<![CDATA[Abstrak: Penelitian ini bertujuan untuk menentukan: 1) perbedaan hasil belajar siswa yang diajari model pembelajaran POE (predict-observe-explain) dengan pembelajaran konvensional pada materi koloid, 2) pengaruh model pembelajaran POE (predict-observe-explain) terhadap hasil belajar siswa SMA 1 Muhammadiyah 1 pontianak pada materi koloid. Penelitian ini merupakan penelitian eksperimen semu dengan rancangan penelitian Nonequivalent Control Group Design. Sampel dalam penelitian ini adalah siswa kelas XI MIA 3 sebagai kelas eksperimen dan kelas XI MIA 1 sebagai kelas kontrol yang diperoleh dengan teknik cluster random sampling. Alat Pengumpulan data dilakukan dengan menggunakan tes hasil belajar dan wawancara. Berdasarkan hasil olahan data, hasil rata-rata skor posttest kelas ekperimen 74,36 dan hasil rata-rata skor posttest kelas kontol 67,03. Berdasarkan uji U-Mann Whitney, diperoleh Asymp.Sig (2-tailed) = 0,023 yang berarti terdapat perbedaan hasil belajar antara siswa kelas kontrol dan siswa kelas eksperimen pada materi koloid. Hasil perhitungan effect size adalah 0,39 yang berarti model POE (predict-observe-explain) memberikan pengaruh sebesar 18,79% terhadap peningkatan hasil belajar siswa. Kata kunci : Model predict-observe-explain, koloid, hasil belajar. Abstract: The aims of this research were to determine: 1) the differences of learning outcomes between students who were taught using POE (predict-observe-explain) model and conventional learning colloidal subject, 2) the effect of POE (predict-observe-explain) model toward student learning outcomes colloidal subject. The from of research was a quasi-experimental with nonequivalent control group design. Classes of XI MIA 3 as experiment and MIA 1 as control were selected by cluster random sampling technique. Achievement test and interview was used to collect data. According to data analysis, average score of experiment and control classes were 74.36 and 67.36 respectively. According to U-Mann Withney test, the value of student of Asymp.Sig (2-tailed) was 0.023 which meant that there was a difference in learning outcome between the control class and experimental class colloidal subject. The result of effect size was 0.39 which meant POE (predict-observe-explain) model had of 18.79% influenced toward student learning outcomes improvement. Keywords: Predict observe explain, colloidal, learning outcomes.]]>
