top of page
Asian Institute of Research, Journal Publication, Journal Academics, Education Journal, Asian Institute
Asian Institute of Research, Journal Publication, Journal Academics, Education Journal, Asian Institute

Education Quarterly Reviews

ISSN 2621-5799

asia institute of research, journal of education, education journal, education quarterly reviews, education publication, education call for papers
asia institute of research, journal of education, education journal, education quarterly reviews, education publication, education call for papers
asia institute of research, journal of education, education journal, education quarterly reviews, education publication, education call for papers
asia institute of research, journal of education, education journal, education quarterly reviews, education publication, education call for papers
crossref
doi
open access

Published: 18 November 2026

Developing and Validating a TPACK-based Knowledge Model of Pre-service Teachers for STEM Instruction

Hong Xuchun, Li Hongzhe, Zhang Yuting

Zhejiang Normal University

asia institute of research, journal of education, education journal, education quarterly reviews, education publication, education call for papers
pdf download

Download Full-Text Pdf

doi

10.31014/aior.1993.09.04.735

Pages: 54-66

Keywords: TPACK, STEM Instruction, Pre-Service Teachers, Science Teachers, Measurement Development

Abstract

STEM instruction in the digital era places complex knowledge demands on science teachers, yet few theoretically grounded STEM-TPACK models have been developed to support pre-service teacher preparation. To address the new requirements of technology-enhanced teaching, this study aimed to develop and validate an instrument assessing pre-service science teachers’ knowledge for STEM teaching based on the TPACK framework. Drawing on the original theoretical framework of TPACK, a knowledge model for STEM instruction with eight dimensions was constructed. Survey data from 195 Chinese pre-service science teachers were analyzed using exploratory factor analysis (EFA) to identify the latent factor structure. A separate sample of 208 pre-service science teachers was used for confirmatory factor analysis (CFA) to examine scale reliability and validity. The results supported an eight-factor, a 28-items model with acceptable internal consistency, convergent validity and model fit. This study provides a model for understanding and measuring pre-service teachers’ multidimensional knowledge for STEM teaching and identifying their preparation needs, thereby informing targeted teacher education activities.

 

1. Introduction

 

To effectively organize STEM instruction, teachers should connect ideas across disciplines and design activities in which scientific explanations, mathematical reasoning, technology, and engineering design contribute to solving a problem (Nipyrakis et al., 2025), which requires attention to disciplinary connections and to how those connections become teachable activities (Stevenson et al., 2026). Examining the structure of teachers’ knowledge for STEM instruction and assessing that knowledge are therefore essential for identifying gaps in teacher preparation, guiding targeted professional learning, and supporting sustained improvements in the quality of STEM education. In China, teachers implementing STEM instruction may face challenges arising from a limited understanding of interdisciplinary STEM concepts and connections, making it difficult to move beyond subject-specific knowledge and design integrated learning activities (Dong et al., 2020).

Technology can address these challenges by making connections across STEM disciplines more visible and supporting learning activities that bring together otherwise isolated subject knowledge. Before digital technologies became widespread in schools, STEM education already treated technological literacy as part of scientific literacy. Today, digital tools offer new ways to visualize scientific phenomena, collect data, and test solutions in STEM instruction (Kruse et al., 2025). Thus, teachers must decide which tool helps students understand a concept and how to use it within a learning activity (Aumann et al., 2023). The technological pedagogical content knowledge (TPACK) framework provides a basis for examining these decisions. It describes how teachers combine knowledge of technology, pedagogy, and subject matter (Mishra & Koehler, 2006).

 

Although recent studies have examined TPACK-based interventions and technology integration in STEM teacher education (Chai et al., 2020; Mansour et al., 2024), limited attention has been paid to developing and empirically validating a knowledge model that explicitly incorporates interdisciplinary STEM content knowledge and the preparation needs of pre-service science teachers. The pre-service stage is particularly important because the knowledge developed during teacher preparation provides a foundation for teachers’ subsequent instructional decisions and their capacity to design and implement integrated STEM learning activities (Dökme & Ünlü, 2024). Therefore, to clarify the knowledge required for integrated STEM instruction, this study aims to develop and validate a TPACK-based knowledge model that explicitly incorporates interdisciplinary STEM content knowledge and pre-service teacher training knowledge. We propose two research questions: (1) What knowledge dimensions constitute a TPACK-based model of pre-service teachers’ knowledge for integrated STEM teaching? (2) What are the relationships among the knowledge dimensions in the proposed model?

