<![CDATA[ Vocational competency assessment in vocational high schools (SMK) plays a strategic role in ensuring graduates' readiness to enter the workforce; however, the quality of assessment instruments used is often not scientifically standardized. Without systematic validation, assessment instruments risk producing biased competency data that cannot be scientifically justified, leading to inaccurate decisions regarding student graduation and certification. This issue becomes particularly critical in the context of SMK especially in the Computer and Network Engineering (TKJ) program as assessments are practice-based, involving complex rubrics that are highly susceptible to rater subjectivity. This study aims to evaluate the validity and reliability of the TKJ assessment instrument at SMKN 1 Sambeng and to design a new instrument based on the evaluation findings. A quantitative approach with an instrument validation design was employed through three phases: (1) content validation by five experts using Aiken's V, (2) empirical reliability testing on 30 students using Cronbach’s Alpha along with item-total correlation analysis as evidence of internal consistency significance, and (3) design and re-evaluation of the new instrument. The results indicate that the original instrument had content validity at the minimum threshold (CVI = 0.82), with six aspects falling below the V ≥ 0.75 cutoff, and inadequate reliability (α = 0.502). The newly designed instrument featuring balanced scoring weights (5 points per sub-aspect) and a three-level discrete descriptor system (0/3/5) demonstrated significant improvement: CVI = 0.94 and Cronbach’s Alpha = 0.844 (categorized as Very High), with all item-total correlations exceeding 0.65, indicating statistically strong internal consistency. This study carries practical implications for teachers and vocational schools: the validated instrument can be directly adopted as a fair and standardized reference for TKJ practical assessments, while simultaneously serving as a replicable validation procedure model applicable to other vocational programs within SMK. ABSTRAK Penilaian kompetensi kejuruan di SMK memegang peran strategis dalam menjamin kesiapan lulusan memasuki dunia kerja, namun kualitas instrumen yang digunakan kerap belum terstandarisasi secara ilmiah. Tanpa validasi sistematis, instrumen penilaian berisiko menghasilkan data kompetensi yang bias dan tidak dapat dipertanggungjawabkan, sehingga keputusan kelulusan maupun sertifikasi siswa menjadi tidak akurat. Konteks SMK khususnya program Teknik Komputer dan Jaringan (TKJ) menjadikan persoalan ini semakin krusial karena penilaian bersifat praktik dengan rubrik yang kompleks dan rentan terhadap subjektivitas penilai. Penelitian ini bertujuan mengevaluasi validitas dan reliabilitas instrumen penilaian TKJ di SMKN 1 Sambeng, serta merancang instrumen baru berdasarkan temuan evaluasi. Pendekatan kuantitatif dengan desain validasi instrumen digunakan melalui tiga fase: (1) validasi konten oleh lima ahli menggunakan Aiken's V, (2) uji reliabilitas empiris terhadap 30 siswa menggunakan Cronbach’s Alpha beserta uji korelasi item-total sebagai bukti signifikansi konsistensi internal, dan (3) perancangan serta evaluasi ulang instrumen baru. Hasil menunjukkan instrumen lama memiliki validitas konten di batas minimum (CVI = 0,82) dengan enam aspek di bawah ambang batas V ≥ 0,75, serta reliabilitas yang belum memadai (α = 0,502). Instrumen baru yang dirancang dengan bobot seimbang (5 poin per sub-aspek) dan deskriptor diskret tiga level (0/3/5) menghasilkan peningkatan signifikan: CVI = 0,94 dan Cronbach’s Alpha = 0,844 (kategori Sangat Tinggi), dengan seluruh korelasi item-total di atas 0,65 yang mengindikasikan konsistensi internal yang kuat secara statistik. Penelitian ini memberi implikasi praktis bagi guru dan sekolah kejuruan: instrumen tervalidasi dapat langsung diadopsi sebagai acuan penilaian praktik TKJ yang adil dan terstandar, sekaligus menjadi model prosedur validasi yang dapat direplikasi pada program keahlian lain di SMK.]]>
