). Similarly, research indicates that students from socioeconomically disadvantaged or minority backgrounds are more likely to be subject to lower expectations, which may influence both teachers’ evaluations and instructional decisions (Doyle et al., 2023). Such patterns indicate that teacher expectations can function as a mechanism through which educational inequalities are reproduced within classroom settings. Earlier work by Ready and Wright (2011) also identified systematic biases in teachers’ perceptions, including the overestimation of girls’ skills by female teachers and the underestimation of the abilities of students from disadvantaged or minority backgrounds, even when actual performance was taken into account.

Beyond their role in shaping instructional practices, teachers’ beliefs can also be viewed as important components of their developing professional identity (Fives & Buehl, 2012). Contemporary perspectives on teacher education conceptualize becoming a teacher not simply as the acquisition of knowledge and skills, but as a developmental process in which pre-service teachers construct and adapt their professional identity through experiences in specific educational contexts (Beijaard et al., 2004). In this sense, examining pre-service teachers’ beliefs is also important to understanding the broader development of professionalism in early childhood education, as becoming a teacher involves not only acquiring pedagogical knowledge, but also developing coherent ways of understanding children, learning, and teaching.

These ways of understanding, however, should not be considered fully formed during initial teacher education. Instead, they appear to develop through ongoing processes of professional learning, where reflection on practice, engagement with peers, and classroom experience contribute to the gradual reorganization and alignment of teachers’ beliefs and practices over time. From this perspective, professionalism can be understood as a continuous, developmental process, where coherence is progressively constructed and reconstructed over time through practice-informed reflection and collaboration. This view is further supported by research on continuous professional development, which highlights how teachers’ beliefs are not only enacted in practice but are also reshaped through iterative cycles of reflection, implementation, and feedback, linking professional learning directly to changes in teaching and learning outcomes (Shi & Chen, 2026).

Within this broader developmental perspective, beliefs about children’s abilities should not be treated as isolated cognitive constructs, but rather as elements embedded within wider belief systems through which pre-service teachers interpret their role, their students, and their pedagogical responsibilities (Buehl & Beck, 2015; Vosniadou et al., 2020). For example, in early mathematics education, research has shown that such beliefs influence how technology is integrated into learning environments and how mathematical concepts are introduced to young children (Alsaeed & Aladil, 2024), illustrating how broader belief systems become operationalized in specific instructional contexts. Recent research further suggests that these belief systems also shape pre-service teachers’ orientations toward equity and linguistic diversity in mathematics education. Fernández et al. (2026), for instance, found that pre-service teachers who held more constructivist beliefs about mathematics were more likely to adopt equity-oriented perspectives toward Emergent Bilingual students, whereas more traditional beliefs were associated with more limited or moderately equitable perspectives. Their findings indicate that beliefs about mathematics, language, and pedagogy form interconnected interpretive frameworks that influence how future teachers perceive students’ capabilities, participation, and access to meaningful opportunities for mathematical learning.

From this perspective, inconsistencies within belief systems may reflect not simply a lack of knowledge, but an ongoing process of professional identity formation, in which pre-service teachers bring together ideas drawn from prior experiences, intuitive conceptions, and formal pedagogical knowledge (Pillen et al., 2013). Empirical evidence illustrates how such inconsistencies can be restructured during teacher education. For instance, recent research has shown that pre-service teachers’ beliefs about children’s participation in early childhood education are often initially fragmented and teacher-centered, yet can undergo significant transformation through targeted interventions during initial teacher education, shifting towards more participatory, child-centered, and rights-based perspectives (Avgitidou et al., 2024). These shifts are also accompanied by shifts in how pre-service teachers conceptualize children’s capabilities, the role of the teacher, and the nature of learning, indicating that belief revision is closely connected to processes of professional identity reconstruction. Understanding how such inconsistencies emerge and are gradually reorganized during teacher education can offer useful insights into how teacher education can better support the preparation of future early childhood educators.

Although a substantial body of research has examined teachers’ beliefs about learning, knowledge, and teaching (Ferguson, 2020; Fives et al., 2015; Kyriakopoulou & Skopeliti, 2024, 2025; Vosniadou et al., 2020), fewer studies have explored how such beliefs, particularly those concerning young children’s cognitive abilities, relate to the development of pre-service teachers’ professional identity. However, teachers’ perceptions of their students’ cognitive abilities and the value they assign to different developmental skills and competencies, appear to guide their instructional choices (Kowalski et al., 2001). Avgitidou et al. (2013) found that pre-service teachers enter their studies with well-established ontological and epistemological beliefs about childhood that can impede a participatory preschool pedagogy which recognizes children as capable and active agents. The authors examine these assumptions through three central dichotomies: universality versus locality, agency versus structure, and being versus becoming. More specifically, ontological assumptions concern how childhood itself is conceptualized, for example whether it is understood as a homogeneous and universal category or as socially and culturally situated, whether children are viewed as active agents or as dependent subjects constrained by adult structures, and whether childhood is perceived as a fixed state of “being” or as a dynamic process of “becoming.” Epistemological assumptions, in turn, concern how knowledge about childhood and children’s abilities is constructed and interpreted. Their findings showed that many pre-service teachers consider preschool children to be cognitively deficient and conceptualize childhood as a distinct, homogeneous and biologically determined life stage, emphasizing age-related maturity over children’s actual cognitive potential and underestimating their own role in fostering such potential. As a result, children’s participation and decision-making abilities are often seen as limited and conditional upon adult guidance. Recent research further suggests that these beliefs are reflected not only in how pre-service teachers perceive children, but also in how they position themselves pedagogically in relation to children’s agency and participation. Melasalmi et al. (2023) found that pre-service early childhood teachers often approached playfulness and pedagogy through teacher-initiated and control-oriented forms of agency, emphasizing guidance, responsibility, and predetermined learning goals rather than relational and community-shared forms of playful interaction. Similarly, Macias and Damjanovic (2026) showed that many pre-service teachers experienced discomfort in inquiry-based and child-centered learning environments, expressing preference for structured and teacher-directed settings with clearly defined expectations and boundaries. Their findings suggest that pre-service teachers’ underlying epistemological and developmental assumptions about children influence not only how they perceive children’s competencies, but also how they construct their own professional roles, agency, and pedagogical identities.