 

2. Theoretical framework of this study

 

This study draws on two complementary theoretical perspectives: TPACK and a situated perspective on teacher learning. TPACK identifies the technological, pedagogical and content knowledge involved in teaching with technology and provides the structure of the proposed knowledge model (Mishra & Koehler, 2006). Korthagen’s (2010) perspective emphasizes the relationship between theoretical knowledge and practical teaching experiences in teacher preparation. Drawing on this perspective, the present study extends TPACK to include preparation-related knowledge, focusing on how pre-service teachers connect coursework with lesson planning, learn from classroom observations, and use feedback to improve their teaching designs.

 

Grounded in the theoretical framework, this study initially formulated eight knowledge categories as first-level dimensions and preliminarily identified 21 knowledge elements. STEM-CK includes both disciplinary concepts and the connections needed to address an interdisciplinary problem. STEM-PK concerns the organization of inquiry, collaboration, problem solving and iterative design, whereas STEM-TK refers to the functions and limitations of digital tools used in STEM activities. The three paired knowledge domains combine content and pedagogy to teach STEM concepts and interdisciplinary connections (STEM-PCK), content and technology to represent and explore these concepts and connections (STEM-TCK), and pedagogy and technology to support inquiry, collaboration, and engineering design (STEM-TPK). STEM-TPACK integrates all three components to design learning activities that connect STEM disciplines in addressing authentic problems. In addition, the pre-service teacher knowledge domain addresses the transitional and contextual demands specific to initial teacher preparation. It comprises knowledge of translating coursework into practice, bridging theory and authentic classroom application; knowledge of teaching and research, emphasizing evidence-based inquiry and sustained professional development; and knowledge of reflective revision, highlighting the continuous refinement of instructional design and implementation through self-reflection. Figure 1 provides operational definitions to guide item development.


Figure 1: Theoretical framework of STEM-TPACK

 

3. Research Method

 

3.1. Instrument Development

 

This study developed a TPACK-based questionnaire to assess pre-service science teachers’ perceived knowledge for integrated STEM teaching. Initial items were constructed according to the proposed knowledge dimensions. Four experts in science education and STEM instruction evaluated the relevance, clarity, and dimensional alignment of the items (Table 1). The scale-level content validity index (S-CVI/Ave) was 0.93, indicating satisfactory content validity (Polit & Beck, 2006). The questionnaire was revised based on their feedback, resulting in 28 items rated on a five-point Likert scale from 1 (strongly disagree) to 5 (strongly agree). Each knowledge element identified in the theoretical framework was operationalized using one or two questionnaire items.

 

Table 1: Expert characteristics and content validity assessment

Expert


Professional title

Research expertise

S-CVI/Ave

E1

Professor

Science education

pre-service teacher education

0.93

E2

Associate professor

STEM education

Interdisciplinary curriculum design

0.89

E3

Associate professor

Educational technology

0.96

E4

Senior science teacher

Science instruction

School-based STEM practices

0.93

 

3.2 Population and sampling

 

Participants were pre-service science teachers from three universities in Zhejiang Province, China. Science teachers play an important role in implementing integrated STEM instruction in Chinese schools. The participants were recruited through random sampling. Participation was voluntary, and informed consent was obtained before data collection. Following data screening, 403 valid responses were randomly divided into an EFA sample of 195 and a CFA sample of 208.

 

Table 2 presents the demographic characteristics of the EFA and CFA samples, comprising 195 and 208 pre-service science teachers, respectively. Female participants accounted for 51.79% of the EFA sample and 55.77% of the CFA sample. In terms of education level, the EFA sample included 94 undergraduates (48.21%) and 101 postgraduates (51.79%), whereas the CFA sample included 105 undergraduates (50.48%) and 103 postgraduates (49.52%). Both samples showed relatively balanced distributions across gender and education level.

 

Table 2: Demographic characteristics.

