<![CDATA[Contemporary early childhood assessment practices remain dominated by structured testing protocols that often fail to capture children’s cognitive competencies as authentically demonstrated in naturalistic outdoor learning contexts. This study developed and validated a comprehensive naturalistic cognitive assessment model for children aged 5–6 years through design-based research integrating mixed-methods procedures over a 14-month period. The Naturalistic Cognitive Observation Protocol (NCOP) was developed to document cognitive behaviors across five domains during outdoor learning activities. Participants included 180 children from six early childhood education centers in East Jakarta, 24 teachers, and 12 expert validators. The model underwent rigorous psychometric validation and an eight-week field implementation. Results demonstrated excellent validity and reliability, with a content validity index (CVI) of 0.93, inter-rater reliability of ICC = 0.86 and Cohen’s κ = 0.83, and internal consistency of α = 0.91. Children demonstrated meaningful cognitive development across all domains, with moderate to large effect sizes (Cohen’s d = 0.54–0.71). Notably, 78% of the observed cognitive behaviors emerged during child-initiated activities rather than teacher-directed tasks, while 89% of metacognitive verbalizations occurred during peer interactions. Most participating teachers (91.7%) reported an improved understanding of children’s cognitive capabilities through naturalistic observation. This study demonstrates that outdoor-based naturalistic assessment can achieve ecological validity without compromising psychometric integrity, providing a validated, developmentally appropriate tool that meets accountability demands and captures cognitive development more comprehensively than conventional assessment.]]>