Preschool teachers do not appear to be guided solely by their beliefs about learning and teaching, but their practices are also shaped largely by personal values and experiences, resulting in varying educational priorities, such as play, socialization, or skill development (Spodek, 1987). This variability is also reflected in more recent empirical research, which identifies distinct profiles of teachers’ approaches to play and assessment, each associated with different pedagogical orientations and instructional practices (Pyle et al., 2022). Contemporary research further indicates that pedagogical beliefs are domain-specific and that teachers often draw on multiple instructional approaches, which may differentially influence instructional practices and the quality of classroom interactions (Wieduwilt et al., 2023). Although teachers often report holding developmentally appropriate beliefs, these beliefs are not always consistently reflected in practice. Research suggests that the relationship between beliefs and practice is complex and shaped by contextual, interactional, and institutional factors, making the translation of pedagogical beliefs into concrete instructional strategies particularly challenging (Pyle et al., 2022; Wieduwilt et al., 2023; Vartuli, 1999).

In addition, studies suggest that preschool educators often prioritize the development of socio-emotional skills and children’s overall well-being over the acquisition of academic skills, such as literacy and scientific or mathematical thinking (Abry et al., 2015; Kowalski et al., 2001; Simeunović & Milić, 2026; Xiang, 2025). Educators also appear to favor child-centered and play-based pedagogies that promote emotional security, participation, and holistic development rather than placing an early emphasis on academic performance (Duff, 2026). Recent evidence further suggests that pre-service kindergarten teachers tend to move from more academically oriented understandings of school readiness toward broader developmental perspectives that emphasize emotional regulation, social interaction, and autonomy as they progress through their studies and practicum experiences (Xiang, 2025). At the same time, contemporary research on professional development highlights the importance of whole-child approaches, reflective practice, emotional competence, and culturally responsive pedagogy in effective early childhood teaching (Shi & Chen, 2026). Moreover, competencies such as attentiveness, compliance with instructions, and task completion, are increasingly viewed as important dimensions of school readiness and later academic adjustment (Geng et al., 2026), supporting earlier arguments that socio-emotional competencies may also function as forms of academic self-regulation (Lin et al., 2003).

Moreover, evidence suggests that early childhood educators may not always fully recognize the academic potential of young children, particularly in domains such as mathematics. Contemporary research in early mathematics education increasingly emphasizes that preschool children are capable of engaging in relatively sophisticated forms of mathematical reasoning, problem solving, and conceptual thinking when provided with appropriate learning opportunities (Elia et al., 2023). Studies indicate that mathematical and STEM learning opportunities in early childhood settings often remain limited and are not always treated as central components of preschool learning experiences (Johnston et al., 2022). Research has also shown that educators’ anxiety and limited confidence in mathematics may further restrict the frequency and quality of mathematical interactions in preschool classrooms (Galeano et al., 2024). Earlier findings further suggest that preschool educators may underestimate children’s mathematical competencies, place greater emphasis on language or socio-emotional development, and rely more on general developmental impressions than on systematic observations of children’s mathematical thinking (Kilday et al., 2012). However, early mathematical abilities appear to be supported by a network of both general and domain-specific cognitive skills. Working memory and phonological skills are, in fact, among the most reliable predictors of the development of numerical competence. Therefore, fostering these capacities in kindergarten can support the early growth of mathematical thinking (Passolunghi et al., 2015).

Despite this, research indicates that preschool teachers do not always fully recognize the extent of young children’s mathematical potential and at the same time research shows that teachers’ perception of what children are capable of learning appears to directly influence their instructional practices. Preschool teachers’ understanding of the mathematical skills typically acquired by preschoolers is linked to both the frequency and the type of mathematics instruction they provide. Teachers with more accurate knowledge of what four-year-olds can learn designed mathematical activities more frequently (Ban et al., 2024). A similar pattern emerges in the area of emergent literacy, where studies highlight that preschool teachers’ limited preparation underscores the need to strengthen their understanding of early emergent literacy processes (Lynch, 2009). Enhancing teachers’ knowledge of the range of skills children can develop across domains therefore appears to be a key factor in supporting more balanced and developmentally appropriate learning experiences.

From this perspective, pre-service teachers’ beliefs can also be examined in relation to processes of conceptual change, as they often reflect the coexistence of prior intuitive ideas and newly acquired pedagogical knowledge (Vosniadou et al., 2023). Research has shown that pre-service teachers may simultaneously endorse constructivist views about learning while reproducing more teacher-centered instructional practices, revealing tensions between epistemic beliefs, educational beliefs, and pedagogical decision-making (Kyriakopoulou & Skopeliti, 2024; Vosniadou et al., 2023). Similar tensions were identified by Pantazidis (2026) in a qualitative study of Greek pre-service early childhood teachers prior to practicum, where child-centered and democratic educational ideals coexisted with assumptions related to teacher authority, classroom control, knowledge transmission, and the successful implementation of learning activities. Pantazidis also found that participants held contradictory views about children’s capacities and participation, simultaneously portraying children as vulnerable and dependent learners in need of adult guidance, while also recognizing them as active subjects capable of expressing ideas, participating in decision-making, and contributing to learning processes. Drawing on the concepts of the “schoolized mind” and “educational imaginaries,” Pantazidis argued that these tensions reflect the coexistence of deeply internalized assumptions about schooling with alternative pedagogical orientations promoted through teacher education. In a similar vein, Avgitidou et al. (2013) argue that such beliefs may function as a broader “framework theory” through which pre-service teachers interpret childhood, children’s abilities, and educational relationships. From this perspective, contradictory beliefs should not necessarily be viewed as isolated inconsistencies, but rather as part of broader and often deeply rooted interpretive frameworks that shape how future teachers make sense of teaching, learning, authority, and children’s participation in classroom life. Such inconsistencies may therefore reflect not simply incomplete understanding, but an ongoing process of restructuring beliefs and attempting to reconcile different ways of understanding teaching, learning, authority, children’s cognitive engagement, and children’s participation in classroom life. In this sense, the gradual reorganization of these beliefs can be seen as closely connected to the development of a more coherent and reflective professional identity.

The Present Study

The purpose of this study is to explore pre-service teachers’ (PTs) beliefs about the cognitive abilities of preschool-aged children, with a particular focus on their views of cognitive functions, conceptual change, and learning. More specifically, it examines how future educators understand the operation of key cognitive processes—such as memory, attention, and self-regulation—the mechanisms through which young children revise and reorganize their concepts, and the ways in which children of this age learn and actively construct knowledge. In doing so, the study emphasizes how pre-service teachers conceptualize the dynamic processes that support the development of thinking in early childhood and whether these beliefs form a coherent belief system or reflect the coexistence of conflicting beliefs.

Building on this, this study also investigates the role of practicum experience in shaping pre-service teachers’ beliefs, by comparing participants with and without practicum experience. Practicum is conceptualized as a situated context where prior beliefs interact with formal pedagogical knowledge. Within this context practicum is viewed as a potentially important context where pre-service teachers construct their emerging professional identity, balancing child-centered perspectives with more traditional conceptions of teaching (Izadinia, 2013).