Parameter

Type

EFA (n = 195) Frequency

Percentage (%)

CFA (n = 208) Frequency

Percentage (%)

Gender

Female

101

51.79

116

55.77

 

Male

94

48.21

92

44.23

Education

Undergraduates

94

48.21

105

50.48

 

Postgraduates

101

51.79

103

49.52

Teaching practicum experience

Yes

155

79.49

137

65.87

No

40

20.51

71

    34.13


 

3.3 Data Analysis

 

Data were analyzed using SPSS. 27 and AMOS. 29. EFA was conducted with 195 responses to investigate the questionnaire’s underlying structure. Sampling adequacy was examined using the Kaiser–Meyer–Olkin measure and Bartlett’s test of sphericity. Principal axis factoring with Promax rotation was used to allow correlations among the knowledge dimensions. Factor retention was informed by parallel analysis, the scree plot, and theoretical interpretability, while item selection considered factor loadings and cross-loadings.

 

CFA was subsequently conducted with 208 responses to evaluate the resulting measurement model. Model fit was assessed using CFI, TLI and RMSEA, interpreted collectively rather than through a single index (Kline, 2023). Convergent validity was evaluated using standardized factor loadings and AVE, while reliability was assessed using Cronbach’s α and composite reliability. Discriminant validity was examined using the Fornell–Larcker criterion and HTMT. CFA latent factor correlations were used to address the relationships among the knowledge dimensions. Internal consistency was assessed using Cronbach’s α and composite reliability. Correlations among the latent factors were examined to address the relationships between knowledge dimensions.

 

4. Results

 

4.1 Exploratory factor analysis (EFA)

 

EFA was conducted with 195 pre-service science teachers to examine the underlying structure of the 28-item questionnaire. The KMO measure and Bartlett’s test of sphericity were used to assess whether the correlation matrix was suitable for factor analysis. The KMO value was 0.956, and Bartlett’s test was statistically significant, χ²(378) = 6384.55, p < .001, indicating that the data were suitable for EFA. Principal axis factoring was selected to examine the common variance underlying the items. Oblimin rotation was applied because the proposed knowledge dimensions were expected to be related. An eight-factor solution was examined in accordance with the theoretical framework. The scree plot was inspected to examine the distribution of the initial eigenvalues (Figure 2). The plot showed a pronounced decline after the first factor, followed by a more gradual decrease. It therefore did not provide an unambiguous basis for retaining eight factors. The eight-factor solution was consequently evaluated through its interpretability and correspondence with the proposed knowledge dimensions. The extracted factors jointly accounted for 61.11% of the total variance.

 

Figure 2: Scree plot of the initial eigenvalues
Figure 2: Scree plot of the initial eigenvalues

The pattern matrix presented in Table 3 showed that the items clustered according to the eight proposed dimensions. The absolute primary factor loadings ranged from .498 to .879, while all secondary loadings had absolute values below .30. Specifically, the primary loadings ranged from .821 to .879 for CK, .664 to .722 for PK, .738 to .769 for TK, .498 to .576 for PCK, .522 to .647 for TCK, .547 to .829 for TPK, .502 to .823 for TPACK, and .560 to .843 for PST. Rotation converged in 11 iterations. Each factor contained three or four items, and all 28 items were retained for subsequent CFA. These findings provided preliminary support for the proposed dimensional structure, although confirmation using the independent CFA sample was still required. In response to the first research question, the theoretically specified eight-factor EFA solution comprised CK, TK, PK, PCK, TCK, TPK, TPACK, and PST, represented by 28 retained items. The first seven dimensions contextualized TPACK within integrated STEM teaching, while PST addressed knowledge associated with pre-service teacher preparation.

Table 3: Factor loadings from the pattern matrix

Item

1

2

3

4

Item

5

6

7

8

CK1

.832

 

 

 

TPK1

.829

 

 

 

CK2

.852

 

 

 

TPK2

.547

 

 

 

CK3

.879

 

 

 

TPK3

.821

 

 

 

CK4

.821

 

 

 

TK1

 

.738

 

 

PST1

 

.560

 

 

TK2

 

.762

 

 

PST2

 

.680

 

 

TK3

 

.769

 

 

PST3

 

.814

 

 

TK4

 

.761

 

 

PST4

 

.843

 

 

PK1

 

 

−.664

 

TPACK1

 

 

−.636

 

PK2

 

 

−.719

 

TPACK2

 

 

−.694

 

PK3

 

 

−.722

 

TPACK3

 

 

−.502

 

PCK1

 

 

 

PCK1

TPACK4

 

 

−.823

 

PCK2

 

 

 

PCK2

TCK1

 

 

 

.647

PCK3

 

 

 

PCK3

TCK2

 

 

 

.522

 

 

 

 

 

TCK3

 

 

 

.606

 

 

 

 

 

Note. Extraction method: principal axis factoring. Rotation method: Oblimin with Kaiser normalization. Coefficients with absolute values below .30 are omitted. Factor signs are retained. Factors: 1 = CK; 2 = PST; 3 = TPACK; 4 = TCK; 5 = TPK; 6 = TK; 7 = PK; 8 = PCK.

