Elementary teachers’ perceptions and readiness: Insights from a computer science professional development in Uruguay

• In Event: Cultural Connections and Global Perspectives in Teacher Training

Proposal

Integrating computer science (CS) into elementary education is increasingly recognized as essential for preparing young learners for future challenges, equipping children with domain skills like problem-solving and critical thinking (Kite et al., 2021; Graziano et al., 2023). Among global efforts, Uruguay seeks to integrate digital technologies to promote equity and inclusion for marginalized groups, aiming to improve learning outcomes and foster innovation and students’ and teachers’ personal growth (Jara et al., 2018; ANEP, 2022). However, formal training opportunities for elementary teachers in CS are still challenging, leading to a deficit in pedagogical and content knowledge, affecting their confidence (Kaya et al., 2019; Mouza et al., 2022b)
Effective PD programs offer teachers opportunities to reflect on their current practices, challenge their beliefs, and adopt new teaching approaches, ensuring they are well-prepared to implement pedagogical content effectively (Darling-Hammond et al., 2017). Research emphasizes the importance of targeted PD to enhance teachers' readiness and self-efficacy (TSE), which refers to teachers’ beliefs in their ability to achieve desired outcomes (Bandura, 1997; Williams et al., 2023). In STEM education, TSE is crucial for effectively teaching computational thinking (CT) and coding skills (Tang et al., 2020). However, integrating those faces barriers beyond instructional challenges. Teachers often need more confidence and familiarity with effective instructional approaches (Margot & Kettler, 2019; Mouza et al., 2022).
In partnership with a US-based research team, a Uruguayan organization localized and implemented a PD session to introduce elementary teachers to CT and coding concepts. Spanning 120 hours over 12 weeks, the program incorporated virtual asynchronous and synchronous sessions structured around a previously translated (into Spanish) version of the Coding as Another Language (CAL) curriculum for ScratchJr, which emphasizes coding as a literacy for the 21st century and fosters positive interactions in technology-rich environments. The modules on EduX included CS fundamentals, localized activities, assessments, and preparation for classroom implementation.
Pre- and post-PD training surveys were administered to assess teachers’ readiness and attitudes toward teaching CS. These surveys, adapted from previous studies conducted in the U.S. and Argentina (Authors et al., 2023, 2024), underwent review to ensure contextual and linguistic accuracy. The pre-training survey focused on TSE and demographic data, while the post-training survey evaluated TSE, readiness, and satisfaction with the PD. The analytical sample consisted of 555 participants—however, the thematic analysis involved data from 491 open responses.
The study employed a mixed-methods sequential explanatory design (Creswell & Creswell, 2018), starting with Confirmatory Factor Analysis (CFA) with Maximum Likelihood to validate the self-efficacy, readiness, and satisfaction with PD constructs. Mean scores from these were then used in path analysis with Bootstrapping to investigate how readiness and satisfaction with synchronous and asynchronous training predicted post-training TSE, controlling for pre-training self-efficacy and demographic variables.
A Chi-square goodness of fit test and other indices (RMSEA, CFI, SRMR) evaluated model fit. Data normality was checked using squared Mahalanobis distances and outlier handling via Winsorization. No multicollinearity issues were detected (determinant = 0.06). The sample size exceeded the recommended guidelines with 571 participants and 76 variables for CFA, 562 participants, and 26 for path analysis. The thematic analysis of teachers’ beliefs about improvements and recommendations for future PD iterations revealed four themes: 1. Meetings, 2. Format and Content, 3. Technology, and 4. Implementation.
CFA demonstrated that the latent constructs effectively measured their intended aspects. Overall, the fit indices (CFI = 0.94; RMSEA = 0.07, 90% CI = [0.07, 0.08]; SRMR = 0.05) indicated the model’s acceptability and absence of misspecification (Hooper et al., 2008). It’s important to note that while the Chi-square goodness of fit test showed significance, χ2(200) = 819.43, p < .001, it is known to be sensitive to sample size. All observed variables had R2 values greater than .20, confirming their adequate representation of the latent constructs (Hooper et al., 2008). Similar findings were observed in the path analysis. The fit indices (CFI = 0.98; RMSEA = 0.05, 90% CI = [0.03, 0.06]; SRMR = 0.06) indicated that the model was not misspecified, though the Chi-square goodness of fit test showed significance, χ2(24) = 54.87, p < .001.
The path analysis showed teachers’ Readiness to integrate CS concepts in their classrooms significantly predicted post-training TSE, B = 0.68, z = 19.42, p < .001. Similarly, Satisfaction with the asynchronous sessions significantly predicted post-training TSE, B = 0.32, z = 6.64, p < .001. However, Satisfaction with the synchronous sessions did not significantly predict post-training TSE, B = -0.04, z = -0.98, p = .325. Pre-training TSE did not significantly predict post-training TSE, p = .171, nor gender, p = .746, or years of teaching experience, p = .752, significantly predicted pre-training TSE. In contrast, participants with a STEM background, B = 0.27, z = 3.53, p < .001, and prior programming experience, B = 0.41, z = 5.83, p < .001, showed significantly higher pre-training TSE. These findings were contextualized with insights from participants’ open-ended responses to understand the results comprehensively.
Under the Meetings theme, 21% highlighted the value of real-time interactions with experts, while 4% suggested more flexible synchronous meeting times, referring to logistical challenges outside work hours. Regarding Format and Content, 10% desired more guided practice and explicit examples throughout the PD. Some teachers sought deeper explanations and ongoing support beyond the course. Technology-related feedback (3%) cited difficulties with EduX and advocated for familiar or improved communication channels. Another 7% emphasized the need for more classroom devices. A positive note (4%) was teachers’ interest in learning diverse programming languages beyond ScratchJr. Implementation feedback (9%) underscored the need for on-site support during classroom integration of PD learnings. Participants valued simultaneous learning and implementation, suggesting it enriched their experiences.
The study underscores critical implications and recommendations for advancing teacher training and research in elementary education. The results advocate for culturally and linguistically tailored virtual PD to enhance accessibility and scalability and accommodate diverse learning styles and schedules. They also highlight the benefits of real-time interactions in fostering community collaboration and enhancing TSE.

Authors

• Francisca Carocca, Boston College, Lynch School of Edu

• Ghaida Alrawashdeh, Oakland University