Based on the literature reviewed, it was expected that pre-service teachers’ beliefs would reflect the coexistence of both constructivist and more teacher-centered orientations regarding children’s cognitive abilities and learning processes. It was also anticipated that practicum experience would be associated with differences in these beliefs, reflecting the role of professional learning experiences in the gradual reorganization of pre-service teachers’ belief systems and professional identity.

2. Method

2.1. Participants

The study included 241 pre-service teachers (PTs), primarily female (98.3%), enrolled in undergraduate programs in the Departments of Early Childhood Education at the National and Kapodistrian University of Athens (73.9%) and University of Patras (26.1%). Participants’ ages ranged from 18 to over 41 years, with the majority between 18 and 20 years old (56%), followed by those aged 21–25 (29.9%).

Participants were recruited using a convenience sampling approach. Students were invited to participate voluntarily during scheduled courses and through institutional communication channels. Inclusion criteria required participants to be enrolled in an undergraduate early childhood education program at the time of data collection.

Participants were classified into two groups based on their practicum experience. The no-practicum group included 116 participants in their first (37.8%) or second (10.4%) year of studies, who had not yet attended practicum-related courses. The practicum group consisted of 125 participants, predominantly in their third (34.5%) or fourth (15.5%) year.

All participants had completed an introductory first-year course related to child development, providing a common foundational background regarding children’s development and learning. In addition, students in the practicum group had attended three courses involving more systematic engagement with pedagogical theories and teaching practices in preschool education. Specifically, these practicum-related courses involved progressively structured engagement with preschool educational settings through classroom observations, participation in teaching-related activities, guided reflection assignments, and forms of supervised teaching participation. Students were introduced to classroom interaction, observation and planning of educational activities, and reflective discussion of pedagogical practice under the guidance of university instructors and cooperating preschool educators.

The study employed a cross-sectional comparative design, examining differences between pre-service teachers with and without practicum experience at a single point in time.

2.2. Childhood and Cognitive Abilities Questionnaire (CCogAb)

To investigate PTs’ beliefs about early childhood and children’s cognitive abilities the Childhood and Cognitive Abilities (CCogAb) Questionnaire was developed. The questionnaire comprised 39 items designed to capture beliefs across three broad conceptual categories related to children’s cognitive development and cognitive abilities: (a) cognitive operations, (b) conceptual change and prior knowledge, and (c) learning processes (Appendix A). Participants had to indicate their degree of agreement or disagreement on a 6-point Likert scale (1. Strongly disagree, 2. Disagree, 3. Somewhat disagree, 4. Somewhat agree, 5. Agree, 6. Strongly agree).

The development of the CCogAb questionnaire was informed by research on conceptual change and prior work examining the coexistence of contradictory beliefs in educational contexts, particularly using the Beliefs About Learning and Teaching (BALT) questionnaire developed by Vosniadou et al. (2020). Previous studies using the BALT questionnaire, including research conducted in the Greek educational context (Kyriakopoulou & Skopeliti, 2024, 2025), have shown that pre-service teachers often simultaneously endorse constructivist and transmission-oriented beliefs about learning and teaching. Within this line of research, beliefs are conceptualized as parts of broader and often internally inconsistent belief systems rather than as fully coherent and independent dimensions. Following this perspective, the questionnaire was designed to capture potentially overlapping beliefs regarding children’s cognitive abilities, conceptual development, and learning processes, rather than to assess fully distinct psychological constructs. The three conceptual categories were therefore used primarily as theoretically informed organizational groupings intended to facilitate the interpretation of different aspects of pre-service teachers’ beliefs.

Consistent with this perspective, the questionnaire included both constructivist-oriented and transmission-oriented statements, as well as positively and negatively phrased items. This approach, informed by previous work using the BALT questionnaire (Vosniadou et al., 2020), was intended to reduce the tendency of participants to respond uniformly in socially desirable directions when reporting pedagogical beliefs.

The items were distributed across the three general categories as follows: (a) Cognitive Operations: twelve items assessing beliefs about strategy use, memory, problem solving, self-regulation and metacognition (e.g., “Preschool-aged children cannot use mnemonic strategies at all to organize material they need to learn”); (b) Conceptual Change and Prior Knowledge: thirteen items assessing beliefs about conceptual changes in children’s thinking (e.g., “It is a waste of time to teach scientific concepts in kindergarten (e.g., from the domains of astronomy, biology, etc.) because children are too young to understand them”) and the role of prior knowledge in conceptual change (e.g., “Preschool-aged children’s prior knowledge may interfere with the way they learn.”); and (c) Learning Processes: fourteen items assessing beliefs about how children learn and the type of interaction with peers, teachers or other adults (e.g., “It is important for adults to provide preschool-aged children with all the necessary knowledge, as they cannot learn on their own”).

The internal consistency reliability of all 39 items was evaluated using Cronbach’s alpha, resulting in a value of a = 0.84. Similarly, the internal consistency reliability for the subcategory “Cognitive Operations” was a = 0.70, for “Conceptual Change and Prior Knowledge” was a = 0.58, and for “Learning Processes” was a = 0.68. For the subcategory “Conceptual Change and Prior Knowledge” the corrected item-total correlations were examined to identify any items that negatively affected the overall reliability of the scale. Three items (Q15, Q18 and Q25) were identified with low corrected item-correlations (below 0.3) and higher Cronbach’s alpha if item deleted values and were excluded from subsequent analyses: Q15 [Preschool-aged children’s thinking does not develop in the same way, even when children encounter the same kinds of problems], Q18 [Lack of prior knowledge in preschool-aged children may create difficulties in understanding new information], and Q25 [Preschool-aged children do not yet possess all the knowledge necessary for learning]. These items appeared to reflect conceptually broader or more ambiguous assumptions that may have been interpreted differently by participants. In particular, some items combined multiple conceptual dimensions, whereas others referred to broadly accepted assumptions about learning and prior knowledge that may not have differentiated clearly between participants’ belief orientations. These characteristics may explain their low corrected item-total correlations and unstable contribution to the overall scale structure. Following the removal of these items, the revised “Conceptual Change and Prior Knowledge” category yielded α = 0.73, whereas “Cognitive Operations” yielded α = 0.70 and “Learning Processes” α = 0.68. Although the reliability of the Learning Processes category falls slightly below the conventional 0.70 threshold, the three conceptual categories were retained as theoretically informed organizational groupings intended primarily to facilitate interpretation rather than to represent strongly differentiated psychometric dimensions.