 

4.2 Confirmatory factor analysis (CFA)

 

Confirmatory factor analysis (CFA) was conducted with 208 valid responses to evaluate the proposed eight-factor knowledge model for pre-service science teachers’ integrated STEM teaching. The correlated first-order model demonstrated an acceptable overall fit: χ²(322) = 660.816, p < .001, χ²/df = 2.052, CFI = .951, TLI = .942, IFI = .951, NFI = .909, and RMSEA = .071, with a 90% confidence interval of [.064, .079]. Although GFI (.813) and AGFI (.764) were comparatively low, the principal incremental and approximate fit indices provided support for the proposed measurement structure. The model fit indices are reported in Table 4, and the standardized CFA model is presented in Figure 3.

 

Table 4: Model fit indices for the eight-factor CFA model.

Fit indices

Values

Criteria

Summary

CMIN/DF

2.052

< 3.00

Acceptable

RMSEA

.071

< .08

Acceptable

CFI

.951

≥ .90

Acceptable

TLI

.942

≥ .90

Acceptable

NFI

.909

≥ .90

Acceptable

IFI

.951

≥ .90

Acceptable

Note. χ²(322) = 660.816, p < .001. The 90% confidence interval for RMSEA was [.064, .079]. Criteria indicate conventional acceptable fit rather than stringent good fit.

 

Figure 3: CFA model of the knowledge model
Figure 3: CFA model of the knowledge model

Convergent validity was examined using standardized factor loadings, average variance extracted (AVE), and composite reliability (CR). As shown in Table 5, standardized loadings ranged from .509 to .954. TCK3 had the lowest loading (.509), whereas all remaining items had loadings above .80. AVE values ranged from .534 to .883, and CR values ranged from .767 to .964, exceeding the reference values of .50 and .70, respectively. These findings supported the model’s convergent validity and composite reliability.

Table 5: Standardized factor loadings, AVE and CR

Path

Loading

AVE

CR

Path

Loading

AVE

CR

CK1 ← STEM-CK

.932

.872

.964

TCK1 ← STEM-TCK

.824

.534

.767

CK2 ← STEM-CK

.946

 

 

TCK2 ← STEM-TCK

.814

 

 

CK3 ← STEM-CK

.920

 

 

TCK3 ← STEM-TCK

.509

 

 

CK4 ← STEM-CK

.936

 

 

TPK1 ← STEM-TPK

.879

.794

.920

TK1 ← STEM-TK

.876

.828

.951

TPK2 ← STEM-TPK

.887

 

 

TK2 ← STEM-TK

.928

 

 

TPK3 ← STEM-TPK

.907

 

 

TK3 ← STEM-TK

.911

 

 

TPACK1←STEM-TPACK

.932

.804

.942

TK4 ← STEM-TK

.924

 

 

TPACK2←STEM-TPACK

.921

 

 

PK1 ← STEM-PK

.933

.883

.958

TPACK3←STEM-TPACK

.809

 

 

PK2 ← STEM-PK

.954

 

 

TPACK4←STEM-TPACK

.918

 

 

PK3 ← STEM-PK

.932

 

 

PST1 ← STEM-PST

.853

.765

.929

PCK1←STEM-PCK

.914

.829

.936

PST2 ← STEM-PST

.899

 

 

PCK2←STEM-PCK

.939

 

 

PST3 ← STEM-PST

.860

 

 

PCK3←STEM-PCK

.877

 

 

PST4 ← STEM-PST

.886

 

 










 

Discriminant validity was first evaluated using the Fornell–Larcker criterion. As presented in Table 6, the square root of each construct’s AVE exceeded its correlations with other constructs, except for TCK in relation to PST. The TCK–PST correlation (.740) exceeded the square root of AVE for TCK (.730), indicating a small deviation from the Fornell–Larcker criterion that warrants further examination. Discriminant validity was additionally examined using the heterotrait–monotrait ratio of correlations (HTMT). As shown in Table 7, HTMT values ranged from .486 to .874. All values were below .90, meeting thresholds. Taken together, the two assessments provided qualified support for discriminant validity.

