Jenis penelitian yang digunakan dalam penelitian ini adalah Research and Development desain Formative Evaluation (Tessmer dalam Rahayu 2008). Subjek penelitian adalah Kelas V. Hasil Penelitian terdapat 30 butir soal berupa pilihan ganda, langkah-langkah awal yaitu dilakukannya validasi oleh tim ahli yang pertama ahli isi sebesar 95% dalam kategori sangat layak, yang kedua ahli kostruk sebesar 77.3% dalam kategri layak, yang ketiga ahli bahasa sebesar 92.9% dalam kategori sangat layak. uji coba one to one dilakukan bersamaan dengan validasi ahli uji coba one to one memperoleh respon siswa sebesar 92.5% dalam kategori sangat layak, berdaarkan masukan dari tim ahli selanjudnya item pertanyaan yang redaksi belum sesuai perlu diperbaiki, setelah itu dilakukan nya uji coba small group hasil uji coba small group yaitu berupa respon siswa sebesar 94% dalam kategori sangat layak, berdasarkan masukan dan komentar small group instrumen diperbaiki dan selanjudnya dilakukannya uji coba keseluruhan kepada 48 siswa SDN 047 Tarakan. Hasil analisis program SPSS21 menunjukkan 21 butir pertanyaan dalam kategori valid, reliabilitas yaitu r11 = 0,823 > rtabel = 0,284 dalam kategori sangat tinggi dalam kriteria reliabilitas. Hasil analisis program AnatesV4 menunjukkan tingkat kesukaran dengan kategori sukar sejumlah 4 item soal (13,3%), kategori sangat sukar sejumlah 1 item soal (3.3%), kategori mudah sejumlah 5 item soal (16.7%), kategori sangat mudah sejumlah 2 item soal (6,7%), dan kategori sedang sejumlah 18 item soal (60%), hasil daya beda rata-rata dengan kriteria baik sejumlah 17 soal (56,6%), dan menghasilkan pengecoh efektif sebanyak 27 item soal (90%).

Type of research used in this research is Research and Development design Formative Evaluation (Tessmer in Rahayu 2008). The subjects of the study were Class V. The result of the study was 30 items of multiple choice, the initial steps were validation by the first expert team of 95% content experts in the very feasible category, the second was 77.3% in the appropriate category, The three linguists are 92.9% in a very decent category. One-to-one test is conducted simultaneously with one-to-one expert test validation of 92.5% of the students in the category is very feasible, based on the input from the expert team so far the question items that the editorial has not been suitable need to be improved, after which the small group test result Small group trials were in the form of a 94% student response in a very viable category, based on feedback and small group comments of the instrument were improved and the whole experiment was conducted to 48 students of SDN 047 Tarakan. The result of SPSS21 program analysis shows 21 items of question in valid category, reliability is r11 = 0,823> rtabel = 0,284 in very high category in criterion of reliability. The results of AnatesV4 program analysis show difficulty level with difficult category of 4 items about (13,3%), very difficult category of 1 item about (3.3%), easy category of 5 item questions (16.7%), very easy category of 2 items The problem (6.7%), and the moderate category of 18 items question (60%), average power difference with good criteria of 17 questions (56,6%), and result effective effectiveness 27 items (90%).