Exploratory factor analysis (principal axis factoring with oblimin rotation) was conducted after the removal of three items. The Kaiser–Meyer–Olkin measure confirmed sampling adequacy (KMO = 0.823), and Bartlett’s test of sphericity was significant (p < 0.001), supporting the suitability of the data for factor analysis. A three-factor solution accounted for 28.59% of the total variance. This percentage is substantially lower than the approximately 50% variance commonly considered desirable in exploratory factor analysis within the social sciences, further indicating that the extracted factors should not be interpreted as clearly distinct empirical dimensions. The first factor emerged as dominant, explaining 17.81% of the variance, whereas the second and third factors contributed substantially less. Inspection of the pattern matrix revealed that several items loaded strongly on the first factor, while the remaining factors were characterized by lower loadings and notable cross-loadings across factors. Communalities were generally modest, indicating that a considerable proportion of item variance was not captured by the extracted factors, suggesting a weak multidimensional structure. The three conceptual categories should therefore be understood primarily as theoretically informed organizational groupings intended to facilitate interpretation rather than as strongly differentiated psychometric constructs. These findings should also be interpreted with caution, given both the exploratory nature of the instrument and the theoretical assumption that beliefs about children’s learning and cognitive abilities may form interconnected rather than fully differentiated belief systems. The moderate correlations observed between the conceptual categories were considered theoretically meaningful within the framework theory approach, according to which partially conflicting or overlapping beliefs may coexist within broader interpretive frameworks (Kyriakopoulou & Skopeliti, 2024, 2025; Vosniadou et al., 2020).

2.3. Scoring & Data Analysis

The items of the questionnaire were assessed using a 6-point Likert scale ranging from 1 to 6, reflecting the participants’ level of agreement or disagreement and each item received a score from 1 to 6. For statistical analyses scores were reversed for 14 items, as more advanced beliefs involved some degree of disagreement (e.g., “All preschool-aged children learn in the same way”). In the present study, the characterization of certain beliefs as more “advanced/sophisticated” reflects the theoretical framework adopted by the authors, drawing primarily on constructivist and developmental approaches to learning and teaching. Accordingly, higher scores reflected beliefs considered more consistent with constructivist and developmentally appropriate perspectives, rather than universally accepted standards of pedagogical correctness.

As noted, three items were excluded from the subsequent statistical analyses. Independent samples t-tests were conducted to examine differences between participants with and without practicum experience across questionnaire items. Homogeneity of variances was assessed using Levene’s test, and when this assumption was violated, results from the Welch t-test were considered. The interpretation of statistical significance did not differ across the two approaches. Given the sample size of each group, t-tests were considered robust to moderate deviations from normality.

For the purposes of data presentation, items are organized into three broad conceptual categories. These groupings are used to facilitate interpretation and do not imply distinct empirically validated dimensions. All questionnaires included in the analyses were fully completed and no questionnaires with missing responses were identified in the final dataset.

2.4. Procedure

PTs were informed about the study’s content, purpose, research procedure and the protection of personal data prior to participation. Those who expressed a willingness to participate provided informed consent. Participation was voluntary, anonymous, and confidential. Participants consented to both their participation and the use of their responses for statistical analyses and the publication of results. The study was conducted in accordance with institutional ethical guidelines for research involving human participants. Data were collected at a single time point through an online questionnaire administered via Google Forms. Completion of the CCogAb questionnaire required approximately 15–20 min.

3. Results

3.1. Group Comparisons of Overall and Subcategories Mean Scores

Table 1. Mean Performance on the CCogAb Questionnaire for all participants and per Practicum Group.

Category of Statements	All PTs (N = 241)	No-Practicum (N = 116)	Practicum (N = 125)	t-Test	Cohen’s d
Cognitive Operations	M = 4.54 (SD = 0.55)	M = 4.40 (SD = 0.53)	M = 4.68 (SD = 0.49)	t (239) = −3.927; p ≤ 0.001	−0.51
Conceptual Change & Prior Knowledge	M = 4.38 (SD = 0.67)	M = 4.07 (SD = 0.62)	M = 4.66 (SD = 0.58)	t (239) = −7.646; p ≤ 0.001	−0.99
Learning Processes	M = 4.66 (SD = 0.49)	M = 4.54 (SD = 0.41)	M = 4.77 (SD = 0.53)	t (239) = −3.616; p ≤ 0.001	−0.47
Total CCogAb Questionnaire	M = 4.54 (SD = 0.48)	M = 4.36 (SD = 0.40)	M = 4.71 (SD = 0.49)	t (239) = −5.880; p ≤ 0.001	−0.76

depicts the PTs’ mean performance across the overall CCogAb questionnaire as well as within each category of statements. PTs demonstrated the higher mean scores in the category “Learning Processes”. Independent-samples t-tests showed significant differences between PTs with and without practicum experience across all three categories and the total CCogAb score. According to Cohen’s conventional thresholds (0.20 = small, 0.50 = moderate, 0.80 = large), effect sizes were moderate for “Cognitive Operations” and “Learning Processes”, moderate-to-large for the total CCogAb score, and large for “Conceptual Change & Prior Knowledge”.

3.2. Item-Level Differences

Table 2. Mean Scores for statements across the general categories of the CCogAb Questionnaire with statistically significant performance differences between practicum groups.