Table 6: Latent factor correlations and square roots of AVE

Dimension

CK

TK

PK

PCK

TCK

TPK

TPACK

PST

CK

—

 

 

 

 

 

 

 

TK

.802

—

 

 

 

 

 

 

PK

.789

.813

—

 

 

 

 

 

PCK

.849

.816

.852

—

 

 

 

 

TCK

.608

.531

.497

.611

—

 

 

 

TPK

.730

.669

.627

.708

.663

—

 

 

TPACK

.797

.783

.688

.831

.570

.693

—

 

PST

.753

.651

.623

.704

.740

.766

.720

—

Square root

of AVE

.934

.910

.940

.910

.730

.891

.896

.875

Table 7: Heterotrait–monotrait ratios of correlations (HTMT)

Dimension

CK

TK

PK

PCK

TCK

TPK

TPACK

PST

CK

—

 

 

 

 

 

 

 

TK

.801

—

 

 

 

 

 

 

PK

.799

.820

—

 

 

 

 

 

PCK

.839

.817

.874

—

 

 

 

 

TCK

.567

.512

.486

.580

—

 

 

 

TPK

.728

.679

.633

.697

.606

—

 

 

TPACK

.808

.805

.722

.856

.557

.693

—

 

PST

.749

.646

.633

.696

.700

.775

.723

—

 

Internal consistency was assessed using Cronbach’s α. As reported in Table 8, coefficients ranged from .754 to .964, exceeding the reference threshold of .70 for all eight dimensions. TCK had the lowest coefficient, consistent with its comparatively weaker inter-item associations. These results supported the internal consistency of the subscales. However, the high coefficients for some dimensions warrant further examination of potential item redundancy.

 

Table 8: Internal consistency reliability of the eight dimensions

Dimension

No. of items

Minimum inter-item

correlation

Maximum inter-item correlation

Cronbach’s α

STEM-CK

4

.856

.890

.964

STEM-TK

4

.801

.854

.950

STEM-PK

3

.867

.892

.957

STEM-PCK

3

.799

.851

.935

STEM-TCK

3

.398

.653

.754

STEM-TPK

3

.761

.842

.920

STEM-TPACK

4

.719

.860

.941

PST

4

.715

.811

.928

 

To address the second research question, correlations among the latent knowledge dimensions were examined using the CFA model. As shown in Table 6, all correlations were positive, ranging from .497 to .852. The strongest association was between PK and PCK (.852), followed by CK and PCK (.849) and PCK and TPACK (.831). The weakest association was between PK and TCK (.497). These findings indicated that the knowledge dimensions were interconnected, but did not establish causal or reciprocal influences.

 

At the item level, within-dimension correlations ranged from .398 to .892, whereas between-dimension correlations ranged from .138 to .776 (Table 9). TCK showed comparatively weaker within-dimension correlations (.398–.653) than the other dimensions. These descriptive associations supplemented the reliability and validity assessments but were not treated as independent evidence of discriminant validity.

 

Table 9: Ranges of within- and between-dimension inter-item correlations.

Dimension

CK

TK

PK

PCK

TCK

TPK

TPACK

PST

CK

.856–.890

 

 

 

 

 

 

 

TK

.618–.721

.801–.854

 

 

 

 

 

 

PK

.611–.728

.611–.743

.867–.892

 

 

 

 

 

PCK

.625–.772

.609–.731

.638–.770

.799–.851

 

 

 

 

TCK

.204–.524

.183–.421

.172–.437

.217–.497

.398–.653

 

 

 

TPK

.520–.702

.418–.637

.493–.600

.445–.673

.138–.598

.761–.842

 

 

TPACK

.602–.719

.591–.683

.559–.690

.615–.776

.199–.480

.468–.683

.719–.860

 

PST

.555–.654

.445–.565

.454–.572

.479–.634

.200–.588

.506–.695

.457–.618

.715–.811

 

5. Discussion

 

This study developed and examined a TPACK-based knowledge model for pre-service science teachers’ integrated STEM teaching. The findings provided preliminary support for eight related dimensions: CK, PK, TK, PCK, TCK, TPK, TPACK, and PST. The CFA model demonstrated acceptable overall fit, and all dimensions met the reference criteria for AVE and composite reliability. These results extend the application of TPACK to a specific disciplinary and developmental context. Whereas Mishra and Koehler (2006) emphasized the interaction of content, pedagogy, and technology, the present model contextualizes these relationships within integrated STEM teaching and pre-service teacher preparation. Previous research has also used TPACK to examine teachers’ perceived knowledge for integrative STEM education (Chai et al., 2019). 

