<![CDATA[This study investigated the interrelationships among teachers' beliefs, perceived knowledge, and self-reported practices concerning formative continuous assessment (FCA) in Gondar City, Ethiopia. Employing a quantitative correlational design, data were collected from 182 grade 5–8 teachers across four subject areas (Amharic, English, Mathematics, and Biology) through a structured, expert-validated questionnaire using a four-point Likert scale (response rate = 91.0%). Pearson product-moment correlation analyses, supported by bias-corrected accelerated (BCa) bootstrapping with 2,000 resamples, revealed statistically significant positive associations among all three constructs. The most robust relationship emerged between perceived knowledge and practice (r = .608, p < .01), followed by the belief–knowledge link (r = .568, p < .01), and a moderate yet substantively meaningful belief–practice association (r = .461, p < .01). Subsequent multiple regression analysis identified perceived knowledge as the primary predictor of FCA practice (β = .512, p < .001), with the overall model accounting for approximately 38.9% of the variance in practice (adjusted R² = .383). Teacher beliefs also contributed a statistically significant, albeit smaller, independent effect (β = .170, p = .018). These findings indicate that teachers' self-assessed FCA knowledge functions as the principal driver of classroom implementation within Ethiopia's newly adopted curriculum, underscoring an urgent need for targeted, practice-embedded professional development designed to strengthen formative assessment literacy.]]>
