<![CDATA[[English]: The study aims to analyze the quality of problem-solving test items and a self-efficacy questionnaire using the AnBuso. The problem-solving test items were developed to measure the students’ cognitive abilities ranging from understanding (C2) to evaluating (C5) in statistical topics. This study employed a quantitative descriptive approach focusing on eight-grade students at a public school SMP Negeri 1 Mlati with a sample consisted of 32 students. The analysis was proceeded in terms of test item characteristics, including discriminative power, difficulty level, expert validity, and item reliability. The findings indicated that 53.33% of the items had good discriminative function, 20% had fairly good function, and 26.67% had poor function. A total of 33.33% of the items were classified as easy, 66.67% as moderate, and none as difficult. The validity was 0.90 (highly valid), and the reliability of multiple-choice items was 0.51 (moderate), while open-ended items reached 0.72 (high). Meanwhile, the self-efficacy questionnaire had a validity of 0.89 (highly valid) and a reliability of 0.83 (very high). Overall, the test instruments and questionnaire showed high validity and adequate reliability. It is hoped that the research finding can serve as a reference for educators and curriculum developers in designing better evaluation tools so that the learning objectives can be optimally achieved. [Bahasa]: Penelitian ini bertujuan menganalisis kualitas butir soal kemampuan pemecahan masalah dan angket self-efficacy berbantuan aplikasi AnBuso. Soal kemampuan pemecahan masalah dikembangkan untuk mengukur kemampuan kognitif siswa pada level memahami (C2) hingga mengevaluasi (C5) pada materi statistika. Metode penelitian ini menggunakan pendekatan deskriptif kuantitatif dengan fokus utama pada siswa kelas VIII di SMP Negeri 1 Mlati dengan sampel sebanyak 32 orang siswa. Proses analisis dilakukan dari segi karakteristik soal yaitu daya pembeda, tingkat kesulitan kemudian validitas ahli, dan reliabilitas soal. Temuan penelitian menunjukkan bahwa instrumen pemecahan masalah memiliki daya pembeda yaitu 53.33% soal memiliki fungsi pembeda yang baik, 20% cukup baik, dan 26.67% tidak baik. Sebanyak 33.33% soal tergolong mudah, 66.67% sedang, dan tidak ada soal yang tergolong sulit. Kevalidan sebesar 0.90 (sangat valid) dan reliabilitas soal pilihan ganda sebesar 0.51 (sedang), soal uraian 0.72 (tinggi). Sementara angket self-efficacy memiliki kevalidan sebesar 0.89 (sangat valid) dan reliabilitas sebesar 0.83 (sangat tinggi). Secara keseluruhan, instrumen tes dan angket yang dikembangkan menunjukkan tingkat kevalidan yang tinggi serta reliabilitas yang memadai. Hasil analisis ini diharapkan dapat menjadi acuan bagi pendidik dan pengembang kurikulum dalam menyusun alat evaluasi yang lebih baik, sehingga tujuan pembelajaran dapat tercapai secara optimal. ]]>