 

Regarding the first research question, the proposed model suggests that pre-service STEM teaching knowledge involves both knowledge for instructional design and knowledge for learning to teach. Contextualizing the TPACK domains within interdisciplinary STEM tasks clarifies the former, whereas PST makes the connection between coursework, observation, feedback, and teaching preparation explicit. This distinction is consistent with Korthagen’s (2010) perspective on connecting theoretical knowledge with practical teaching experiences. However, PST remains a proposed extension whose conceptual boundaries and empirical distinctiveness require continued examination. The present study contributes a complementary approach by retaining the seven TPACK dimensions and incorporating preparation-related knowledge.

 

The inclusion of interdisciplinary content knowledge addresses a central requirement of integrated STEM teaching: understanding both disciplinary concepts and meaningful connections across disciplines. Within this model, CK encompasses scientific understanding and connections with other STEM disciplines; integration is therefore not treated solely as a teaching strategy. This conceptualization is consistent with Kelley and Knowles (2016), who situated STEM integration within authentic contexts connecting scientific inquiry, engineering design, technological literacy, and mathematical thinking. More recently, Portillo-Blanco et al. (2024) documented both shared principles and continuing diversity in definitions of integrated STEM education. These perspectives suggest that interdisciplinary knowledge should be explicitly defined when adapting teacher knowledge frameworks to STEM. Nevertheless, the present CK scale should not be interpreted as a comprehensive assessment of knowledge across all four STEM disciplines. Its coverage requires further examination through expert review and performance-based tasks.

 

To address the second research question, the positive correlations indicate that respondents perceived the knowledge dimensions as interconnected aspects of STEM teaching preparation. The close relationships of PCK with CK, PK, and TPACK are consistent with its conceptual role in connecting disciplinary understanding with instructional decisions. PST’s associations with the TPACK-related dimensions further suggest that learning from coursework, observation, and feedback is connected with perceived readiness for STEM instructional design. Nevertheless, these relationships neither demonstrate mutual influence nor establish that growth in one dimension produces growth in another. The strong associations involving PCK also require attention to whether the corresponding dimensions are sufficiently differentiated. For teacher education, the findings suggest that subject courses and teaching methods courses could be connected through shared lesson-design tasks. Pre-service teachers could identify the scientific concepts underlying an engineering problem, anticipate students’ misconceptions, and plan appropriate scaffolding and assessment. This emphasis is supported by Berisha and Vula (2024), whose collaborative STEM course helped pre-service teachers develop integrated lesson plans but also revealed continuing difficulties with formative assessment.

 

The technology-related dimensions highlight the different forms of knowledge involved in using digital tools for integrated STEM teaching. TK provides a foundation for selecting and operating relevant technologies. TCK connects these technologies with disciplinary content by emphasizing how digital representations make scientific phenomena and relationships accessible to learners. TPK concerns how technology supports inquiry, data analysis, and collaboration, while TPACK coordinates these uses with interdisciplinary content and pedagogical goals in integrated lesson design. This differentiation complements Chai et al.’s (2019) examination of technology-related knowledge in science, mathematics, engineering, and integrative STEM teaching. It also highlights the value of preparing pre-service teachers to evaluate the scientific accuracy of digital representations and justify their instructional choices. Teacher education activities can support this preparation by asking students to compare simulations, interpret sensor-generated data, and incorporate appropriate tools into STEM lessons addressing authentic problems.

 

The addition of PST represents the model’s focus on learning to teach during initial teacher preparation. Its items address applying coursework, learning from classroom observations, and revising lesson plans through feedback. PST therefore makes preparation-related processes visible alongside the knowledge required for instructional design. This focus corresponds with Zhang and Zhu’s (2023) identification of interdisciplinary preparation, assessment, and technology integration as important directions in STEM pre-service teacher education. Nevertheless, PST’s conceptual status warrants further clarification: some items may capture reflective learning capabilities rather than knowledge itself.