General Category	Content of Statement	No-Practicum (N = 116)	Practicum (N = 125)	t-Test	Cohen’s d
Cognitive Operations	S4: Teaching strategies	M = 4.78 (SD = 1.08)	M = 4.45 (SD = 1.42)	t (239) = 1.998; p = 0.04	0.26
	S22: Problem solving *	M = 4.29 (SD = 1.35)	M = 4.73 (SD = 1.22)	t (239) = −2.618; p = 0.01	−0.34
	S23: Attention focus	M = 2.45 (SD = 1.15)	M = 2.88 (SD = 1.23)	t (239) = −2.800; p = 0.01	−0.36
	S29: Metacognition	M = 3.70 (SD = 1.21)	M = 4.33 (SD = 1.17)	t (239) = −4.083; p ≤ 0.001	−0.53
	S31: Teaching strategies *	M = 4.91 (SD = 1.28)	M = 5.26 (SD = 0.90)	t (239) = −2.406; p = 0.017	−0.31
	S35: Attention focus *	M = 3.89 (SD = 1.30)	M = 4.78 (SD = 1.08)	t (239) = −5.804; p ≤ 0.001	−0.75
	S38: Memory strategies	M = 4.36 (SD = 1.11)	M = 4.66 (SD = 1.02)	t (239) = −2.193; p = 0.029	−0.28
Conceptual change and Prior knowledge	S1: Prior knowledge within domains	M = 3.13 (SD = 1.34)	M = 4.22 (SD = 1.14)	t (239) = −6.754; p ≤ 0.001	−0.87
	S3: Prior knowledge	M = 4.54 (SD = 1.33)	M = 4.95 (SD = 1.11)	t (239) =−2.590; p = 0.01	−0.33
	S9: Teaching scientific concepts *	M = 4.57 (SD = 1.23)	M = 5.32 (SD = 0.972)	t (239) = −5.256; p ≤ 0.001	−0.68
	S13: Prior knowledge *	M = 5.13 (SD = 1.08)	M = 5.49 (SD = 0.867)	t (239) = −2.847; p = 0.005	−0.37
	S19: Knowledge about the social domain	M = 2.95 (SD = 1.31)	M = 3.62 (SD = 1.28)	t (239) = −3.977; p ≤ 0.001	−0.51
	S20: Language-Literacy *	M = 4.30 (SD = 1.28)	M = 5.14 (SD = 1.08)	t (239) = −5.465; p ≤ 0.001	−0.70
	S30: Number concept *	M = 4.20 (SD = 1.32)	M = 5.08 (SD = 1.02)	t (239) = −5.789; p ≤ 0.001	−0.74
	S34: Understanding scientific concepts	M = 4.01 (SD = 1.36)	M = 5.00 (SD = 1.18)	t (239) = −6.028; p ≤ 0.001	−0.78
Learning Processes	S2: Direct teaching *	M = 2.72 (SD = 1.47)	M = 3.21 (SD = 1.67)	t (239) = −2.377; p = 0.018	−0.31
	S5: Adult contribution	M = 5.00 (SD = 1.03)	M = 4.68 (SD = 1.28)	t (239) = 2.121; p = 0.035	0.27
	S14: Direct provision of knowledge *	M = 4.13 (SD = 1.38)	M = 4.84 (SD = 1.26)	t (239) = −4.169; p ≤ 0.001	−0.54
	S27: Pre-existing skills	M = 3.98 (SD = 1.13)	M = 4.27 (SD = 1.13)	t (239) = −1.980; p = 0.049	−0.26
	S33: Direct teaching *	M = 2.75 (SD = 1.35)	M = 3.88 (SD = 1.50)	t (239) = −6.116; p ≤ 0.001	−0.79
	S37: Processing information	M = 4.20 (SD = 1.24)	M = 4.56 (SD = 1.11)	t (239) = −2.376; p = 0.018	−0.31

* Items with reversed scoring: For reversed-scored items, lower scores indicate stronger endorsement of the original transmission-oriented statement, whereas higher scores indicate disagreement with the original statement and greater alignment with constructivist-oriented views.

presents the PTs’ mean scores for the statements within each general category for which statistically significant differences emerged between the two groups.

Within the “Cognitive Operations” category, significant differences emerged for several items. Participants with practicum experience reported higher scores in items related to problem solving, attention, metacognition, and memory strategies (e.g., S22, S23, S29, S35, S38), indicating more sophisticated beliefs about children’s cognitive capabilities. Effect sizes for these differences ranged from small to moderate-to-large (|d| = 0.28–0.75), with the largest effect observed for beliefs about sustained attention (S35, d = −0.75). In contrast, participants without practicum experience scored higher on the item related to the role of teaching strategies (S4), although these effects were small (d = 0.26). Notably, for both groups, a low mean score was observed for S23 “Preschool-aged children can remain focused on an activity for an extended period of time”, whereas a higher mean score, indicating disagreement and aligning with a more sophisticated belief, was observed for S35 “Preschool-aged children cannot sustain their attention on an activity”.

In “Conceptual Change & Prior Knowledge” category participants with practicum experience consistently demonstrated higher scores across all significant items, including beliefs about prior knowledge (S1, S3, S13), understanding of scientific concepts (S9, S34), and domain-specific knowledge such as literacy and mathematics (S20, S30). However, both groups appeared less likely to endorse statements related to the development of social cognition and theory of mind, as indicated by the low mean scores in S19. Effect sizes in this category were generally moderate to large (|d| = 0.33–0.87), indicating substantial differences between groups. The largest effects were observed for beliefs about children’s prior knowledge across domains (S1, d = −0.87) and their ability to understand scientific concepts (S34, d = −0.78).

In the “Learning Processes” category, significant differences were observed in items related to instructional approaches and learning processes. Participants with practicum experience showed higher scores in items reflecting more sophisticated conceptions of learning, such as information processing (S37) and the role of pre-existing skills (S27), although these effects were small to moderate (|d| ≈ 0.26–0.31). Notably, both groups demonstrated relatively strong agreement with reversed items reflecting direct instruction (S2, S14, S33), with non-practicum participants scoring significantly lower, particularly for S33 (d = −0.79), indicating a large effect and suggesting endorsement of transmission-based views.

In the preceding table, of particular interest within the “Learning Processes” category are the statements that conceptualize teaching as the direct transmission of knowledge—S2 [Preschool-aged children learn best when the teacher provides them with knowledge and information], and S33 [Preschool-aged children learn best through direct instruction by the teacher]. Both items yielded relatively low mean scores, indicating agreement with more transmission-oriented views of teaching, in which the teacher is positioned primarily as a provider of knowledge. In contrast, both groups reported high mean scores for statement S7 [Preschool-aged children learn best by actively interacting with their environment] (M = 5.34, SD = 0.76 for non-practicum group and M = 5.23, SD = 0.94 for practicum group) which reflects a more active role for the child in the learning process.

To further investigate this discrepancy, we examined the possibility of coexisting transmission-oriented and constructivist-oriented beliefs across selected item pairs related to learning. For this analysis, Likert-scale responses with values from 4 to 6 were classified as agreement (or disagreement when items with reversed coding), whereas responses with values from 1 to 3 were classified as disagreement (or agreement when items with reversed coding). As shown in

Table 3. Co-existence of beliefs consistent and inconsistent with constructivist beliefs.

S2 [direct provision of knowledge] vs. S7 [student as an active agent]	All PTS (N = 241)	34.9% of total Disagree to S2 96.3% of total Agree to S7 62.2% of total Agree to both
	No-Practicum (N = 116)	29.3% of total Disagree to S2 97.3% of total Agree to S7 68.1% of total Agree to both	χ2(1) = 3.27, p = 0.071
	Practicum (N = 125)	40% of total Disagree to S2 95.2% of total Agree to S7 56.8% of total Agree to both
S33 [direct teaching] vs. S7 [student as an active agent]	All PTS (N = 241)	44.4% of total Disagree to S33 96.3% of total Agree to S7 53.9% of total Agree to both
	No-Practicum (N = 116)	26.6% of total Disagree to S33 97.3% of total Agree to S7 70.6% of total Agree to both	χ2(1) = 25.25, p < 0.001
	Practicum (N = 125)	60.8% of total Disagree to S33 95.2% of total Agree to S7 38% of total Agree to both
S14 [direct provision of knowledge] vs. S7 [student as an active agent]	All PTS (N = 241)	79.7% of total Disagree to S14 96.3% of total Agree to S7 20.3% of total Agree to both
	No-Practicum (N = 116)	71.6% of total Disagree to S14 97.3% of total Agree to S7 28.4% of total Agree to both	χ2(1) = 9.10, p = 0.003
	Practicum (N = 125)	87.2% of total Disagree to S14 95.2% of total Agree to S7 12.8% of total Agree to both

, a substantial proportion of participants simultaneously endorsed both direct teaching and active knowledge construction. Across the total sample, 62.2% of participants agreed with both S2 and S7, while 53.9% agreed with both S33 and S7. However, the proportion of participants simultaneously endorsing both positions was considerably lower for the S14/S7 pair (20.3%), particularly within the practicum group.

