Tujuan dari penelitian ini adalah untuk mengetahui apakah terdapat pengaruh yang signifikan terhadap pencapaian nilai siswa kelas xi sma negeri 3 tarakan dalam menulis teks narasi setelah siswa diajarkan menggunakan 6+1 trait writing model. penelitian ini menggunakan desain quasi experimental. sampel dalam penelitian ini yaitu kelas xi ipa3 dan xi ipa4 yang dipilih menggunakan purposive sampling. setiap kelas terdiri dari 33 siswa sehingga total sampel berjumlah 66 siswa. pretest diberikan kepada kelas xi ipa3 dan xi ipa4. setelah memberikan pretest terhadap kedua kelas tersebut, peneliti memberikan treatment menggunakan 6+1 trait writing model pada kelas eksperimen atau xi ipa3 dan four square writing method pada kelas control atau xi ipa4. selanjutnya untuk mengetahui hasil pencapaian siswa pada kedua sampel tersebut, peneliti memberikan posstest, dimana siswa diminta untuk menuliskan karangan narasi nonfiksi. peneliti menganalisis nilai siswa pada pretest dan posttest menggunakan spss versi 22. hasil penelitian menunjukkan uji normalitas dan homogenitas pada kedua kelas lebih tinggi dari level of significance (>0,05), jadi ho diterima dan ho ditolak. hasil ini megindikasikan bahwa nilai siswa dikelas eksperimen dan kontrol terdistribusi secara normal dan homogen. sementara itu, hasil perhitungan formula t-test pada pretest menunjukkan probability value lebih tinggi dari level of significance (0,453>0,05) dan t-test lebih rendah dari t-table (0,755

The objective of conducting this research was to find out whether there was significant effect of using the 6+1 trait writing model in increasing students’ achievement in writing narrative text at eleventh grade students of sma negeri 3 tarakan in academic year 2016/2017. the research design was used quasi experimental design. the samples are xiipa3 and xi ipa4 were taken by using purposive sampling. each class consists of 33 students. thus the total of the samples were 66 students. pretest was carried out to class xi ipa3 and xi ipa4. after gave pretest to both of the classes, the researcher gave treatment using 6+1 trait writing model in experimental class or xi ipa3 and four square writing method in control class or xi ipa4. furthermore to measure students achievement in both of the samples, the researcher carried out posttest, where the students were asked to write their personal narrative essay. the researcher analyzed students’ score by using spss version 22. the research findings showed the normality test and homogeneity of variance in both of the classes were higher than level of significance (>0,05), then ha was accepted and ho was rejected. it means pretest and posttest scores in both of the classes were normally distributed and homogenous. meanwhile, the result computation of pretest showed that probability value was higher than level of significance (0,453> 0,05) and t-test was lower than t-table (0,755