 

Practically, the model offers a provisional framework for organizing pre-service STEM preparation around connected knowledge domains. Teacher educators could use the dimensions to structure activities that combine interdisciplinary problem analysis, technology selection, lesson planning, microteaching, and feedback-based revision. These are proposed applications rather than effects demonstrated by this study. Because the questionnaire relies on self-reports, its scores primarily reflect perceived knowledge and capabilities; they should be supplemented with lesson-plan evaluations, classroom observations, and knowledge tasks. Moreover, the correlated first-order CFA does not establish a higher-order knowledge factor or justify interpreting a total score as a single construct. Longitudinal and multi-institutional studies are needed to examine how the dimensions develop during teacher preparation and how they relate to the quality of integrated STEM teaching.

 

6. Implications

 

This study offers a framework for understanding pre-service science teachers’ perceived knowledge for integrated STEM teaching. Theoretically, it contextualizes the seven TPACK domains within interdisciplinary STEM instruction and incorporates preparation-related knowledge. This extension draws attention to how pre-service teachers connect coursework, classroom observations, and reflective revision with instructional planning. The model therefore provides a basis for examining both the knowledge involved in STEM lesson design and the preparation processes through which prospective teachers learn to apply it.

 

Practically, the questionnaire could help teacher educators identify perceived strengths and preparation needs across the eight dimensions and plan targeted learning activities. The positive relationships among the dimensions suggest the value of creating opportunities to connect disciplinary knowledge, pedagogy, technology, and reflective learning. Science content, STEM teaching methods, and educational technology courses could incorporate shared tasks in which pre-service teachers analyze an authentic problem, identify relevant disciplinary connections, select suitable digital tools, and develop an integrated lesson. Microteaching, mentor feedback, and peer discussion could then support revision of these plans. The dimension-level scores should be interpreted as indicators of perceived knowledge and supplemented with lesson-plan evaluations and teaching observations.

Teacher preparation should also emphasize students’ perspectives. pre-service teachers need opportunities to anticipate difficulties learners may encounter when connecting concepts across STEM disciplines and to plan appropriate scaffolding. Analyzing classroom examples and students’ work could help them select accessible representations, formulate guiding questions, and monitor students’ reasoning during inquiry and engineering design. Together, these activities offer a practical way to use the proposed framework to organize coherent, learner-responsive STEM teacher preparation.

 

7. Conclusion

 

Preparing science teachers for integrated STEM teaching requires attention to disciplinary connections, pedagogical decisions, technology use, and professional learning during initial teacher education. This study developed a TPACK-based knowledge model and a questionnaire encompassing eight related dimensions for pre-service science teachers. Expert consultation, exploratory factor analysis, and confirmatory factor analysis provided preliminary evidence supporting the proposed structure, convergent validity, and composite reliability. The model contributes by making interdisciplinary content knowledge explicit within a STEM teaching context and incorporating knowledge associated with pre-service preparation. The findings also indicate that the dimensions are interconnected, although the boundary between TCK and PST requires further clarification. The questionnaire offers a provisional tool for examining perceived preparation needs and informing teacher education activities. Further validation using diverse samples and performance-based evidence is needed before drawing stronger conclusions about its capacity to assess actual teaching knowledge or predict instructional quality.

 

 

Author Contributions: Hong Xuchun: Conceptualization, Methodology, Formal analysis, Writing original draft. Li Hongzhe: Validation, Review & editing. Zhang Yuting: Supervision, Project administration.

 

Funding: This research is funded by Zhejiang Philosophy and Social Sciences Planning Project in 2026, Special Project for Reform of Basic Scientific Research Expenses in Universities, No. 26NDJC013YBMS

 

Conflict of interest: The researchers declare no conflict of interest in this study.

 

Informed Consent Statement: This study was approved by the Ethics Committee of Zhejiang Normal University.

 

Data availability statement: Data will be made available on request.

 

Declaration of Generative AI and AI-assisted Technologies: No generative AI tools or technologies were used in preparing this manuscript.