To further examine whether coexistence patterns differed between groups, chi-square analyses were conducted for these selected item pairs. Significant associations were observed for both the S33/S7 pair (χ²(1) = 25.25, p < 0.001), and the S14 [It is important for adults to provide preschool-aged children with all the necessary knowledge, as they cannot learn on their own]/S7 pair (χ²(1) = 9.10, p = 0.003). Participants without practicum experience showed substantially higher rates of simultaneous endorsement of contradictory beliefs than participants with practicum experience for both S33/S7 (70.6% and 38%, respectively) and S14/S7 (28.4% and 12.8%, respectively). Although contradictory beliefs remained evident in both groups, their coexistence appeared significantly more frequent among participants without practicum-related experiences.

3.3. Correlations

Table 4. Pearson r analyses of correlation between the general categories of statements of CCogAb Questionnaire for all PTs and per practicum group.

	Category of Statements	Cognitive Operations	Conceptual Change & Prior Knowledge	Learning Processes
All PTs	Cognitive Operations	-
	Conceptual Change & Prior Knowledge	0.630 **	-
	Learning Processes	0.556 **	0.630 **	-
No-Practicum	Cognitive Operations	-
	Conceptual Change & Prior Knowledge	0.526 **	-
	Learning Processes	0.307 **	0.537 **	-
Practicum	Cognitive Operations	-
	Conceptual Change & Prior Knowledge	0.671 **	-
	Learning Processes	0.691 **	0.670 **	-

** Correlation is significant at the 0.01 level (2-tailed).

presents the Pearson correlation coefficients between the three general categories of the CCogAb questionnaire for the total sample and separately for the practicum and no-practicum groups. Statistically significant positive correlations were observed between all three categories across the total sample. The strongest correlations in the total sample were observed between “Cognitive Operations” and “Conceptual Change & Prior Knowledge” (r = 0.630, p < 0.01), and between “Conceptual Change & Prior Knowledge” and “Learning Processes” (r = 0.630, p < 0.01).

In the no-practicum group, correlations ranged from r = 0.307 to r = 0.537, whereas in the practicum group correlations ranged from r = 0.670 to r = 0.691. Overall, correlation coefficients were higher in the practicum group across all category pairings.

3.4. Cluster Analysis

Separate cluster analyses were conducted for the practicum and no-practicum groups using participants’ mean scores across the three conceptual categories of the CCogAb questionnaire. Initially, hierarchical cluster analysis was performed using Ward’s method and squared Euclidean distance in order to examine the grouping structure of the data. Inspection of the dendrograms and agglomeration coefficients suggested that a three-cluster solution provided the clearest grouping structure in both groups, prior to larger increases in the agglomeration coefficients.

Subsequently, k-means cluster analyses specifying three clusters were conducted separately for each group. The resulting clusters were interpreted as representing relatively lower (“Initial”), intermediate (“Intermediate”), and more advanced (“Advanced”) belief profiles. Across both groups, mean scores increased progressively across the three conceptual categories from the Initial to the Advanced clusters. As shown in

Table 5. Cluster analysis of mean scores for PTs in the Practicum and No-Practicum groups across the general categories of the CCogAb Questionnaire.

No-Practicum (N = 116)	Initial (N = 24)	Intermediate (N = 56)	Advanced (N = 36)
Cognitive Operations	3.69	4.44	4.84
Conceptual Change & Prior Knowledge	3.34	3.97	4.72
Learning Processes	4.27	4.42	4.92
Practicum (N = 125)	Initial (N = 27)	Intermediate (N = 57)	Advanced (N = 41)
Cognitive Operations	4.04	4.60	5.22
Conceptual Change & Prior Knowledge	3.82	4.71	5.17
Learning Processes	4.08	4.72	5.30

, participants in the practicum group demonstrated overall higher scores across all conceptual categories compared to participants in the no-practicum group. In both groups, differences between clusters were particularly evident in the “Conceptual Change and Prior Knowledge” category, suggesting that beliefs related to conceptual change and the role of prior knowledge differentiated the profiles more clearly than the other conceptual categories.

To further examine the internal differentiation of the cluster solutions, one-way analyses of variance were conducted separately within each group. Significant differences were found across clusters for all three conceptual categories in both the no-practicum group [Cognitive Operations: F(2,113) = 77.75, p < 0.001; Conceptual Change & Prior Knowledge: F(2,113) = 103.41, p < 0.001; Learning Processes: F(2,113) = 36.89, p < 0.001] and the practicum group [Cognitive Operations: F(2,122) = 100.53, p < 0.001; Conceptual Change & Prior Knowledge: F(2,122) = 136.07, p < 0.001; Learning Processes: F(2,122) = 135.63, p < 0.001]. These findings suggest that the identified cluster solutions captured meaningful variation in participants’ belief profiles rather than reflecting arbitrary classifications. In addition, the relatively balanced distribution of participants across clusters (24/56/36 and 27/57/41) further supports the interpretation that the identified profiles reflected meaningful variability in participants’ beliefs.

4. Discussion

The present study investigated the beliefs of pre-service preschool teachers regarding the cognitive abilities of preschool-aged children, focusing on three categories: a) cognitive functions, b) conceptual change and prior knowledge, and c) learning processes. The results revealed statistically significant differences between PTs who had completed three practicum courses and those in the first years of their studies with no practicum-related experiences. An important finding of the study is the coexistence of contradictory beliefs within PTs’ responses, reflecting partial coherence in their understanding of how young children think and learn and pointing to an ongoing gradual process of professional identity development.

