<![CDATA[Abstrak Berpikir komputasional relevan sebagai bekal memecahkan masalah tidak rutin. Meskipun siswa berkemampuan rendah sering kesulitan mengerjakan soal TKA akibat hambatan kognitif. Penelitian ini bertujuan mendeskripsikan profil CT siswa kelas V berkemampuan rendah serta scaffolding yang diperlukan. Pendekatan kualitatif studi kasus melibatkan delapan siswa dari tiga SD di Cirebon. Instrumen soal berbasis dekomposisi, pengenalan pola, abstraksi, dan algoritma, observasi, wawancara. Hasil menunjukkan siswa mampu mengenali informasi numerik, namun hambatan terlihat pada abstraksi, penguraian masalah, penyusunan langkah sistematis. Kesalahan utama: memilah informasi sangat penting, kebingungan prosedural, kesenjangan lisan-tulisan. Dua pola revisiting: kembali ke reduksi data atau penyajian data. Scaffolding kategori: struktural, metakognitif, teknologis. Penelitian memberi panduan praktis guru merancang pembelajaran CT bertahap, eksplisit, adaptif terhadap pola revisiting siswa. Abstract Computational thinking (CT) is increasingly relevant as a key skill for solving non-routine problems. However, low-achieving students often struggle with CT-based tasks due to cognitive barriers. This study aims to describe the CT profile of fifth-grade low-achieving students and identify the scaffolding needed to overcome their difficulties. Using a qualitative case study approach, this research involved eight students from three elementary schools in Cirebon. Instruments included CT questions based on decomposition, pattern recognition, abstraction, and algorithms, complemented by observation and interviews. Results showed students could recognize numerical information, but obstacles appeared in abstraction, problem decomposition, and systematic step formulation. Main errors included inability to filter important information, procedural confusion, and gaps between oral and written explanations. Two revisiting patterns were found: returning from conclusion to data reduction or to data display. The scaffolding needed includes structural, metacognitive, and technological categories. This research contributes a practical guide for teachers to design CT learning that is gradual, explicit, and adaptive to students' revisiting patterns.]]>