References

  1. Aumann, A., Schnebel, S., & Weitzel, H. (2023). The EnTPACK rubric: Development, validation, and reliability of an instrument for measuring pre-service science teachers’ enacted TPACK. Frontiers in Education, 8, Article 1190152. https://doi.org/10.3389/feduc.2023.1190152

  2. Berisha, F., & Vula, E. (2024). Introduction of integrated STEM education to pre-service teachers through collaborative action research practices. International Journal of Science and Mathematics Education, 22(5), 1127–1150. https://doi.org/10.1007/s10763-023-10417-3

  3. Chai, C. S., Jong, M. S. Y., & Yan, Z. (2020). Surveying Chinese teachers’ technological pedagogical STEM knowledge: A pilot validation of STEM-TPACK survey. International Journal of Mobile Learning and Organisation, 14(2), 203–214. https://doi.org/10.1504/IJMLO.2020.106179

  4. Chai, C. S., Jong, M. S. Y., Yin, H. B., Chen, M., & Zhou, W. (2019). Validating and modelling teachers’ technological pedagogical content knowledge for integrative science, technology, engineering and mathematics education. Educational Technology & Society, 22(3), 61–73. https://www.j-ets.net/collection/published-issues/22_3

  5. Chai, C. S., Jong, M., & Yan, Z. (2020). Surveying Chinese teachers’ technological pedagogical STEM knowledge: A pilot validation of STEM-TPACK survey. International Journal of Mobile Learning and Organisation, 14(2), 203–214. https://doi.org/10.1504/IJMLO.2020.106181

  6. Dong, Y., Wang, J., Yang, Y., & Kurup, P. M. (2020). Understanding intrinsic challenges to STEM instructional practices for Chinese teachers based on their beliefs and knowledge base. International Journal of STEM Education, 7, Article 47. https://doi.org/10.1186/s40594-020-00245-0

  7. Dökme, I., & Ünlü, Z. K. (2024). An effective intervention using STEM instruction for improving pre-service science teachers' self-efficacy. Research in Science & Technological Education, 42(4), 930–951. https://doi.org/10.1080/02635143.2023.2209847

  8. Korthagen, F. A. J. (2010). Situated learning theory and the pedagogy of teacher education: Towards an integrative view of teacher behavior and teacher learning. Teaching and Teacher Education, 26(1), 98–106. https://doi.org/10.1016/j.tate.2009.05.001

  9. Kelley, T. R., & Knowles, J. G. (2016). A conceptual framework for integrated STEM education. International Journal of STEM Education, 3, Article 11. https://doi.org/10.1186/s40594-016-0046-z

  10. Kline, R. B. (2023). Principles and practice of structural equation modeling (5th ed.). Guilford Press.

  11. Kruse, J., Voss, S., Easter, J., Kent-Schneider, I., Menke, L., Owens, D., Roberts, K., & Woodward, L. (2025). Preparing students for the modern information landscape and navigating science-technology-society issues. Journal of Research in Science Teaching, 62(3), 792–824. https://doi.org/10.1002/tea.21972

  12. Mansour, N., Said, Z., Çevik, M., & Abu-Tineh, A. (2024). Science and mathematics teachers’ integration of TPACK in STEM subjects in Qatar: A structural equation modeling study. Education Sciences, 14(10), Article 1138. https://doi.org/10.3390/educsci14101138

  13. Mishra, P., & Koehler, M. J. (2006). Technological pedagogical content knowledge: A framework for teacher knowledge. Teachers College Record, 108(6), 1017–1054. https://doi.org/10.1111/j.1467-9620.2006.00684.x

  14. Nipyrakis, A., Stavrou, D., & Avraamidou, L. (2025). Examining S-T-E-M teachers’ design of integrated STEM lesson plans. International Journal of Science and Mathematics Education, 23(2),537–560. https://doi.org/10.1007/s10763-024-10474-2

  15. Polit, D. F., & Beck, C. T. (2006). The content validity index: Are you sure you know what’s being reported? Critique and recommendations. Research in Nursing & Health, 29(5), 489–497. https://doi.org/10.1002/nur.20147

  16. Portillo-Blanco, A., Deprez, H., De Cock, M., Guisasola, J., & Zuza, K. (2024). A systematic literature review of integrated STEM education: Uncovering consensus and diversity in principles and characteristics. Education Sciences, 14(9), Article 1028. https://doi.org/10.3390/educsci14091028

  17. Stevenson, E., van Driel, J., & Millar, V. (2026). How to support teacher learning of integrated STEM curriculum design. Journal for STEM Education Research, 9(1), 59-84. https://doi.org/10.1007/s41979-024-00133-0

  18. Zhang, Y., & Zhu, J. (2023). STEM pre-service teacher education: A review of research trends in the past ten years. EURASIA Journal of Mathematics, Science and Technology Education, 19(7), Article em2292. https://doi.org/10.29333/ejmste/13300

bottom of page