A closer examination of the item-level results further supports this interpretation. Although participants with practicum experience generally demonstrated more advanced beliefs, the observed differences were mostly small to moderate in magnitude, indicating gradual rather than fundamental shifts in their conceptualizations of learning and development. Interestingly, variability within the practicum group was also somewhat higher in the “Learning Processes” category compared to the no-practicum group. This greater heterogeneity may further suggest that practicum-related experiences do not influence all pre-service teachers in the same way or to the same degree. Rather than reflecting a uniform developmental trajectory, exposure to practicum may be associated with diverse and uneven processes of belief reorganization, consistent with the interpretation that conceptual and professional development are gradual and non-linear processes. At the same time, relatively strong endorsement of items reflecting direct instruction (e.g., S2, S33) was observed in both groups, with a particularly moderate-to-large effect for S33. This pattern suggests that differences associated with practicum experience do not necessarily reflect a simple replacement of transmission-based beliefs with constructivist ones. Instead, pre-service teachers appear to reorganize their existing belief systems, allowing elements of both perspectives to coexist. In this sense, the findings are consistent with the view that belief change in teacher education is not a process of substitution, but one of restructuring, in which prior and newly acquired conceptions remain active within a developing, yet not fully coherent, belief system. Indeed, Skopeliti et al. (2025) found that pre-service teachers’ beliefs can change through explicit conceptual change instructional approaches, although such changes appear to be gradual and partial rather than immediate.

PTs with practicum experience achieved higher mean scores on the overall questionnaire as well as across the three categories. They appear to hold beliefs more closely aligned with constructivist and developmentally appropriate approaches to learning, although various misconceptions remain across the different categories of statements. This pattern is also consistent with evidence comparing Greek and Cypriot pre-service preschool teachers, two contexts that share important similarities in educational traditions and teacher education practices, where stronger alignment with developmentally appropriate practice (DAP) coexists with residual misconceptions—similarly to the coexistence patterns observed in the present study—and where beliefs significantly predict self-reported practices (Sakellariou & Rentzou, 2012).

In the “Cognitive functions” category, PTs without practicum experience tend to underestimate children’s ability for sustained attention and the use of mnemonic strategies. They also believed that preschool children lacked the ability to reflect on their own learning and were less likely to support teaching learning strategies to children of this age. Apart from the persistent belief about children’s inability to remain focused on an activity, PTs who had completed their practicum demonstrated higher scores in the remaining items of this category, highlighting the belief that these children can learn how to learn and benefit from instruction in learning strategies. However, two items in this category showed the opposite pattern, with participants without practicum experience reporting slightly higher scores regarding the role of teaching strategies (S4) and adult contribution (S5). Although these differences were small, they may suggest that movement toward more child-centered and constructivist-oriented perspectives does not necessarily occur in a linear or fully balanced manner. In some cases, greater emphasis on children’s active role in learning may coexist with a relative underestimation of the adult’s scaffolding role in supporting cognitive development. This pattern further supports the interpretation that pre-service teachers’ belief systems are undergoing gradual and uneven reorganization.

These differences are consistent with findings showing that DAP beliefs and DAPs are positively correlated, whereas DAPs and developmentally inappropriate practices are negatively related, supporting the association between developmentally appropriate beliefs and instructional decisions more consistent with constructivist and child-centered approaches (Sakellariou & Rentzou, 2011, 2012; Wilcox-Herzog et al., 2014).

In the “Conceptual change and prior knowledge” category, PTs without practicum experience often underestimated children’s cognitive readiness and considered age as the main criterion for understanding. Specifically, they tended to believe that young children could only learn superficially and lack prior knowledge in domains such as mathematics and biology. They also failed to recognize that children’s pre-existing knowledge can influence their learning, as well as that its absence can cause difficulties in understanding new information. PTs with practicum experience recognized that children of this age possess sufficient prior knowledge in various domains, yet they too struggle to understand the role of prior knowledge in learning.

These findings confirm a broader tendency among PTs to regard biological age as a primary indicator of cognitive ability and to underestimate preschool children’s abilities considering preschool children to be cognitively deficient (Avgitidou et al., 2013). It may also explain preschool teachers’ tendency to underestimate their students’ mathematical skills (Ban et al., 2024; Kilday et al., 2012).

The findings suggest that participants in the no-practicum group were less familiar with basic assumptions associated with contemporary theories of cognitive development. More specifically, participants appeared to recognize assumptions commonly associated with constructivist and sociocultural approaches, such as the role of the social environment and the importance of collaboration and guidance from an experienced adult or more knowledgeable peers, as well as the active role individuals have on their environment. However, they seemed less familiar with more recent approaches arguing that humans come into the world endowed with some core knowledge which enables them to deal with important aspects of their physical and social environment (Gelman, 2015; Spelke, 2000).

In the “Learning Processes” category, it is interesting that PTs in both groups expressed the belief that preschoolers learn best through direct instruction by the kindergarten teacher. This is a finding also evident among PTs who had completed practicum courses and studied current constructivist theories. Indeed, we observed a coexistence of initial and more advanced beliefs, with a significant proportion of PTs simultaneously endorsing statements recognizing the child as an active co-creator of knowledge and statements supporting the transmission of knowledge by the educator to the child.

These results align with our earlier studies showing that PTs’ beliefs about learning and teaching form a complex system in which conflicting beliefs often coexist and shape instructional practices. Previous research has shown that these beliefs lack internal coherence. Beliefs emerged that are simultaneously consistent with a constructivist epistemology and a student-centered approach to learning and teaching (Kyriakopoulou & Skopeliti, 2024, 2025). Although the present study was not specifically designed to capture the coexistence of conflicting beliefs, it is interesting that this finding nevertheless emerged highlighting the resilience of the knowledge transmission model, according to which learning is primarily understood as the direct transfer of knowledge from teacher to child, even after systematic exposure to constructivist approaches in practicum courses. Similar tensions were also identified by Pantazidis (2026) in a qualitative study of Greek pre-service early childhood teachers prior to practicum. The study showed that participants frequently expressed child-centered, caring, and democratic pedagogical ideals, while at the same time reproducing assumptions associated with control, transmission, teacher authority, and the successful delivery of planned activities. Pantazidis argues that democratic and relational aspirations often coexist with assumptions tied to authority, transmission, control, and performance, rather than fully replacing them. This interpretation is consistent with the present findings, where constructivist and transmission-oriented beliefs also appeared to coexist within PTs’ responses. The transition from a teacher-centered to a child-centered, constructivist model of learning therefore appears to be a challenging and gradual process. This coexistence of constructivist and transmission-based beliefs reflects partial coherence in pre-service teachers’ belief systems potentially associated with an ongoing process of professional identity development. At the same time, the stronger correlations observed between the three conceptual categories in the practicum group may indicate not only differences in individual beliefs, but also a greater degree of coherence across broader belief systems. Although contradictory beliefs were still present, participants with practicum experience appeared more likely to organize their beliefs in more internally connected ways. The cluster analysis further supports the view that pre-service teachers’ beliefs do not form a fully uniform or internally coherent pattern. Instead, distinct belief profiles emerged in both groups, ranging from relatively initial to more constructivist-oriented profiles across the three conceptual categories. Importantly, even within the practicum group, not all participants were classified within the more advanced profile. This pattern is consistent with the possibility that professional learning experiences and practicum-related exposure may coincide with gradual processes of reorganization and partial integration within pre-service teachers’ belief systems.

