Exploratory regression models showed both technical dimensions were significantly associated with active-learning perceptions, while only active learning was significantly associated with self-reported knowledge of programming and machine learning¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r1 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Model 1 predicted active learning from knowledge of programming and machine learning (β =.423, p <.001) and block-based programming and game engines (β =.434, p <.001), accounting for R² =.558. → Susana Sánchez Castro 2026 q2 i? Model 2 predicted knowledge of programming and machine learning, with only active learning significant (β =.540, p <.001); the block-based predictor was not significant (p =.095). → Susana Sánchez Castro 2026 (2)

Evidence¶

Susana Sánchez Castro 2026¶

Susana Sánchez Castro, Mª Ángeles Pascual Sevillano, Rogelio Buceta Otero, Javier Fombona Cadavieco. (2026). Developing computational thinking in teacher education: a brief experience with coding, robotics, and machine learning. Education Tech Research Dev. https://doi.org/10.1007/s11423-026-10744-6

q2 · i? · causal · r1

Exploratory multiple linear regression on experimental-group data (n = 105) with active learning as the dependent variable; both predictors showed similar positive associations. R² and β are reported as printed; no standardized effect size beyond these coefficients is given.

"This model accounted for a substantial proportion of variance (R 2 =.558; F (2, 102) = 64.31, p <.001), with significant associations with knowledge of programming and machine learning (β =.423, p <.001) and block-based programming and game engines (β =.434, p <.001)."

Susana Sánchez Castro 2026 (2)¶

Susana Sánchez Castro, Mª Ángeles Pascual Sevillano, Rogelio Buceta Otero, Javier Fombona Cadavieco. (2026). Developing computational thinking in teacher education: a brief experience with coding, robotics, and machine learning. Education Tech Research Dev. https://doi.org/10.1007/s11423-026-10744-6

q2 · i? · causal · r1

Second exploratory regression on experimental-group data with knowledge of programming and machine learning as the dependent variable; the block-based programming predictor was non-significant (β =.166, p =.095 as printed in Table 6).

"This model was also significant (R2 =.435; F(2, 102) = 39.33, p <.001), although only the active learning dimension showed a statistically significant association with self-reported knowledge of programming and machine learning (β =.540, p <.001)."

Discussion¶

Related Claims¶

• Among workshop participants, the three questionnaire dimensions were positively and significantly intercorrelated, with the highest association between block-based programming familiarity and active-learning perceptions — related
• Favorable overall self-reported perceptions coexisted with lower self-reported familiarity in specific technical areas: machine learning, prior block-based programming experience, and game-engine programming — related