<![CDATA[Penelitian ini bertujuan untuk mengetahui karakteristik butir soal materi penggabungan badan usaha pada mata kuliah Akuntansi Keuangan Lanjutan I. Penelitian menggunakan pendekatan kuantitatif dengan desain deskriptif eksploratif, melibatkan 40 mahasiswa sebagai subjek. Instrumen berupa tes pilihan ganda terdiri dari 30 butir soal dianalisis menggunakan teori tes klasik dan model Rasch. Hasil validitas isi melalui indeks V Aiken menunjukkan 24 butir (80%) sangat valid, dan 6 butir (20%) validitas sedang. Secara kualitatif, butir soal memenuhi aspek materi (70%), konstruksi (79%), dan bahasa (97%). Berdasarkan teori tes klasik, 60% butir berada pada tingkat kesukaran sedang, 67% memiliki daya pembeda baik, dan 41,67% pengecoh berfungsi optimal. Analisis faktor menunjukkan 25 butir (83%) signifikan, dengan reliabilitas instrumen tinggi (Cronbach Alpha 0,835). Terdapat satu butir mengalami bias (p < 0,05). Dengan model Rasch, 28 butir (97%) sesuai model dan tingkat kesukaran berada dalam rentang sedang (b antara -2 hingga +2). Temuan menunjukkan sebagian besar butir mampu memberikan informasi akurat bagi mahasiswa dengan kemampuan menengah. Simpulan, instrumen soal menunjukkan kualitas baik baik dari sisi isi maupun konstruk. Item characteristics of business entity merger material in Advanced Financial Accounting Course I Abstract: This study aims to examine the item characteristics of test questions on business combination material in the Advanced Financial Accounting I course. A quantitative approach with a descriptive-exploratory design was employed, involving 40 students as research subjects. The test instrument consisted of 30 multiple-choice items analyzed using Classical Test Theory (CTT) and the Rasch model. Content validity using Aiken’s V index showed that 24 items (80%) were highly valid, while 6 items (20%) had moderate validity. Qualitative review revealed 70% compliance in material aspects, 79% in construction, and 97% in language. Based on CTT, 60% of items had moderate difficulty, 67% demonstrated good discrimination, and 41.67% of distractors functioned effectively. Factor loading analysis showed 25 items (83%) were significant, with a high reliability score (Cronbach's Alpha = 0.835). One item showed bias (p < 0.05). The Rasch model analysis indicated that 28 items (97%) fit the model and had difficulty levels within the moderate range (b between -2 and +2). Most items provided accurate information for students with average ability levels. In conclusion, the test items demonstrated good quality in terms of both content and construct validity.]]>