These findings can be viewed from the perspective of emerging professional identity. Although professional identity was not directly assessed in the present study, the observed coexistence of contradictory beliefs can be interpreted through theoretical perspectives that conceptualize beliefs as one component of emerging professional identity. More specifically, the coexistence of contradictory beliefs may reflect underlying tensions in the process of professional formation. The presence of both constructivist and transmission-based beliefs may indicate not only a lack of conceptual coherence, but also a transitional stage in which PTs are in the process of constructing their professional identity (Beijaard et al., 2004; Pillen et al., 2013). During this phase, individuals draw simultaneously on prior school-based experiences, intuitive beliefs about teaching and learning, and newly acquired pedagogical knowledge (Flores & Day, 2006). Within this process, practicum may function not simply as a context for applying knowledge, but as a learning environment in which these beliefs are tested and partially restructured through engagement with classroom practice (Izadinia, 2013). The persistence of such contradictory beliefs, even among PTs with practicum experience, suggests that professional identity development is gradual and non-linear, requiring sustained opportunities for reflection and support in connecting theory with practice.

Additionally, the persistence of contradictory beliefs even among PTs with practicum experience, can also be interpreted through the lens of conceptual change theory, which provide a theoretically cognitive explanation for how individuals revise, reorganize, or resist changing their pre-existing beliefs and intuitive understandings when confronted with new information or scientific knowledge. According to the Framework Theory approach of conceptual change, individuals form initial framework theories or belief systems early in life based on their intuitive experiences and the information they receive from the sociocultural environment, which often differ significantly from scientifically accepted theories. These initial framework theories are resistant to change, since information consistent with them is more easily assimilated than information that is inconsistent which requires conceptual restructuring (Vosniadou, 2013, 2019; Vosniadou & Skopeliti, 2014). Avgitidou et al. (2013) also argue that PTs enter university with a “framework theory” grounded in specific ontological and epistemological assumptions that shape their reasoning about childhood. Conceptual change intervention methods, such as the use of refutational texts and instructional analogies, may support PTs in recognizing contradictions within their existing beliefs system and promote reflective restructuring of their beliefs (Skopeliti et al., 2025).

Although participants with practicum experience appeared more likely to endorse constructivist-oriented beliefs, practicum experience alone did not seem sufficient to resolve these inconsistencies, further suggesting that both conceptual change and professional identity development are gradual and non-linear processes. Similar findings were reported by Avgitidou et al. (2024), who found that pre-service teachers’ views changed during a systematic intervention in initial teacher education. The changes involved the ways participants understood children’s participation and abilities, as well as the role of the teacher and classroom decision-making. The authors also highlighted the role of reflection, discussion of prior beliefs, and engagement with examples from educational practice during the intervention. Any potential contribution of practicum-related experiences may depend on the extent to which it is accompanied by chances for systematic reflection and explicit connection between theory and practice. Thus, more explicit and targeted support may be required to help PTs recognize and critically examine their own assumptions in light of scientific knowledge and research findings.

These findings also highlight the importance of how teacher education programs prepare future educators at a broader level. If PTs enter the profession with partially coherent belief systems, this may influence not only individual instructional practices but also the overall quality of early childhood education. Supporting the development of more coherent, research-informed belief systems during teacher education may therefore be an important component of strengthening the professional foundations of the early childhood education workforce.

University programs should incorporate targeted interventions that help PTs become aware of their initial beliefs, critically examine them, recognize how these beliefs shape their instructional decisions, and actively work toward their revision. Conceptual change interventions suggest that pre-service teachers are more likely to reconsider their beliefs when they are encouraged to reflect on contradictions between their existing assumptions and contemporary scientific or pedagogical knowledge. Activities such as the use of refutational texts, instructional analogies, guided reflection, and structured discussions may help PTs recognize inconsistencies in their thinking and gradually develop more coherent and research-informed understandings (Skopeliti et al., 2025). Equally important is sustained and in-depth engagement with contemporary theories of cognitive development and recent findings from neuroscience, combined with extended practicum experiences in order for PTs to develop research informed, constructivist understandings of how young children think and learn. In parallel, PTs should be encouraged to broaden their knowledge of child development both in general and at the level of individual trajectories, as well as of the cultural and social contexts within which children live. Such an approach can support the development of coherent, research-informed, and reflective professional identities, while enabling pre-service teachers to construct more consistent understandings of how young children think and learn.

5. Limitations

The findings of the present study should be interpreted in light of several limitations. First, this study employed a cross-sectional comparative design, which does not allow for causal conclusions regarding the role of practicum experience in shaping pre-service teachers’ beliefs. Differences between the practicum and no-practicum groups may also be associated with differences related to year of study, age, exposure to pedagogical courses, and other educational experiences, making it difficult to isolate the specific contribution of practicum experience itself. In addition, the sample was drawn from two Greek universities and consisted predominantly of female participants (98.3%), which may limit the generalizability of the findings to other educational and cultural contexts or to more gender-balanced populations. Moreover, potential institutional differences between the two participating universities were not examined separately in the present study.

Regarding the CCogAb questionnaire, the relatively low proportion of variance explained by the exploratory factor analysis further suggests that the three conceptual categories should be interpreted primarily as theoretically informed groupings rather than as strongly distinct empirical dimensions. This consideration also applies to the cluster analysis, as the identified cluster solution should be interpreted cautiously and viewed primarily as exploratory in nature. Further research is therefore needed to examine its psychometric properties of the CCogAb questionnaire across different samples and educational settings.

6. Conclusions

In conclusion, this study highlights that pre-service teachers may hold incoherent belief systems regarding children’s cognitive abilities that are shaped by prior theoretical frameworks. These belief systems may influence their educational practices and can be linked to broader processes of emerging professional identity formation. Practicum-related experiences, together with broader teacher education experiences, may be associated with processes of belief revision and reorganization, as long as they are supported by activities that promote reflection and critical examination of their initial assumptions. Finally, these findings underscore the importance of designing teacher education programs that explicitly address pre-service teachers’ initial beliefs and provide structured practicum experiences that may support the development of coherent and research-informed professional identities. Future research could further explore how different practicum designs may influence both belief systems and identity formation, providing additional insights for the preparation of early childhood educators.